1.0 it 1124al: 2.2

IL-I111: IIIII 2 0

1.1

2.

11111

.25 111 1.4 11.6

MICROCOPY Pi 4,11IInN WIT CHAR I

DOCONENT BEMIRE

ID 174 644TM 009 464

TITLE The Effects of Coaching on Standardized AdmissionExaminations. Staff Memorandum cf the Boston RegionalOffice of the Federal Trade Commission.

INSTITUTION Federal Trade Commission, Washington, D.C. Bureau of

Consumer Protection.PUB DATE Sep 78NOTE 192p.; Data tables may not reproduce legibly because

of the poor print quality of the original document;The technical appendices were nct publicallyavailable as of August, 1979; Fcr related document,see TN 009 465

AVAILAFLE FROM Public Reference Branch, Room 130, Federal TradeCommission, Washington, D.C. 20580 (no charge)

EDES PRICE MF01 Plus Postage. FC Not Available from EDRS.

DESCRIPTORS Achievement Gains; *College Ent-.ance Examinations;Graduate Study; Higher Education; ImprovementPrograAs; Legal Education; Predictive Validity;

*Program Effectiveness; Racial Eifferences; *Scores;Senior High Schools; Testing Problems; *TestWiseness; *Tutorial Programs

IDENTIFIERS Coaching Schools; Law School Admission Test;*Schold:=tio Aptitude Test; *Test Coaching

ABSTRACTA non-experimental design was used to determine if

scores of students enrolled in specified major coaching schools were

significantly higher than scores of comparable uncoached groups.

Score increases at two Scholastic Aptitude Test (SAT) coaching

schools and Law School Admission Test (LSAT) schools were compared.

Over 1,400 SAT examinees and 31,000 LSAT examinees, administered

these tests between 1974 and 1977, were involved. Comparisons were

reported by study subgroup: uncoached first-time examinees; coached

first-time examinees; uncoached second-time examinees; second-time

examinees coached berore +he first test; and second-time examinees

coached between test adrinistrations. Results indicated that coaching

was significantly effective in raising SAT scores. LSAT coaching was

marginally effective, possibly because of the abnormally large-

control group increases and the relatively low co-,:relation between

grade point average and LSAT scorns. However, the Federal TradeCommission acknowledges that there are "several major flaws in the

data analysis, making the results unreliable." (The history of both

tests, an analysis of coaching research; a description of the sales,

tuition costs, and claims of major coaching schools; extensive graphs

of test score data; and case histories nt 17 coached LSAT examinees

are included). (CP)

***********************************************************************Reproductions supplied by EDAS are the best that can be made

from the original document.***********************************************************************

MCTICE TO RECIPIENTS OF THEBOSTON REGIONAL OFFICE REPORT ON THE

EFFECTS OF COACHING ON STANDARDIZED ADMISSION EXAM.

U S DEPARTMENT OF HEALTH.EDUCATION S. WELFARENATIONAL INSTITUTE OF

EDUCATION

THIS DOCUMENT 1145 BEEN REPRO.DUCED EXACTLY A5 RECEIVED FROMTHE PERSON OR ORGANIZATION ORIGIN.ATING IT POINTS OF VrEev DR OPINIONSSTATED DO NOT NECESSARILY REPRE.SENT OFFICIAL NATIONAL INSTITUTE OFEDUCATION POSITION OR POLICY

The attacheC "Staff Memorandum on the Effects of Coaching on

Standardized Admission Examinations," prepared by the Boston Regional

Office of the Federal Trade Commission, has several major flaws in the

data analysis, making the results unreliable. Brief explanations of

each of the major flaws are described below.

In this report standardized exam scores (Scholastic Aptitutde Test,

or SAT, and Law School Admission Test, or LSAT) are compared for groups

of coached and uncoached students without controlling for differences

which may exist in personal and demographic characteristics of the

students in the two groups. Subsequent analysis has determined that there

are important differences between the coached and uncoached groups, making

it impossible to attribute any observed differences in SAT scores to

coaching.

To illustrate, students who are coached for the SAT have, on average,

higher grades than students who are not coached. It is not possible to

know, without controlling for grad,-s, if SAT score differences between

the two groups are due to coaching or to the differences in ability

reflected in such factors as past school performance.

In addition to failing to use the data available to control for

differences in personal and demographic characteristics of the students,

the results reported cannot 1.2 evaluated because the report fails to

provide tests of statistical significance which are necessary to interpret

the results. Because the report fails to provide these important statistic:,

one cannot determine if the results reported are meaningful or are

merely due to chance.

411S141111/1110

Another defect in the report concerns the method used to present

the findings from the data analysis. This is especially true for the

cribanalysis of the LSAT performance. In that section of the report, the

analytical technique employed is misused, causing flaws in the presentation

111121

of the results. For example, in a number of places, findings are presented

for students who have a grade point average of 1.0 (a D average).

Because there are no such students in the sample, it is not legitimate

is predict the behavior of this class of students. Moreover, because

virtually no students with a D average would ever take the LSAT exam, it

is inappropriate to discuss the findings interms of this group instead

of in terms of students with more typical grade point averages.

Finally, this study utilized a nonexperimental design, Therefore

all the limitations associatedwith this type of study apply to this

/ report.

Bureau of Consumer Protection

0, May 1379

DATE:

REPLY TOATTN OP:

September 11, 1978

Arthur E. Levine, AttorneyBoston Regional Office

UNITED STATES GOVERNMENT

memorandumProgram Code 116

OBJECT: Unnamed Test Preparation Services, Pile 772 3000Stanley H. Kaplan Educational Center, Ltd., File 762 3033Educational Testing Service, File DB8 0013

TO: Federal Trade Commission

VIA: Bureau of Consumer Protection

I am transmitting herewith the staff memorandum of theBoston Regional Office on The Effects of Coaching on StandardizedAdmission Examinations. This memorandum encompasses thethree above-captioned cases.

Buy U.S. Savings Bonds Regularly on the Payroll Savings Plan OPTIONAL roRm NO. 10

4(REV. 7-70)OSA VPMN (tt GM) 1014t.010104 IS

ACKNOWLEDGEMENT

Dr. Conrad Strack, of C.A.C.I., who gave invaluable assistance

in the preparation of some of the technical aspects of this

memorandum deserves special acknowledgement, as do lay col-

leagues in the Boston Regional Office, especially E-ancis

X. McDonough.

ii

TECHNICAL APPENDIX

Si. Uncoached 1-Time Takers of SAT

S1-1. PSAT verbal and math score means

S1-2. PSAT verbal vs math score table

S1-3. SAT verbal and math score meansS1-4. SAT verbal score histogram51-5. SAT math score histogramS1-6. SAT scores for PSAT verbal in 30's

S1-7. SAT scores for PSAT verbal in 40's

S1-8. SAT scores for PSAT verbal in 50's

S1-9. SAT scores for PSAT math in 30'sS1-10. SAT Scores for PSAT math in 40'sS1-11. SAT scores for PSAT math in 50'sS1-12. SAT verbal vs PSAT verbal tableS1-13. SAT math vs PSAT math tableS1-14. SAT verbal = f(PSAT verbal) regressionS1-15. SAT math = f(PSAT math) regression

S2. Coached 1-Time Takers of SATS2-1. PSAT verbal and math score means for

all, Stanley H. Kaplan EducationalCenter (OO1A) and Test Prep Centers(022)

S2-2. PSAT verbal vs math score table

S2-3. Enrollment counts for Stanley H. KaplanEducational Center (OO1A), College En-trance Testing Service (010) and TestPrep Centers (022)

S2-4. SAT verbal and math score means for all

S2-5. SAT verbal and math score means for

Stanley H. Kaplan Educational Center(OO1A) and Test Prep Centers (022)

S2-6. SAT verbal score histogram for all

S2-7. SAT math score histogram for all

S2-8. SAT verbal and math histograms forStanley H. Kaplan Educational Center(001A)

S2-9. SAT verbal and math histograms forTest Prep Centers (022)

S2-10. SAT scores for PSAT verbal in ?O'sS2-11. SAT scores for PSAT verbal in 40'sS2-12. SAT scores for PSAT verbal in 50'sS2-13, SAT scores for PSAT math in 30'sS2-14. SAT scores for PSAT math in 40'sS2-15. SAT scores for PSAT math in 50'sS2-16. SAT verbal vs PSAT verbal tableS2-17. SAT math vs PSAT math table

S2. . Coached 1-Time Takers of SAT (cont'd)

S2-18. SAT verbal = f(PSAT verbal) regressions

for all, Stanley H. Kaplan EducationalCenter (001A) and Test Prep Centers (022)

S2-19. SAT math = f(PSAT math) regressions for

all, Stanley H..Kaplan Educational Center(OO1A) and Test Prep Centers (022)

S3. Uncoached 2-Time Takers of. SAT

S3-1. PSAT verbal and math score means

S3-2. SAT first score, second score, andscore gain means

S3-3. First SAT verbal score histogram

S3-4. First SAT math score histogram

S3-5. Second SAT verbal score histogram

S3-6. Second SAT math score histogram

S3-7. SAT score gain histograms

S3-8. PSAT verbal vs math table

S3-9. First SAT verbal vs PSAT verbal table

S3-10. First SAT math vs PSAT math table

S3-11. First SAT verbal = f(PSAT verbal) re-

gressionS3-12. First SAT math = f(PSAT math) regres-

sionS3-13. Second SAT verbal = f(first SAT verbal)

regressionS3-14. Second SAT math = f(first SAT math)

regressionS3-15. Second SAT verbal = f(PSAT verbal)

regressionS3-16. Second SAT math = f(PSAT math) regression

S4. 2-Time Takers of SAT Coached Before First Attempt

S4-1. PSAT verbal and math score means

S4-2. SAT first score, second score, and score

gain meansS4-3. First SAT verbal score histogram

S4-4. First SAT math score histogram

S4-5. Second SAT verbal score histogram

S4-6. Second SAT math score histogram

S4-7. SAT score gain histograms

S4-8. Enrollment counts for Stanley H. Kaplan

Educational Center (OO1A), College En-

trance Testing Service (010), and Test

Prep Centers (022)'

S4-9. PSAT verbal and math score means for

Stanley H. Kaplan Educational Center(001A) and Test Prep Centers (022)

S4-10. SAT first score, second score, and

score gain means

vi

S4. 2-Time Takers of SAT Coached Before First Attempt (cont'd)S4-11. SAT first score, second score, and

score gain histograms for Stanley H.Kaplan Educational Center (OO1A) andTest Prep Centers (022)

S4-12. PSAT verbal vs math score tableS4-13. First SAT verbal vs PSAT verbal tableS4-14. First SAT math vs PSAT math tableS4-15. First SAT verbal = f(PSAT verbal) re-

gressions for Stanley H. Kaplan Educa-tional Center (OOlA) and Test PrepCenters (022)

S4-16. First SAT math = f(PSAT math) regres-sions for Stanley H. Kaplan EducationalCenter (OO1A) and Test Prep Centers (022)

S4-17. Second SAT verbal = f(first SAT verbal)for Stanley H. Kaplan Educational Center(001A) and Test Prep Centers (022)

S4-18. Second SAT math = f(PSAT math) forStanley H. Kaplan Educational Center(OO1A) and Test Prep Centers (022)

S4-19. Second SAT verbal = f(PSAT verbal)for Stanley H. Kaplan EducationalCenter (OO1A) and Test Prep Centers t022)

S4-20. Second SAT math = f(PSAT math) forStanley H. Kaplan Educational Center(OO1A) and Test Prep Centers (022)

S5. 2-Time Takers of SAT Coached Between AttemptsS5-1. PSAT verbal and math score meansS5-2. SAT first score, second score, and

score gain meansS5-3. First SAT verbal score histogramS5-4. First SAT math score histogramS5-5. Second SAT verbal score histogramS5-6. Second SAT math score histogramS5-7. SAT score gain histogramsS5-8. Enrollment counts for Stanley H.

Kaplan Educational Center (OO1A),Cellege Entrance Testing Service-(010) and Test Prep Centers (022)

S5-9. PSAT verbal and math score meansfor Stanley H. Kaplan EducationalCenter (OOlA) and Test Prep Centers(022)

S5-10. SAT first score, second score, andscore gain means for Stanley H. KaplanEducational Center (OO1A) and TestPrep Centers (022)

vii

S5. 2-Time Takers of SAT Coached Between Attempts (coned)S5-11. SAT first score, second score, aria

score gain histograms for Stanley H.Kaplan Educational Center (OO1A)and Test Prep Centers (022)

S5-12. PSAT verbal vs math score tableS5-13. First SAT verbal vs PSAT verbal tableS5-14. First SAT math vs PSAT math tableS5-15. First SAT verbal = f(PSAT verbal)

regressions for Stanley H. KaplanEducational Center (OO1A) and TestPrep Centers (022)

S5-16. First SAT math = f(PSAT math) regres-sions for Stanley H. Kaplan EducationalCenter (001A) and Test P:ep Center (022)

S5-17. Second SAT verbal = f(first SAT verbal)regressions for Stanley H. Kaplan Educa-tional Center (OO1A) and Test PrepCenters (022)

S5-18. Second SAT math = f(first SAT math)regressions for Stanley H. KaplanEducational Center (OO1A) and TestPrep Centers (022)

S5-19. Second SAT verbal = f(PSAT verbal)regressions for Stanley H. KaplanEducational Center (OO1A) and TestPrep Centers (022)

S5-20. Second SAT math = f(PSAT math) regres-sions for Stanley H. Kaplan EducationalCenter (OO1A) and Test Prep Centers (022)

SW. SAT Window AnalysisSW-1. Coached 2-time takers of SATSW-2. Coached 3-time takers of SATSW-3. Uncoached 2-time and 3-time takers

of SAT

SDQ. SAT SDQ and ZIP ProfilesSDQ -l. Uncoached 1-time takers of SATSDQ-2. Coached 1-time takers of SATSDQ-3. Uncoached 2-time takers of SATSDQ-4. 2-time takers of SAT coached before

first attemptSDQ -5. 2-time takers of SAT coached between

attempts

viii

9

LW. LSAT Window AnalysisLW-1. interval from coaching to exam for

1-time takersLW-2. Coached and uncoached 2-time tdkers

LW-3. Coached and uncoached 3-time takers

LW-4. Coached and uncoached 4-time takers

M. Testing and Coaching Histories of Multiple-Time Takers of LSAT

LH. LSAT HistogramsLH-1. Uncoached 1-time takersLH-2. Coached 1-time takersLH-3. Uncoached 2-time takersLH-4. 2-time takers coached before first

attemptLH-5. 2-time takers coached between attempts

Ll. 1-Time Takers of LSATL1-1. Uncoached 1-time takers of LSATL1-2. All coached 1-time takers of LSAT

L1-3. Stanley H. Kaplan Educational Centers(Puerto Rican) (001AP)

L1-4. Stanley H. Kaplan Educational Centers(001)

L1-5. Amity Review Course (002)L1-6. John Sexton's LSAT Preparation Center

(003)L1-7. Law School Admission Test Review Course

(Evergreen) (004)L1-8. Remaining schoolsL1-9. Regression for uncoached whitesL1-10. Regression for uncoached non-whitesL1-11. Regressions for coached whites and

coached non-whitesL1-12. Regressions by coaching school for

coached whites

L2. 2-Time Takers of LSAT Coached Before First Exam

L2-1. Uncoached 2-time takers of LSATL2-2. All 2-time takers of LSAT coached

'before first examL2-3. Stanley H. Kaplan Educational Centers

(Puerto Rican) (OO1AP)L2-4. Stanley H. Kaplan Educational Centers (001)

L2-5. Amity Review Course (D02)L2-6. John Sexton's LSAT Preparation Center (003)

L2-7. Law School Admission Test Peview Course(Evergreen) (004)

L2-8. Remaining schools

ix

Jo

L2. 2-Time Takers of LSAT Coached Before First Exam (coned)L2-9. Regressions for uncoached whitesL2-10. Regressions for uncoached non-whitesL2-11. Regressions for coached whites-and non-

whitesL2-12. Regressions by coaching school for

coached whites

L3. 2-Time Takers of LSAT Coached Between Exams

L3 -l. All 2-time takers of LSAT coachedbetween exams

L3-2. Stanley H. Kaplan Educational Centers(Puerto Rican) (001AP)

L3-3. Stanley H. Kaplan Educational Centers(001)

L3-4. Amity Review Course (002)L3-5. John Sexton's LSAT Preparation Center

(003)L3-6. Law School Admission Test Review Course

(Evergreen) (004)L3-7. Remaining schoo7sL3-8. Regressions for coached whites and

coached non-whitesL3-9. Regressions by coaching school for

coached whites

I. SUMMARY

It is the established policy of the United States to

provide equal educational opportunities for all citizens.

Rcognizing that the Nation's economic, political, andsocial security require a well-educated citizenry, the

Congress (1) reaffirms, as a matter of high priority,

the Nation's goal of equal educational opportunity, end(2) declares it to be the policy of the United States

of America that every citizen is entitled to an educa-tion to meet his or her full potential without financial

barriers.1/

The present investigation discloses that standardized

examinations which are utilized as important decision making

elements in the admission process to undergraduate and

graduate colleges and universities are susceptible to the

short-term preparation (often called coaching) offered by

numerous commercial entities throughout the United States.

Analysis reveals that there is a statistically signifi-

cant difference between the score increases obtained by

coached and uncoached individuals. (However, the score in-

creases are variable both between and within examinations.)

Even more importantly, score increases resulting from coaching

have a practical, educationally meaningful, effect in that

coaching can be the determining factor in deciding who is

admitted to undergraduate and graduate colleges and univer-

sities. The availability of coaching is positively correlated

1/ General Education Provisions Act, 20 U.S.C.A 51221-1

(1978).

12

1

to the ability to pay the tuition at coaching schools, which

can be as high as $250. Therefore coachable standardized

admission examinations create financial barriers to educa-

tional opportunities in direct conflict with our Congression-

ally declared national education policy.

The commercial test preparation schools do not perform

uniformly. Some are quite beneficial while others appear to

provide marginal benefits. Moreover, the coaching schools

have a universal propensity for engaging in making unsub-

stantiated advertising claims, in some instances making false

and deceptive advertising claims, and engaging in other unfair

and deceptive marketing practices, all in violation of Section

5 of the Federal Trade Commission Act.

Nonethele.s the existence of only one coaching school

(and there is more than one) that can materially increase

individuals' scores on standardized admission examinations

such as the Scholastic Aptitude Test and the Law School Admis-

sion Test reveals the lack of reliability and validity of

these examinations. The test makers themselves tell us that

standardized admission examinations should be used to help

predict the academic performance of an individual in under-

graduate or graduate school. Yet, since short-term prepara-

tion can increase scores, but has a questionable long-term

effect, the true predictive value of the standardized exam-

inations is suspect.

2

13

The standardized admission examinations are discrimina-

. tory in a number of ways. They discriminate against any indi-

vidual who either: (1) can not afford the cost of commercial

preparation or (2) elects not to attend a commercial prepara-

tion course even if he can afford it because of acceptance of

',)e dogma promulgated by the test makers, test administrators,

and test users over the past twenty years that coaching is

valueless.

The examinations appear to discriminate on the basis of

race. Certain sub-populations may receive a lesser benefit

from coaching than others. (However as grade point average

approaches the maximum there is an apparent inverse effect as

minorities show slightly greater score increases due to

coaching.) This result may be attributed to the failure of

coaching schools to develop materials and techniques aimed at

these groups or it may be a manifestation of an inherent bias

in the examinations themselves. Whatever the cause, the

effect is to put minorities with low grade point averages at

a greater disadvantage in the quest for positions in our

schools of higher education by widening the already existing

score abyss.

The economic and social benefits flowing from admission

to undergraduate and graduate colleges and universities

3

(especially the more prestigious) are axiomatic. The suscepti-

bility of standardized examinations to coaching, the unavaila-

bility of coaching to all on an equal basis, the discriminatory

effects befalling some groups who benefit less from coaching

than others, and the failure to rectify this situation, con-

stitute violations of Section 5 of the Federal Trade Commis-

sion Act or are indicia of serious flaws in both the post

secondary school admission process and in our national edu-

cation system.

Accordingly, we propose for Commission consideration

some or all of the following remedial actions to alleviate

or eliminate the unfair and deceptive acts or practices

we have discovered:

(a) Publication of either . staff report or formal

Section 6(f) report to Congress;

(b) Initiation of formal investigations of the Educational

Testing Service, College Entrance Examination Board, Law

School Admission Council, and American Bar Association

(These investigations would seek to ascertain the

extent of our jurisdiction over the investigated parties

and we would simultaneously seek to negotiate consent

agreements. Complaint recommendations, if appropriate

and necessary, would follow these investigations);

(c) Additional studies to supplement the statistical

analysis we have completed and which ultimately would

4

15

lead to an attempt to create uncoachable, more valid

and reliable standardized admission examinations;

(d) Initiation of a trade regulation rule proceeding

regarding the coaching school industry or alternatively

individual litigation against the coaching school with

the most egregious violations of Section 5.

These and other alternatives are explained in greater detail

in the recommendations section, infra.

II. INTRODUCTION

On October 13, 1976, the Federal Trade Commission issued

a resolution authorizing and directing its Boston Regional

Office (BRO) to conduct an investigation:

To determine whether or not various examination and testpreparers, review courses, coaching schools and otherpersons, partnerships, or corporations, may have been,or may now be engaged in unfair or deceptive acts and

practices in or affecting commerce in violation of Sec-tion 5 of the Federal Trade Commission Act, as amended,including but not limited to false, misleading and de-ceptive advertising, non-disclosure of material facts,unfair and deceptive point of sale misrepresentations,and other unfair or deceptive acts and practices inconnection with the advertising, offering for sale, saleand distribution of the services and products of varioustest preparers, review courses, and coaching schools.The investigation is also intended to determine whetherCommission action to obtain consumer redress wou1:1 be

in the public interest.2 /.

2/ See Category I File 772 3000, Vol. I, p. 1.

5

An FTC news release 3/ dated November 2, 1976 publicly

announcing the investigation indicated that the investigation

would be designed to provide insight into such issues as:

. whether test preparation centers have had or now have areasonable basis for claims of score increases at thetime such claims were or are made;

. whether test preparation centers have the capacity toincrease or maximize scores on certain standardizedtests that are often prerequisites to admission toundergraduate or graduate colleges, universities, pro-fessional schools or profession's;

. whether test preparation centers have the capacity toincrease scores on certain standardized tests to thedegree advertised;

. whether test preparation centers are engaged in unfairor deceptive point-of-sale practices; and

.whether Commission action to obtain consumer redresswould be in the public interest.

Prior to commencing the nationwide investigation announced

in the November 2, 1976 press release, staff had been investi-

gating the Stanley H. Kaplan Educational Center, Ltd. (File No.

762-3033).4/ After much cerebration and consult.ation with

experts in the field 5/ staff concluded that a national in-

3/ See Category III File 772 3000, Vol. I, p.88.

4/ The investigation of the Kaplan organization formallycommenced on August 18, 1975. Its original premise wasthat Kaplan could not substantiate certain advertisingclaims at the time they were made and accordingly hadviolated-Pfizer, 81 FTC 23 (1972).

5/ See Category III File 762 3033, Vol. I, p.83-84, 100-101;Category I File 772 3000, Vol. I, p.47-48; Category IFile 772 3000, Vol. II, p.1-2, 14-16, 34-39.

6

17

vestigation of the test preparation industry would be far

more' cost efficient and better serve the public interest

than an investigation of only one coaching school.6/ The

investigation of Stanley H. Kaplan Educational Center,

Ltd., although retaining its own 7-digit number has therefore

been incorporated into the industrywide inquiry.

III. A HISTORY OF THE SCHOLASTIC APTITUDE TEST (Shr), THE LAW

SCHOOL ADMISSION TEST (LSAT), THEIR USES BY COLLEGES,

LAW SCHOOLS, AND OTHERS

The SAT, LSAT and other standardized admission examina-

tions are intermediate predictors. They do not purport to

measure how an individual will perform in a given profession

or field. Instead it is suggested by the administrators

of these examinations that they be used as one criterion in

predicting how well an individual will perform in the first

year of academic study.7/

The need for a universal predictor of first year academic

performance is indicated by:

(1) the variance in meaning of grades from different

schools and curricula;

(2) the inability of admissions offices to cope with

large numbers of applications on a personal basis;

6/ See Category I Fill 772 3000, Vol. I, p.9-10.

7/ Guide To The Admissions Testing Program, p.21, CEEB,

iir(977).

7

8

(3) the inability of graduate scnools (and some under-

graduate schools) to expand their enrollments to meet

demand; and

(4) the desire to minimize the attrition rate;

(a) because of the psychological and economic hard-

ships on students who fail; and

(b) because of the economic disutility of a high

attrition rate to academic institutions.

On November 17, 1900 the College Entrance Examination

Board (CEEB) officially commenced operations. Seven months

later 973 students wrote essay examinations on subjects in-

cluding history, english, mathematics, science, and foreign

languages. The grades received on those examinations were

sent to participating colleges to be used as part of the

admission procedures. By June of 1925 College Board examina-

tions were being taken by approximately 20,000 candidates.

In June of 1926, 8,040 candidates for admission to

college received the first administration of the Scholastic

Aptitude Test (SAT). The SAT had been developed for CEEB

by Carl B. Brigham, a professor of psychology at Princeton.

Its initial purpose was to supplement the regular series

of CEEB examinations.

As the number of applicants for admission to colleges

rapidly rose (excepting a regression during tIe.depression),

the numbers alone made the essay format of the College Boards

8 19

impractical. In 1946 some 46,000 applicants took the College

Boards. Although traditionalists grumbled, the multiple-

choice format of the SAT was a natural substitute.

In 1947 CEEB, in cooperation with the American Council of

Education (the Graduate Record Exam) and the Carnegie Founda-

tion for the Advancement of Teaching (which administered the

National Teachers Exam), founded the Educational Testing

Service (ETS).

ETS' initial client was CEEB, and as a result it

administered and continues to administer the SAT and college

achievement tests. Since its inception ETS has diversified

and now administers, researches, and refines numerous

standardized admission and vocational examinations including

the LSAT, GRE, GMAT as well as the SAT and achievement tests.

Today approximately 1.5 million students per year take

the SAT. The SAT coupled with ETS' other testing activities

accounted for about 87% of ETS' $70 million revenue in

fiscal 1977.

The specific use to which the SAT has been put in the

past decade has varied from college to college and from year

to year. In general the SAT was most important in the

admission process during the post-Sputnik era when competi-

tion for college admission became intense. Most colleges

9

20

responded by raising their standards rather than increasing

enrollment, and test scores therefore became critical.8/

With decreasing applications and enrollment expansion the

SAT today is generally less critical in the admission process

than it was in the 1960's. However it is still crucial for

those seeking admission to "high prestige" schools. It

appears that for schools other than, the highly selective the

SAT is most important at the bottom line. That is, failure

to obtain a certain minimum score will result in rejection.

For example the California State University and Colleges

System requires a minimum combined score of 2500 on the

SAT plus three achievement tests and a B average. If a student

fails to attain a 3.1 high school average (B-) the University

of California/Berkeley requires a minimum combined SAT score of

1100 plus a minimum score of 1650 on three achievement tests.

Oregon State University requires a minimum combined SAT score

of 890 or a high school average of 2.50. Florida State Uni-

versity requires out of state students to obtain a minimum.

score of 500 on both the verbal and mathematical sections of

the SAT while Florida Technological University requires out of

state students to have a combined minimum SAT score of 900.

The University of South Florida requires out of state appli-

8/ See generally, 2 The_Encyclopedia of Education, p.211-213, The Macmillan Co. and The Free Press 1971;Ravitch, The College Boards, The N.Y. Times Magazine,May 4, 1975, p.12-22; Ssiencel-alest, DEaTer 1976,p.39-45; Forbes, November 15, 1976, p.94.

10 21

cants to obtain a combined SAT score of 800 with a minimum

of 400 on the SAT verbal. North Texas State University re-

quires a minimum combined SAT score of 675 and the University

of Mississippi requires a combined minimum SAT score of 680.

Princeton University, Smith College, Stanford University,

Wellesley College, Brown University, University of Chicago,

Yale University, and Massachusetts Institute of Technology,

among others, all accept freshman classes in which over 90% of

the admittees score over 500 on both the SAT verbal and SAT

mathematical sections.9/

As is evidenced by the figures above some universities

prescribe definitive minimum cutoffs and at others the fail-

ure to obtain a certain minimum score means almost certain

rejection.

In some instances where high schools refuse to send

grades to colleges or where a high school is relatively un-

known to a college admissions office the SAT becomes the sole

criterion for admission. Furthermore New York State has

substituted the SAT for its traditional Regent's Exam to

determine scholarship recipients.l0/

9/ The sources for SAT score data are Comparative Guide toAmerican Colleges, Cass & Birnbaum, Harper Row, New York,8th Ed.-, 1§f7; and The College Handbook, CEEB, N.Y. 1977.

10/ See Category III File 772 3000, Vol. I, p.87. This is

an interview report with the Chief of New York State'seducational testing wherein he states that if coaching

is only minimally effective it would distort to someextent who obtained Regent's scholarships.

11

22

Generally while the SAT is not as important a factor

in the admission process as it was several years ago it

is still a very material factor. A reversion to a national

attitude similar to that of the early 1960's would make

the SAT even more critical in the admission process.

It appears that the first law aptitude examination was

offered at Columbia Law School in 1921.11/ Prior to 1948

at least seven specially constructed legal aptitude tests

were used by various institutions as one means of selecting

applicants to their law schools.12/ In 1925 Merton L.

Ferson, Dean of the University of North Carolina Law School,

and George Stoddard, Assistant Professor of Psychology and

Education at the University of Iowa, announced the creation

of the Ferson-Stoddard Law Aptitude Examination.13/ The

Ferson-Stoddard examination was remarkably similar to the

current day LSAT containing sections on: (1) capacity for

accurate recall (memory); (2) reading comprehension; (3)

reasoning by analogy; (4) reasoning by analysis; and (5).

skill in pure logic.14/

11/ Ramsey, Law School Admissions: Science, Art, or Hunch,12 J. L6Tif (960).

12/ Id.

13/ Ferson, Law Aptitude Examinations, 5 Am. L. School Rev.563 (1925).

14/ Stoddard, Person and Stoddard Law Aptitude Examination- Preliminary Report, 6 Am. L. reffool Rev. 71 (1927).

Za12

In 1941 an experimental edition of the la Apti-

tude Test appeared. Apparently dissatisfied with the r.!xamina-

tions available, a group of law schools approached the College

Entrance Examination Board in 1947 regarding the possibility

of producing a national aptitude test for the legal profession.

The result of these negotiations was the birth of the LSAT.15/

The LSAT has grown in importance to the point where as a

prerequisite to accreditation the American Bar Association

(ABA) requires law schools to require their applicants to take

the LSAT.16/ It is not only the demand of the accrediting

authorities that makes the LSAT prominent in the admission

process. Although intensive examination of an applicant's

educational and life record might well yield more satisfactory

admission results 17/ most law schools were and are thinly

15/ See n. 11 at 513, supra.,

16/ "All applicants except those physically incapable oftaking it, should be required to take an acceptabletest for the purpose of determining apparent aptitudefor law study. A law school that is not using the Law

School Admission Test administered by the EducationalTesting Service should establish that it is using an

acceptable test." ABA Standards for Legal Education,

Cha ter V, Section 503. The AmeriFin Assocrit37F---lof Law

Schoo s (AALS) has i-iimilar requirement. In fact there

is no other test that purports to measure aptitude for

the study of law. Goolsby, A Study of the Criteria

for Legal Education and Admission to the Bar, 20 J.

Lega Ed. 175 (1967).

17/ Rosen, Equalizing Access to Legal Education: Special

Programs for Law Students Who Are Not Admissible piTraditional CITEeria, U. of Toledo L. Rev. 321 (1970).

13

21

staffed with admissions personne1.18/ As a result, law

school admission decisions, excluding special admission

programs, are based almost exclusively on the numbers.

Even a vigorous opponent of this conclusion notes:

At some point of course it does become essentially impos-sible for personal factors to compensate for uncompromis-ing objective credentials for the over-all correlation ofLSAT scores and undergraduate grades with law school per-formance can raise questions of survival.19/

While there is a great diversity in approach the most

common method used in selecting applicants for admission to

law school creates an index score by adding the LSAT score to

a number equal to 200 times the grade point average (Index =

LSAT + (GPA X 200]).20/ This procedure attempts to give

the LSAT and the GPA approximately equal weight In the

admission decision making process.

18/ Schmidt, Admission to Law Schools: Not By Computer,

Not a Change, 10 Tulsa L.J.-111 (19717.

19/ Winograd, Law School Admissions, A Different View, 59

A.B.A. J. 862, 865 (1973).

20/ See Category III File 772 3000, Vol. XXXI, p.183.Although this information is obtained from a coachingschool it is highly credible in that the Board ofEditors consists of professors from Harvard, Loyola,Wayne State, University. of Illinois, University ofTexas, and University of Southern California Law Schools.

See also Category III File 772 3000, Vol. I, p. 84-86,Admissions Practices of Temple Law School; and Pre LawGuidebook of Boston College, p. 65, wherein it 1i stated:

Almost every law school will weigh certain variableswithin the structure of a specific formula. Many

law schools multiply their LSAT score times twoadding this result to your Grade Point Average (GPA)and the rating of your undergraduate institution.

14

25

The importance of the LSAT score in the admission pro-

cess can not be underemphasized. Not only does the LSAT

share approximately equal weight with the Grade Point Average

(GPA) but the failure to obtain a minimumn LSAT cutoff score

at some law schools means automatic rejection.21/ While only

a few law schools admit to having absolute minimum LSAT re-

quirements others admit to having a "preferred" minimum LSAT

score.22/ While the official guide to law schools informs

us that "Some schools do follow automatic qualifying or dis-

qualifying credentials such as minimum gradepoint averages

(GPA) and LSAT scores, but most are not wedded to such cutoff

scores and averages," 23/ a review of that publication indicates

that the use of a minimum cutoff score is not infrequent.

For example the following law schools among others did

not admit any students with LSAT scores below 500 for the 1977

graduating class (the number in parentheses is the number of

applicants applying with an 11,SAT score below S00): University

of Denver (196), George Washington University (384), University

of Illinois (199), University of Kentucky (199), Mercer Univer-

sity (227), University of Minnesota (82), University of New

Mexico (187), University of Wisconsin (107). In summary, of

21/ Lunneborg & Radford, The LSAT: A Survey of ActualPractice, 18 J. Legal Ed. 1313 (1966).

22/ Id.

23/ Prelaw Handbook; AALS, LSAC, ETS (1974).

15

the 1581 applications received by these law schools reporting

an LSAT score below 500, none were accepted.

The following law schools also did not accept any appli-

cants with an LSAT score below 500 except for those students

admitted pursuant to a special admission or minority program

(the number in parentheses again reflects the number of appli-

cants to the institution with LSAT scores below 500): University

of California/Berkeley (151), University of Colorado (240),

University of Georgia (272), Marquette University (264),

University of North Carolina (279), and the University of

Oregon (96). In summary, of the 1308 applicants reporting an

LSAT score of below 500, not including those admitted pursuant

to a special admission program, none were admitted to the

above-mentioned law schools.24/

The above survey is not exhaustive. Not all law schools

have furnished data and we could see no benefit at this time in

attempting to acquire admission standards from every law school

in the country. We believe it is clear that minimum cutoff

scores are prevalent.

Even if it is conceded that law schools are not using a

minimum cutoff score, it is evident that the use of a formula

in effect creates a threshold LSAT score that must be sur-

passec to obtain comsideration for acceptance. This becomes

24/ I.

16 27

even more evident when it is recognized that the range of the

grade point averages of applicants is narrow (as a result of

both grade inflation and increased applicant ability) thus

limiting the grade point average component of the formula to.1,

a range of about 400 points rather than its theoretical 800

points.

The LSAT is being used (abused) for purposes other than

admission to law school. We have had personal experiences of

emp_ayers requesting LSAT scores on applications for attorney

positions. In fact until recently the Commission requested a

job applicant's LSAT score on the supplemental attorney's form.

We believe this is a practice that is not uncommon in both the

public and private sectors.

IV. REVIEW OF PRIOR RELATED RESEARCH AND LITERATURE

Considering the imporance of the issue, and the constant

student and parent request for its resolution, there is a pau-

city of prior research and literature concerning the coacha-

bility of standardized admission examinations.25/ In fact we

find virtually no material dealing with the effects of coach-

ing on any standardized admission examination other than the

Scholastic Aptitude Test (SAT) and no material at all concerned

25/ For the most thorough and most up to date review of thisentire area, see Pike, Short-Term Instruction, Testwiseness

and the Scholastic AkAitraiTest: A Literature TieVreC4wiTh--Research Recommendations, Educatioilial Testing siiTriFe,IFTREgi3n, New Jersey, 1978.

17

23.?

with the commercial coaching schools as they presently exist.

We also find no material on the effects of coaching 1-time

test takers, only test repeaters who have been coached

in the interim period between test administrations. There

is however considerable material dealing with the general

topic of test-wiseness. (TW).

Two research reports deal extensively with the current

state of knowledge of test-wiseness.26/ Test-wiseness has

been defined as the ability to manifest test taking skills

which utilize the characteristics and formats of a test and/or

test taking situation in order to receive a score commensurate

with the abilities being measured. Alternatively the term has

been defined as a subject's capacity to utilize the character-.

istics and formats of the test and/or the test taking situa-

tion to receive a high score...and is logically independent of

the examinee's knowledge of the subject matter for which the

items are supposedly measures.

Some aspects of test-wiseness include effects of practice

on tests, familiarization with item types, following instruc-

tions, using time efficiently, knowing when to guess, not

26/ See generally, Fond V., Everything You Wanted To Know

About Test-Wiseness, 1973 (Eric Document ReporneEr5gtTEVia-ifd 093 912); Jongsma, E.A., Wershaver E.,

The Effects of Instruction Test-Making Skill Upon

Student Performance on Standardized Achievement Tests,

99775- (Eric Document Reporduction selirEgigirrICIWT.

18 21)

hesitating to change an answer, and having a planned strategy

for answering examination questions.I

Test-wiseness is recognized by most experts as a source

of variance on educational tests. Moreover it is generally

agreed that test-wiseness is cognitive. It would then not be

surprising to find that coaching schools have the capacity to

increase scores on stnadardized admission examinations, at

least to the extent that test-wiseness is a score component,

especially where they are dealing with test-naive students.

Generally the few prior studies investigating the effects

of coaching on the Scholastic Aptitude Test have concluded

that coaching is ineffective. Based on these studies the

trustees of the College Entrance Examination Board have stated

that "score gains directly attributable to coaching are of

such a small magnitude that it is unreasonable to expect

coaching to affect admissions decisions".27/

However, a 1972 study conducted by two Educational Test-

ing Service psychometricians, which was commissioned by the

College Entrance Examination Board, severely criticizes theie

prior studies.28/

27/ College Entrance Examination Board, Effects of Coachingon Scholastic Aptitude Test Scores, New York, CEEB,1§65.

28/ Pike L. & Evans F., Effects of Special Instruction forThree Kinds of MatheiiigErFrAptitude Items, CEEB, NewYork, Tin:

19

3o

The authors state:

... the particular question of the effects of specialinstruction on different mathematics item formats is notadequately answered by previous research. First, themajority of studies were designed to determine whethercoaching schools gave their clients an unfair advantageon the SAT. The instruction provided in these studieswas, where its nature can be ascertained, rather scanty.There appeared to be little or no systematic attempt toidentify the skills needed to perform well on the testand to develop materials to meet these needs. Sincemost previous research on this question involves sub-jects at the extremes of the ability range, generaliza-tion to the more heterogeneous population of candidatescurrently Reeking admission to higher education becomeshazardous.29/

The authors also stated that:

... many of these studies were conducted in the 1950's

when the SAT candidate population was more homogeneousthan it is today. Consequeatly, much of this researchinvolved subjects enrolled in private prepatory schoolsor in public schools whose student bodies were wellabove the national average in tested ability. One canprobably assume that the subjects in these studieswere therefore less likely to benefit from special in-struction than were students in the more general popula-

tion.3C/

It is also important to note that students were assigned

to experimental and control groups in some instances without

regard to their desires to receive coaching, although in other

instances the assignment was on a voluntary basis.

The major focus on the 1972 study was to determine if

a new mathematical question format (quantitative comparisons),

which the College Board was considering using on the SAT mathe-

matical section, was susceptible to coaching. As a by-product

30/ Id.

20 31

the authors also sought to ascertain if the SAT question types

then in use, regular math and data sufficiency, were suscep-

tible to coaching. The authors concluded that "... each of

the three item formats was susceptible to the special instruc-

tion directed toward it."31/ They further stated that,

"Certainly the gains achieved by the quantitative comparison

and data sufficiency instructed groups on the app-opriate

items are of practical value."32/

As a result of these findings the College Board declined

to use quantitative comparison items on the SAT at that time.33/

Nevertheless these item types are found on the Scholastic

Aptitude Test, Law School Admission Test, Graduate Management

Admission Test, and Graduate Record Examination today. More-

over the Scholastic Aptitude Test still retains data sufficiency

and regular math item formats.34/

In a most recent paper Pike 35/ recognized ten components

of observed SAT test scores:

31/ Id.

32/ Id.

33/ Id.

34/ See generally the bulletins prepared for applicants

for standardized examinations by the administrators

of the examinations. File 772 3000, Physical Exhibits

I, J, K, and L.

35/ See n. 25, appra.

21 32'

A. "True score" components: e.g., verbal aptitude, math

aptitude.

1. A composite of underlying knowledge (e.g., vocabu-

lary, elementary algebra) and reasoning ability,

developed over a long period of time. (Long-term

acquisition, long-term retention.)

2. A state of being well-reviewed, so that the per-

formance to be demonstrated is in line with the

individual's underlying developed competence.

(Short-term acquisition, short or medium-term

retention.)

3. Integrative learning, overlearning, consolidation.

(Short-term acquisition, long-term retention.)

4. Learning criterion relevant, analytic skills (e.g.,

how to identify the main idea of a paragraph; how

to simplify complex quantitative terms before com-

paring their value). (Short-term acquisition,

long-term retention.)

B. Primary test specific components.

1. The match between the domain of developed ability

(including the various score components listed

in A above) and test content. Mismatches may occur

as gaps in such areas as skill in locating infor-

mation in reading passages and ability to work

with the algebra of inequalities.

2. General TW--Test familiarity, pacing, understand-

ing of general directions, general strategies for

using partial information, and so on.

3. Specific TW--Components similar to B2, but in

reference to characteristics of specific item formats

(such as verbal analogies and quantitative-comparisonitems) and other item characteristics.

C. Secondary components influencing test taking.

1. Level of confidence.

2. Level of efficiency--The ability to use avail-

able knowledge and reasoning ability quickly,

with a relatively low rate of errors resulting.

from working rapidly.

22

33

D. 'Error." Fluctuations in attention, sampling error,variations in luck in guessing, etc.36/

Pike concluded that eight of these items (except Al, a

composite of underlying knowledge, and D, error) were, based

on his review of prior research, susceptible in varying

degree and situations to some type of preparation (either

coaching, short-term instruction, or intermediate term in-

struction).37/

This is evidence that test preparation services may

have the capacity to increase scores. Pike stated in considering

the discrepancy between past studies on coaching that

in principle a single study showing substantial positive

gains cannot be countered or refuted by any number of studies

failing to get positive results."38/

V. INVESTIGATIVE METHODOLOGY

Test preparation services are those commercial 39/ enter-

prises that in a relatively short time, by employing a variety

of methods, attempt to aid an individual to increase his/her

scores on standardized examinations that are often pre-

requisites for admission to undergraduate and graduate colleges

36/ Id. at 3-4.

37/ I. at 38-42.

38/ Id. at 35.

39/ See our comments on the non-profit segment, infra.

323

and universities.40/ Their intensive programs are directed

specifically towards the examination that an individual is

preparing to take and any residual long-lasting benefits

that may result are purely secondary. The terms "review

course", "coaching school" and "cram course" are synonymous

with and used interchangeably with the term "test preparation

service". Our initial investigatory tasks were to identify

these organizations and ascertain the nature and structure of

the test preparation industry.

Surv.ys of the Education and Career Development Directory

of the Nw..7 York Times 41/ and the education sections of other

natc.onal Sunday newspapers proved fruitful in identifying in-

dustry members. Once an industry member was identified it was

invariably asked to identify its competitors. We also closely

scrutinised the telephone directory listings under tutoring

and coaching, and visited numerous campuses to view any adver-

tisemeat3 appearing in college newspapers or posters placed on

bulletin boards by coaching schools. Needless to say we

asked students if they had any knowledge of who was offer-

ing coaching courses.

ON..11111Ms.

40/ This definition is similar to that chosen by theCollege Entrance Examination Board (CEEB) to definecoaching. See, Effects of Cooaac_hing on ScholasticAptitude Test Scores, CEEB, N.Y. ( 9Z5) p.4.

41/ See Category III File 772 3000, Physical Exhibit 4,p.50-51.

24 35

Perhaps our most successful technique for identifying

industry members was to mail four hundred thirty-two (432)

questionnaires 42/ to pre-law advisors in universities in the

metropolitan areas of Boston, MA; Chicago, IL; Los Angeles,

CA; and New York, NY.43/ We received 185 responses to

these questionnaires (42.8%) and 69 responses named either

an individual, commercial enterprise, or not-for-profit organi-

zation that was offering or had offered test preparation.

Our previously acquired knowledge of the industry alerted

us at the outset to the impossibility of identifying every

industry member in the United States. Accordingly we decided

to focus on major metropolitan areas that have a large number

of universities, a high student population, and a high number

of coaching schools.

Stanley H. Kaplan Educational Centers, Ltd., the largest

coaching school in the industry in terms of annual revenues as

well as total number of students taught, informed us that its

42/ These questionnaires were mailed after receiving clear-

ance from the General Accounting Office as required by

the Federal Reports Act, 44 U.S.C. SS3501 - 3511. See

Category I File 772 3000, Vol. I, p. 49-59.

43/ The pre-law advisors were selected by reviewing Directo-

ries of Colleges from California, Illinois, Massachusetts,

and New York. These directories are incorporated in

File 772 3000 as Physical Exhibits M, N, 0, and P re-

spectively. The responses to these questionnaires are in-

corporated in Category III File 772 3000, Vols. IV - VI.

25 36

largest. centers were located in the metropolitan areas of

Bostoh, Chicago, Los Angeles, and New York. We also de-

termined that all the national coaching schools operated in

at least one of these cities. We therefore decided to limit

our initial inquiries to these four metropolitan locations. We

have since decided that close scrutiny of other geographical

locations is unnecessary at this time.. The statistical

analysis that we have performed and its conclusions can be

validly used to draw parallel inferences concerning the entire

nation.44/

We also decided to limit our investigation in time to the

period from October 1974 through December .1976 inclusive.

This decision was based on several factors. The number of

students coached in this period gave us large sample sizes

which is beneficial to the accuracy of a statistical inquiry.

The record retention of the coaching schools deteriorates

rapidly prior to that period and identification of now defunct

entities that existed prior to October 1974 is extremely dif-

ficult. Data maintenance for a greater time period and more

44/ In fact, this is what the mathematical discipline of

statistics is all about. Statistics is the science of

inferring generalities from specific observations. Al-

though we never know for sure in any particular case

whether a hypothesis is true or false the procedures of

statistical inference make it possible for us to state

precisely what the probabilities are that we will accept

a false hypothesis or reject a true hypothesis. This

statistical proof is the basic form of proof used in

the investigations of all sciences.

7?26

extensive geographical locations becomes extremely cumbersome

-and costly as time frame or geographical location is increased.

Not only have we limited our specific inquiry to time

and location but we have also limited it, for the purpose of

testing the effectiveness of coaching and the truth or falsity

of coaching school claims, to two standardized examinations:

the Scholastic Aptitude Test (SAT) and the Law School Admission

Test (LSAT). The justification for the limitation of the

specific.inquiry to these two examinations is the same as that

for limiting the specific inquiry to four geographical loca-

tions -- viz. we believe valid inferences about the coacha-

bility of other examinations can be drawn from the specific

results we obtain for the SAT and LSAT, or at the very least

place a strong burden on the parties to whom any inferences

are adverse. Furthermore, the SAT and LSAT are the two most

heavily coached examinations administered by the Educational

Testing Service and therefore we have had less difficulty in

developing a significant data base for their analysis than we

would have for other examinations.

One further limitation on our specific inquiry must be

noted. It came to our attention that there are many high

school teachers and others coaching high school students for

the SAT. For practical reasons (time and resources) it is not

possible to ferret out all these individual entrepreneurs and

we have not attempted to do so. This fact, coupled with the

27 38

extremely extensive and voluminous SAT computer files main-

tained by the Educational Testing Service caused us to de

cide to limit the specific SAT inquiry to major coaching

schools in the metropolitan New York City area. The failure

of some of these organizations to maintain any records,

legible records, or to have taught an appreciable number of

stude0-s during our relevant time frame has reduced our SAT

statistical analysis to two major SAT coaching schools.

This limitation in scope of our statistical analysis has not

affected in any way our industry analysis. Furthermore this

limitation of our statistical analysis, if it is a bias, dis-

favors the coaching schools actually analyzed, in that we can

not be sure that our control groups have been culled of all

coached students. This and other biases are discussed, infra.

The statistical analysis to which we have been referring

was designed to determine the effectiveness of coaching schools.

As a by-product it also determines the truth or falsity of the

coaching school claims as well as the truth or falsity of the

claims of the test makers. It is discussed in more detail,

infra.

In order to develop data for computer analysis, enrollment

foems reflecting coaching school enrollees had to be matched

with the ETS.computer tapes. As individual industry members

were identified they were requested to submit information to

us pertaining to their operations. The most important

28 39

element requested was the student enrollment or registration

forms. Although most test preparation services had this data

in some form their record retention and record legibility was

not always exemplary. Therefore each of the more than 14,000

enrollment forms we obtained was inspected for legibility and

. for any apparent inconsistencies or omissions. If for example

a name was partially illegible we checked the appropriate

telephone book in an attempt to ascertain the correct name

spelling. If a street address was omitted we followed the

same procedure. In literally thousands of instances zip codes

were omitted from addresses on the enrollment forms and these

were manually supplied by employing the National Zip Code

Directory. This procedure resulted in our matching over 80%

of the coaching school enrollees with individuals on ETS'

computer tapes. Part of the unmatched percentage can be

attributed to a failUre by the student to actually take the

SAT or LSAT.

In order to ascertain the completeness of the coaching

schools' enrollee records we determined the expected number of

enrollees by dividing each coaching school's annual income by

its tuition cost. Although this procedure was unavailable

in some instances due to integrated financial statements, and

was approximated in other instances due to tuition fluctuation,

we conclude that we have acquired over 91% of the total LSAT

enrollment forms we sought from the coaching schools involved in

29

40

our statisticalanalysis and in excess of 99 of the SAT

enrollment forms we sought from the coaching schools involved

in our statistical analysis.

Once the enrollment forms had been acquired and any

illegibilities or omissions were rectified, the names, ad-

dresses, and other available identifying information, including

the dates when the student was coached, contained on the enroll-

ment forms were typed onto plain bond paper (only students

attending the same coaching school appeared on the same page

with a code corresponding to the coaching school where the

students had been coached appearing on each page).45/ These

student enrollment lists were then ready for computer matching

with the ETS computer tapes and for statistical analysis.

VI. INDUSTRY STRUCTURE

It is fundamental to recognize that the economics of the

coaching school industry is merely a subsidiary consideration

in ascertaining the public interest of this inquiry. The

social, psychological and economic benefits that may be

gained by indivichials Olo attend a coaching school transcend

the current stature bf the coaching school industry. Not-

withstanding, the following industry analysis is offered.

45/ This procedure was followed to ensure that Only FTC

'--- personnel could determine which results corresponded

to which coaching schools until such time as the data

is publicly released.

30

A. COMMERICAL

In the metropolitan areas of Boston, Chicago, New York,

and Los Angeles from October 1974 through December 1976 a

total of in excess of 15,000 students were commercially

coached for the SAT and LSAT.46/ We estimate that 50,000

individuals are being coached annually nationwide for all

standardized admission examinations.

For 1975 and 1976 combined, the twenty-one commercial

entities that we have identified and included in our analysis

had total annual sales of $9,403,615.47/ We estimate that the

total annual sales volume of al] commercial preparatory courses

nationwide is $10,000,000.

The twenty-one entities considered here expended $400,000 48/

on advertising in 1976. Their tuition ranged in price from a

maximum of $250 (plus a $50 refundable deposit for materials)

46/ We have in our possession the names of 11,906 LSAT coacheesand 2286 SAT coachees that were obtained for use in thisstudy although not all have been used. All SAT coacheesare from the metropolitan New York City area.

47/ This figure includes revenues obtained from all standard-ized admission tests, not only the SAT and LSAT.

48/ Since the coaching schools are advertising in effect toa captured audience they can reach all potential enroll-ees by using low cost techniques such as posters andschool newspapers. The Stanley H. Kaplan Educational Centerestimates 801 of its students attend due to referrals notadvertisements. See, Transcript of Investigational-Hear-ing of Stanley H. Kaplan, dated May 6, 1976, p.40. Al-though many coaching schools advertise in large newspaperssuch as the New York Times, we perceive this is an attemptto reach parents more than students.

31 4 2

to a minimum of $40. Their courses ranged in duration from 40

hours (10 sessions at 4 hours each) of classroom indstruction

plus unlimited use of library tapes and homework to a minimum

of a one-day 8 hour seminar. Not surprisingly, the most ex-

pensive courses had the longest duration, and the least

expensive courses had the shortest duration.

All preparation courses offer written materials as well as

lectures. Their materials consist either of the commercially

available review books 49/ or materials designed by the coaching

schools themselves. By and large the majority of the materials

consists of numerous examples of question types appearing, or

expected to appear, on the various standardized examinations.

One coaching school has an extensive tape library where. over

200 hours of additional preparatory instruction is available.

The larger coaching schools have staffs of researchers

who are constantly updating materials and writing practice

examination questions. One school is so vigilant in its quest

to keep current its materials that it pays individuals to take

a standardized admission examination for the sole purpose of

49/ We have not endeavored to analyze the effects of mere-ly using one of these review books without attending aformal coaching course. Such use by subjects in ourcontrol groups would of course bias our,findings againstthe coaching schools.

32 4:1

*remembering the examination questions.50/ The value of this

practice is questionable. A study was commissioned by the

College Board because it was believed some operators of

coaching programs obtain some items that have been included in

past editions of the SAT. The coaching materials consisted of

actual items that later appeared on the SAT the subjects would

take. The author concluded that "...even if it were possible

to coach with as many as one-tenth of the items that would be

used on a future form of the SAT -- a virtual impossibility

-- the effect on the scores of students coached with these

items would be negligible."51/

There is much variance among the techniques used by

coaching schools. They cover a continuous spectrum. . The

simplest approach is teaching solely test-wiseness which is

devoid of substantive content and mainly concerned with exam

taking techniques. A more elaborate technique is short-term

instruction (STI). This is a relative term which includes

both elements of test-wiseness and substantive content. The

50/ See, Transcript of Investigational Hearing of JohnSexton, February 9, 1977, p. 41-42.

"I wouldn't want ETS to knoW our operation, but letme say that we do have people who take the tests forus which obviously is a way of monitoring."

51/ French, John W., An Answer to Test Coaching, CollegeBoard Review, No.-ITTFin Tg55, p. 5-7.

33

Ii

duration of instruction is short compared to the amount of

time the test administrators would have us believe it takes

to develop the abilities (over a lifetime) that standardized

admission examinations measure. The most elaborate teaching

technique employed by the coaching schools is intermediate-

term instruction (ITI). ITI is of the same nature as STI

although somewhat longer in duration. It is however of

relatively short duration when compared to the amount of

time generally considered necessary for substantial changes

in test scores to occur.52/

Although there is no bright line that serves to determine

where one of these approaches ceases and another begins, both

the nature of the materials used by test preparation services

and the length of instruction leads us to conclude that the

majority employ short -term instruction techniques. Organiza-

tions offering formal instruction for a weekend or less engage

in teaching solely test-wiseness. The Stanley H. Kaplan Educa-

tional Center, Ltd. (provided the 200 hours of library.tape

are used) may be viewed as offering intermediate-term in-

struction.

Almost all teachers at the coaching schools have obtained

or are in the process of obtaining graduate degrees. Another

52/ See, n. 25, at 4-5, supra.

34

almost uniform requirement for teaching at a coaching school

is scoring extremely high (minimum of 650) on one or more

standardized admission examinations.

Although we have confined our initial inquiry to LSAT

preparatory curses in the metropolitan areas of Boston,

New York, Chicago, and Los angeles and our SAT inquiry to

preparatory courses in the metropolitan New York area, we are

aware of commercial coaching schools existing in every state

and in virtually every major city or college campus throughout

the country for all standardized admission examinations.53/

Although the coaching school industry may not be large

in terms of annual sales by some standards, the industry

structure already exists which enables it to reach any of the

approximately 2,500,000 individuals who annually will be

53/ A list of some of the locations of noaching schools

follows;

Brooklyn; Buffalo; Manhattan; Syracuse; Albany; Ann

Arbor; Atlanta; Augusta; Austin; Birmingham; Boston;Champaign; Chicago; Charleston; Cleveland; Columbus;Dallas; Denver; Detroit; Durham; East Brunswick; Gains-ville; Houston; Indianapolis; Jacksonville; KansasCity; Los Angeles; Louisville; New Orleans; Omaha;

Palo Alto; Philadelphia; Pittsvurgh; Portland; SanFrancisco; Seattle; St. Louis; Tallahasee; Tampa;Washington D.C.; Toronto; Lugano, Switzerland; Euclid,Ohio; Clinton, Iowa; Knoxville; Boulder; Fort Worth;Baton Rouge; Baltimore; Montreal; Columbia and Greenville,South Carolina; Miami; Harrisburg; Oxford, Jackson,and Columbus, Mississippi; Orlando; Tuscaloosa, Opelikaand Mobile, Alabama; Lexington, Murray, Maehead, andBowling Green, Kentucky; Fayetteville; Little Rock

and Jonesboro, Arkansas; Phoenix; Tuscon; Milwaukee;Madison; Salt Lake City; and Utica.

35

46

taking a standardized admission examination. It has the

potential to grow quite rapidly 54/ and to exceed annual

sales of one-half billion dollars.55/ Furthermore there

are almost no entry barriers to this industry.56/

The charts appearing on the following pages represent

for each of the entities we have specifically examined, name

of the coaching school, total annual sales (1975-1976), number

of students enrolled (Oct. 1974 - Dec. 1976), advertising

expenditures (1976), cities where courses are offered, dates

when courses have been offered up to December 1, 1976, costs

of the course, and length of the course.

54/ Harcourt Brace has recently acquired a coaching school

WAT METHOD). This acquisition by a $300 million

conglomerate is further evidence of the potential

for rapid annual sales expansion of this industry.

55/ Assuming each of the 2,500,000 students who takes a

standardized admission examination each year enrolled

in a coaching school at an average cost of $200, the

industry would have annual sales of $500,000,000.

56/ Franchisees of the Evergreen course, for example, must

pay only $200 upon execution of the franchise agreement

and then 33 1/3% of the gross up to $90,000 and 15%

of the gross over $90,000. See, Category III File

772 3000, Vol. XLIX, p.45. In this vein it is of interest

to note that there are numerous high school and college

teachers coaching small groups of students for standard-

ized exams on a part time basis, and we are unaware of

any state imposed licensing restrictions.

436

NW OF

ORGANIZATION

TOTALNUMBER OP

ANNUAL SALES STUDENTS

1975-1976 10/74-12/76

1976

ADVERTISING

EXPENDITURES

Dollars

Stanley R. Kaplan7,834,093 1/ 4915 2/ 209,357

Educational15041/

Centers Ltd.

Amity Review267,000 I/ 677 4/ 26,001

Course

1500 1/

CITIES

WHERE OFFERED

DATESCOST OF LENGTH OF

OFFEREDCOURSE COURSE

over 50 cities Prior to

in U.S. plus10/74 -12/16

Toronto and

Lugano, Swat,

$250-LSAT10 weeks at

$225-SAT4 hrs./wk.

over 20 cities10/75 -1 2/16 $115 or 30 hours

in U.S.$125 or wlekend

John Sexton's511,184 6/ 3020 6/1/'

40;000 8/

LSAT Review

Course

NY, Bos., DC, LA, Prior to $95

Houston, Tampa,10/74. 12/76 $150

and others$250

1111.0~1000161111160.1111101~1111.14.1~1~1..~WIMMIMAMANIM..A

Low School236,547 9/ 1738 12/

Admission Test43,661 IS/ 1216 II/

Review Course9,220 II/ 367 TO

(Evergreen)

187 IT/

30,000

11 hours

28 hours

10 hours

In excess of Prior to $125 30 hours

60 locations in 10/74-12/16$250 40 hours

U.S. and Canada 2/75-12/76 10/

1/ Nationallyfor all exams

1/ Sou, NY, Chi" LA (LSAT)

1/ NY only (SAT)

Oos,, Chi., NY,

LA only (LSAT)

48

2/ Nationally (LSAT)8/ Estimated,

6/ Includes NY, franchisees 9/ Includes $64,270 received

in LA and NJfrom franchisees as commissions,

does not include MCAT revenue

2/ No records Of students

exist prior to 1/75 10/ LA franchisee

37

13/ Due to'poorrecord retention

this representsnumber of

names in our possession.

Approximately 15 are miselog

for 12/76, 75for 2/76, and

538 for 4/15 and 7/75

11/ Chi, francdiee

12/ LSAT for NY/NJ

area only. Does not

include 200 scholar-

ships per year

MANS Of

ORGANIZATION

TOTAL'

ANNUAL SALES

1975.1976

Ad Educational 33,000 14/

Services

NUMBER OP 1976

STUDENTS ADVERTISING CITIES DATES COST Of LENGTH Of

10/74.12/76 EXPENDITURES WHERE OFFERED OFFERED COURSE COURSE

(Dollars)

186 15/ unknown NY/Boston 10/74.12/75 $239 20 hours

$179

ON.MOIMOMPOOM~AMONIPilkilinmiNolaOrs4pomameW00000~60=1MNIMosaliesealNdrINwPiba%mandayilMI001001POIMINO

Rutgers Review 11,941 14/ 126'15/ . 33,000 NY/NJ

Center (ARC) 60,166 11/ . 417 Tglsot,

LSAT

NEMO

10/75.12/76 $85 20 hours

$115 (weekend)

$115, 7 sessions

21 hours,

44 hours

(11 sessions)

aormosofkoamommmoodwomboaftftraftarrieftftearalessalimilWeirMill.

15,000 283 17/ 10,000 LA.,SLISD

oppl, 485 Ti/Salt Lake City 10/15.12/76 $40

OMB 011111011011110111111..wel

1 day

(I hours)

Western States 42,526

Review

Test Prep

Chicago

14/ GSAT only

IS/ LSAT only (also offered

courses for other exams)

43,471

115 17/ 5,000 LA, SF, Seattle

500 IV est, Honolulu

171 unknown Chicago

10/75.12/76 $85 li hours

(2 days)

2/75.12/76 unknown unknown

1

5 0.

16/ All courses total18/ All cities total

17/ LA only

38

NAME OF

ORGANISATION

TOTAL NUNOE4 OF 1976

ANNUAL SALES STUO46 ADVERTISING CITIES

1975-1976 10/74 -12/76 EXPENDITURES WHERE OFFERED

DATES COST OF LENGTH OF

OFFERED COURSE COURSE

.1400.0060,..4.0.1001011411.01.6/{nol..~141~020111C2311. AL 6.040.00~0000baa~dass~ .1.0.6*"

prporAnarokea0asowatmaraartaohowsoraweb~a

Test Prep

Boston

32,400 4 +319/

205

2660 Boston,

Amherst, MA

9/75 -iii 7h 1119

IftolliaaM11Mignalki1110100/110b11.0..4,2~.10*~0,46M1r0~b0~11000.~0.

,Columbia Test 62,600 514 21/ 9000 NY

Preparation

Institute

Law Board 22/ 34,901 102 22/ 3000 est, thruout United

Review Center

States

Mlio,d0601.1111dPIWIMI1011V.011

Creative Educe- 3,960 38 23/ 1200 .LA., SF

tional Services

9,500 10 24/negligible NY

Law Boards

Institute.

6 meetings

(24 hours)

5/75-12/76 $135 6 meetings

(21 hours)

Prior to

forwareplarireiblharromerielrbollaribellalMilill

10/74-12/74 $85

10/75- 7/76

10/76-12/76 $60

$160

10 hours

(one weekend)

7/75- 7/76 $154

$88

13 hoUrs

24 hours

30 hours

16 hours

Americo.)

Tutor Int'

Service

firalftgarmlIMAIWW.12"0

Guidance

Center

5,400 50 unknown NY 10/74-12/76 $100

$175

8,745 9 25/ 8375 LA7/7i.10/76 $125

- 20 hours

(5 sessions)

20 hours

19/ LSAT only

10/ TOW all exams

52

21/ All exams,

Doesn't coach LSAT

22/ Including franchises

in NY area only

23/ LA only

24/ Estimated

25/ LSAT only

39

TON. NUMBER OP 1976

NAME Of ANNUAL SAM STUDENTS ADVERTISING CITIES DATES COST OP LENGTH OP

ORGANISATION 1975-1976 10/74-12/76 EXPENDITURES WHERE OFFERED OFFERED COURSE COURSE

Dollars

Dulac 46,952 26/ 10 27/ neg ) iq ible NY

~Iltommlrr.,6.

Prior to $110 - 10 hours

10/74-12/76 $135 (private)

A. Mandell negligible 16 28/ negligible GA

Education

Center

McBurney

YMCA

unknown 18 negligible

160 29/

College Entrance 14,200

Testing Service

142 30/ unknown

.0.1.11.1dlyfteillni....11.0~00limablybspmollmo...16

College Skills unknown incomplete negligible

recordsCenter

Test Prepara-

tion Center,

Inc, NY

88,777 595 30/ 4723

unknown unknown unknown

NY

orIONINOYOMOmmlobaidliftwO"IIIOMMI.0.11....

7/75-10/75 unknown unknown

NY

prior to

4/75-11/75

4/76-11/76.

$100 30 hours

NY prior to

10/74

$195

$65

30 hours

6 hours

NY unknown $75 24 hours

26/ Includes all courses

27/ LSAT

28/ All courses

29/ Total 1973 - 1975, 142 occurred before relevant time frame,

30/ SAT only (Some names may be missing),

40

Ski ,

B. NOT- FOR - PROFIT

We have been aware from the outset that the not-for-

11B

profit segment of this industry was one which required our

attention.57/ There are a large number of undergraduate

colleges and universities that offer preparation for admission

tests to graduate schools.58/ It is ironic that colleges and

universities offer preparatory courses. If they truly believe

that coaching is effective for graduate standardized admission

examinations then it seems logical they would concede coaching

is effective for the SAT. Yet these very schools employ the

57/ See Circulation by Commissioner Dixon, Category I File

762 3033, Vol. I, p. 85, wherein he stated:

I have no objection to this investigation, but I would

urge staff as it seeks to formulate standards for the

advertising of this respondent to ask itsemrhow the

not-for-profit segment of this industry would stack up

against those standards. While two wrongs (if that is

what is involved here) do not make a right, we should

be careful in areas such as this, where the underlying

issues may involve some measure of controversey and

novelty, that the effect of our actions is not merely

to create a double standard, disfavoring the entrepreneur

without eliminating the source of the ailment.

58/ See generally, Category II File 772 3000, Vol. I,

p.27-50. Among those offering such courses are Columbia

University (LSAT), Bentley College (LSAT), Fordham

University (LSAT), California State University (LSAT),

University of Michigan (LSAT), Boston College (LSAT),

American University (LSAT), Cornell University (LSAT),

Puget Sound (LSAT), Monrovian (LSAT), Colby (LSAT),

Adelphi (LSAT), University'of Mass. (LSAT), Amherst

(LSAT), Boston University (LSAT), Scranton (LSAT),

Mt. Holyoke (LSAT), Chatham (LSAT), Washington Uni-

versity (LSAT), Trinity (D.C.) (LSAT), and University

of Houston (GRE).

SAT as a prerequisite for admission and as a predictor of

academic success. Furthermore if these colleges believe the

SAT is coachable then they are employing an unreliable, in-

valid predictor, and they believe it is not coachable then

by analogy the graduate admission examinations are not coach-

able and they are wasting students' time and money.

Tne not-for-profit segment of this industry can attempt

to rationalize its offering of preparation courses for gradu-

ate schools in a number of ways. First, the not-for-profit

courses are usually offered at no charge or at a minimal charge

to cover overhead.59/ Second, some members of the not-for-

profit segment of the industry offer coaching as a means of

diverting students from the commerical courses. They'believe

that coaching (inoluding their own) serves as no benefit but it

is better to serve no benefit gratis than it is for up to

$250.60/ Third, the not-for-profit coachers may believe

59/ See Category III File 772 3000, Vol. I, p.43. Statement

of Fordham University. (no charge).

60/ See File 772 3000, Physical Exhibit Q p.4, and

Category III File 762 3033, p.33-36. See also,

Pre Law Guidebook of Boston College, Chestnut Hill, MA

TT974) p.64e where it is stated you can prepare for

the LSAT "... to the extent of becoming familiar with

the types of questions asked. It has been my experience

that a thorough perparation with the questions that

are offered in the Pre-Law Handbook, the Admissions

Bulletin and the $5 Arco or Barrons preparation book

will eliminate the need for expensive professional

pre-law courses." Yet, Boston College offers an LSAT

preparation course for a fee of five dollars. (The

fee is to cover costs of overhead and printing).

42.

that they are effective. Realizing the importance of stand-

ardized admission examinations in the graduate school process

they may feel that by increasing the scores of their under-

graduate students they increase these students' opportunities

for graduate school admission and therefore enhance their

own undergraduate school's reputation.

Not only does the not-for-profit segment of the industry

charge minimal or no tuition but they also engage in

virtuall} no advertising. Furthermore their enrollments are

rather sma11.61/ The not-for-profit segment of the industry,

although existing at many schools, has a small total en-

rollment 62/, charges de minimus fees, and does not advertise.

There is one notable exception to the general description

of the not-for-profit segment of this industry that we have

discovered. Recently, for a fee of $125.00, St. John's Uni-

versity in New York began advertising for an SAT Prep Course.63/

Not only was this fee comparable to that charged by some of

the commercial entities but the literature distributed was

61/ Columbia U. enrolled 40 students from 1974-1977.

See Category III File 772 3000, Vol. I, p. 28.

62/ Probably not more than 5000 students nationally.

63/ See Category III File 772 3000, Vol. I, p.70.

43

"commercial" in nature.64/ This is the only instance in which

we found a university offering a coaching course for the very

examination it uses in determining, at least in part, who

will gain admission.65/

We have not discussed the not - for -profit coaching that

occurs in public and private high schools across the United

States for the SAT. These are largely unadvertised, voluntary,

non-fee courses and as such would not cause any economic harm

to the student taking one, although they would be a waste of

time if valueless, but would result in social and economic

harm to those who are denied access to them if they are bene-

ficial. Due to the nature of the statistical analysis of the

SAT the inclusion of these entities in our inquiry is not

necessary but we will as a by-product make recommendations

concerning means of offering a gauge of their effectiveness.

64/ Id.

65/ For the full text of an interview with a dear. of

St. john's University regarding its SAT preparation

course and hi.; rationalization for its existence,

See Category III File 772 3C30, Vol. I, p.71.

44

59

VII. STATISTICAL AND COMPUTER METHODOLOGY

The reader who has little interest in the complexities of

statistical and computer methodology may bypass this section

without a loss in. continuity. The following section, "Statis-

tical Findings" details the analytical results of this inquiry.

A. Purpose of Study

The statistical study determines whether commercial coach-

ing courses can significantly increase students' SAT and LSAT

scores. If commercial coaching were found to be effective, a

secondary study purpose was to determine whether various commer-

cial courses vary in effectiveness. Particular study questions

were as follows:

(1) Are SAT and LSAT scores of coached students signifi-cantly higher than the scores of comparable studentswho are not coached?

(2) If coaching is effective, what is the estimated sizeof score increase attributable to coaching?

(3) If coaching is effective, does its impact vary amongdifferent types of students?

To the extent that coaching can improve students', scores

on the SAT and LSAT, questions may arise concerning test de-

sign, interpretation of test results, and perhaps the precise

nature of what such tests are measuring. Although intriguing,

such questions are beyond the scope of this statistical study.

The sole purpose of this study was to'estimate statistically

the impact of commerical coaching upon SAT and LSAT scores.

45

60

B. Alternative Approaches.

Two major alternative approaches exist for estimating

the effects of coaching upon SAT and LSAT scores. The first

approach is to conduct an experiment. The second approach

is to treat an existing situation as if it was an experi-

mental result.

Conducting a coaching experiment requires defining two

comparable groups of students. Members of the experimental

group enroll in coaching courses, while members of the control

group receive no coaching. The purest methods to acquire

these two groups would be to deny access to commercial coach-

ing on a random basis to one-half of the potential coaching

school enrollees, or treat the control groups with a placebo.

If the experimental and control groups are in all other ways

similar, then SAT and LSAT score differences between the groups

would be attributable to coaching.

Analysis of existing circumstances requires identifying

students who voluntarily enroll in coaching courses. Also

identified are students who choose not to enroll in coaching

courses. These two groups correspond, respectively, to the

experimental and control groups of the first approach. As

with the experimentalapproach, if the enrollee and nonenrollee

groups are in all other ways similar, then SAT and LSAT score

differences between the groups are attributable to coaching.

46

61

Both approaches have their benefits and detriments. The

.experimental approach is preferable because it avoids self-

selection of enrollees into coaching schools and because it

assures that the presumably uncoached students are in fact

uncoached. Self-selection can be important if students

attempt the SAT or LSAT under severely competitive conditions.

Under such conditions, aggressive individuals intent upon im-

proving their test scores might be more likely to enroll in a

commercial coaching course than their less competitive or less

affluent counterparts. Also, students whose first SAT or LSAT

attempt produces unexpectedly low scores may be more likely to

elect commercial coaching than students whose first exam score

more closely matches their expectations. If aggressive atti-

tudes lead to better exam scores even without coaching or if

coaching school enrollees are mostly students receiving unex-

pectedly low first scores, then score gains or differences

attributable to coaching may contain a self-selection component.

That is, self-selection may lead to an overestimate of coach-

ing benefits.

Conversely, a nonexperimental control group may lead to

an underestimate of coaching benefits. That is, presumably

uncoached students may in fact receive some form of coaching

other than formal enrollment in'a commercial coaching course.

They may for example attend courses given by the not-for-profit

segment of the industry, engage in extensive self-preparation,

47

or attend a commerciil entity surreptitiously due to a desire'

to receive coaching. These unmonitored efforts, if they occur

and if they are effective, should tend to increase the average

test scores of the allegedly uncoached students. These in-

creased scores, containing a component properly attributable

to coaching, tend to shrink the apparent benefit from commercial

coaching.

On purely theoretical grounds the experimental approach

may be preferable. However, the experimental approach is ex-

pensive, lengthy, and in the real world almost an impossibility

as it would require denying access to commercial coaching to

subjects who want it, or to treat them with a placebo. Also

it would have conflicted with our investigative goals in

that a sufficiently large sample fo: statistical purposes is

unlikely to remain a secret from the coaching schools. Coach-

ing provided in awareness of an experiment may differ from the

coaching normally provided. Since the purpose of this investi-

gation was to evaluate commercial claims made for "normal"

coaching provided in the absence of awareness of an experi-

ment, and because of limited time and study resources, we

used nonexperimental data.

C. Overall-Study Design

Given the practical considerations of time, cost, and

security, this study used a nonexperimental approach to

48

3

estimating the effects of coaching on SAT and LSAT scores.

The major study elements were as f011ows:

(1) definition of a study group of SAT or LSAT-takers;

(2) separation of coaching school enrollees from nonen-

rollees;

(3) establishment of subgroups with similar backgrounds; an'

(4) comparison of coached and uncoached students within

subgroups.

The major study assumptions were and are as follows:

(a) the study group is representative of SAT andLSAT populations;

(b) SAT and LSAT exams are comparable among adminis-

trations;

(c) coaching school performance is consistent duringthe study period; and

(d) the effects of enrollee self-selection, if any,

and of coaching of presumably uncoached studentsoffset one another.

A brief discussion of the study elements follows:

(1) Definition of study group. The study group defini-

tion began with coaching school enrollment lists obtained from

the coaching schools. These enrollments contained student.

names, addresses, and course dates covering the testing years

1974-1975, 1975-1976, and 1976-1977. This 3-year study window

was defined by several factors including:

(a) incomplete, illegible, or missing coaching schoolenrollment records prior to 1974; and

(b) a desire to avoid any confounding influencefrom long-term gradual declines in average SAT

scores.

49

As previously indicated SAT coverage was confined to two

coaching schools in the New York metropolitan area; LSAT

coverage included coaching schools in the Boston, New York,

Chicago, and Los Angeles metropolitan areas. Inspection of

the student addresses allowed definition of the primary mar-

ket areas served by the surveyed coaching schools in these

metropolitan areas. These market areas are compact, contigu-

ous areas which generate most of the coaching schools' en-

rollment. At the 3-digit level of Zip Code geography, the

SAT market area for metropolitan New York was:

. 064-069 Connecticut

. 070-080 New Jersey. 085-089

. 100-127 New York

The LSAT primary market areas were:

.

.

.

.

.

010-029 Massachusetts ) Boston

070-080 New Jersey )

085-089 ) New York City)

100-127 New York )

463-466 Indiana )

)

530-538 Wisconsin ) Chicago)

600-619 Illinois )

900-937 California ) Los Angeles

Given these geographic market areas, CEEB, LSAC, and

ETS then provided the test records for all persons located in

these areas who attempted the SAT or LSAT during the study

50

6.5

windoW. This large group of persons, containing both coached

and uncoached students, constituted the preliminary version

of the study group.

(2) Separation of enrollees from nonenrollees. Separat-

ing enrollees from nonenrollees began with identifying coach-

ing school enrollees within the large group tested during the

study period. Essentially, this task consisted of taking a

coaching school enrollee's name and address and searching the

testing data until that person's testing history was found.

The goal of this task was to establish for each person a com-

pact data set containing the person's testing history, coaching

history if any, and relevant biographical information.

For a variety of reasons, this proved to be a fairly

formidable undertaking. LSAT statistical analysis began with

11,906 coaching school enrollments and 157,982 testing histor-

ies during the 3-year study window. Computer matching using

zip code, surname, and first initial located 9,479 testing

histories for coaching school enrollees. Subsequent manual

matching, for enrollees whose testing histories were no found

by computer search, located an additional 2,869 testing

histories; the manual search was required because of the use

of nicknames, misspellings ,,nd the like.

For the 11,906 coach...n, school enrollments, we had now

located 12,348 testing histories. Clearly, some persons

had participated in more than one LSAT testing year. The

51

resulting search for the same persons in successive testing

years shrank the total LSAT study group from a presumed

157,982 to an actual 132,043. Of these, some 7,981 had taken

all their LSAT tests prior to test year 1974-1975; these per-

sons were dropped from the study.

This left a group of 124,022, of whom 8,660 had a total

of 9,029 coaching school enrollments, implying that the study

contained at least a few people who had been coached more than

once. In addition, scanning sample testing histories indicated

that many people had attempted the LSAT more than twice. To

make sense of this situation, the following table was produced:

Number of times coached for LSAT

0 1 2 3 4 5 totals

number of 1 81,575 3,722 69 0 2 0 85,368

LSATattempts 2 26,660 3,172 113 4 4 2 29,955

3 6,093 1,146 69 7 2 3 7,320

4 785 223 26 3 2 0 1,039

5 190 56 9 0 0 0 .255

6 42 13 3 0 0 0 58

7 13 5 0 0 0 0 18

8 3 3 0 1 0 0 7

9 1 1 0 0 0 0 2

totals 115,362 8,341 289 15 10 5 124,022

Inspection of testing histories suggests thal-,these groups

tend to perform rather differently on the LSAT. This finding

52

67

in turn suggests that comparability of enrollef:s and uncoached

students required some assurance that apparent 1-time takers

are in fact 1-time takers and contained no persons for whom

the presumed single LSAT is really the first of two attempts

or the begining of an extended LSAT taking career.

Because proper analysis of all combinations would have

entailed research design complexity sufficient to overwhelm

available sample sizes, and because 115,129 or 93% of the LSAT

study group had the "normal" LSAT experience of fewer than 3

LSAT attempts arl fewer than 2 coaching school enrollments,

persons coached more than once or with more than 2 LSAT exams

were excluded from the study.

The apparent 1-time takers, 2-time takers, and 3-time

takers were used to conduct a "window analysis". The purpose

of this task was to discover how far to narrow the study win-

dow so that students thought to be 1-time takers were indeed

likely to be 1-time takers, and students thought to be 2-time

takers were also likely to be 2-time takers. The actual

analysis consisted of constructinc' time profiles of 2-time

takers and 3-time takers. The ti profiles indicated how

many intervening LSAT exams passed between successive attempts.

Because the. latest LSAT coaching school enrollments were for

December 1976, and because this group contained a considerable

number of coached persons, we were extremely interc,sted in

53

the effect of using that LSAT as the new boundary for the

study window. The results were as follows:

percentage of LSAT repeaters with 3 or fewer intervening

LSAT exams between successive attempts

2-time takers 3-time takers

1 to 2 1 to 2 2 to 3

uncoached students 75.8 '86.8 55.6

coached before 1st 89.0 96.1 63.0

'coached before 2nd 82.4 86.1 65.6

coached before 3rd88.0 56.0

Assuming that these time profiles of LSAT repeaters are

stable through time, and because three SAT administrations

followed the December 1976 LSAT during the 1976-1977 test year

(the test year commences in October andterminates in July), we

can estimate that about 76% of uncoached 2-time takers whose

first attempt occurred in December 1976 would, if they are

destined to remain uncoached 2-time takers, appear for their

second uncoached LSAT attempt before the end of the test year.

A similar interpretationapplies to the other percentages..

It would have been more comforting to work consistently

at a level of 95%, but this would have required such a narrow-

ing of the study window that all 197('-1977 test year data

would have been discarded. Because the study group was about

to shrink drastically, reasons described infra, it was

decided to leave the LSAT study window at October 1974 through

December 1976 inclusive. in any event, the LSAT study group

54

6,9

was then partitioned into five subgroups based on coaching

experience:

(i) uncoached 1-time takers;

(ii) coached 1-time takers;

(iii) uncoached 2-time takers;

(iv) 2-time takers coached before first exam; and

(v) 2-time takers coached between exams:

A similar adventure marked definition of the SAT coached

and uncoached groups. An additional complication arose be-

cause the SAT testing histories arrived in fragments; that

is, one person's data was likely to be found in several iso-

lated pieces, even within a single testing year. As a result,

the entire SAT data file was sorted alphabetically and a list

of "candidate matches" was generated for each surname in the

coached group. Because of use of nicknames, apparent misspell-

ings, and variations in format of street address, manual search

of these candidate matches was-used to find the SAT testing

histories of coaching school enrollees. Testing histories

were located for 1777 of the 2286 SAT coaching school enroll-

ees. Part of the failure to locate testing histories for all

enrollees in both the SAT and LSAT files is of course attri-

butable to the failure of a coached individual to utlimately

take a standardized admission examination. Because the SAT

data consisted of so many elements and was so scrambled, a

sample of uncoached persons was felt to be the most reasonable

55

70

and efficient way of establishing a control group. The target

sample size of 2500 was chosen to yield a control group com-

parable in number to the coached group. Systematic selection

with random start was used for two reasons:

(i) unknown and apparently unequal fragmentation of

individuals' testing histories removed the equiprobablity

advantages of simple random sampling, which would ordi-

narily be used to allow statistical evaluation of results;

and

(ii) selection of coached students was also nonrandom,

since the coached group contained all students enrolled

at the investigated coaching schools during the study

period.

Choosing an integer between 1 and 150 from a table of

random digits and then selecting that data record plus every

succeeding 150th data record yielded a sample of 2597. The

same type of "window analysis" as used for the LSAT showed that

a December 1976 cutoff allowed working at the 95% level. As

with the LSAT part of this study, the SAT study group was

divided into five subgroups:

(i) uncoached 1-time takers;

(ii) coached 1-time takers;

(iii) uncoached 2-time takers;

(iv) 2-time takers coached before first exam; and

(v) 2-time takers coached between exams.

(3) Establishment of subgroups with similar backgrounds.

The reason for establishing subgroups with similar backgrounds

was to avoid bias either for or against coaching schools. The

56

71

reasoning was that intergroup ethnic, academic, or sex dif-

ferences might lead to corresponding SAT or LSAT differences

and, possibly, to differences in coachability. Therefore, by

an extensive sorting of LSAT test scores we attempted to dis-

cover whether significant LSAT score differences existed,

coaching aside, among groups defined by sex, ethnicity, cumu-

lative grade point average (GPA), and combinations of these

attributes. This sorting revealed that LSAT scores:

(i) vary moderately with GPA;

(ii) are markedley different for different ethnic(and racial) groups; and

(iii) display only slight differences for males versus

females.

Therefore, as described infra, analysis of the effects of

coaching was confined largely to whites.

Restricting attention to persons for whom biographical

information was available further reduced the LSAT study group

size. For instance, using a group whose LSAT exams fit within

the study window of October 1974 through December 1976, whose

GPA was known, and for whom no score irregularities were

recorded produced the following study group.

coached uncoached

1-time takers 2,272 30,459

2-time takers 2,090 10,832

Requiring information on sex and ethnicity shrank the

study group still further. Excluded here were persons whose

57

sex or ethnicity were unknowl or whose records contained

contradictoryinforaiation, such as being black one year and

oriental the next year. The resulting SAT study group thn

appeared as follows:

1-time takers

2-time takers

coached uncoachee,

1,571 22,136

1,259 6,651

Indians and orientals appear to perform differently on

the LSAT when compared to other non-whites. Therefore, and be-

cause these groups comprise only a small part of the uncoached

group and a miniscule portion of the coached group, they were

excluded from further analysis. The resulting final definition

of LSAT study groups is shown below; as used here, non-whites

include blacks, chicanos, and puerto ricans.

white nonwhite11,1T

uncoached 1-time takers 7 2,448

coached 1-time takers

uncoached 2-time takers

1,395 147

5,729 781

2-time takers coached before first exam 672 54

2-time takers coached between exams 1,113 128

Fortuitously, the SAT study subgroups proved to have

very similar socioeconomic profiles, so that no further prun-

ing or subdividing was necessary to assure comparability. The

only qualification for inclusion in the final group, for rea-

sons described infra, was a Pre-Scholastic Aptitude Test

58

(PSAT) baseline score prior to the first-SAT administration.

The resulting final SAT study subgroups were as follows:

uncoached 1-time takers 384

coached 1-time takers 109

uncoached 2-time takers 494

2-time takers coached before first exam 248

2-time takers coached between exams 246

(4) Comparison of coached and uncoached students. The

basic approach to detecting coaching effects was quite straight-

forward. For both the SAT and LSAT portions of the study, a

baseline was established to compare coached students against

uncoached students. From this baseline were generated expected

average test scores for coached students and for uncoached

students. Direct comparisons of these average scores were

then used to evaluate the effectiveness of commercial coaching.

PSAT scores furnished the SAT baseline, while undergraduate

GPA provided the LSAT baseline. Both baselines were chosen be-

cause of availability, convenience, and apparent reasonable-

ness. Although GPA is clearly inferior to PSAT on the basis

of reasonableness, because of the universal uniformity of

PSAT scores, it appeared to be the only available base-

line for the LSAT when analyzing one-time takers.

For both the SAT and LSAT analyses, the technical approach

was the same. Ordinary least-squares linear regression was

used to generate conditional average exam scores for both

59

74

coached and uncoached students. Estimated impacts of coach-

ing were then inferred from comparisons of these conditional

averages.

The statistically sophisticated reader will immediately

note the absence of fiducial lines around the linear regression

estimates. This omission is deliberate. Strictly, fiducial

lines can be drawn only where the underlying data is the pro-

duct of sampling where sample selection probabilities can be

calculated; because the calculations are convenient and well-

known for simple random sampling, this approach is normally

used. However, investigatory requirements and constraints

of time and cost have prevented this statistically desirable

approach. Fortunately, the difference between coached and

uncoached student scores is sufficiently large that the

absence of this formal statistical test appears to be a minor

defect.

At this point it is of value to focus on the terms

"control" or "uncoached". They have been used in the preced-

ing discussion as well as in other parts of this memorandui

and are used on labels on the charts and graphs appearing

infra. These terms are somewhat misleading. We have not

determined to a certainty that the "uncoached" or "control"

groups have in fact been culled of all coached subjects.

Some subjects may have attended a coaching school we have not

identified, they may have undergone extensive self-preparation,

60

7.3

they may have been coached by the not-for-profit segment of

the industry. For purposes of convenience we continue to

use the terms "uncoached" or "control" remembering that due

to the probability these groups are not pure we will have

underestimated the benefits of coaching.

VIII. STATISTICAL FINDINGS

The following discussion analyzes with statistical

methods the effects of coaching on the Scholastic Aptitude

Test and Law School Admission Test.

It is important to keep in mind that although this docu-

ment is meant to be self-contained, it does not include all

of the different approaches and methods we have employed in

attacking the problems we faced. For example missing from

this discussion are histograms, frequency tables, and analysis

by stratification (although this is in essence included in the

regressions).

In order to fully understand the breadth of this undertak-

ing it is essential to make an in-depth review of the techni-

cal appendices. However such a review io unnecessay to compre-

hend our findings.

The charts and graphs appearing in this section that

analyze individual coaching schools are labelled with either

a numerical or numerical/alphabetical code in lieu of a

coaching ichool's name. These codes are present due to the

convenience they afford in labelling. The codes corresponding

61

76

to the coaching schools names are found parenthetically in

the text of this section after the mention of a coaching

school. The full set of codes which must be utilized to

decode coaching school names in the technical appendices

are attached hereto as Appendix A.

A. THE SCHOLASTIC APTITUDE TEST

(1) SAT 1-time takers

Actual SAT scores for 1-time takers appear on pages

65-67. The first of these pages displays verbal and math

scores for uncoached students. The second page shows verbal

and math scores for students coached at the Stanley H. Kaplan

Educational Centers, Ltd. in New York (SHK) (OO1A). The

third page shows verbal and math SAT scores for students

coached at Test Prep Centers in New York (022).

Pages 68 and 69 compare average SAT scores for coached

students against uncoached students. These comparisons indi-

cate that SHK (OO1A) is extremely effective in coaching 1-time

SAT takers. Average verbal SAT increases attributable to

SHK's (OO1A) coaching range from 40 to 76 points above the

uncoached group. Students who score higher on the PSAT re-

ceive the greater benefits. Average score increases for the

math SAT are 40 points above the gains of the uncoached group

over the entire PSAT range. Students of SHK therefore who

are coached in the period intervening the PSAT and the SAT,

62

and who take the SAT only once, show a combined net increase

of between 80 to 116 points above a' matched control group.

Although generally beneficial, Test Prep Centers (022)

appears to be somewhat less effective than SHK !001A) for

1-time takers of the SAT. Average verbal score margins for

students coached at Test Prep Centers (022) range from an

increase of 60 points for low-PSAT students to a decrease of

33 points for high-PSAT students compared to the uncoached

group. Conversely, average math SAT score differences range

from an increase of 8 points for low-PSAT students to an

increase of 40 points for high-PSAT students above the gains

of the uncoached students. The total net benefit for 1-time

SAT takers at Test Prep Centers (022) can be calculated to

be between a range of -25 to +100 points compared to the un-

coached group.

To amplify on the preceding analysis SHK (001A) shows a

gross verbal SAT increase from PSAT verbal of between +84 to

+42 points from the low-PSAT to high-PSAT verbal scores. The

corresponding control group shows an increase of +44 SAT

points at the lowest PSAT verbal score to a decrease of -34

points at the highest PSAT verbal score. Therefore the net

benefits to SHK (001A) students range from +40 SAT points

at the low end of the PSAT range (+84-+44) to +76 SAT points

at thg,high end of .the PSAT range (+42- -34).

63

78

Correspondingly SHK (001A) shows a gross math SAT in-.

crease of +68 to +39 SAT points from the low to the high end

of the PSAT math range. The matched control groups shows an

increase of +22 to a decrease of -3 SAT math points from the

low to the high PSAT range. The net benefit therefore to SHK

(001A) 1-time takers for the math SAT is roughly 40 SAT

points over the entire PSAT range (+68 - +22 = +46 to +39

- -3 = +42).

This type of analysis can be done for each of the scat-

tergrams presented here by merely inserting the appropriate

figures into the formulas representing the regression lines.

It is important to note that the figures recited in the

preceding analysis are conditional average figures based on

the equation of the regression line. The actual score in-

creases or decreases may be considerably larger or smaller

as can be seen from the plots of the scattergrams.

6479

SAT 111.11

Ver.4

MAI

f2201

111.40

Was

&ism

4:2.22

11140

Map

Mao

210.00

1

1

1

1

I

2

I.I

I

:

I

IIt.3

SATV=PSATV(8.7)

N=384 4.I

11.:

:

I

2

0 .I

0224, eso

1 . I 1 .1 3

....

1 3

22

2 2 2 4 22 42

Se

2

2 2 3

+ 70.4

000000 ........2

I.

. ..

1 27 2 2

o w . 2 2

1 I . 1. . 2j . 2 2

2 4. 3

. 2 2 2. . 2... .......q. ....

I 1

I 2f f

1.

1.

.2

2

. ..

1

.2

3

i3

3

a

I

1

1

I

I

2

2

I

:

I

Is

Z

Os

1I

I1

I

I

I

I

1

I

1

1

1

1

1

1

1

1

1

1

1

1

1

20.00 23.30 30.40 11.10 41.10 44.30

SAT verbal and PSAT verbal coreefor uncoached 1 -flee takers of SAT

27.10 42.44 47.70 71.c

PSAT Verbal Scores

170.10SAT'lath

Scores

....... o.1

I

21

o. .

114.10

I

a : 1

1

1

411.10

.1 3

. I f

412.11l

I 2

..... ....... .... ..2--

Ior f .1 Ili

1

242.0e 2 4 2 42

1

1

I s a 2

:122.11 .6 a a Ia

2 1

1 1 1

I 1 f 2 2 1

11.111 2

I2 I 2

I

11

1 2

2144: a 1

II 2 1

I I I 1

I f 1 1

SIMS 3 I

I 31

2. 4 1

8 I t 13I

1

14441 -60

1 ,41.,ii,r 3co

2

212.12

141.11 11.11 33.40 11.41 43.21 41.111 11.41 13.41 41.40 67.1 7140

PSAT lath ScumSAT Math and PIA? oath .cfor unecachd 1-tloa thl of SAT

VerbalScore*

e.

1

1

1

1

tsg.le1

1'SATV= PSATV(9.3)A

1N=72$114.112

416.011

MAI1

a

124.02

424.00

a

+ 98.2

1..

MAO

Si...,3

st,:ZIA*

tO.00 2s.co

a

30.00 33.00 40.00 45.00 50.00 $5.00

SAT verbal and ?SAT verbal score.for 1-Clce taker. coached at 001A

740.00

7s5.e0 SATM= PSATM(9.5)

2112.11

4711.110

N=72

+ 78.4

1421.11,

IS4.110

a

1

1

317.0. .

1

1g.

MA,

al

a

. -2

40.00 45.20 20.

PSAT Verbal S

2

' mee 23.1112 32.20 37.10 41.4O 47.15 48.1e 11.70

SAT rash and rSAT nath 00000for 1-11.4 Were coached atinA

66

11

Olaf 47.20

SLAT nath Scores

a

3

".5

SAT /MgtTselLal

Stow'15.0

41/.110

SATV=PSATV(7.11N=37

+ 162.4 .

1141.111

...... 40

211.21

all0.$11a

434.01

a

112.11

144.0

gtg.1

.RIS.1

...

10.40 24.) 16.48 32.10 31.70 4140 44.1 110.10

.SATMathScores

SAT verbal and PSAT verbalfor 1-time takers coached at 022

..... «g.s

s

' SATM=PSATM(10.1 + 35.4sg).

smog

ssg.

.

N=37

IMO,

s

La..a

Sl.le

SUM

IW$

A." IA',

.

.

MAO $41.71) 113.

TSAT Verbal Scores

# ...... 0

la..: /1.71 15.15 $4.11.4 4940 ,Lee Slag ISM SI il.S It

8 9SAT usth sod rsAT cot..0s, $ilas gales, teethed el 022

PIAT bath Scores

i . . . ' - . .1.. :.: . .... . .

200 1 ! ; . : i ; : I : . ;

1 : : ; ; : . 1

- ------- I : .1I : I

. : I 1.:_ ..1- ; ..!. .1 1-400 . ;.. .._...._.: 8_: 4- 1. 1_

; . 8 : I :

1 ;

!- --1-; -I- I II I I ! ; I

300 rT7-1.-:-___.:.._:....__._ ;.... .._._ ._......_. /...... .1----- . .1

-: 1-7- I ----.-- 1

.--1I

- ine.va Ned

1

. I

400

200

I

I o

..,....-7--1...-1.. .........................1- - - i .

1.8 1_Joo la;. ...1. I _ i _44-1 .1 _ _12 Li : _. Li ol'I':1

20 430

SAT t0f3 rLathicor as

700 --

SAT mint verbalfor 1 -time triccre.

tudente

50 60 70 tO/SAT Verbal Score.

*cores a (unction of rsAr verbal scoresres:... 001A enrollee gainat. uncoached

00Ut

;

: I 1

--% I

I

; .

600.1

'1

1

! I _;--

471Incoached

I I ;I , ,

Sooi 1

: :g ! I

.

400 ...____,-I I-_. --1_4--_.. - ---.......-!..._ _4....;__ . 11

o s . ....1

I1 , ; I , I : : I.

; J ___: ___. . __.. :.... : j . 8.4_ '... .. _J

___I ' -i 1 .--71 -----

- 1...* .... . -... . -.-._. , ........1-1-:I

200

I_1

-riI'

I1 J1

1 1_

1717120 30 40 50 60 70 l°/SAT Kith licore

SAT ge oath eor f unc t tan of /SATearth for 1 -tine takers. elapartng OblAenrollee. getnst ,sleoser.e.f tvdents

.

; .

688 3

'term/ /

; :

.L-4,

SAItat,Scot %

700

600

II i L. _ _I . -.. _ ____.1_. _1 I , .

1 - i !' .1 i talc . octiod..

:-

H .i : i 1 ,

-;-i-._ ! .

;4---- -FIT--- 1

I :. . :

. .'.--I

I tr -t ,

1 , , .. ; ;

; :, '' i -I i i--- - ..-- -.--

.ti, 1- . -:-I--i.-400 - ------1 _4

_., _ __I

1,..., l

.1 .... ....! ...... .1 ---......

1

I

. I . I II i

I

--4

-.. : I

1 . . : .. L . .. ..' I

I ..1 1 '1 7- ..r".......-....

1__1 .il__11' il . . 1

---,--. ---- ---- - ..._._.___. ._ . , _ _ .

1.-T-j- -I ---7-7-7-7 71/ I 1_ '

-1

1+1 _l__I ;_ i_i_i_ l_i_i_l . __ _L41_11_

20020 30 40 50 60 70 EJ

PSAT Verbal Scores

SAT verate verbal scores as function of SAI verbalscores for 1 -tire taket tn:-partng 022 enrolleeage ltat tconched tudents

20 30 40 50 60 10 BOrSAT Math Score

SAT varra euith acorn its fur tfon of PSTIN4th et0f41$ (or 1.tir, talfo, tor .arin$ 022enrollee, ttlnt uncoacheJ student,

698

(2) SAT uncoached 2-time takers

Actual SAT scores for uncoached 2-time takers of the SAT

appear on pages 71 and 72. The first page displays first-

SAT and second-SAT verbal scores. The second page displays

first-SAT and second-SAT math scores.

Page 73 compares average SAT scores for the students'

first and second attempts. For both verbal and math scores,

SAT improvement for uncoached students appears to be pro-

portional to PSAT scores; a similar improvement pattern

is shown later for uncoached white 2-time takers of the

LSAT.

The precise meaning of this improvement pattern is not

clear. Possibly, two factors are at work:

(a) Persons with higher initial test scores tend to con-tinue to develop further than those with initiallylower test scores; and

(b) Persons with higher test scores find the SAT somewhat"learnable" even without coaching. The fact that asimilar score growth pattern occurs for uncoachedwhite 2-time takers of the LSAT, who are presumablyalready mature and unlikely to exhibit the rate --Jfaptitude growth of younger SAT students, tends tosupport this hypothesis.

Sc3

70

&AT

Sccres

411.4

SATV1=PSATY(8,1h

I

411.111

1

.N=494. et 1

1

1

c a 2

1

121.101 2

12

11

.2

322221 2 t

) 3 3

433.301

I 2 2 4 .22 I3

a 3 1 2 2

1

1

2 - 2 2 2

3

2 2

31 1 3 2 f g

5'.

2 1 t

..4- .1-4.4.

13

)/43112

$ 2 1.3 2

3a 3 3

3 2 I

. 2 2 1 1

2

1

1

1

MAO1 2 2

3

1. 3

1

MAO 3

ur 130.0

lerb.1S-ore

1.4.40

20.90 31.03

a33

1..e

11)11

ADD01

432.0

m

10.30 31.10 30.00 31.00

S.

-:9

a

I.

1.0.00 ms.o

LSAT and 111-6t-SAT verbal scoresfor uncoched 2-tlmc takers of SAT

1

ShT72= PSLTV(8.7D + 82.7-4=479

40.00 45.00 I

?SAT Verbal Scores

2

4

2 1

1 3

13..

2 2 1 2 1

20

1

3 ) 2p

u p

3

2 3 I

S.3 3

a)

g 3

1

S.II .3 7. 3 ;

Iso; g .m,. 01

2 n 1 - .0," 3 33 , . s

w ; .

12

,,0:004d/Ism I

I 52

, t D

3

a

13.01

22 4 8

3

1 A21 Si

3

3 1

3

11.11 23.111

71

4.10 41. 1.11 11.2 00.10 45.0*

UT Verbal S

PSA7 and second-SAT oralfor wrocochd tler of SAT

86

10

222 360.66

haft&We a

1:41.99a

SATM1=PSATM ( 6'4=454

. 9) + 7' 7. .

161.66It /

1

41241

2 2

a

........

1/J 1

SSI 2 3 2

I.

2* 3 f we 3

...al 322 3 4 2

. 2

a 2 3 3

11116.66 2 1 2 $

1 2 1 2 2 33

144.21

3 2 3 f2

a

1

1 2 2

3

3 3 . .farmirrmwm -1-28

5 2

3.1.50.1 1

II I 2

12 3

2 so 2 2 -

2 1531.611 1 ..

1:

121

1 2 a*.

1

274.00. t

1.120.00

. . o... .

21.00 24.10 21.20 14.10 4..40 44.10 11.40

PSAT and first -SAT oath storesfor uncoached 2ttne takers of Si

54.70 61.40 44.10

PSAT hoth Score.

( 9 . 68 . 3 0

8

SAT 7111.40

Mathscore.

733.4{

1 . .-

1 SATM2=PSATMN =479

1) 1-

as 228414.0!

2

12 12 3

2 S.

3

111.662 2

/2 3 )

a . - .-..-......- ......

14 2 ) 1 2

12 2 2 I.

561:6,12 1 3 4 3 2

33 1

22

2 as

4 1 3

661.66 2 2 4 4 3 2 4 1

12 4 1

12 2 Z

1 2 2 3

1 4 2 2 5. 2

241.0 11 1 1 1 8 1 -

2 1

/ 4 1 I 1 IIlermammma .4ra-r. I.. . ...... .m ....... . w

1

MO.. 41. 2 a

1 8 1

WA,1

2

1

221.861

1

2. '

2211.0

21.66 16.16 11.26 14.11 44.42 16.16 61.66 16.10 61.16 114.10 71

/SAT Path Score.

72

PSAT and scond.SAT Nth Cafefor ongochd 2 -tire takers of SAT

7

SA

TT

ost61S

cores

SA

TS

core

114RI

700600

300400

: , . - .7 i I , ! i

- ! i : I .: : .1..; : .1 I ;

- ,i..-i -

I

e.--/

t--

_I__ .! costr1

SA

I

L-.

1 .

I 1 -;.. . . -.

: iI I --:-.

1

Iii

II I l' ! : I ;. . ..4

- 1 I- 4 '

, :. .:.........:-.-1-1-

-1 : i- - - ."-tit

et SA

T -1

-.

-:.-1-...!-. -

! I - --, .s I

, _..i._!_

_I.

_I _

!_ i. :

.

_IF-1 --11 .. ..

_:...._,. 1._..........

i .. : .1I :- -

:......1L

:Li..1.__

I .;_l

- ?. .

. ..' I

-:

I

I ;

:_ 7

__J

3oo .

! I " ItI-i --,

-i-t---1--,-t-r

7-17

7-1

200

1 -1 1-1..1

1-F--

_

n t! I! i

30

4020

30 60 70 tO

PS

AT

verbI

Score

SA

T

verte

verbal

scores

as

function

of PS

AT

corc

for

ancoAched

2-tthe

takers

603

r , ---,-

--1---

.---I-

---,-1. :. I : ' II--

-..i.---

__.-.____...1: ' I. . i 1--

".-/

: -

- ,

H

1 .__'

I. - . .

I.

!..:__I..1

:-

4700

,ilf.;; , 1

.....-..ncSA

T.

'i'IL1

r. 1II

L.1 1 . ._ 1

L . I , , I . I 1 ' 1.

i I : : I

CI

. .

-1 1 L

.., ._

_____._:___I: ,-r-

r- 11:1-t

. , ..- first

SA

T

I

6.7.

- ; ' I- 7-71-

.1. - Ill.------

i I I

' .. f.-i-'

-1--- .....-

I. ', I :

: 1

5°D

.

400200

- 1.___

I, 1 i ,

t i: I

_

-4 ) -rr ._

20

30

I 1-:1- ----!

40

Sal

. I

______._

_ _ _ ___

60 70 60

TU

T

)1.th

Scores

SA

T

r*.**

sNtIl

sror

full:sitar%

of PU

T

for

untocher7

2 -tins

takr

73

88

(3) SAT verbal scores for 2-time takers coached beforefirst exam

Actual SAT verbal scores for 2-time takers coached be-

fore their first exam appear on pages 75 and 76. The

first page displays first-SAT and second-SAT verbal scores

for students coached by SHK (OO1A). The second page shows

first-SAT and second-SAT verbal scores for students coached

at Test Prep Centei:-.s (022).

Pages 77 and 78 compare average verbal SAT scores

for coached and uncoached students. The comoarisons suggest

that SHK (OO1A) is clarly effective for all students and that

Test Prep Centers (022) is effective for low-PSAT students,

but possibly counterproductive for high-PEAT students; how-

ever, Test Prep Centers' (022) high-PSAT stunts do recover

much of their lost ground on their second .9T attempt.

SHY (OO1A) students show average verbal conditional SAT

score increases above a matched control group ranging from

56.7 to 32.7 '.:Yrits from low to the high PSAT scores. Test

Prep Centers' '022) students reflect gains of 34.4 SAT verbal

points at the low end of the PSAT range but Losses of 55.6 vu.'

verbal points at the high end of the PSAT verbal range The

second SAT verbal score for Test Prep Centers (022) reflects

losses of 22.0 SAT verbs' points at the high end of the PSAT

scale compared to the matched control ltoui.:.

74

SATVerbal&r

. 4.. .

- ""*-- **

1111.11 =). ±.1583 .

- - -. ..

1

118.681

81

410.11

1

1

424.11. . 1

1

1

174.0211

0000. t232.0 I

WAS.

140.40

00.00

.

2

C

e

3141

34ItaftliSAT 227.00

Vrbal2

178.5 SATV2

434.15

514.0

1112.1

11

31110.48

412.88

1-

411.08

844.18

1141.11

1

810.01

=

.2 a

. II

2

2

2

30.40 34.40

a

a .

-

32.40 43.00 47.20 11.40 13.40 39.00

/SAT verbal and first-SAT verbal scoresfor 2-tioe takers coached at CO/Abefore their first SAT

1

PSAL V (8.05 153.4'1=139

.

2

a

..... .4. ...I

. .

I.

M

2

0

2

di'

5

PUT Vce ,1 Scores

4....0. .0 ....I

MOO 14.19 $11.611 34.45 Mee 41.05 42.12 11.4 11.6 52.211

/SAT Verb,' Scores

V4111401 and second-SAT verbel scarcefor 2-tin tars vlachod it UOIAbefore their fleet SAT

SATVerbalSevres

WAS1 . .

1

1

1114.11

1

SATV.1 = PSATV(6.N.+611.1

1 l' N=1:07.

1111.141

. ,

..... ......1

2416.41

ss

4631.1101

3

1

3

a

1

31!.003

1

158,0_a

1

a

a

1

. ....

. 3

Ofte.aad

13.911 14.13 311.4111 14.410 17.20 41.00 41.10 41.11 17.4e a1.: is

PiA7 Verbal arme

Secorase 76/.66

Verbal&erre.

1611.611

AMU

134.1

"am

VW verbal and fir at -SAT verbal scoresfor 2-tine takers coached at 022before their first SAT

SATV2= PSATV(7.t7) + 139.9N=107

I .. .....

1

1

.....13.1e 11.44 JOA 34.6 1446 411.1$ 61.111 46.6$ 43.46' S.T.ta 6

SAT Verbal Wiese

76

PUT verbal and second-SAT retailfur 2-lim takers coached At gm6416r their flask wr 91

ITIPt

SATVerbalStem

700

'Co

500

400

200

Second ELIO

SATVerbalScore.

. : 2..,......._4._, .

..

. . .,Iiii 'i'L...,,.....,_.1 .

.,.:._t_.),-1-:,-; 1 --:--!....-,......,

r.--- ;...,:...>,------

.........--, .: 4 ...:.1.,i .;..' I '.` 1--r---1--1

- i; t 1

j Imo!_

. ... I

. '..---

t Til i__......1,.-1 -I . s_...-,.._...,

i .....1. . 1

...L. _ _._..: . .. - -:----r_d-i-.- -:-s-i- , :, ......._

.\--:. .-.0.4:d.:=11-1.-_,T1

i '.-r: 7 1.- .... . -,

- --.. : ..1

''' -.,.. I .! - - I.. ;_,1

I ''; .

\_ ...------- -- . --__II 1

1 - !

:-'.-1__I__:,_ I. , ,..----- ' ---: 1

' I I' .'- , '- : '' s 1-;'"*----- '''

-77-1--: ":,,,,...--

:---------1--------...;:-.21--'s

7....1--- -.-/-* -2 -.-

..-.--- - : .....-' ', .,,,,.""--------1,-

.---- 1 '4

N... -----'7"- ; ---;

---;---------.. ' .-- f: ,-"' '. ..--........ ..:

-_,_-, 1 :---................--4----... : , i-. -f _. ,

.

t____, ...:_.; _____.............:

1 ' I ''.1 i :............." ... :.. #I

I LI j I H!--1. ,--i -1-7, -- k.,,,-- .-- -:- s ' 1 - ;--..I -: ..--...J

----- ..,-' \ ..--''

20 30

SAT verase vericor:::1.e,

of PSia verbal

---i.,---.--.-J'IvyJ---,-'II --1.-

40SO

./.--

'foc.,t. 1.1i1,isA,7T2: a

likens

ea

in$ 2-1113e t foiction

60 \,..,..i...J__3- ' i I i ;

, _-

before their fIt." 5AT

tbrer oglin" uiltikr, cosched

1t 00114

f.. ! ...

# ..

I . i . I

./....l

.

--..- : , ............. r'-: " .......s...! "...I- . ..... ,..j..... . : ...., .......,../

: :s.......''....--.1 , ,,.. 's! .

: i i ii ; ; '', '' . .I

0014....tI

1

.,. . -

200

.

:.... -, :

1" ; ;- ;1,

,

s_.

,

- .

- ..1...;..!

, _ . ,

20 30' 40 ao70 SO

SAT t f In. 5%

rAST Verbal s c ,,,,

se varlA I at6r,,oar ena-SAV

ofI$AT

verbal "I"' t '°,, Int ,. l,,,- g iting11011

At 001A Woof 4".1. titt%r:Al °II" -itt% 'i "chedt ..-

th toNI 2,Liovo tokens

77

SATVelblScores

I

,5

1 0

. ....... :; ; .5 i t1- .. .- .. - 4....--,;.................

.--

Second 6°.-1

SATV erbalScore

700

400

500

-400

200

:0 30 40 50 60 70 60PSAT Verbal Scores

SAT overage verbal scores f or first -SAT a functionof PSAT verbal scores. Ce"TArifIg 2-tine takers coachedat 022 before their first SAT with uncoached 2-time taker

_

r .: 1 -i'-'-'7-!-71-7 . .

t , .:

.:

i I : : --.----.. 1

[ ; ' '1- -.1 !-1.-:- -.:i I

. :..

i..1-,--1--- ...-T---:1-' !

., -11 1- i I!

1-71.-1---.! -1. i_i

L-' i-.! :

... .:_ ,-i--1.- .

1 . ;. :. -..---i--,--,--7-1

I ' I . -:7.- i f-7:-171:1_1

1 1.. !-f !-I .........-.:.--.I

1: ._

i s : . , .t.! ji, 022-7.._.... .

:. _ _ _ .i___.... i_

,. 1 .

.,...

! ;

1

1 .

I ; :-....1

--- ___ _...-f -7---7- I -T.--

. .r --if--

- II 1 4 1

:1-

1 r---t-

I 30.F40 SO 70 110

verbal s

SAT eeeee g verbal icor for ond-SAT trunotionof PUT verbal scoria. emparinr. 2.410. taker, coaciridat 022 before their Urfa SAT with uncoached i-ttoaZjaaro

20

I

78

(4) SAT math scores for 2-time takers coached before firstexam

Actual SAT math scores for 2-time takers coached before

their first exam appear on pages 80 and 81. The first

page displays first-SAT and second-SAT math scores for stu-

dents coached by SHK (001A). The second page displays first-

SAT and second-SAT math scores for students coached at Test

Preis Centers (022).

Pages 82 and 83 compare average math SAT scores for

coached students against uncoached students. These compari-

sons suggest that SHK (001A) coaches eff- .:ively for 2-time

takers, with the greatest benefit going to low-PSAT students.

SAT average score advantages range from 57 points for low-PSAT

students to 16 points for high-PSAT students. Comparing second-

SAT scores reveals an average difference of 74 points for low-

PSAT students and practically no difference for high-PSAT students.

Test Prep Centers (022) also appears to offer initial

coaching benefits favoring low-PSAT students. This school's

higher-PSAT students make considerably greater improvements

than uncoached students on their second SAT. Test Prep

Centers' students show improvements of 42 SAT math points

at the low end of the PSAT range and decreases of 54 SAT

math points at the high end of the PSAT scores. However

their second SAT math scores increase 77.6 points from their

first SAT math score at a high end of the PSAT range compared

to an increase of 36.6 SAT math points for the control group.

79

94

F-ut rr SAT--Isr.tia 1

1

14th 1

Score, 1

1WM1

SATI41 .= PSATM (7 ..13) .150...0012.00

3 N=141

621.441

1

1

1

$144.1

1

1

*

313.0.1

1qg

471.001

1

. 1

1

427.001

4

1

1a

370.01. 1

."". 11

1

fte.t1

1

1

1

230.02 .. ....14.50 32.46 2.40 41.22 41.40

t44..SATMathScores

08 a . 4 1.8..

"

II

I

.,

./. .....: s

a .

10.00 97.q St.55 41.10 47.41 !:

?SAT Kith S

?SAT and first-SAT oath ocnrifor 2 -cisc taker. coachea

before their (Sect SAT

. ........

110.001

11

11

1 SATM2 = PSATM (7: 9) + 166 :1N=139141.90

612.60

I

441.00

1a

1

114.01

'11

2 3

1

S4S.101 2

1

I

let.t.a .

- -3 -.. .1

3 a 1

2 1

1

40.05

3,..001

$411.00 .2

ISOM. 4.. .

f

24.0. stoke 14.10

an

...

41.4 15a. 14.45 24.40 62.30 64.44 /1.

'SAT Math Seer..

?SAT and .cond-SAT math *cotefor 2-tlia. taker coached at OglA

before their first /AT 90

.S4 46.4 411.0 4.8 9

SAT meth 'mod rsAT siTII Sc.,..Sew 1111Aos 1.1.1 coached et 022

MICROCOPY Ri',0(11TION IL ,

Stem1

478.80

100.111

Sat.,

St0.0

fe.101

410.0$

311.111

S$0.0$

1

310.0 .

$70.20

11.11

SATM1 = PSATM M 9) +. 15 7.

Nag107

17.13 35.t8 21.40 1. 7.00 51.0 SS.00 s.0

[SAT and fire; -SAT meth scoresfor 2-time Were coached at 022

before their first SAT

SitCtita ..... .... .--....... ..... .....

11A7Ile.8

I

Math 2 SATMZ = PSATM ( 8 6) + 94.6Stereo I

H0.0,I

I

1

I

414.00t

I

I

e

I

.0II....... ............ ..... ............................ . ...... ..... .

1

1Il Il 8

1I,A0 .1

I

ta

t:

411.11

N=107

$

I

400.1111

.1

51440 1

1

100.011

aft.o.I .

IISOM as.....L........................... ..... ......04..........00.00 ..... 0.0.0 . 0

Mee tt.im 35.15 I1.00 31.01 41.11 4E00 11.11 SSA@ SIAS AS

PUT rdith S

00.00.00000000

ria7 Math Scores

a

81

tit? end stand.SAT ',silk ters*ler 2.ttes tehor coached at OZZIslets theta tIrst SAT

96

tlrotSATMathStereo

SecondSATIlath

Scores

:5

room

.-- ..,... ..1

I : ; . : -; : -1700 I 'I

.

; i 1!i

s :

.. I..1 -s. ... . $ : . 1

... _.._:__ :- : i .1 i

. 1 . I '.. I. I . . 1 . .;4 1 1

.. $ . . ..... . *.. . .

i . ..._..

..:-.................:- -.1- a...4 e :i. 1. . I

.!.....................,J.-...: j !- . I1 I

s j .. 1... : .. s t .

...ow.....1 . '.: / II , 1 . ;. 1

1 11.1;..- 1 i : t 1 1.

.....I . . 1 1 ; . 11400!

: __:I-2.i_ i-.I __-__A

. : i I I

"..-

'. : '-!' -1.

4---uncoseLi --7-1..":

g ::. ; 1

200

I - . I

--1---1-1,____ ____ ....... :--r-i: ,-, :-:--1 .

1 _: 1# I

. .

1

1 H____; i;" ..,1:....i....1--.

-aI I

- .. : 1 L : ..1.--. ------ --1 .L.... 1

1 1

I i

-17-20 30 40 $0 60 20

MethSO

SAT to oath scores for first ..SAT as functionof PSAT r ..th comp.ring 2 ticoe takers coachedat GOLA before their first SAT agslost uncoached 2 -Liebe taker.

SAT average oath (or cend-SAT as function

t t2AT oath scores. a voila* 1 -tine taker coachedet GOIA lifer. theft Mist MT against uncochod 2-tiss takers

8297

/intut"RathUses*

SecondSATStith140C*6

100

'Co

. .I

. . .. .

.1..:! o . ,

-

.... -

. . I. . .- ... J...

I i . - ! 6 t . ! II

. . i . - ; i

i

;.

i i t !e ; I ! I

!.. :4:ii.. 1. ! I. 1 1 I . : .

1 -.....1.- i .1. .. I . s 1-

-16.---%!.-i -11-!.!-- -1-.7-- ,

1

1-1 -..--;7-.-

i 1 : : .1 .. ---"1"1 ! -1 '14: .. s : ..;', s........i-4. i.

! 1 I --1-i-4 4-.1:.:._: s. .

1 .......i ..; : 1 - e

-:-.1-:. , ; ' . ' I : i I

............:...-,-........ --i lel s II

..-. -..-.-- : .-. - -.. .- - . !.-.

. ' '

-4. I

s I. 54

: I1

'

- .:.ai.......... :...-_______1............4...... .............. .. .

! : i 1 . - i

r,

__-1 ' I

1 I i I ;!r 1. --- "I'"7

..m.'; I I i . . r:..1T1-:-1

i . J.; I I : ,

I . 9

'.............L..".. ."; "I 1 1 r--, .---- ;--... . . . :

1 1

__ .-f---,--:

--I i-I --I -1±1 ,_;_,......_:... , : i

.........,i :! I

3010 70 CO

?SAT Math Scores

SAT to oath scores for first -SAT as a /uncialof IrSAT oath scores. cor.paring 2-time toter coachedat 022 before their first SAT with uncoached 2 -clue takers

.122 -.. 0.--., . .

1

200

too

20 40 30

r: : : i i . .: i .... I : ;- ..! . . ...t ir".:..JT_....._u_

. . 1 . : , . :

!

1

. ... --_._:. I _..z...--.__._P?st..

:1; ; : :III ; l'i , I 1022700 1 1 ' '- ' ' " gi 1 1 .i."'

:

. : 1 .

...,- 'l

:- I "-:-*Ci I 1 11. I.

..._:

I

I ; i

. i ! .-I-1.1 .1- ..!" ... -..-: : I; , .:--1__L___

L ' ' -. i500 - -""7-i .!--t--1:::- 1-

; --F-1-7:.--1-17: !.. :_: ._1_

,

.1._,__:. 1 ;. . .. , :.: 1..:..i.:_t...:_i..:_ , ._ ..._:_i_.; 1 !_:.....,_!

a 1, , :

f ..1- !

- "!! 1 i !-- 7:-t ! i

: isTr--!.-i I - :1 .. 1

.....4

OM

..±.1±1.... . .

--r I, ------";.-7-7.71-s . 1

sit, ,

--t' ----;1. .

1

1I ."7-`"-

'---s.

300

...ow

.1-.an 1: Ulm....

120

SO 70l'SAT Math S

SAT oath acorns for second-SAT a /unction64 11141 Path acorns, Co4rsflPg Islio0 lia..01111 Csalthad

I 027 Worm their that SAT with encaechad 2 -tine tater@

30

8398

-(5) SAT verbal scores for 2-time takers coached between

exams

Actual SAT verbal scores for 2-time takers coached be-

tween exams appear on pages 85 and 86. The first page

displays first-SAT and second-SAT scores for students coached

at SHK (OO1A). The second page displays first-SAT and

second -SAT scores for students coached at Test Prep Centers

(022).,

Pages 87 and 88 compare verbal SAT scores for coached

and uncoached students. These comparisons suggest that

SHK (OO1A) is effective although Test Prep Centers (022)

offers minimal benefits.

SHK (OO1A) shows a fairly constant increase over the

entire PSAT range with average conditional SAT second, verbal

scores increasing from 43 to 39 points. Test Prep Centers

(022) shows increases of 1 to 7 average conditional verbal

points over the entire PSAT range.

99

84

11111Cii. Verbal&coco

1.

SATV1 = PSATVii.5/ + 133.1

N=16q

2

471.111

3O

3

43$411a

35

2 0-

ISS01

a

341.10

3

I

0 a1 a

S 0

if

I

S a1

III SiO II

a . 3 0I 4 1

a 1 a a

2...,.:

Is, _..-- ............ ..,......... - ....... ..

I

MOO ZIA* 33.20 3300 at 40 tt ....00---- ..................7p;41..$7.22 "'SA? Vcrbl 5"re

VSAT and first-la

c°4c1:t1 :ctoGgebetween SAT ersto

SWAM/ ...-............---...

SAT

. ..............--*----1

...........

of115.

. .- ..........

Verbal I.

.

.I

Score.t .

S

I7

S

t

S

414.10 SATV2 = PSATV(8.d) + 128.1I

I

3

.

I

3 N =155 1 a I

I0{1.11

,

1

1

1

1

/

1

s a

008.1S

I

S

1.....-.... ....... -.....-_ ---. ....... ...........-- ...ie.,. . I ..

I

e

1

a2

S44.00

a 0I

I

I

1

I

II

a I I1

1a

I

01043II

11

I

4170.18 0 0

1

I e. 1 If

3...............................:..4 .. ... ......"-. .

.... ...... ......................-

410.00

.............I

422.8113

11

I. t

I

ISSAS1

I/I

1

I1

WAS'oba. 1

1

200010

SSOSS.4aw .....

MIS 33.28 01.4112306 ' 11;19 .. As.00 ".......... '"

42.$Z ..... . ..42.40

!SAT et"' $"1"a

85

LSAT and gerund.&5A: "bet sicrso

Cor 2"1" -41-ges114cbid 2 00IA

between :AT 'sake

Scores 1

0 1

SIMS 1 I 1

t

151.0 ISATV1 =PSATV(9.0)

oia ! N=78

4/4.10

1

3 1

4115.5 1

354.00

1

I St1.00

1

1

231.00 1.

r4 1-

230.00

+ 87.2

f

1

.

a

t . . 1

f

a a

1 .

t

T

..... ...... ................

a

..

ISM 24.50 20.00 33.33 35.20 43.34 43.01

0SAT and first -SAT verbal score.

for 2 -tie takers eoached at 022

between SAT exams

................... 31.35 31.00 40.10 45

tut Verbal Scores

itc.oree)

. .

Verbal SAT CMO

1 '

1

....... ....... ., 4... ...I.

1 SATV2 = PSATV(8:6) + 84.9

Scores 2 1

MAO s N=75 s

s

s

3

1 t

. . 1

222.12, 1 ..

e s

I 3

. I

s s

.. .:

t 1 t .

:

S3S.O 1 1

.

/

1 o

j 1 1

...«.-......----.......-1«:

1

« ..... «« ............. «....: aft

4,11.0 1

S I

; 1 a

1 1

1 1 a

t 3 418 1

I 1

1 0

1 3

1

1

443.11$ o 1

1

1

1 1

1 1 * 1

1 1

a s . s.

s

I s a s

s

400.00 e ; 1

1

1 1

1 1

1

1«...«......-7-«

1

I o«.......

...................«......«««««««««

........ «...«.......1

SSS.O$ S 1

1

I

I

s

I

sts.to

:

1 a

s

.

2 1

1

1

1

122.9.1 t

.

le

1

. 1

1

1

1 '

1

111.11 w /

1

.s........1......................... .......... ...................................

............... .

111.10 82.1 21.111 '.2.11 16.1, 4/.111 42.2111 $1.21 10. 112.11,

Score.

PEAT sad orond-SAT verbel scores

ti.0 Verbal

for 2-Ifo tiare :7chd

ma 02:

IrlUveft SAT exams 0/ 5

rivet

'CO.'UT

Scores

SccdSATVerbalScores

..

J

700

6C0

400

.....I

! i i. i _1. !---;i---.*--- 1. -4

; I.4 , ., .; ..- 1. I l......-1 . i

. 1 ,. z '-11 !!. -1..1 '. ' I.

i r . . :. . . - . . i. ...: ,

i

.-11 ..---7...: : ..il .. i . . . 1 1. : - . !

........ ..- -L.. :... .-.:.i ' 1

. I z i

7 i-1.1:1 :

. 1 ' .

_,... 4. __

7obu: '

- .

. !,--1

...-t - -.

._1

"' -1 -1 : -----a _, I. f

i . .

I _t_111.-.. -T. L : ...-.........---.......

1 1 I ri t i / : 1 I

20 30 40 50 60 70 sorSAT Verbal Scores!

SAT se verbal scores for first-SAT . function

of LSAT verbal scores. comparing 2 -tire taker coachedat 001A between SAT ex... with uncoeched 2 -tine takers

2C0

r.i

100

500

400

200'0 40 50 60 70 BO

?SAT Verbal Scores

SAT Re vestal scores for creitt-SAT Aka firietlonf ?SAT vertal n.or. comparing 1-ttria tekr coachedtit 0014 Lstweer exams with tosrosched j-times tellers

I 02

SecondVATVerbalScores

owvI

700

!

400.

500t

400

200

1v0

700

600

600

300

coact,

20 I30 40 I 5O 60 70 so

?SAT Verbal Scores

*SAT average verbal scores for first-SAT a. !unction

V ,of PSAT verbal score.,comparing 2-tide takers coached

at 022 betwcen SAT exam.with oncoached 2 -time takers

w co

I _ .

20020

5.40 60 70(SAT Verbal

IICAT.serene verbal scores for second-54T function

f PSAT verbal scoria,comparing 2..ttoa toter. coached

at 022 betweenSAT saws with oncoachod 2 -floe takes

88 1 03

16) .SAT math scores for 2-time takers coached betweenexams

Actual SAT math scores for 2-time takers coached between

exams appear on pages 93 and 92. Page 91 displays first-

SAT and second-SAT scores for 2-time takers coached at

SHK (OO1A). Page 92 shows first-SAT and second-SAT scores

for 2-time takers coached at Test Prep Centers (022).

Pages 93 and 94 compare average math SAT scores for

coached and uncoached students. As was the case with verbal

SAT, these comparisons suggest that both coaching schools

are effective, and that SHK (OO1A) is the more effective of

the two schools. SHK (OO1A) shows second average conditional

math increases between 22 to 76 points from the low to the

high range of PSAT scores whereas the control group shows

ir.aeases of 1 to 35 points. SHK therefore shows a net bene-

fit for second-SAT math scores of between 21 and 41 points.

Comparing results for those who were coached before their

first SAT and those who were coached between exams suggests

that coaching and SAT experience confer sliybtly different

benefits. Coaching without SAT experience appears to help

low-PSAT students. Exam experience with or without coaching

appears to help high-PSAT students and coaching after an

exam appears to help everyone. Possibly, this may mean that

commercial coaching tends to stress basic skills and remedial

learning, while not emphasizing the more difficult portions

89

104

of the SAT; however, high-scoring coached stutie:its with prior

SAT experience may be able to integrate their coaching lessons

with the requirements of the more difficult parts of the SAT.

1.J5

90

SUMO'U.S.

47/.111

MAIO

S60.06

Mal

4$0.011

4t1.0

340.00

21448

243.00

1

1

1

1

1

1

1

II1I

I

II

I

II

3

I

. .

SATM1 = PSATM('13.0)

N=168

0

P

a

2

+'114.8

3

OSa 0.7

aa

a ;r a

SATMath****

710.e

736.00

405.00

6)7.80

1186.00

814.1110

433.11

362.10

331.11

ROM

74.00 31.50 34.40 31.13 45.30

3

a

- .. --

5t.s0 55.80 61.10

PSAT and first-SAT math scoresfor 2-t1me takers coached At 001A

between StT exams

06.40 71.70 73.

PSAT With Scores

« . + O. . . ..

3 SATM2 = PSATM(8.19) + 118.7 0

0 0

N=155a

1

a

I. 00

= 7

21

a

1 7 el I

2

1

C

r

I

1.1

1 so

I

1

Ima .....

14.68 11.38 14.64 30.10 13.20

Q1

30.00 13.10 61.10 46.4 71.70 7

PSAT Huth Score.

PSAT and second-SAT math seemfor 2ttm. Ws,. coached at 001A

between SAT exams

1

1

MATMethSmile

4211.11,

412.14SATM1 = PSATM(9.F) +.26.7

4

4411M

$22.80

438.07

4312.86

314.83

33430

412.80

r

3

2

1

3

e 1

:. 1

270.80.7

....... e ... ... . ...

22.80 21.20 32.00

8 a

37.00 $2.00 47.00 51.00 14.40 87.00 07.00 7:

rsja and first-SAT rath scores

for 2 -rice takers coached at 022

between SAT exams

SAT Y10.88

RathScore,

847.04

1

3

SATM2 = PSATM(96)

N=75

2

+ 51.8

Si'..., .

2111.80 1 3

......0

3

23..ee 2

3

411.11

422.11

........

481.12

3

284.14

a -

1sla.s

'

Ina*1141

,

1731 31.91 arm 43.11 *1.19 Si... 11.10

92

rSAT end eecond.SAT oath SCOT,.for Mime tiler. coached at 022vows SAT mail ()

rsAr haul Scores

....41.11 01.11 72

'SAT Roth Scores

Malistbala.assesSMS14106111111:""11..:..e.a.olait.V.:.4..-3;ftm.i4elfetSAT _1 _I.11.0

Wimp mo

111:

30 40 SO 40 70

P147 !lath S

SALT oversee oath for firet-SAT so a functionof PUT oath scores. comparing 2 -time takers coachedstoves 447 eases with uncosched 2 -tin. takers

SO

20. 70 TO 10PM? Math Scenes

SAT ..state oath (640. for eeeend.ST ea a functionOf PUT oath eerr. tverstift laeln wge (44m0664.1 cm between SST *...o silk umeache4 Satkoe Wore

93 108

SUN'SATOathSCR'S

SeemedSATMathScam

OIN1

.11. 00 lab 00

aI I

400

T00

SOO

TOO

600

SOO

400

100to

20 30 40 So GO 20 SOt$42 Rath Seam

SAT average oath tar first -SAT as a Imistiomof Pm? &sit% stores. to:marina 2.tiala takers seethedat 022 betimes SAT eases with toe...chat 2 -time takers

M 1"-ri

.

...

II

I I

1

.

..

.

.....

-

.

_....

1

-.

-

-....

......-.

r..

..

-1

i

I.

1-1-..,..-

..

.

-

:

........022

1.1

.

-...+-

..

t

.

. .

.

..8

0

IP.

......

- -.

.1.

1...

. ...

. -.

0

0

. r.0 . I

,...tot..

I

a

i I.

.

I

I

.

i &

.t

.

*

.

ow ..

...

1

.

ow

.

I11

_

iOM

I I I.

IF

._

.

.I-

4......

1

.1

:'

So 40 30 GO 00PM? Moth Isom

SAT oath acme. too aetood.347 as a tom:Umcat PSO oath .taste. canoatifti 2.1goe 'Naos .sashedat 401 betootat Ste ONO. with UMCO1004 &lift Salome

.9 4 00

(7) SAT 2-time takers coached between exams

This final section of SAT results displays and compares

second-SAT scores as a function of first-SAT scores. Actual

first-SAT and second-SAT scores appear on pages 97-99.

Page 97 displays verbal and math scores for uncoached 2-

time takers. Page 98 displays scores for students coached

at SHK (001A). Page 99 displays scores for students coached

at Test Prep Centers (022).

Pages 100 and 101 compare coached and uilcoached students.

Each diagram in this comparison contains a "breakeven line"

representing no change between first and second SAT exams.

In effect, for this set of diagrams, the study baseline has

shifted forward from PSAT to first-SAT. Once again. SHK (OO1A)

is clearly effece.,e, although the case for Test Prep Centers

(022) seems somewhat marginal.

Using the first SAT score as the independent variable the

following results can be shown:

(a) SHK (OO1A) students' second average SAT verbal

scores increase from 42 to 24 points from the low

to the high range of first SAT verbal scores above

the control group;

(b) SHK (OO1A) students' second average math scores in-

crease from 20 to 44 points from the low to high

range of first SAT math scores above the control

group;

(c) Test Prep Centers' students (022) show second ver-

bal SAT scores changing from -3 to +15 points com-

pared to the control group from low to high first-SAT

verbal scores; and

95

I to

(d) Teat Prep Centers' students (022) show second

average math SAT scores changing from +7.1 to

-5.4 points compared to the control group, from

the low end to high end of the first-SAT math

score range.

./

96

I 1

Sttem4SATVerbal

Stone

SeccdSATMethStores

119.1011I

m 1

.o.

1 SATV2 =I SATV1(.9) + 52.91

.

s

4114.22N=494 a

t..

1I I

II

222.21I

I

I

I

522112

I

I

1

441.te3

1

I

I

411.11.I

I

I.I

$51.11I

3

1

I221.20

I 13 81

1

8111.11 11$

I

122.111

I

2

11 I

I

86

.

8 /1

OS I 1 1

8 I1 1

2 2I

/8 1 1

1

I I 1 2S

I si- 2 :15111 at 8

iI / 8 1

6 4 S V 2 I1 : 6

1 P. ) 8.1

1 *1 2 3 2I : :

1 s 4

2 / I.1.1....1........S.1 .

, 1

I

8

1

i

11111 171.110 184-ee 271.03 4211.00 10.01 MAO 181.55 16,09 age.t7 71'

Met SAT Verbal Sc

..

2111.2

SATM2

212.

1

4.21.1

212.114 1IMO.

1 IS I 1

MAO I 3 1 41 51

1f I I 8

I 411, 4

I I

1

g s 03 I

441.113 t I a

M.

/ I 4. 3

3 1

11116111 so

i 8g.

8 1 t

tolog

SAT verbal for

umoched 2-time takers

** ...

= SATM1(.9A) + 49.6

N=494

t

1 1

2222 3

6 g t

1 IR4 1 i

I / 3 1 ..

3 I 2 3

i 2 g I

8 8 t 8 ird.

g 8 Rs8

23.211

I

1113.04

w 3

1

3 14

It1.111 e...

104a* 1141.21 111.64 Mae 4.4.11 MAO MOH 41240 664.11 114.11 71:

Mgt SAT Hath S

97

SAT math stereo for

unteethed taker.

1 1 2

Second1411.

1

SATTorbal 1

PAD. 11911.

s

:

SATV2 SATV1(,813)+4$1.1

1

1 N=168

11.1

100.7

1

s a

1

...........429.1 070.0

.

I

a

a

it

.11

sas.ss

II

.4

0

111.00 010.03 11$.01

lia.s.0

$44.1

1

SI.SO1

1

1a

1

471.41$a

1

11

1

4111.0

1

340.60a

1

1

1

240. .

Ag

341.00 ..... .....

114.0 210.0 1110.10 140.0

locoedSArMath;cote.

first SAT Verbal S

SAT verbal scores far 2-tl. takerscoached at 001A between SAT woe

.6.,......., ...... .......................41044 e

I

II

SATM2 = SATM1(i98) + 61.91

1st...s N=168 . .

1

s

1

1

I

1

1

1112.01

. .

I_ .......................-........

1

a a

I.

42 2

. .

155.00'I1

I

I

521.001

a

1

1

I

4.0a

1a

431.05

wale

I a8SS8

efoo0

4

2

a 2 a a

I

1110.10112.10 0.50 MAO

. 98

II

oo

06.1 013.03 70.011 1246.11 it

?lest SAT Math S

CAT oath score. tar Ntlne takersCoached at 001A btween SAT smoll

1 1 3

os OM Lit Ig4

" ti'Oo, ".991 46.bIt 10"1!- 9::!%b

wee, west erase oriel well ......... ....... ........ ... ..................... .... wes

..... , 1141

1 's zto es lottilloat

44ass. twat An

66

.... ......

a

..... "'".

- arm

I 10ft

'Etat .

11 1

1

1 111

I

II a am

8

! wolf

1

8L=N 8

4'99 + (b81TWIVS=ZWIVS 8

Will

8

1 t.

............ .........

r .

g 182.A 1, 114 '11,1 0181 2,1,

ta., ...... ,s.,6t was: writ ee-11/ wets

--"" .... - . ............... - .............. ...........

0wra "

I13 -" 1114..1,1 zzo 3 paSpOS

St" C 461 442 (A In

. . .

I

.......... ... ... .. z

.. . ... ..... . .......... ...

1

1

.

ZzS +.

8

1

'rift I ITaa.

'11

8

! 848

, 1111'111

gem,

Witt

8L=N s

(E6')TAIVS=ZAI 411 VSI."

1181

orate

_

. 0...00.

1,f 41,11

LYS puass

soma, ma.% LYS

SecondSATVerbalStereo

Secenct

SATMath

300 A00200 500 600 700 SOOTarot SAT Verbal Stereo

SAT e,,erage verbal score. for second-SAT as a function if

first -SAT verbal . comparing 2-time takers coachedat 001A between SAT exams with uncosched 2-ttre taker.

200

700

600

$00

400

I ! I .:-;-; :-:-- -!' i' ;_. 1...

i

1

---: 12 i ,_13 ...I.-1-1--- I. - :f/1--"-';;. .. ' I 1 ' A .

I ; ; : I I i I ; ;... 1i....

1..li , ;____.0016.-3- Z ._ ..akeachad

1

i I - ii ! I

i i i / .. I I .; I : 1 ./I .1 i

..i. ...:

/ "--: -,;.-r/ ... , . ... , . J. ...._./- .1.ti

../ : .. .;....1._ .......1..,............,1 /4z; : Li.: ,,!.. .,..k.v,.12......./ . ,-, ... ... = I alp' 1 ..... 9 0i

1.- --, T- ''"; i : t- . i 1

e. i I._ ......T......_ ...... . OM. all . a al..

...'. ; *

1._e 1 , , 1

i ; ;.!- I 1" :..T. !1 1'

--e -- L - :. :-.1- -;-.....--"-

--1

....1 ......... I . . :. .-f-ii : .

I--;

-.. --..1 I :.4_,..

. .

. ..

11.

:00SOO 500 400 Soo 600

/fretSOO SOO

et SAT Ilath Sews.

1001 1 3

Second&ATVerbalStereee

too

700

SecondSATbathScores

GOO

I.

-

1

:

.

022---.-.:_:.. .`:-..nreechett...,1

.1 I: 1

. I - I

.. , ....-. - -....1. , .

--: I, .. ...L.-. ..._ .....!...1 ..!--..--

, . .I 1

.. 1 : , ............-. -. .....1....- I - ...... - .- . - -... - - v--. -1 - - .- I .-: -

! 1 1 I I

.4. I I..- . - . .4 1 1 _1 1 1

...A... '1.'..

..:7Tbreakven line-J-4- 1-.-,.-1-4 - Iv. .

/ ' i t 1-1; !I l'.. -..- - i-....-1

: iH -i ..d

-1- / 1 i

-!... .,< . _ .. : . -1- .'... 1.--.. ---: . . .. :--1-71.7-11

.. _ _i_.i....,--L-A-.:-_1.- ..41 _ . ...--Hf-;

.--:: .. - ..;I. .4. ;..... -; ' 1

11300 -.-

200200 300 400 500 600 700 800

rirst SAT Verbal ScoresSAT average verbal ocores for second-SAT as functionof fIrst-SAT verbal scores. comparing 7-tire takers coachedat 422 between SAT eican with uncoached taker

800 - r f I I.4.1. I . i .. A :. . _ ; ..! .,...1........ I _,.: _1..r . r s " -- -:

_ 1_......1__ . i 1- -;_..: - .- -: - 1- !

I _I i _._ 1: . . _..1...:.;.; .._

1. , 1 1 . : . ,

7001 1 1. Mt cvaclied.-----..

-.-1-.7-1 t_ ...o

. ; . . / :

400 ; i / .. : o V -!-----!-- -.. / I

. 1 :,... ........ _ _ 1 _ _ _1........a.-L-1

-- ...1..._ 1 . i ir 1 e : . ; -.......-.. .. - r.....,....... .

. : . i 1 1

SOO ___, ..t.-:- -- -14.---"-.-..1-1-. .. .....I

.1._... _... ..] ,/. 1..-. .... .... .---.......7.... L I- _I- : ...,-.

- /.-i- I -A--r-1-.. . _...._,...._. 1 . ". ___. ----t--.

400 . .,--.1.....,..._.

IT-......._......._,

-.-i--;--.-.-:-; ;-

..... : Ir r

-:--

I-1I

i: 1. 1i1 -. 1:1: 1

I ..1

.

__

: 1-

i

1

...

. . I..;1

200700 400 300 400 700 700

first SAT Math Scores

6AT is svath scor for ec...ad-SAT lunette* atliget-SAT 1 t 1. court. c.sweeln 7.414. takers coachedat Olt between :ell Ith wocosebd 2-tla ta

17101

1

B. THE LAW SCHOOL ADMISSION TEST

(1) LSAT scores for white and non-white 1-time takers

As a prelude to the analysis of LSAT coaching, this pre-

liminary section indicates why the effects of coaching are

estimated using LSAT scores for whites only.

The first part of this section displays LSAT scores for

uncoached white, coached white, uncoached non-white, and

coached non-white 1-time takers of the LSAT. Comparison of

average LSAT scores for these groups reveals two major dis-

tinctions:

(a) Whites, whether coached or uncoached, perform con-siderably better than non-whites, whether coached or

uncoached, having a similar GPA.

(b) Among whites, LSAT coaching appears to benefit pri-

marily those with a lower GPA; among non-whites,

LSAT coaching appears to benefit those with a higher

GPA. This LSAT coaching impact for whites appears

repeatedly, suggesting that LSAT coaching operates

by refining or renewing some basic skills. But if

this is so, it is not clear why higher-GPA non-whites

benefit more than lower GPA non-whites. In any event,

there appears to be a distinct difference in coacha-

bility between whites and non-whites.

Amplifying on the foregoing, uncoached whites with a

GPA of 1.0 (on a 4.0 scale) have conditional average LSAT

scores of 436. Uncoached whites with a GPA of 4.0 have

conditional average LSAT scores of 640. Uncoached non-whites

with a GPA of 1.0 have conditional average LSAT scores .of 313

while uncoached non-whites with a GPA of 4.0 have conditional

average LSAT scores of 531.

102 117

Coached whites with GPA's of 1.0 and 4.0 have condition-

al.average LSAT scores of 458 and 634 respectively. Coached

non-whites with GPA's of 1.0 and 4.0 respectively have condi-

tional average LSAT scores of 301 and 556 respectively.

Uncoached whites score on the average between 123 and

108 points higher than their uncoached non-white counterparts

with identical GPA's. Coached whites score between 157 and

78 points higher than their coached non-white counterparts

with identical GPA's. Coaching prior to taking the first

LSAT appears to widen the white/non-white score gap at the

low end of the GPA spectrum but narrows the score gap at

the upper bound of the GPA range.

103

y43 lleteee3

/1 1 11'1 IC1

1

.

..... e 1

1 I

5.. OS I .8.0 1 I t 8 ups f

I ..... Si. es 611

1 1 1 gi

ob see / I

I C 110 ...IC I

I 2.2 /I 11:1.1211 I 111 11 t/// t

... 111 %/1.011t. "It.

1 6. /* Met Met ./ 1 .1. .. 11:ttC :

1 1. ' 1 .C.C:.1C.1.tt.0 1-.....-1.1 . sotoCtt0g1t..ittli101......t/O.00

1. ea .tt C 1.lttC11 v.

-- / 1 t .12..1111.1. itt11412/ttItt SIC. .... t I.L./MC.111

1 OS.. e.' u/ Mi./ .11// 1 1. I too MI ..i (ft St.

laj V of4m/ / 1/ f 1 /1/ g II lio 1 is PI

1 /

gS 1

1 OS of 1 is met 0

L. 1 I

4 I

I I

I

.

I

I

...-----

ya0 4211144%*

0511 tat Wt .........---..---............----...-- t

I

I I I S *

1 1161,11. 1

11 5 1 .. .. t 1

I

I .. C.::tt :///4 I .C .. ...1 I / t me.t.teltilloiw:(

is //..r.:i:1.1.1.1g1:111i111:15:1111 1 et 1C1tittt:ttt:Ctlit

1.2.:C/otCgtotIttgt.t1;it.:Ol1111 I ...t.tttttitC(C(ttl:COASPItOtt: 1 1. C.1t 4,1C.IS.St;

Si . :1411254 I . 1Ilti1y111.::i:1114:%4W66:T6I0.6116. I 1 t/W.c1tta2; I

1 /::,:t/Ii$.t/ .... 1 ::11;g1 I , sct,..1,61.: 1 10,1,441: 1 11101141. 1

51 .. 1 It12 I 6

1....... 1, 1

1 9. y. I ,

1

1 t. :

. . ..

1. 1 14110.51... 4 666666666 L11 1 1

1 CM 6666 V 666666666666 06%,%4046*

I 11:t11.61:Itt0.,C, St:ttttS I 1t1/,./ft :.t:%t:st:/.C

I. // .:.....:.,:. 1.14 I . /0

0. 66666 1 t m .

1,.)

sell 66 *

. se

w .1 I 1

gf fa

C. 1.11

Ctt

t.t.

(ii

CCCCC/.

I. Ito

10'1

1.111 1.:.

. . ..tc, tc

i..sc Ntl tt :tt..(...tisO11/..:./i::.1oZO

..... ZST4S:4114SO44

ttr,...:setl.:.:11.1-1sstit.C.:c,..

II.

66666 .i$4:$S1401

1.. 1 /

t.t

r V13 8 u011,155; . sitim pyaps

tt St I.

SOS

IS 18

Sm... I

1 Id . /5 0..6

S OS 1 ID. 5 1

a 1 ; 1 /. .1 f tell 1 11

f el. it SIC C ls .15

I I

/s tr a .10 d /0 t 1111

.../C/ of. 1. I

Ir. t t .1. //g /1 . t. 1

., C. . ti. 1 Is

1.1 / 1 1 ,

t 5 / Ire O MO f Si / 0

'

.. .......

siting/ 2(012-1 paglopun

Itt Ct 9101

..... -..--- .---0---0 .1

mg Se

S. is 4 f SO 45511

: 5 stay ass / goo 1 111 1 , 4.4 I ca ill off a,. 1.

:C .::C.:. i:/ :(t: Lt 11... C. f

tt:Z.12?:1 vt:/ tt..c::. flott..:1..C1t

?1t1 1 ltititt2 :1t

CITO:2:tf:t.I.:t :t ill

101tgl.:460;v:ito,Stt.t. 111,411S11/0.

. 1c1C

: ::v4t1:1t1.1.1 44.1,111S :10.

:36761:664.41..1,Ittr :117..0- ..::*

...VI;

1.616ei2:1,1 14,111

11,11111,i1/ :6;:tZttt:t

14:114S4Ctilt11:11 666666 , s, $ . , o I t : i

../. Z...;1:,:.1

41t04:S1:11 .1....5;;;.2%..0.22:,:11

: 1:st /..././.

:11,601 1t:6/,/ :i.111:.SCItt:, 1:trt 0....,,4t,..t.. I 4 11,014:1.11 111* .1. it

1 weld vt .1.11 Is ff i of

tt/ i :.ii 66666 "( "

s.

OS OS

tr,/ JO;

11'1

Of

.

OS 1

...._

/ 11

/ MO SO

owils8 leti

11.11 11.1

11111

. ... 410. 0 WWII

1

1

.-.-...... .......... 1

1 1W111 I

. , U.S ...... .. .. I

a' 1 11.1110

OS . sIt / ...............L_. .2. . ...

I .1. le

..... ___,1 .......... 1

.. ........... CI'lli

.

I

... ...-.--.. g

- .

1 g 1

2 Ira it - . ... ----.. t.. ----

. /

OS ... .....1

. . I

111111 1

e 1

1

1

S 6 £ T=N t """ I

............._,

_ 4 6 6 C. + :. st:sva

'78 ---i-Vci0=4.11Srl : . / 11421

.. I ITS1 011111,

-6 . . ........... .

vdo jo uolilurg v114A 20j 41530) 1111

1/1 Ill WI Ot --"0-7--.!-_:-:!:---

. (L

- 6_666

0

.. 2

16

1 . 1 1.

I

.1 11

..1 (.t

1111 1. 111

.t It

. i

f-f t

.. I

c. 1

WOW' .

- ...---- I

. - . . . ... I

. ........... ................. I

00171 1

1

I

_ - -- - - - ----- Ordli- I

I

.. ---_..- : ',M.,.

Welt 1

-, - ... - ........

I I

101 I

1 II 211;- I

- . . ... tell I

1 l'

I

. WOW. 1

I 1

11 1

1 GA

is 1

.- _ MIND An.* M .

..... 00 1 ei/.

1 i

. - _ - ' - - - 10E +

1 1 -TiFE§3 '

op elt 1

1.

1

M 6 0

146014LSOkol

141.42

411.411

114.2

$21.00

41.04

426.44

1/40.

241.42

313.10

114.00

1110.40

LSAT=GPA(72.8)+ 240

N=2438

1

1

: s.t,me 3 . ...

I 4 3 1.21,e1.4 2.. a. 24

. f 2 0

1-1..12.22...---172

BOv.. f

s. aef 4e. f 31 1

f 11.. em 2

1 r 3 al 2 l2. .41.1 fed of. f

4,:.1. 12.11:1 2 212

2 "22422. .2 11,1 ego12:2.2.12.1 1

342 ...... .0221. 3.33.4: 2 .1

1...1.111722.41412. 211

3.031 df11233.1 232...

...Tssr. 1 "If...1 $-12.362.1 .2

1::::4:21173..31:e. .31211 ..

. 7 11.1 4.1.3.421:1.1

:30 .1 .2 121.

.2 2 .2222.1. 3.

32.13124 7.3:2 m 1

211 1 t:. .

3 2. . .

20 03.1 1

4 222 .11.... 2 I...

2: 2

2. a.

1 2.

.

-.......-..........

2010 1.11 3.03

.

2

1

.3

I

I

I

41,1

,1

1

1

1

1

1

1

1

1

dot

11

1

11

1

1

1

1

1

1

1

1

I

1

4.00

1

1

1

:

1

1

1

1

1

1

1

1

1.---

1

1

1

1

1

1

1

1

.......11.2,

._..............._.._...-._.......--.--......

0.10 1.01 1.43

2 .1 141.... 11141.1.4041..

go 2 04:36 4 2.1.2)1.1 I 2 ./4....1212441

022 21 1.1 2222222.2.1.212.2322222se . . :2.14122 14.04

22 114 . : .. 221.44.

1.4. 17 2.2 ..to

2.2 242 6 7423.bog .2 e. )31::.....31:0:30.:

a .122 :7 a7.:124.8 2 24

..) 2 2411444 .

S .1 1`. ..:a11141" .23:331:003.7)

3. .. .3.21.1:31 -44):333.1-1-1) $ 722.5.5. .32413111S:2222:s 22 1342.11.12.123

1.2 S2221.62 3.3613. 120 72 .:: .7A7.312.

1. 2.7 25.33325 274.3 .5..03.3511 22224.42 1 2.322.341.

...IS.) ..... .2 26242.2..3 3 ea..

1.

... .0. 22212. 771 7

.3. .3 .3 2.12 .... 3..Moab 1 3 2

11 4.2.332 .2 .3.2..2. .6. esse 2. 0.02.

1 Ob. 2

v. 2 1 ..

.1 as. ..1.1

..... -...-. .............

1.60 1.10 1.11

CUMUtiVe CPA

LSAT scores for uncoached nonwhite1 -tine takers a. a function of CPA

471.10LSATScores

451:10LSAT;---CP-A-(84..._-{:..2.161..

..

--9-)--

I

s

a

IP

2

a

a

2

a

.....

....a.

"

O.

I.

I

1

1

1

1

1

1

1

11

1

1

1

1

1

1

A

1

1

1

1

1

-5

5"g°- * 42-NIEL

-44/.40

11

411.54 --

I

I

?11:411I

.12717-1.0411.211'

21

I .

1

...-........

Is- 1

114/114'

.........

1

1

1/11.111

2.11. ...1014 1.14 1.11

............... ..........

1.11 1.63 1011 3.21 1.44 1.44 1.42

/SAT :,, for eeehd1.11w 4.2234 4 imictiu of &TA

10512o

CuraO1,1111

=111 .

I

ZC3 I 111 0 : t I:

i I ilow. w ' ,II

I 1 1 / wow' =01 I

1 0 t

700

1 I I

; I

1 ;

.11

I =111,

-1- I 1

OE 1.=1111. Ni

600

L. I «1

44! I ;

t

50 ,.........

110

1

I

=1111, own/WM..0 *Ewe

white uncoached

white coached

. - ... . i... .....; -.. . ea on!. . 00 .

a 1 / / nonwhite coached. .. 1

,{ i ' ; I

1

.4. : 0...1.. 1 .44 4. : ... e..°..4 .. 1 nenuill te untioa::"C

' I ! 1

' , i 1 ; 1

11 !LI...1.HW+ OD

I I I11....

11'..L. ' ... v. ; I

'0 I

h . I 0. /11, 'O.*. ..

I ."WM =1111=1NN

*. 11: .1...4'. 4.. !

10Jill

.;

,

- masalw 1. wit.. le o Moo.:

`

121

400 1.1-1--. .L. .

I

.I. I '......L...i. I .I I ! ?...VD 411=10 ..I...9

,

I I1 1 i ! 1 1 I i 1 1

I I, 1 .-. . 1 '

.....:..... ..1........,.. . 01 so 0 . 6 . 111. 0 ON . 0 0.r aI I i

o 1

1 : I. I 0

I .j I 1 I.

' ,

. .. ........... L.........' 1 I '....r a dm. I w.011. : A w a .. 0 o' ow I 0 .

I I II; i I / '/ ; . : , .

I . . .. , m I . IM....11.10, ....... # ..... ; ; li lo f ; .

0 '0

0 I 0 I I

0 ' 0

0 0

1

,.., ,.. 1 . Lgwl.. .. am i 1 oma MD. .1M1 1 H. OW i /WM II M. WO, . - .09 III MIR dr Mb

I I

/4 I .0111/ rd... #.= 1. M .1 111. 0 4 0 1 I14 . .4too

I I I I I

1

I I II . I I ,

I 11. si. i . t so I I . . ' i .0 0 0 ', .0., 0

I I

I

il r 1 I 1 i c .

.

.. _ i .. ... . ..... ..... ............ .., . . .

1 *a 411' V. 0 .110. i 11 III r ' i MI 4 . 0 . ... M

200

II. 11

.j I ,

1/

1.0 1.5 2.0 2.5 3.0 1.5 4.0

Cnactiv, .,

I 42

(2) LSAT scores for uncoached white and non-white 2-timetakers

This part of the white/non-white LSAT analysis displays

LSAT scores of uncoached white and non-white 2-time takers.

Comparison of average LSAT scores for these groups leads

to the following conclusions:

(a) On the average, both whites and non-whites displayimprovement on their second LSAT attempt.

(b) This upward shift, which is larger for both groupsas GPA is higher, may represent a "learning effect";that is, actual LSAT experience provides informationand exam familiarity that are translated into higherscores on the second attempt.

(c) Whatever they actually measure, GPA and LSAT appearto measure somewhat different things; this differ-ence is implied by the relatively low correlationbetween GPA and LSAT scores. However, increasingGPA presumably bears some relation to demonstratedability to learn; the same ability to learn whichleads to higher GPA may also lead to the higherscore gains associated with a higher GPA. This lineof reasoning suggests that the LSAT is "learnable"even by uncoached students. It also suggests thatstudents with greater ability to learn in generalwill find the LSAT more "learnable". Finally, be-cause intergroup score differences grow with in-creasing GPA, it may be that whites find the LSATmore "learnable" than do non-whites.

Unccached white 2-time takers of the LSAT show an

average LSAT score increase of 9 points at the lower bound

of the GPA range and an increase of 50 points at the upper

bound of the GPA range.

Uncoadhed non-white 2-time takers show average increases

from their first to second LSAT of 13 to 36 points ranging

from the lower bound to upper bound of the CPA.

107

123

... .,118.42 .

LSAT i3

Stoves II V

8 LSAT1=GPA (48.2) - a8 1 .. 2

2 g # s s .2

212.2lb 01 3 .2. ...I.. .. .

.+ 363.8 , 1 $ a 2 66622 6 ole 2

1.2 .2 U. .1 22622. l l Ise 1. l

II .2.1. 02 Moj e 1:6..6 22 22 .4.2

Ni--57zg---..'.... . 2 .. 2 l PI 1 :55320 l:4 I 40.22.2 $$ .. 24...2..2

1 2 48 - .

2 :, .ts I. 14::t 34/2$45. 2.. 262. 2I

so 2. solo 2 I/ l2.121.48 4242. *: 244.22222 1i 2 2 / :: :....2 22:11:21,22. .26462216.6 ; :6:2 22 .2

l 1 12 Ill 8232.23 l.* 2SI81 I34/65:22:%%.326.42 2:1

1 I 146:61.6- :36:r :44. iiiii 1.161161:421 ....:..1

821:28 - -. . . - 2 1 22 4:1$4::113.12$s3 22.0221.3.3:2412 iiiiiii 1;t% it 221.

1 I 2l .1 2.242.4. :62:61:.1: :1662.6226162:16.62:.:2

2..............--.--......---..-..--.2--11.32-4:26.:2621.2622.:22666.22611:2.1611,.. ....1

2

2

211.18II

I

1

46162

e .2.

-45 2I

1.2 2

)

1

is.. .4:,14./.4.2to.221.102:111.t:111.,fla... ..,.:62. :6 i

..1 3:: 2.2:22:422.3.2232oilot 20/Soe. .114221$5.42$4.. .3

13: Il3.:2:41:111:0 :21 2 2 122 2:2lo 2

12.11:1:242/.:3',.'...3;.I.4 :4' -43.c :7.:434:38 22 8. 1

M... 62 "...WI. . s.2 :628:2133's i 1612:22 22 2

2 2 ...2 As li'1s$'26 26,62,616 :866e6.22.22.. 2 I

.... SlIll 3 6:612S: 2 6446:1144 47..224 42 24. 2 /

.4.26:. 2 So 4..421:: .1.6.2411144311:2/164 2S2o:24.22 3 .. 2438..1. :4:214S)4::33636242.22 y 3

. 2.11 2:66:1.63:31.0.2 2.2.74 :Sls.S3S., v 42222 1

4.1... Isl Sloe .. 42.:2222:1:.:122:SletS:12:.SS.::24:3.22., 2.222 s 2

ISOM

a .1 6: 6:si 46:2.::$.0.1:2221::.2.22 Si .22.11 . .2

1

1

.0

6.

13...-.:-. -....

3

I

1.

1

11

32

11

Sal lag 1.12

1....31 .:311 IS $2.43 :4:::48 .16.:66:.2:.2 2. 2..

el 2 2 'SW :..r. 23:3:6.:462:2.16: 2:2::31 4 1

2 .1.2212 lel: ooll 2.4 2 22 1

22222.2..2.2..242222 ......... ...*..........2.......6............2

1 lab 2 2. 2.1.2.2.. .12 2 se 1

es 2.. IS "I I : .20 $

II i 2 . le es see I

222 "2.2... U. 81 3

12 . e 22

t es ease 2 as242. se

so I

. ..2

aa.

1

1

1

1

......--- .. .....2.10 2.27 2.24 1.32 2.12 3.44

LSAT first -score. for uncoached white 2 -tin. takers

Oa function At crA

3.71 4.3:

Cumulative CFA

0......m.00.......ma0................400.0..................................41...+0............w.....................0.1 .

LSATSUM I

3 1

1.

SCOVC3 .8 1

LSAT2=GPA(61.4) . a 3:1 8. .. 1

MAO + 359.6 Ise I 2 . 2414

3

a' l Ie 2 es. 2 I

/6 1 1,. 2 ..1.34.1 2212 22. 2 .1

..- :..1.1=5729. _ 2 2 2. S. 2...32 .11.2412 32I2 ...... 24 I el. 1s 'WWI 11..11.22 lael.ff. so .4. 1

2AS11.21I 4.2. .2.22 62 .222266:.:2.261 .. S R.

12. 2.2 ll.. ...... 422.22313S4421sI2III.. SI. ..... .2 el

I . . - ..lo 12 S. M.2 3$ 2464Si/el/Malt/ ........322312 2.2. 2

22 2 it loos

221.4.2213.4 ...... 22622m:6;61 2410. 2 21 1

.

.162115.4.76,224112222.. ol

-210.22.1 es 2 1333.1.411 I3S212.13.421112124

13.-.--.-

2 I 221 S./ 212 4 .....1....1 .l -.

33812

12441;:: ::::::ti i

.11411 *1 1.

31. 1 8 2 1:1161

F.*.

22 420 220.200:226622:.4245340:0411121 ........1;;6;440:4:42042:3 1

$110.611g 81.1. 14241234I3 222 og.

o 18284 :SoSPI 42114$8:41I41.. .o.1411 SSSSS ZI)S.:414) 82.2. 1

1.122.1. 1. I

3 1 .1212 22.4232::S3O22 .

2.2 1...2."

I22.3 /Sal SSSSSS ISS SSSSSSS :12oSS2411 134101 1

1 q 06 122.2.1 S. 2$11.4'"

I ......2,34,: SSSSS )261122262 SSSS :1:::1::11::.;:,::;:,

.3

$91.15a 2. .. 22/142.:$2242Itsl SS

. . .....1....................so 3 ,221.1 222 1,22:14:2422414,l2JO2424Sg244.42,412121122.22

I

.4 211$244SSS2.4. 3...). t 8 1I

I2 1 . 2.2214 1,44 441043.44M:02422 44444 212:4 24) 3s. .32I" SS" 4401ISI4.,/I44.1 I

II. . .4.. e2 22..410 0$222112lll2 el 2 1

446.8 18 I .1 2 13. 331 313 2..24. 3242222 224..2 S..) . 2

.'1 I1 .222 470. 311:48 2.60. :41444 2282.2 2: 2. .222 1

... ...

--

2 2 1. 1 33 .14.82.1.81 .1)131 64.3.42. 824. I I.. ... 1,;,-------.;;-...---------g----2--1.-2-s"J'y 4 2..21324.22..22.2.22,2 1

$a 8. 8 42 a.. s

MIMS 8 to 2. I .32133. 1...4 .2 2.2 Pogo l

s. 1 2 2. 1 24...2 2 . 11.1

.. 020 2 0.0 2 .4, ia I 00I

I

. s

sa I # se me Ii I I

---ri 4141 . a a s

Os le es $ Os2 OS2t....

8 1 .. 2.54 1 e

ss

2 I

sI

s

ss .

s

sum' " 3

s

I

I $s

s 1

1

.s

2$

Iss.sss

I

. 6

4.21 1.11 1.13 202 1.22 1.44 2.11 1001 1.44 1.22 4.1

Cumulative CFA

LSAT second. far 3nge8c1121 01212.1fime WW2at 141%2214in 1 crA

108

a

LSATti111.1

124.11

1112.11

403.2

468.1

3.2.4

6.4

344.31

114.1

20,.e.

i LSAT1=GPA (45.7)i + 283.8'

-; -N=781-

SO

3

aaa

SS0./

aaaaaa2 2

3. a ..12) 2 SS

.3 3 .22 aI am .22de 2 s.f 211 a de

01 I 1 ...Zoe

.. .3 a

las S. 3 4.0. 22 2go 22 .S2

II.. I. fa

seel

.1 pV

I we sow

2 en

ell

2

1.27 lap

2

as2.

2...222

2

.led 61

J. 8.132 2...

..

.1el. 12

a. a.

..............

)se

I

7.94

3.3

3

2.I.2

1 a

3 00

2.. 21

2

2

2

3

3.17

1

1

3

I

I1

1

.

s

I

1a

2

1

I

.8a

1

3.2 1

1

Ot

3

2 1

.....01.....................21

3

1

13

1

I

3

...

1.10 1.7) 3.14

23

31

1

3

lI3

S.3

3

I--

I33

1

IS1

I

1

1

3

3

7.27

6 11i

3a.

el 2 .2

8 1 e

a . 2 63).2 .2.

2 Gs

SO a le'5. 2 ..

etft3 .1 212

a 4

a le

2 1

1

a

se

1.14 2.11 1.24Cumulative CPA

LSAT first-scores for uncoached nonwhite

2.41.14 t aaaaa function of CPA

LSATScot

:,............................................................6114.11

1

3 *LSAT2=GPA (4 5 . 7 )a

+ 302.SSS3.14

1

1

1-.. . Nil Z-- ' .

122.1

8

3 43

V

614.4 11

2 I3....... 2.........e

1

3o

$33.$33.4., .

31

885

1I 17

4381.118 0 143 3

1- - SG

3a

I.

111.2 .122

I61.4 1 a a.....

11

::

iiI

: es

1

II 3 SS

8 aISe 0

I a3 Se es fee........1...... ...

3 Si

3 35 11I. 2 2 eve as 23.

WI SS 2 a 1. 2

2 3 Jell 1 2.21 a. 22i 2..2.2 2 .. a 3

a a I.1 11

22002 112 al..

3 mu lila 1..1. al a.

as...a .aSO

e ...... ......a2 Sed00.4.2 .2.2 ... we

2

.22 I

1lie I

a Se OS u 2

1de

. 1. a

1

a 1f

1-7.

2.71 3.1 2.84

....................

3 2 2

100

22g

1

1 a

1IIt P

as1

S4

se

SO

.. e

3.3/

........1

a31

a1

1

1

2

.a 1

.0.1.0.000.02

al 1

1

1

1

.1I

1

1

1

a . 1

1

Se I

II

21

1.S 1.7) 3.52

2

I. . el.

Mae al1

ego ...!

11

I .S. SOS

I

Mae s

1all.

1

1

1

MIS,11I

I

loom .... 0 aaaaa ....0

1.41 1.1 2.91 /O

LSAT gerund. few unattached nhtt2 -lim takers e function 01 CPA

1091 25

Cuviliv CIA

LSAT

Scores

.12U

7777!

honwhite LSAT 2

nonwhite LSAT 1

200

1.0 .1.5 , 2.0

1

;2.5 .3.0 ; ;3.54.0

Cumulative CPA

LSAT average scores for uncoached 2-time takers as a function of CPA,

comparing whites first-LSAT scores, whites second-LSAT scores,

nonwhites first-LSAT scores, and nonwhites second-LSAT scores.127

(3) LSAT scores for white and non-white 2-time takers coached

between exams

This part of the white/non-white analysis displays LSAT

scores for white and non7white 2-time takers coached between

exams. Comparison of average scores of these groups, and

also comparison against uncoached white and non-white 2-time

takers, reinforces the conclusion reached earlier regarding

the 1-time takers that LSAT coaching tends to have its

greatest benefits for low-GPA whites and high-GPA non-whites.

Despite coaching, marked LSAT score differences persist

between whites and non-whites.

For example, whites with a GPA of 1.0 coached between

administrations of the LSAT show score increases of 40 points.

Their non-white counterparts show score increases of 2 points.

Whites with a GPA of 4.0 coached between exams show conditional

average score increases of 51 points. Their non-white counter-

parts show conditional average increases of 65 points. The over-

all effects of coaching for 2-time takers of the LSAT based on a

comparison of this section and the preceding section may give

th, impression that at least in this category (2-time takers)

coaching has at best marginal benefits. However as we will

demonitrate when examining individual coaching schools in

an unaggregated form, this conclusion is not warranted for

for all schools.

se 04 OOOOO 000000.... ........,

That 1113.80.

I

$

It

LSAT I

t

icre4 II

1

(11.3$: LSAT1=GPA ( 41 .1 + 370.

1 ) SOIs1 OS

1 . ss

:1

WagI

; N=441 44 2

SeI 2

SIM:

001

1.

I2

el 411 j 4 8 .2I. 1 / .1 4

I

1

II so as f 2 . I

541.4$0 14 0

10 2 f 4

1.................o.. ....... .....0.00 I - . 2..... I

11

2 2 .1 el

Ij 7 S

4444.26:

es 2.. . . 1. 2

1I. sob 01 1 .. 1

1lb. 2 sl 2 1

I.1 Of I I

II es 2 a e. Of :0 2. f si 1 1

441.822 1 I

...4 2/ el 0 f: SO 1

1 4 I 2 I. I

II a 3 1 I

s1 2. s t

1121.411I . 2 I.

aI

sof I . t

II I 2I....-- ...... ........... t

I

I

221.18 2

1is

I

1

I1

1

2

1

355.452

11

1

224.202

1

I1

211.t0...... -.....----1.5 2.2 2.24 2.44 2.41 2.1 3.12 .2.24 8.54 3.20 4.10

Ctesultive GrA

LSAT first-score as function of CPA for whits

2 -tine taker. coached between LSAT cum.

Second 740.00LSATScore.

LSAT2=GPA( 44.9) '2. j

417.40 + 406N=441

1

1

.4 .016 1

1124.10I

I1 a 1

1P. 2

I j

1 2112.20

.2 2 /a .2 22 I

1 2 1 .2. 2 2

2 1 3 / 2

241.40Pe 2 2 0 1

1 a 1 : e 2

11$ '. 2 Is I

a I. j I

I es 1 1 1

/27.00e. 2 f $ see. ad $ Se

i 1 4 I

00;2

SO IS le.. I 1

I 4 1 I .. 1 1

22 me I

4$440 1 ma 22 I I

11

I

I. a I. I

a: ....-. ... 1

tI 1

4414 2.

A 22

2I

2111.24a

2

I2

22

451.111 1

1114.11,50..... . ........ .....4.........

1 2.8 2.24 2.44 8.4111 1.11 2.11 1.14 2.24 4.29

LSAT second.cors as function of CPA forlAttt 2-tins takers cochd between LSAT

112 129

eumuttirs CPA

'featLSATScores

SeccodLSAT

126.16

Mae

4116.66

4444P1

4116.66

01.66

144F16.66

111.11

216.66

10.66

106.66

626A0

543.4.

341.1.

Was

.412.11

4117.66

. .

.31....

- .

317.66

111.11

621.66

$66.66

ed.. 000000 .0.........o.........S.

3

.

LSAT1=GPA ( 31+ 301

N=74

2

2

2

2

2

3

3

2

2

I

I .........

1

1

1

1

1.60 1.12 2.17 2.311

LSAT first-AcoresNonwhite 2-time takers

s

t

; LSAT 2=GPA ( 52.4 )+ 282

I

iN=74

1

I

I

I

I

:I

I

2

I

:

I

::.--------- .... ................-....-.-.....................--..........

I

I

3

11

i

1

1 is

1

I ..2

11

3)

3.16

aa

2

3.22

1

.

oII

I

2

2

2.67 3 6Calculative CPA

......I

$I

I$

I

$I..

I

2

:2I

2

1I

:

:

:

s

we

11

..

e.........

.

2.26 2.71 2.22

se a function of CPA forcoached between LSAT exams

g .....--....-..........-------..------------.....-11

1

..-.....-I

3

3

3

33

3131o

2

2

2

I

.....6 ..... swOollo.m.............. ...... 0 ......... 4 ww.......om.. m..0imp 1.22 8.11 1.74 1.24 2.711 1.11 2.16 7.22 2.42 7.66

LSAT second. 000000 a function of CIA forNomd.its Mims tiler. coached botwo LSAT stoma

113 130

Cumulative CIA

LSAT

Scores

!fit) . Of 4., 11r OM

I ! I ! ? 11 E' 10 1i 77'.i1

. .0.1 op * , , ,.... ....Wm polgodMM .iI NW 0.0 maw Om.

L. 1 I .I .

t I 1 i ; I;III 1, ;61 1 ,

It own op Nis 11= .0 iklim 1101 0 aim. p.m. wool .IMS

I 1 L 1 11 i ii :i ,

1

1

i! 1 1

.... ....._41....1........,. .. O. 4.1.. 0 , ,,', ' ,... b....n...9 1mb. 0 10 fa ..

1! I I I I 11 I

I I

! I

) I 1

I

.11

I

I I

... GM O. .; . ow i at IN ...m. ., N,11. I a. MO MO m,m

1 II i .

1 1, i I

.,...........

700 41, me ... I.", I ....4 ro ...

I 1

f i

1 I I

1

I '...a I- 41 en ma. t

mem*

11 O. rif ...Molls. 111114 ,I11.00 LEI ...II OM I I ==I

I I II1

0 ..... Po,....I.I II

1 411; ...J.." . 11. ,. - I . . !.. a... I glowI '

I II

r iI

. I

600 ...,

...,...14

I

IMN.

if J.....1,... j....,.... .....

,1 1

1-1- . ,

LI.

500

titt

400

300

200

1.0

131

, NOWI I

nito, whites LSAT 2

all OM IMO Illy a 11 r704

1 11 t 1

1....jaami11. MM.. , OW ....I: ..

4.

c w 14.

1 MI AgoI I i

O 0.. L......... .

whites LSAT 1

..,, nonwhites LSAT 2

[ 1 I .1%.nonwhites LSAT

=1 .6.

1.5 240 2.5 3.0 . 3.5 4.0

Cumulative CPA

LSAT average scores for 2-time takers coached between moo

with LSAT scores as a function of CPA,

comparing whites and nonwhites .

1'32 .

14) .LSAT first and second scores for white and non-white--I-time takers

This final part of the white/non-white LSAT analysis dis-

plays and compares second-LSAT scores as a function of first-

LSATscores. Shown are LSAT scores for uncoached whites,

uncoached lion-whites, coached whites, and coached non-whites.

The first set of comparisons displays differences be-.

tween uncoached whites and uncoached non-whites and between

coached whites and coached non-whites. In both comparisons,'

whites outperform non-whites, with greater differences appear-

ing at higher first-LSAT scores for the uncoached groups and

at lower first-LSAT scores for the coached groups. The second

set of comparisons reveals that coaching is more beneficial

for whites in the lower score ranges than for non - whites with

similar first-LSAT scores.

This study does not attempt to explain these persistent

differences in LSAT scores. We appear to have an effect with-

out an identified cause. However, to clarify subsequent com-

parisons among LSAT coaching schools and between coached and

uncoached students, non-whites are excluded from the remaining

analyses. Such exclusion allows subsequent analysis to com-

pare relatively more homogeneous groups. In addition, rela-

tively fewnon-whites enroll in coaching schools, so that data

are too sparse to rigorously evaluate individual coaching

schools regarding the extent to which they benefit non-white

115

133

students. Uncoached whites with an initial LSAT score of 200

show score gains of 63 points compared with - average gains of

.96 points for their coached counterparts. Uncoached whites

with initial LSAT scores of 800 show theoretical gains of 9

points compared to theoretical decreases of 2 points for their

coached counterparts.

Uncoached non-whites with an intitial LSAT score of 200

show average gains of 50 LSAT points while their coached

counterparts show average gains of 71 points. Uncoached

non-whites with theoretical intitial scores of 800 show average

decreases of 10 points whereas their coached counterparts

reveal average decreases of 32 points.

..

116

2.1.10

1

.

. ;O 2 00,

1

0 1.0 I

1

ft SO

148.80 to

.. 2. 2.. . 2..

LSAT2=LSAT1 ( . 91) . *441 srs see .,P le 1 444: 0... I

.. 1: ....... . 80, 6 . .... . .44 .1.1 :.1.1 2111.:; Z... 4 s

Is. iss.:1 %%%%% ss:sts:.: 7. le l

15.10 9 7 29 a 4.Zig.g:::.1::::sf:::1011 11 : s

1

I ittwts..::,..

t

181 all $$$$$ 4.14:41.104. ..e.1. 2 4 1

"I

. g 21:411. : il$IJ: 2

20./.

1

181/.811

8..........................................$$:::21:4!":%1'1

r1

1 1 1ter tttttttt ttttttt .1 I

1I. :: ttttt 12141. 4111! . tttttt 10.442. 2 / 1

DJI

I 1. 111:8I e..4 0.848 ge

OO2.

4224S0

1asses 22:4201? 11$10.1,ell$214 ..1 I

181 41421I . :s I

I1 4 ttttt IIII/Itet 4181 ttttttt 2/ISIIS 1

Mae 11.1.141. 4 ...81.. 8

16l 1.1.1.4..1 a 1II II./ 1 1

I ..1.0.11 . IIM .. a .. 1422.1411401 .V444.41/14441114. .1 I 1

..I. .. 0.1.2.1s. :I. 7V.1:.41111 21.1 ...

I

1

I2.42 14 . ttt :7121.401.2. II" 1

44242/ ..?7.5 191111):1 l I

11 3.1: .. ttttt 124:42...44.42

1

I84 it' .1:118812 111.71:6:81.1 1

I

8 ...........8. ..2zsft531: O ..00. OMe

I. .: el:III :1 S... e 1

8.2.18le ) S...11 2.12$.72. 1 es

1 1 8 :1 : 88 8... 8.I

I222..2 es 1.11.1 .

1

1 0 ,2 2 40 II

"1 01 112.: I .I

/MOO DOI MS

I. j *Jo..

I

1 4 11

1II SO

I

1 OS 0 1

'ags.seOS

11

I

.

Ia

I

.11.4,4PUS:SS11. 1 S a Slio ...

I a

ISOMa

11110.00 131.11 101.111 111.4 411.20 414.11 811.8 140.411 421,0 911,29

First LSAT

LSAT second .tort e se s function ad (trot-LSAT 0014for wLit ncoschod 2.tis caters

.0............ tttt....1...........LSAT

1046.10I

11

1

&Item 1

1

, LSAir 2 =LSAT' ( . 9 Oi .S"''

. o

1114.111 + 70.2..

.

I I

I. . .

. I

y

eI

' '1.1781s 4 4 1

:1164518.15

I

1

40 III

1

3 'Nub I I

1 ! 5 1 el1

..... am. or a. woo. da.....m... a ea.. ii see Ilo Is.

'22240I

* 1a/ .... I

--.......-1

1--...-.**,.....

28: lees . .1ttttI

I

I a I I It II/ I . I

.5111.411

.2 I ill il I I a

I541)3.2.4.4242 el woo I

1

8.1.45

1

I go I. I. I ell all

I0, , 8. w. 1

I.11 1 141.11 lo I

I

458.1111o wow. es 1 41 P ll peas j

I1.....0.1 ~of I

I

I""- ' es .3 1 II dm. ale,

I

2is 3111 14 lk a. w. dm

I

8810a 12. .2. I 1.4 p 33.

4I

es .. 88... 3 I i.e.. 83

1..I. 1 tt Mir

i

............1

1

I 8a. ...SO. S

............................ /.. 1191.- 8.8 ... ..... ..........

I 2 .8 .21 8 11

Mae 1.18.1 1 1;

8 1 108 8.8 lI

I4 I. 1

I

Iwo . 53 1 . I.I

as a I OS 4.

I

seel . so I 1

1 5 s 2 8

I

11. el I

1

t 8

I

I.a

1

mar...4....raomaraall

ISS.SS

.11

1101.4148..,..................4444ttttttttt 4040....000.040.

8118.10 844.111 114.0 See.40 111.010 44e.10 1118.10 $14.55 411.15 221.4./

LSO flft

184/ wood states es lanstIeo of 101.SAT Ken

souldto easousiod ?.t1.. 4above 1.35117

LSAT M.OSous.Store.

LSATSecondScores

...........-

I0

11.40

8 LSAT2=LSAT1(.841+ 129ma. I

1 N=4416224

1

142.6

1

It7.1.

464.46

16.000.01....... 088800.0.0.11.0.0.11.00..0000.....S

I

I. a

6 I

0 0 12. 10

418 0 0 O. .0I

S

I. 2I .. 1.1.1 1

II ..I. I I t I

II 1 *I 2 10III I

...... ..........880.411000 080.41.1 Se jo

O. 14 I... I ..

Moot I lost

1.2 1 I

111

II 1 of 1

O.1. 71 1 2o oi I Use

j

2 i .0 I 0 2. . 102. 2.

II

I.0 2 2 O

. .. ...

O. 100 .0

2 1001

00 W.S. O.0

2

1.

3

121.20

MAO1

"Vv.%

114.21 .

122.00 121.1.

No

212.40 171.12 442.00 4s4.2 .241.20 2111.40 35.15.

LSAT second tanctian of first-LSAT scaresfor white 2wttoe takers coached between mos

LSAT Tirol S

1

1

3

3

1

3

1

1

2

*MOO

.......... ....................................- ......0..................0 -...---...--.......---.................

ilT2.2..

I

I

I

I

1

1

""

I

+ 1061

I

I

I

1

1 I

1.0.60 . N=74 .

1I

I1 1

I3

1

I

11.1e42

I2

I- .--. - ..- -10..... - ..-... - - -.. ...... ...... 1

I

1

I

I

5.110421

4 4 1

111

1

1

1I

1 LSAT2=LSAt1(.83)

421.0

$1

122/.40 1111

a

. :. . . 1. :

. 1

2

.......... ....................... ............. .................w..........s.......I

1tla, 2.54 112. 420.611 4,14.110 4118.118 212. 214.11

LSAT /trot Stew..

LSAT ...nd seem as tunettow of (SAM? este,ie. 2202.4.2be 2121e. tofu'. twselhod letimoso mums 3 r.;

118

(.1

llooso4LSATStereo

'CG

I ; I .

.i.1_!. . -

I / ; "1

- I -I -I

!

I , lottoched outdo -t.I i..i....i. i J..: .. ;.I

; ..r , .

I t. 1 i I- 1 t .1

.. .....11...1...L_:1../-i...;...t-&_.

. I ! : : . 1 !.

-; I- I.

, . 4 , :

o - .I i /1 -..-

O 1.j "7loreakerro 1 InsI-1.

I : 11. . _

- ,- , . : . : - .- :-, -. i r

' L '. I . .

t... ...

-fi

I

I---bocoocood nonotat.

I

F

400

200

&eam, 300LSATSeem

700

400

300

400

300

200. 202

200 300 400

1-r-

:._;_!_._ J.-r-

.4I. , _ _J___, I, : Fri ; I

. _1

1-, 1--.- i_

.....,__

' 1 i i. I . _ ri .. _E_-

soo 600 200 LOO

first LSAT 1

LSAT overage for trocoelled 2-11m Jokerswith second-LSAT (unction of Ettot-LSA2,tooporInt whites and nanolitt

NM

s.p. sit_yitig.4:7R- t,7

oak14gill

!lirIp:d-11

I :,e

111L

. a

-:11 I 7 I: L--. aidu. , A ;air" ...mi......

,4

pliyi.v_,/- ,..1"

IIII

ii

_.,Ger,-.--. ...moo001,-

i-i-

iyenitt.,

A-iiiumimil :..:

mai

Li INI -:51:1:

white

300 400 300 600 .200 KOfret LSAT Scioto.

LSAT ay for Milos taws tomchoillkit000n Im,Iowilog ron4.1.1T story. fttottlen of ftruLSAT coro,toorottns blottobe and soorIstir

11.9 137

Siscfm6LSATScores

700

600

SOO

400

300

200. 700 300 400 700 400

first LSAT S

LSAT aver.te scores for 2-tice tette, shooing second-LSAT

scores ea lune tion of first-LSAT scores.comparing wncoacheJ whites with whites coached between LSAT ems.

Second SOO

LSAT

so

700

600

SOO

400

300

WS.

200200

300 600

-

t.

.

_' L-

...]:_f_

i

__

AIMI_

#

..1 '

/I1/ ...4....1 I

43-T/ . i a INr1

' iI

17-riI

. N

eoltom_linc1

111

he

.

eW ..7:tar /./'4.0/.. j

7! ' EE. .

colaci,e4 triages IIt MM

IIi.-/

.-'.

, ;--1 EKE am._,"-e-pMINIMpMUMNII

11

. :HI.....v

P "'I .

' ill !gal---hil--r_____ iiiirsIm

300 600 300 700 600'glut LSAT Uwe.

1SAT overags scores foe 2.ttest liter dloswifte ddoodLUT

*corn se function of ftreC.LsAT scores. comports/;meschml sweAltes with fowsbitwo so.chsi Wiwi* LSAT saws*

120 d9

600

f

(5) LSAT scores for 1-time takers

This section displays actual LSAT scores for I-time

takers coached at various schools against uncoached 1-time

takers. These comparisons reveal that:

(a) Stanley H. Kaplan Educational Centers (001)

provides effective coaching for the LSV';

..(b) All coaching schools seem to help low-GPA students

more than high-GPA students;

(c) The Amity Review Course (Amity) (002) and John

Sexton's LSAT Preparation Center (Sexton) (003)

appear to be marginally effective at best; and

(d) The LSAT Review Course (Evergreen) (004) and a

composite of small coaching schools (005-021)

appear to be effectivl primarily among low-GPA

students.

The following table reflects average conditional first

LSAT scores as a function of GPA for those individuals coached

before taki.-..g the LSAT and for the uncoached group.'

GPA 1.0 4.0

Uncoached 436 640

001 474 648

002 407 619

003 444 625

004 485 614

Others 478 610

The greatest net benefit is shown by Evergreen's increase of

49 points at the 1.0 GPA level. At the 4.0 GPA level only

SHK shows an increase and that of only 8 points.

121

139

447Sew.

LSAT$

1

'Atm LSAT=GPA(58.2)1 + 415.5

ISO ; N=6081

1143.411Is

I.aaaa3

1.1Si

..0 ...... ..0r -

11 VI

/4 a

t1 1 .1. e me

a I I

fee s: ow es/ 3 11 1.

t I I 1

.1 41. . t 1 /

Oa 11 .2. .111 1 S. 91 : /

.. 7.

1

a

t

a.

af

I

Se / 1

so.

I . 1 31 .3 .1. 1 .

11 3 4 Soo ... s

MAO I. .61 / 3 3 es

Ile 1 / 33

I I a 8 . I: 1 1 I

1. 1 .I. 2 S

Ii 1

MAO 8I

I o t I

I1 g. Se

go., II es

II

41.14so

I a Ia

e.

I.

IAS

1

11

3

11110.43

3

3

I

141.83

a

I

34

I

Map

200

700

WO

.$00

400

co

1.76 1.11 3.22 2.41 3.6 2.I 3.13 3.34 3.35 S.s

Cumulative CPA

150.7 scores as A function of CPAfor White 1-tioe limiters coached at 001

-1 1i

-i- j I i I ! ;....,

t : t

I

-1-'

1

1-'

;

---t-1-1

-17-_L--

1 i :

I -1`i 1 1 1+1

--6 ...t.tte...z_ !

.1 I 1

.

....

a

.

I

a

--I

a --

"7--. ---- 7I.;0licoethed-Yhite---.. .

1 i I 1

-1 ;.al ..1..ialaral

a''''..

A_11---

r-

1 T--

..4....J.._i

. i - ...

1.3 2.0 3 0 3.3 0

Cvmdialvim CPA

Wit ..... stereo for 1.tio 6 r a flotillas a GrAl,colorlog while. 440.4 sit 001 oltlowatoeck464 oreltes

122 4 0

2.3

.

OAS

OW. Ieri.,01......

LSATISOM

Score' I

111.111 8 LSAT=GPA (70 . 8)+ 336.2

N=106 .ISOM I

i

f$1.1@ !

Ia

0

0We

00000 ..................1

Ia

II OS

a

.

a

II

.........

a

re-

t

1Se1

1

1

11

1

.1/

114

11

11

1

. 1

1

11

3

1...............1

;17.I@

I1

1$1.11@1

11

1

411.4,1

1'AMU

111

.354.41

1

3

141.7@

3

1

Issas

LSATScovs

'CO

700

600

SOO

400

300

.4...........................................e........e....... ...... .....-...4-..... 4............. .4.

1.11 245 144 Paa 3.7 1.11 3.11 3.11 3.17 1.71 ss O0

Cumulative CI'A

LSAT for I -tIs taker se function of CPA

for Mite, cach4 at 002

; .

I

I ; s I

I

1- -1:-.- I 1+1

I ;

3001.0 1.3

r2 0 LS 3.0 .3.5 A 0

Cfautattve CPA

LSAT refgo scares ref 1 -Um* tablets so bastion of CPAteAtatlfts whItso tacked at UCZ with unseat/64d MI.11 4

123 41

2beurl. *siZat0/A0AY*Job:46...iAi :aucan.64....aA.&...::

ucL1142

Scores

010.11010.111

eLSAT=GPA(60.4)o 383.6

cum ; N=3501.

1

5114.7111

a

4.0.

10. 1

1 1..... .....

I

OM 0110

2 a

a

. ......

041

O0

7

.

I

7

IO

3

a

le

8 Aa

I

4.7 a

a

a. ..

JI..2-2

asae a

al ..

as

.......

a

8

..

e

0

2

I.1.,.

filwoodwft...

.1!

I

I

I

I

i7

I

I

4

I

i

1

o

212.II

ft

I'4111.$

1000.40

400.30

11$1.2

'MAO

2$1.0

LSATScore.

100

700

SOO

$00

400

300

100

1.73 6.17.

2.2 2.22 2.$1 2.1$ 3.00 3.14 LAS I.73 3.18

Cuoulativa CPA

LSAT scores for 1 -tine takers as (unction of GrA

for whites coached at 003

47 1 01-1.-4 1.-14--:--

wm_Aba&vidteez...-i 1

.

--1 i

.! j i-.

_...1 1 ; 1_14.1I

1

1

-1 1

11-1-1-

-1

1.0

,.r ;

I !

!

i_lI...... ........

_ -4-J-1--:._1 I i

-1_111._ _ I _J:::::. .. -.L. -I-,

,

I

1 5 2.0 2.3 3 0 3 5 4 0

Cumulative SSA

ISO averse. cote. for 1.tio. toter. as function et CPA

tOertIns %Mies cumchd WI 003 Oth.unco0..61 vhill'

124X42

OPPONIM....0.................M.M......... 11 a. er. .......... .

LSAT MAIa

. I

1,4.1.1..SATGPA (430.0)+ 441.8

s .11 1.9D41,134.13 = ..

I111:111

- ...-----a

I

St.

fe

..

1.'1.11a I MMUS.IMO

111.33

. -

113.31I

als.11

Is

770.10

'113:30-

33111.16

15AT 600

700

600

/.ft

$.

1

1

1

1

1

a

II SO I

40 II OS II .01Oa 4 I

411

I

1

1.1 1

1

1

1

11

1

t .If&

7.17

1

r

2.17. a.aa 3.11 I.4 1.41

LSAT core for 1 -tire takers a foutIon of CPAfor white, coached t004

rdrri- rrTTIT1:

11 II I it/LI

J iT1

I %mimicked 11

-1-1-, , `-004 tail to.. 1 ii 1

--I 1.-i.--1-: 1-1-7/1i I I

I

7.44 3.11

Cumulative CPA

1 - -4--,i

1-7 '

. I

- -

T-1-1-I

_i____,L_1 1

1.0 1S 20 2.$ 30 3.3 40Cir/laity. crA

LSAT F. Roues for 1-11.. takers fiattIo of 463tokrreng ioldt coached se 14.4 via encivched whit 4

125

a

4.4

LSATStores

LSATScores

111.14

1

LSAT=GPA(44.2)1 + 433.6

0.

Phi"

N=141. t

iist:se --

-

$11.111

471.11

a

;S1.30 .....1.54 1.40 1.11S 3.21

te0

700

a

a

...

O 111

a I1

a I

I

a I

a 1

a iii I

a a I

ai... 1

i

eas I

ai

as S I

a I

a I

I

I

a

. .6. ...... ...... .....am

2.5s 7.71 3.1 1.77 3.5I 3.71 4.0t

Cusulative CPA

LSAT score. for 1-time takers as function of CPAfor Olt.s coached in other schools

----i=athlAas in other school.;

100 ..1...-.j! I I

!

2001.0 2.0

1

3.02.S 3S A0cvmuttv. CPA

LSAT awros score' for l-tio, talkers a function of CA.

CooarIon whites coached in othr hooi with lancoachd whIte

126 J44

(6) LSAT scores for 2-time takers coached before firstexam

This section displays actual first-LSAT and second-LSAT

scores for 2-time taker's coached before their first LSAT.

These scores, together with comparisons between coached and

uncoached students, are shown for individual coaching schools.

These results suggest that:

(a) LSAT coaching is primarily effective for low-GPA(white) students;

(b) SHK (001) produces better-results than the otherschools; and

(c) It appears that the benefits of coaching may besomewhat ephemeral; in most cases, the uncoachedgroup improved its relative position on the secondLSAT exam.

127

145

rfrt 414.11

LSATScorn

444.4111

411.11

sts.ts

ss).4

Cti..1.

4sr.c

10.110

413.11

814.110

304.10

;1

1

1

sc

I

II

LSAT1=GPA(30.8)

N+ =34 1136.2.

.

...

2

I

3

4.4...

. .

a .$a .4.4 .8

.I

tI

SO

e

ii

........... .......

IS

118

.4.--

1

1

O.SO 1

1 di4. Os I

I.11: Ir.

1.......8 a:

ll. 0

PO 8 I

I 3 1

1

see 1

1

1

3 a 1

...., 1

4 1

h 1

... .........q 8

1

1

.3

1

11

a

1211oP

IN1

.....;.7 1.08 2.22 2.44 1.1

........................

1. 3.88 3.33 3.31 3.77 3.77

LSAT score. for 2.ttme takers as a function of CPA,hewn first-1SAT score. for whites coached at 001before their first 15ar

Carsulative CPA

....... 41.11............................ ........... .......1......41......0.....40.

Second411.0

t

. 1

LSAT I .

1

Scares 1

8 LSAT2=GPA(47.8)e 1

1

)811.1ks8 + 399.2

8

1

a 3

t

I 1

487.48t N=313

3

1

. 1

1

. . 1

10 OM s

I.

* 1

424.08a .

I4 .. 9 1

... ........................ft....0 11.. a. .... I

8...............-----1

. .4 am 3

13 gm k a 1

$33.30 .1 ph .02 .

1.110 e 1 wire se I

3it

gee $ 1

1 Ia

3

t1 I

sehli a -

2 3 I 4P

t

.

. I es 1

11 aim ell 8 a a 1

1

. 1

I1.

8 I

41.411$ se

11 Os OS 1 41 1

ell" : 4 11. I

1- ........wm.dompwm... ..........0...8

1OM

444.48a .

1

1

8

1

40.488 a

18

11

1172.s1

1a

818.08

81

3

0.84 Oat 8.11 2.40 1.44 1.111 3.t1 1.33 3.41 1.17 1.88

CalmulatIve CPA

LSAT 'eaten for 3 -tine '.tars a function of CPA.

stowing &IAA? e1 fur 1/1111. cotited at COI 1 4 fbolero 'lett that LSAT 1711_

4

a

.

1

1

1al

1

1

1

first 4.44

kW111.10

"'"

w.

1

;LSAT1=GPA(68.3)+ 291.4

$14.11!N=41 . 1

GM. Were NM 1

1

414612 OP 1

1

1

1

asa.a 1

1

1

1

431.401

NIWOM

111.11

1

2414.111

MAO1

11

1

1

1

311.1111

0... .. .

1.11 1.11 1.44 1.60 1.14 SAO 1.14 3.1 1.34 3.41 3.01

Cumulative CPA

LSAT scores for Mime teker as e function of CPA

showin& first -LSAT scores for whites coached at 002

before their 11rt LSAT

7''7 ..

.

Second SUMLSATScores

sei.S1LSAT2=GPA(61.0)*

1

1

1

+ 337.51

1

a 1

1

1174.11 ; N=41U.

1e 1

a 1

1111.11 1

1

1

117.201

1

1

1

1111.5

1

2a,smea

....... 4.--....- ......

3

1

II

1

411.4

I

Uhell

111.55 0OOOO O

LIS 1.44 1.40OO

1.11 1.11 3.01O

1.11 1.14 . 1.4 1.04

47 1.241 score, for 2.tiee satire si fuor1lon of CPA,

hinflog eocend.LSAf store. far whites coached at 002

**lose Shelf tarot LMT 129

CtolillaUf crA

. _

rivet1.14T

1tas

811.44

9$0411

$4..4

Mae

41$.$

417.11

1175.56

113.4$

111.33

341.$6

,0 * *

4LSAT11=GPA (42,9 )

+ 364.7N=192

t.

a

1a

3. e

1a

1

Weem...

.......

s

OS

to

a

.8/

I.1

3

1

tom

s

I

.........

.

t

OS

.

SSa

SS

3

$ ..4

a

SIi

1

1e.. II

8 m s. 1

1 ..OS a

ipa

1

es ao . /

...... I. ....I1

a

II

e ill IO a

1a II

a1

10

a11I

I

3

3

3

.ffi.O

SecondLSATScores

..... ......

106 Sae 1.11 2.41 2.65 2.64 3.3* 3.21

LSAT scores for 2-time takers as s function of CPA,

showing first-LSAT scores for white. coached at 003

before their first LSAT

-141.$0

LSAT2=GPA ( 65 ; 9)MAO + 324.3

;N =192_1131.11$

3

I

ISO444.4$

11114.30

3

I2S.*

.473.113

0.va

431.31- - 3

SIMI3

1

$411.211

1

SII.s

.. 0

a

1.31 3.71

Cumulative CPA

3 11

.....

3

I

soass

1. ess 1

ess s

as

I

I

1.

It II CO I OSI a

ee . a

a / 41. a

4 a

aI

00 TOi Se a

: . aa

. 1 s

as ils p

a a 1 .. a

s

O ..... ...... ...I

a 3

a I

11

....-

1

......33

II

III

....... 6I

.

a

tad 1.141 8.17 1.11 SS 3.114 S.'s 341 0.$11 1,71 343

ClatIve CIA

LSAT scoria for 2-ti tabors as tictim 1 CIA.

Shoving ecn.AAT for *lt toothed at 003

blot Ott, gil.t tsif 130 148

SecondLSATScores

Met 4$11.9

UAT

1I2.16

MAO

.0..... 00000000 ..0

114.19 $LSAT1=GPA(37.6)

8 381.81

ma. N=891

1

Issas

311.40

2

448AI2I

la

421.48-.4.%,

426.2822

2

I

36O.SO

......

132.88

a

a

2

2

216.gal 8.24 1.41 8.61 2.42 1.81 2.1 1.11

LSAT scores for 2-ttne Weer. as function of CPA.

shoving first-LSAT scores for *bite coached at 004

before' their first LSAT

...

LSAT2=GPA(39.3)+ 411.8

N=89

MAO

MAO

Ise

4118.12

atlas

114.40

SO

s

3.2s 3.7 1.

Cumulative CPA

......

a-

I

t

al

I

8.81 844 8.42 841 2.1 1.13 1.113 2.14 1.2 LAI/

() LSAT stores fur T -tine tailors es function 4 CPA.

showing ron.J.ISAT enr Int whites coschJ t 1104

111

Cnultivo CIA

611.01MiltLSAT1401e0

111.11

sLSAT1=GPA(39.1)

el

1115.41 + 387.2

s N=378

SHAD

....... ........... .... ...

I

I

I

II

3

II

fOA.S .

411

I

I

I

I

II

I

3

40141 . 3

1---......- ........ -- --- ----

------.------..... ----.--

.........-.....................1

4$7.1 3

73

3

I

434.13

70

1

61840

7

507.1

5.01 240 1.31 147 2.76 1.13 7.16 3.33 141 3.70 3.11

SecondLSATScorn

076.4a

117.31

t

Cumulative CPA

LSAT scores for 2 -tire takers as function of CPA,

showiny first-LSAT scores for whites coached at other school

before their first LSAT

LAT2=GPA(72.6)4 314.9

Nrg;37

1...- .....

le

8

.

Iel

I

.....

III Sat tat 8.13 .16 SAI 1.111.1 Sat

Cumultiv CFA

LSAT to P.1141 111,40 fonetin of OA,lmatod t.61 *coes for whits. twiched at othetystl.00ls

Wilms thole Mot tIAT /

132

Viral r' ILSATUwe'

700

500

coo

I. : a- , , ,

A - i . I 1 - 1 - 1 - ; ;-- . - I-I--; . . . _ _.2. __I'. . - - - ! : I . . . I

-'I 1

. 1 I . 1 . .1 _ . i . : .1 - . . ,.1 . - .! . - . . --- . .1- . . I .1 .i__ ! I . -I..1 _ 1.. I I', I 1

-------1 .1-7,...4...,- 4- 1 .--.---1.1 " I __ ---- -1--- , . ,

1I. a - . ..,- .i .,- ... I. 1-..- i 1 .. ,

. i ....-. 1 . '. - 6---1

11 I I i I.1 !_j I 1 :

___._ i____:,[... .1 ;.:t1- ,--1 . . .i

. . . ..:.:,1 .. i . : :

, . 1 : : .Ti 1.. r _.,..) ri. i,I r"1

1_1_11_t /If 1

1 evItteethen selalt,ceI a I

tol--.41Tit.

..... ...... ; . ' -J I .1,1.....t. J. ! 1 8

1 1 1

, : i-L -f(t 1 i I I i .....1

i L. I " n ! j 7 i , . : . :

_H..]1

: : . : i i1 1 ! I t 1 :

. I 7-7-=_FI:

3C3

d-i... -i- -. __i I..__' 1111-71: -.1!1 CL___ : ! F 'rn. : 1 : 1

: I 1 ! i 1

L-11ii ;

1 6 I 1 1

:-Ij --.--/ . Lll 1 1 i ; i1 I I ..!--/.1-- . 1 . I-1 .___-__,-T: -4_!__

200J'il I IIi i

1.0 1.5 2.0 2.5

Second "°'LSATScores

700

600

500

400

300

200

3.0 .3.5 4.0Cumulative CPA

LSAT ge cores for 2-tInc takers as function of CPA.

shoving first -LSAT score, and comparing whites coached at 001

before their first LSAT against uncoached vhi tea

-; I .- -[.S 1 1_L1 1 :

LI1

1-17 I I I 1

1# Iil l'r7 1. 1 I I

A

1--7-4--Hi i

lam i , , , :. ,

i I I. 1 1 I! -1 1 I i

.21_ _ UMI

I

1

I 1 I :111

1. I i 1 ,

1II1 I 1

i

----001 wit

;

-. -tI

tncoecled eltte-%I I

i 1 I '.......L...

i 1 !

-,1 I

, 1 I-.1

I F- --1- i-_,_,_

i-t-I1.

i .i ._ -t

_ .

1

;

1

T----: I

'

--71-1......--,L..4

-1-1-1- i

1.0 2.0 2 5 3.5 4.0

CueglatIve

5 , IISAT ttttt o store. Cm 2.tIe cater. w (unction of CPA

,ahotv 1 hp ..eld.L:AT wrote, cvspag 1..ft whits,. coached t U0Ifetur Weir first Ilef soloist encroached ladies

133

15 30

!tee

firstLSATSt0e90

SecondLSATStores

t

4C0

4 4 -; :

11 MO I .1. . 11+,,, I _,_r--- 4

1--.

anceiched 14 1 te s

200

8C0

700

600

$OO

400

300

200

002

. . !

_.1....d : ! . ! . : ; .

, III

i I

.._r_!_.__!_; ...4__ , . i ! 1.

1 1 r, L --r ! -1 , .

_ -.3.-.17.[_i ----,._c.-__-_:_, j : ,----1 ---i-Tr_Qi ;----.__t- . . . , ,

i_H____ _LI 11,11,,L, .,,_IL - _1_4..1__,..1_ . . _

1_1 1 14 H 2T:I : _iT-r- I ! I I

I H-i"-:--I Ili 1 I r ±

-IF-1

1.0 1.5 20 2.5

1

3.0 35 4.0

Ctraulative C?.14

1.Skr average scores for 2tioe taker. a. function of CPA,.

showing first-L.5AT scores and tor:paring whites coached at 002

before their first LSAT against %stanched whites

t

I

3.0 2.0 2.5 3 5 4.0

Cuftwlatfa. CVA

LSAT eeeeeee cor.. far 2.tiral taker. as a function Af CPA.

hawing second-L:411 scares and ca.pnlins. whites coached el 002

Were. their first VAT against wet...abed *hits.

134

1.5 3.0

1 52

FirstAILSAT

Scores

Second

scare.

COM

TOO

430

SOO

'CO

200

5C0

700

S00

S00

400

'200

:J4::I -1-T g_. _g , .

_ 1

ems, 1- F . t .

rI t ,

;

----1---*...1_1_._.;

i 1 -4-1-

l sncosched w4itsimmn:,

..- _,.':-..003 vtlite,__

f____. . : r-, : : :.

1---- --- i ,. .

A .

111..+--

-- ^- 1 1 I -.---L ..-.1-71 -.4 --J.-T-11-- 1 1!--1I

I I : ;1 s

.---, Fr----7. .--1-":1 : :

.1---- .1r- "--.--".; . : '

-+--

i

L. 11 I! 7---- 11

I. 44--ii

i

i . : . i I I

I_j_--1 1 -1 I

--i I i I 1

1

i, 1 H : -1-FH:

1.5 2,0 2.5 3.0 ,3S 4.0

Cumulative CPA

LSAT to scores for 2-time takers as function of CPA,

showing first-LSAT scores and camparIng whites coached at 003

before their first LSAT against wcoachad whites

1.0

2.0 1.5 2 03 S 4.0

Cumulative CPA

LSAT ge o for 2-tine later se a function of CPA,

showing their second -LSAT COre1104PCCOperinA VhitOS coached at 003

before their first LSAT agaluat tmcoached Nate,

3.0

15.3135

Met 100

1:47Scam

SecondLSATScore.

700

400

00:1.0; il5

SOO

:

700

. _

un Ord wlitg000

whit s

2 0 2 5 3 0 3.5 4.0

Clasuiskitvc CrA

LSAT to scores for 2-time takers as function of CPA.showing first-LSAT scores and co-raring whites coached at 004before their first LSAT against uncoached whites

to

00

00f 1.0: 1 5 2.0 2 5 3 0 3 5 4 0

cumulative crA

tsAr AAAAA tel scoria for 7-11. takers se a function of CPA,allowing cunt.i.wr eeeee on4 a...paring Witte. cnochod at 004Wawa their first LSAT ageinat uneooched whit.%

136 -I ° 1

fleetLsATSee's.

SecondLSATScores

I

700

S00

.300

2001.0

SOO

700

;

SOO

I

00

200I 2.0

L

'-.whites apached in other school.

i t 1 i "I "i i

2-.. 1......-J

_I IJ.:=1"/-1

1 S 2 0 2.5 3 0 :3.5 4.0

Cumulative CYA

LSAT ge scores for 2 -tire tater. as function et CPA.

showing first-LSAT scores and coopering whites coached at other school*before their that LSAT against uncoached hites

-a

chid wi i it a

-II-

1 Iff

ites:cnached at other schools

-

1 S 2.0 2.1 3 0 3$ 4 0

Cumulative CIA

LfAT &&&&&& score. for 2rio tater. as a (unction of CM,shoving acund.LUT Genie. and c.avarIng white. coached at other school,

before their first LW &vinal uncuached whit,

55 137

(7) 'LSAT scores for 2-time takers coached between exams

This section displays first-LSAT and second-LSAT scores

for white 2-time takers coached between exams. Comparison

of average LSAT scores for coached and uncoached students

reveals the following:

(a) Amity (002) appears to offer the greatest benefit

for high-GPA students, whereas the other schools

seem to be marginally effective with Sexton (003)

showing a score decrease.

(b) Evergreen (004) appears to be ineffective for

low scoring GPA students whereas all the otherschools appear to be extremely effective.

The following is a table of first and second LSAT scores

as a function of GPA at the 1.0 and 4.0 levels for the coach-

ing schools we have analyzed. The uncoached group is also

included:

GPA1.0 4.0

001 LSAT1 410 534

LSAT2 464 586

002 LSAT1 302 580

LSAT2 364 645

003 LSAT1 467 524

LSAT2 502 568

b04 LSAT1 417 515

LSAT2 414 569

Others LSAT1 354 536

LSAT2 402 591

Control LSAT1 412 556

(Uncoached) LSAT2 421 605

The composite group of small coaching schools shows gross

average conditional LSAT increases of 55 points at the 4.0

GPA level, Amity (002) shows a gross average conditional

138 1 5 U

increase of 65 points. The corresponding control group shows

increases of 49 points indicating a net benefit of 6 LSAT

points at the 4.0 GPA level for the composite group of

schools and a net benefit of 16 points for Amity (002)

students. Although these increases are small compared with

the dramatic increases reflected in SAT scores part of the

apparent minimal effect is accounted for by the extra-

ordinarily large increases in the control groups.

At the 1.0 GPA level Amity (002) shows score increases of

62 points, SHK (001) shows score increases of 54 points and

the control group shows average conditional score increases of

9 points, indicating a net benefit of 53 score points for

Amity (002) students at the 1.0 GPA level, and a net benefit

of 45 LSAT points for SHK (001) students at the 1.0 GPA level.

Met 801.8LSAT t

tSeot t

t

841.18LSAT1=GPA(41.2)+ 369

$.411

00.000..0.4.00.00 ................. 0....000110

571.1

411.58

451.10

411.10

114.40

147.50

110.30

N=155

.

a

1

.

- a af

1

1

SO1

1

1

1

1

I.1

as1

1

f

atI3

.5 a

1

3.

a

a

2.54 1.73 an 5.09 3.17 1.40 3.44 7.2 4.10

Cumulative CPA

LSAT first Acores for 2-tio takers as a function of CPA

for white coached at 001 between exams

Second 740.00

LSAT 1

Scores

721.10 LSAT2=GPA(40.7).+ 423

`"*" ; N=155.

11.43

I

I

3 1

a

a 1

1

SO

1

01M .... 0.

Ia 5

5

Ia :

I

ill II

I5

e s

a 5

as I....,I

al1

1 , 4 0I

:

:gI

0I

I:I

......... ..........1.71 1.9 1.t 3.41 b.8 4,151.01

CUm=8* CPA

LSAT cond scorn for P -118.0 takers function of CPAfor mitts. coathd at COI Initusn egaa

140i 0

t.O 1.11 0.84

,

PIC' .04...Geo 00000000 .4104444..44

LSAT 1114311 1%W 1

1

SecondLSATfumes

LSAT1=GPA 92.8 )1.11

+ 209'soar

Fq.4 4

sm.a

442.4

1

1

411.1

..21.6

1

21.11

111

tim.1

1

1

1

153.41

1

1

1

111.

1.1 106 1.3 1.5

$

21

1

se1

1

S.1

a

1

$

11

1

1

1

1

1

1

1

1

1

1

...... .. .. ..

1.76 7.17 3.17 3.37 3.37 3.77 3.2/

Cumulative CPA

LSAT first ..... for 2-ttoe taker as function of CPAfor Atte/ coached at 002 between exam.

ts.

670.66

63.

$21.1

177.$

212.66..

413.4salt.

.

21.6

$111.1.

211.1

116.

..... . .

t r .

LSAT 2=GPA ( 93.4 )+ 271

N=44

1

1

1

1

1 a '

1

1........1

1

1

1

1

1

1

1.1 1.1 1.16 1.16

do. O.

OMB.

7.16

4. .0.. ...

. 4.

O.MMMIP..w.O.m.m..0...1,0MOMIPMAI..MO..

o .....

8.71 1.11

O.

a

1.11 1.11

.

3.17

$

11

11

1

1

1

1

1.1$

59LSAT conf for 2.4104 takers as function ffor whites coached at 002 ....

141

CUA6/41141 CPA

CPA

Stet asSJOLSATrn ISco

1

4154533 LSAT1=GPA(18.9Y1

+ 4481

11 -113974.111 ;N

1

1

&mond1SATScores

***

1141.5e

1

:110

Om. . U.

I

31

a . 1I

3

1

1

.. 1

1

1

35. .......flemmw.m. 1

1

0 1

el1

OS I S' el 1

......

MAO ..11

1

355.55

1

1

2317.60.. ....

1.31 2.07

. ---701.50

1

1

1

I LSAT2=GPA(21.9)1 + 4801

{MOO1

N=139lo3

3

11

WAS1

1

sa

11

I

1

3

11

1

1

1

1

..... -

2.10 1.11 1.12

3

1

1

2.7 2.60 2.03 1.12 3.72 1.14

Cumulative CPA

LSAT first scores for 2-tine taker. as function of CPAfor whites coached at 003 between exams

15 .15.

:es 1 ..

1 :

. I :

a: 1

1 1

:.. 3

. 3

.... ....1 .... .............. ....... ..... . .............................:3

3

1

s 3

11

1

1

11

..0.

*am.

I

I

I

1

;?-

s.

P si

I

I

1

. 1

13

1.11 1.5$ 1.611 lag SAO 3.11 la, S.IS 1.1Cumulative CIA

for utilise toschd et 003 betwn ...for -tine toter. as (unclean of CPALSAT earonJ

14260

rivet...... 4.......... 4. .......... ............, . -..

LSAT14.

2

Seats.

LSLT1=GPA(32.7)a

2

1

111.11s + 384

I a

; 1

!0;

1 4 1

104 2

511.10 N=58..

I

I

1 I

521.41.1

1

..... ... ....... -.................. ...... I

1

I a 1

I I

411.1I.

I

II

I

. 1 I

4141so

...I

I

III

1

I

411.8

1

1

321.1I.

1

1

1442.4011

a

1

311.21

1

1

12.1.40

1.1111 2.00 2.20 2.40 2.40 2.$t 3.11 3.21 3.41 3.41 5.131

Cumulative CPA

LSAT first scores for 2-time takers as a function of CPA

for Cites coached at'OCA betwecn exams

Ind 774.00LSAT 1

Scores I

111.00 LSAT2=GPA(51.4)a 363

10a

.11s N=58

a

ste.s

I. .....

.

1114.110a

......... ..... ..........- ................. ..... -- .4

1.11 /A0 3.35 1.45 1.41 1.41 1.01 1.11 1.411 3.41 3.01

Cumulative CIA

.

.

LSAT ccund score. for 2-tio taker, es function of CPA

for Atte* soilchod at 004 aveect exams

. O.

I

8

11

1

1

....

Of

161143

/fretLSATSeers'

185.88

SA.11

1131.11

L9AT1=GPA(60.4)+ 294

1

1

1 N=451

s'

11.»»

Mat .

sa

4711.0

1

417.101

1 e..

wre.1

17.10

317.0

1

1

337.101

1

327.10 . . e .

SecondLSATScore.

104.20

114.10

1

1

1

I

I

1

111.70

317.40

1

Ass.s.

1.11 2.2, 1.11 2.13 2.10 2.17 3.111 . 3.32 3.41 3.11 3.113

CUaulatiVo CPA

LSAT first score. for 2 -time takers a a function of CPAfor whites coached at other schools between exams

LSAT2=GPA(63.0)+ 339

N=45

..

43140

411.11

111.00

a

11

1

8

.

1.111 2.41 1.40 1.17 3.111 an 1.4. 3.14Cumulative CPA

LSAT cond scorn for Ttime taker, (unelln of CPAfor Whites coached at ether schools betemoett

144 162

!fleetISAT&cocoa

2.2i., -

.--I -1-1- - i-1

; ..- i - , . i

F.

--- I "i I iI-i.1 '

i?CO

,. , .

Uncce ched e .-.I t

, --t---ti---, - .--..-4-4-..!--.-i-14.1.- i 1 1

... - -1.--------------:,'1 . ,

! i'i _..L II; . ..414 .111:1::t=.11, t 2 1 I l' 1 i 1 I T -:-. ,.- . i , j 1,1

1_1_1 1_ _;. . _ , . :.4____) . ......_ ._.1.!_.1.....1_ __171_.L.

_.., rf-------1- i' 7 ; I- 17-- -1 i i1 , . 1

11

300 l- l4- ,

". I- I t i ir_l_...._.

_ _I-i- 1_m -

_

'_

1 i 1 I j -: 1

1

r

! -!- tI 1 1t1:200 L_

_1 . -___....._.. -- .

Second E04LSATScore'

100

3 0 1.5 2.0 1.4 7.S .1 0Cuullative CEA

Vef !co. 11. it 1.41 , of tsCocv4cl.r.5 I .A1 .. 'n 70ot' C nt 001 '3etiolurn examss6a lost f le t- U.Ar 4....c.tr, vtiite

r---r T7, i ;

-,. -1.:!._ _,_ _ ...---1

-i I 71 TI'I:L.1-1 I

i ! I . ,

'._..-._7. Dr_l__T17.19

, _ I 1 _ I_____,,..1. _..L...,_

,--i" a

I 1--;--t -1- --1'-'' .---i----- 1 1121 -1-1- Er -'

...I. _..1._:___1____:_,_.__; i.,_ : ____;1_4_ -, r-1! , ±_,

-77 I 1_ 1_4 .i Ll...1 1 . .

_. . . - i ' 1- 1 . . 4._ ,_.: _t_i -I_ -1,...7,.... ""-

, , : . .--_-.----- - __Lill i_ _Ili_ _ ._. .: l_r___, j>.,...._7_,....1_4_1+,

1 I 001 14.1 It .. -. ' .-...-- __,.... 1 1 11iT --i ---'' 1 1. l'li1 '

-_- -.....:.,.0.,,, ,..,, ...., I ,,_, .1_4_4j....

1 -r-ri--1-1-7--E 7 I-1111 j1 ii , , I

l'":- ,--1.--I-7-7- I- . -1,--4, -1- _J--1--:---t -r-T----1--- -,---1--, 1l'!It,-, k .,II,.

i ..._ 1 __I _ ,------- , . _ _ . i

i I I

fr_i_:,_.__;___.:......, l_j_,4 ...i. i !-

...;_i____:,._:_i__, _1 ____-____i____,

1 0 2.5 :'.$ .!-C :. 5 40Ctralete'_,I.yo C?A

LSAT ***** a c-4 far .!'-t t41 te1e a .4 . fonettot of CTA./ (per ring lGOni-L:.A.1' AC I, e i : -r ....

. foot ce4-.J c:.res Cot ar,.."...1.44 ...alt...t -c ..,^0.4 .1 V.#1 tebtletwo lawn

1.S

145

Mat $00

LSAT&coos&

200

400

35;)

200

d\Second400

LSAT&cores

200

GOO

$00

e1 et- 400

300

102

3 0 1.5 2.0 2.5 3 5 4.0

Cumulative CA

LSAt ge score. for 2-tire takers aa a function of CPA.

empertne first-LSAT scores for vhites coached at 002 between !ROCS

gatnst first-LSAT scores for uncoached 4hitea.

3.0

!-------1-7- ;-14-1 ! lip- 1

Illim"ri riN°!

.

--I T , .

III ± ; PPP--T-E-i-............111 es--.4

1

'-'wached

. -7 -1 'ill um_1_

_ _.

,addt--002-

:=1111.116

11011

a p

--.

-7--7

.. iilliiiman11_ milmal

lams

.1ili

am.2.52.0 3.5 4 0

Cumulative CPA

LSAT average 000000 tee 2.th.. takers of lametion 41,1 CTA.

0100$0,TIA seeeed-tSA, serves ler whttee teethed at 042 babes.* e

sides% se-con4.13AT fin unceetted whites

1.0 1.5

146

30

MACLSATSu res

700

400

4400

300

200 II -IL;.

"1

,. 1

--T-i--:-- -;----;--1-------' -1- --..

i_

_.._it..._i........_!...1___:_! -

f . 1 I

:-

-;-1-..- -";---- -.1--. 4

orirosaed

Secoul COO

LSATScore.

700

SOO

. W.4

400

200

10 2.5 3S L0Cuesast I ve CPA

svvaEe scores (or 7-tire Ulcers. as function of C/A,

eortartor first-VAT scores for whites couched at 003 between enema

against (fret -LSAT scores for uncloached sisttes

1.5 2.0

LSAT

3 0

1 0 1.5 2.0 2.5 3 S 4.0

Cimltiv. CrA

LSAT Ve101. re... for 1-11.. takers e (unction of CrA,

coortp.6 ecrod-LIAT Zvi,* for white. earthed at 003 between exams

tsInt weeendLSAT steles for uncostl.ed %bits1 6.3

30

147

$OO

LSATStores

700

...1a. A.

'Co

200

Second' P00

LSATScores

100

. 400

200

1 0 1.5 2.0 2.5 30 3$ £0Cumulative CPA

2SAT evetece .toree for 2-ttme tia-ers so function of CPA,

compartnr. firet-1..\T ecores for .4,fte coached at 004 between exams

against fir6t-LSAT scores for vr40ached weatell

I

1-- --1--i '-. : 1'_

1

: _s_

-T.F14-1-11121- --I-1-1 untocf,ed whites

77...

3.0 2.0

Jr.

2.$ 3$ 10Cr*

LSAT average scores Per 2.ttfte tele,. is (unction et C,A,compering ees..J-LT score. for %%Iles coached at 004 between emotegatatt acondASAT scores toe wnteachad %Mine

3 0

148

166 .

rirfa'ISATStores

200

700

NA-

secood

Scores

333

200

00

700

GOO

a400

300

srlf.,H.; 1_ ,

_ .. ....1...1._:- r r!"1" : .

. 4- f,-- ,--.-- ---i-4 I-

1 ;

--1-1-tmc..04.d.whikes

7.wrates in.otber sctssils

F., r.

F-1-77.7.1f 1

i-47,

0 2.5 3 0 3 S 4 C

Cumulative CIA

LSAT Averste scores for 2-time takers as function of CPA,

comporlPt Tf:st-1.;;J scores for +elites coached oL other schools between examsagainst first -LSAT scores for uncoochod whites

7.5 :.0

2.0 3.5 - 2.0 2.5 3.0 3 5 4.0

Cumulative Cr*

67 LSAT ,core, inr 2-tiro takers function ef Crik,

tomn!kring second-1.NA? scar., for whites coached et other schools between sane

egoicet olicon4.1.SAT score. (or hecoached ehites

149

(8) LSAT first and second scor:s for 2-tine rakers coachedbetween exams

This section displays second-LSAT scores as a tqnction of

first-LSAT scores and compares average LSAT scores by coaching

school for coached aLd 61coached students.

All the major coaching schools UK (001), Amity (002),

Sexton's (003), and Evergreen (004) -- appear to benefit pri-

marily students with initially low LSAT scores. The composite

of small schools appears to offer fairly uniform benefits.

The influence, if any, of self-selection is somewhat obscure.

If self-selection were dominant, then we would expect to see

greater gains for higher initial scores for SHK (001), Sexton's

(003) and Evergreen (004). Instead, the average benefits are

larger for lower initial LSAT scoring students. Similarly, the

composite of small schools shows an almost uniform average benefit

while self-selection would imply a greater benefit for lower initial

scoring students. Only the pattern for Amity (002) is consistent

with the theory of self-selection.

This section is an alternative method of demonstrating

score increases for 2-time LSAT takers to the preceding

section. Here the first LSAT score is the independent vari-

able and the second LSAT score is dependent on it. In

the preceding section the GPA is the independent variable

and both LSAT scores are dependent variables.

150

The following table shows average conditional (theoreti-

cal) score changes assuming an initial score of 200 or 800

for the coaching schools and for thb control (uncoached) group.

200 800-49VO16- +12

002 +133 -5

003+91 -11

004 .+92 -28

Others +81 +21

Uncoached +63 +9

As this table indicates all the coaching schools increase

scores at the lower bound with Amity (002) improving scores

on the average by 70 points above the control group and by a

gross increase of 133 points. The least effective organiza-

tions at the 200 initial score level are the composite group of

small schools that show a net increase of 18 points and a gross

increase of 81 points.

At the 800 level of initial LSAT score coaching would

appear to serve little benefit'or in some instances be a

slight detriment. Of course 'ur concerns are not at this

highest level of initial score because individuals receiving

the maximum score on their first attempt neither retake the

examination nor enroll in a coaching school.

Second ,,..ppLSATSeems 1

'MOOLSAT2=LSAT1(.86+ 124

foal

' N=155.11.10

1

1

1

1194.es

sasses

601.01

3

3

I

487.70

101.0011,

331.00

Second8

LSATScores

I

a

e

.. . ......

S.OS

00

2 es 8

.

. .

110.CO )47.30 364.40 41.10 451.10 414.513 S33.30 311.00

LSAT second score. for 2 time takersas fwIctim of first scores for whitescoached at 001 between LSAT eras'

006.40 443.70 ess.io

first LSAT Scores

300 100 700 100

'trot LSAT Sc

LSAT ...... second scours for 2.ttewe tolurr

as fenctlon of fleet scores for white

coached at U01 between ItAT uses'.

Compered with uncoacbed Lit 2.tiews takers1 7 o

152

SecondLSATScores

I.... ..... 000000000lt1.0

LSAT2=LSATL(.77)+ 179

N=94

II

$11.114

511.11

411.4tO

431.10

211.28

1111.110

..

1

1

111

1

1

I

a

11

1

1

1

1

1

1

Second aoo

LSAT

700

400

SOO

400

200

333.43 115.00 131.20 313.40 433.00..

443.40

LSAT second scores for 2-ttoe taker.as function of first core for whitescoached at 002 between LSAT exam.

sir... 100.30 312.0 533.23

large LSAT Scores

' 1

....

.. r 1 1.

-1

I 1.- I

T riEn1111""', i

1 1

-I -Fi1 1 1.........4 i.,_

! l

.1. ..,......,

I L....... / ,i !..1._, I....

; . I I I

I

_,1 I I I '

I

I lelI

. ..,!!...11 LI I i i .

i 1 6,I

.,!

1_ _

,

', ,

_

. : ; 1

-i-1?:.

1

, . . 1

.

.4.1.,_LIZ:vr.

.,....

r1

!.:.:..

,.

,

.

, , WIN.

i NENPActiscLuh

;4._

I ; NM MIREz 1 -1 t Ii __L_t

1

gill+I.---E- Ism.

IlwANI

...

203 3G0 400 SOO

LSAT ge prowl score. for 2-tto taker,

Imoctifet of (trot scot. for whitesteethed et GO: between INAT wee.tempered wth uncosehd wills tebere

1 7 153

Ileac LSAT S

:7

ftee. o ###

Second /aAsLSATS I

"" .LSAT2=LSAT1

$15.111

N=139ma.

1I-II

ItS.ss

1

$211.141

1

1

1

SILOS1

1

1

1

diSr.ss1

1

1

Cu...1

40,0I

111.SI

I

HIM330.05

+ 125( .

421.10

0

412.00

.

4111.5 511.0

a

se

.541.1111

.....

034.0

a

404.50

O.11 01

I

1

1

1

1

1

I

1

1

1

1

1

1

11

1

1

1

11.110

.0.1

oo310.10 MAO

StrandLSATScores

600

700

COO

SOO

£00

300

..

200

LSAT second .,ort for 2 -tine takersasA funct4un of first scores for white.coached at 003 between LSAT

r-

1*-

1 ---1 1 -Ty 1 _

1.

-1 i_ii,:.

.

.

1_ _ _...

7- .

J......... .

:-. _. .. 1

I 1 '

.- _.1...)...1..1

incoachertA4iitias.. ;

-H_I . . i . sum

I- ---1. Elm__!. i Al.go

_ T s/...2. rinir

IIIn

l7--- t;TfitrsiLin

: z.i_ 1

/, I LL 1

i

1 Irmo____

1 1

-+-/ --r- 1 1 1 U1111

IFRan

.1 1

71rnaRIM

E91 ail -1

r.:PS11:1 : .I

200 . 300 400 500 50

LSAT avrss icnd tee 1-ti. taker

ad function of first 00000 for Milts*

leeched at 003 between LSAT vases,

compared with uncosched whits 2.116.4 come.

154

1 1ov

first LSAT Scores

first LSAT S

SoundLSATSews,

611.1

I

LSAT2=LSAT1 ( .14o )+ 132

Whet

414.61

I

4111.011

Iv

424.6

321.11

1.

354.01.1

Second "0LSAT&core

'Co

600

SOO

400

200200 300 400 500 IWO TOO 100

First LSAT ScoresLSAT go 'wand o for 2411.6 takersas function of first score. for whitesCoached at 000 btwon LSAT s.a..

' sempArd with wmcoached whit. 2.4i.. takers

-

so

a

o

11.............

r - -

105.25 152.40 371.40 414.50 ASt.CO 411.20 !34.50 345.110 103.51 453 00

Pleat LSAT ScoresLSAT second scores for 2-ttos takersso a function of first scores for whitescoached at 004 b LSAT 'macs

4- 1-..!_i_.. A

I

_LI.-

i7

-r......I

1

1 'IT;I-1-_)'.

!1:1-I_

impschreithiti.i, "'

____

'

_.-7-1-1

. !

aili

i___

1

MT

_il-1I I.

-.1--I

I

, /41111AinA1111111AM

. , 1 1

__i

--T-L.11.1.:

, ___ .

....,

;T:b....--

I

iI

,

' 1

, .1-

111

IIIT: 1

vLd_LI

--1

-1-

,_/_:__,e.,_

1

.,,: 1

"L7,

-TrlinI.

sli =nom_ . ...: .N...:

17 #3a.

SecondLSATScroces

.....................................ells./1

I

1II

III

1 . 1

454.9e LSAT2=LSAT1 ( . 90bt' + 101

: N=45

1

1

1

WAS11

1

1

135.151

11

1

1101 AO1

1

44.11

11

431.:1111

1

511.1S1.1

1

1M% 1

;MoO .. . .. ...

MAIO 117./0 107.04 417.00 40.4

11

1

1

1

11

1

1

1

11

11

1

1

1

1

111

1

1111

070.0 05I.li 010.40 044.00 106.10 12s.00Mist LSAT Scores

ScondLSATScores

v...

LSAT second sccres for 2-tin takersas function of first scores for whitecoached at returning schools between LSAT exams

S.

6120

)00

$00

SOO

too

300

4%.

IL%

200200 300 400 $00 $00 200 SO

met LSAT Seers

_

-r 1 1--r-I-T-1

! il-,-,

.1 1 1

i--f- 1

. .

' 71-..../1. I_:. .1..

--i

....) +. 7;L

.

__Li1 1

..,,ii td xi ..0t1 er.schcol

Iri:'unto

.1.1 5ti.!,; ,-. -Illchod uhrt

1 1 1 ; 1 / MIMII i L I /I !

-

131Er I, 1 1. I . . Z Ti . . WI:tr. 1 v.: *:-1.resteven. line

! 1 ' 1 1 1 IIImg II

....._y

,!: r I , .

. .

_III-

;111141

3, . ; 1 Mifflin"'4RIMRN

71,1--1+ !

II

III!II MIIIII-1111RE 11111111111:11

/SAY second for 2 -[tee takers

as function of first SCOgt0 for Matescoached at other schools between LSAT *silos.

tooested with uncoached vl.ite 2.11.. takers

156

1 74

.There is no one easy generalization that characterizes

the effects of coaching for all the analyses we have dis-

played. Furthermore the analyses exhibited in this document

are only one segment of the.approaches we have undertaken.

As we have already indicated, exploration of the technical

appendices is mandatory for total comprehension although this

docuient can and does stand alone.

The results of our analyses reveal that coaching is

dramatically effective for the SAT. The LSAT, although

susceptible to coaching in a general sense, does show areas

where the effects are marginal. As we have indicted we be-

liev part of the apearent lack of susceptibility of the LSAT

to coaching is related to abnormally large control group in-

creases and the relatively low correlation between GPA and

LSAT scores. Furthermore the lack of consistency of GPA's

both between and within undergraduate universities may be a

contributing factor. Additionally we believe the gross

score increases found when analyzing 2-time LSAT takers

and using the first LSAT as the independent variable may be

applicable to 1-time takers coached before the first LSAT.

This would greatly increase the benefits received from LSAT

coaching. We hope to look into some of these areas in the

studies recommended, infra.

157

175

G. CASE HISTORIES

The case histories that follow are drawn from Technical

Appendix M (Miscellaneous). Appendix M contains relevant data

concerning all individuals who were either coached more than

once or who took, the LSAT more than twice. Technical Appendix

M reflects testing histories of some individuals who have

taken-the LSAT five or more times and those of some individuals

who have enrolled in multiple coaching courses, occasionally

simultaneously.

Technical Appendix M is composed of the testing histories

of those individuals who have probably suffered the greatest

psychological trauma from the law school admission process.

They have in some instances spent over $1000 in application

fees over a three or four year period only to be rejected

by every law school to which they applied.

The probability of acceptance to law school appears to

diminish with the length of time one remains in the "admission

process". Therefore any acceptances to law schools of

individuals whose data appears in Technical Appendix M

are not only surprising but emotionally gratifying.

The case histories that follow all have a portion of

happiness as all the individuals ultimately were admittee to

law school. They all were admitted after they had been

rejected, gone to a coaching course, and increased their LSAT

scores. Our study demonstrates that coaching is responsible

I71158

at least in. part if not in full for the ultimate acceptance of

all these individuals to law school. However, the fact that

these individuals waited until they were into the admission

process to enroll in a coaching course, rather than having

taken a coaching course at the outset, may have caused them

undue anguish and expense. They appear to have enrolled

in a coaching course primarily out of desperation.

The case histories that follow ?erhips articulate better

than any other mode, although in anecdotal form, the persorcl

costs and benefits of coaching.

STUDENT A

Student A, a history major at Bryn Mawr College, with a

GPA of 2.10, took the LSAT in December 1974 receiving a score

of 531, and in February 1975 receiving a score of 565.

Student A applied to New York University Law School, Columbia

University Law School, Fordham Law School, Rutgers Law

School, Brooklyn Law School, Hofstra Law School, and St.

John's Law School, for the class commencing in the fall

of 1975. Student A was rejected by all these law schools.

Student A then enrolled at the Stanley H. Kaplan Edu-

cational Center in New York for the course prior to the July

1975 LSAT. _After being coached Student A took the July 1975

LSAT and received a score of 665.

Student A then applied to Columbia Law School, New York

University Law School, Fordham Law School, Rutgers Law

159

177

School, Brooklyn School, Hofstra Law Schc31, and St.

John's Law School. Student A we's placed on the waiting list'at

Rutgers Law School and was adMitted to bc1-. Brooklyn Law

School and St. John's Law School. Student A registered at St.

John's Law School for the class commencing in the fall of 1976.

STUDENT B

Student B, a major in social work at use University,

with a GPA of 2.27, took the LSAT in July 1973 scoring 518

and in December 1974 f..ng 511. Student B then applied

to Fordham Law School ,r;z: the class enterins in the fall

of 1975 and was rejected.

Student 3 then enrolled in John Sexton's LSAT Preparation

Center prior to the October 1975 LSAT and received a score of

616 on the October 1975 LSAT.

Student B then applied to Fordham Law School, Temple Law

School, American University Law School, and George Washington

University Law School.

Student B was admitted to Fordham, Temple, and American

University Law Schools and enrolled at Fordham Law School

for the class commencing in the fall of 1976.

STUDENT C

Student C, a history major at McGill University, with

no reported GPA, took John Sexton's LSAT Preparation Center

course in New York prior to the July 1975 LSAT but did not

take the LSAT until October 1975. Student C scored a 480 on

160

78

the October 1975 LSAT and a 516 on the February 1976 LSAT.

Student C Aid not apply to any law schools for the class

entering in the fall of 1976.

Student C was coached at the Stanley H. Kaplan Educational

Center in New York prior to the December 1976 LSAT and

received a score of 632 on that exam. Student C then applied

to 18 law schools for admission to the class commencing in the

fall of 1377.

Student C was accepted at Fordham Law School, Syracuse

Law School and at the University of Pittsburgh Law School where

Student C enrolled.

STUDENT D

Student D was a psychology major at the John Jay College

of Criminal Justice with a GPA of 2.88. Student D enrolled in

John Sexton's LSAT Preparation Center course for the December

1975 LSI.T but did not take the LSAT until February 1976.

Student D received a 437 on that exam.

Student D applied to Brooklyn Law School, New York Law

School, New York University Law School, St. John's Law School,

Fordham Law Sc; and Pace University Law School for the

class entering in the fall of 1976. Student D received no

acceptances.

Student D then enrolled simultaneously at John Sexton's

LSAT Preparation Center and the Law Boards Institute prior

to the July 1976 LSAT. Student D scored 581 on the July

161

1 79

1976 SAT. Student D reapplied to New York Law School, St.

John's Law School and Pace Law School. .Student D was admitted

to Pace for the class commencing in the fall of 1977.

STUDENT E

Student E, a political science major at Wellesley College,

with a GPA of 2.82, took the LSAT in October 1974 and December

1974 receiving scores of 402 and 424 respectively. Student E

applied to 8 law schools for admission to the class commencing

in the fall of 1975 and was rejected by all. Two of the

schools where Student E was rejected were Georgetown

University Law School and New York University Law School.

Student E enrolled in John Sexton's LSAT Preparation

Center course prior to the December 1975 LSAT and receivr,d a

score of 503 on the December 1975 LSAT. Student E then applied

to 8 law schools including New York University and again was

rejected by all 8.

Student E enrolled in the Stanley H. Kaplan Educatio.111

Center prior to the October 1976 LSAT. Student E to tit*

December 1976 LSAT and February 1977 LSAT scoring 529 crud 527

respectively. Student E then applied to 13 law schools.

Student E was accepted at Rutgers Law School,

University Law School, and New York University Law School f r

the class commencing in the fall of 1977. Student E

registered at New York University Law School.

162

I 80

STUDENT F

Student F, an english major at Manhattan College, with a

GPA of 3.26, took the LSAT in December 1974 and February 1975

scoring 512 and 518 respectively. Student F then applied to

5 law schools for admission to the class commencing in the

fall of 1975 and received no acceptances.

Student F retook the LSAT in December 1975 receiving a

score of 530 but made no applications to law school f-,r the

class entering in the fall of 1976.

Student F enrolled in John Sexton's LSAT Preparation

Center course prior to the October 1976 LSAT. Student F scored

582 on that exam. Student F then appliel to 5 law schools for

admission to the class commencing in the fall.of 1977 and was

accepted at Pace University Law School.

STUDENT G

Student G, an economics major at Queens College, with a

GPA of 3.18, took the LSAT in April 1975 and scored 536.

Student G applied to St. John's University Law School and

was not accepted.

Student G then enrolled at John Sexton's LSAT

')reparation Center prior to the July 1976 LSAT but did not

tako the exam. Student G next enrolled in the Stanley H.

Kaplaa Educational Center prior, to the December 1976 LSAT

whtct, dent G scored 649. Student G thenapplied

to 8 law scnools ±or the class commencing in the fall of

163

1977. Student G was accepted at 7 of these law schools

and enrolled at St, John's University.

STUDENT H

Student H, a psychology major at Brooklyn College, witha

GPA of 3.54, took the LSAT in December 1974, scoring 496.

Student H applied to three law schools for the class

commencing in the fall of 1975 and received no acceptances.

Student H enrolled in the Stanley H. Kaplan Educational

Center prior to the October 1976 LSAT. Student H scored 621

on the October 1976 LSAT and applied to 8 14w schools.

Student H was rejected by 3 of these schools (Hofstra,

Fordham, and New York University). However Student H was

accepted at Brooklyn Law School, New York Law School, St.

John's University Law Soh(' 1, SUNY at Buffalo Law School and

Rutgers Law School.

Student H registered at Rutgers Law School for classes

commencing in the fall of 1977.

STUDENT I

S4-ulent I, a sociology major at Hofstra University, with

a GPA of 2.99, took the LSAT in December 1974 and February 1975

scoring 514 a.d 424 respectively. Student I applied to 8 law

schools for admission to the class commencing in the fall of

1975 and was rejected by al'. One of these schools was New

York Law School.

164

Student I enrolled in the Stanley H. Kaplan Educational

Ceiiter prior to the July 1975 LSAT. Student I took the July

1975 LSAT scoring 549 and the October 1975 LSAT scoring 590.

Student I applied to 2 law schools for the class

commencing in the fall of 1976.- Student I was accepted at New

York Law School.

STUDENT J

Student J, a philosophy major at Harvard University, with

a GPA of 2.98, took the LSAT in July 1974 and December 1974

receiving scones of 560 and 573 respectively.

Student J applied to 7 law schools (New York University,

University of Pennsylvania, Cornell, Harvard, Fordham,

Georgetown, and University of \_rginia) for the class com-

mencing in the fall of 1975. Student J was rejected at all 7.

Student J enrolled in the Stanley H. Kaplan Educational

:n NPw York prior to the October 1975 LSAT. Student

J sk..t.t.1 618 on that exam.

Student. 3 then applied to 15 law schools for the class

commencing in tie fall of 1976. Student J was placed on two

waiting lists (Fordham a;ld Rutgers) and was accepted at The

Benjamin Cardozo School of Law.

STUDENT 1:

Student K, an economics major at Fordham, with a GPA of

3.15, took the LSAT in December 1974 and February 1975 scoring

482 and 518 respectively. Student K applied to 5 law schools

165

i 8 8

for the class ente,!-: ir the fall of 1975 and was rejected by

all 5 including New York Law School and Seton Hall University

Law School.

Student K enrolled in John Sexton's LSAT Preparation

Center prior to taking the December 1975 LSAT. Student K

scored 569 on that exam.

Student K then proceeded to apply to 4 law schools for

admission to the class commencing in the fall of 1976.

Student K was admitted to two of these schools, New York Law

School and Seton Hall University Law School.

STUDENT L

Student L, a social sciences major at Seton Hall Univers-

ity, with a grade point average of 2.79, took the LSAT in April

1973, December 1974, and February 1975. Student L received

scores of 482, 430, and 429 respectively.

Student L made 6 applications for the law schoc_ class

ertering in the fall of 1975. Student L made 7 applications

for the law school class entering in the fall of 1976. All

13 of these applications were rejected.

Prior to the October 1975 LSAT Student L was coached by

the Law SJnool Admission Test Review Co.Irse (Evergreen).

Mysteriously, Student L did not take the LSAT until December

1976, s,.oring 586. Student L applied tc six law schools

for the crass entering in the fall of 1977 and was accepte

at the Internatione School of Law.

166

STUDENT M

Student M, a business major with a 3.42 GPA at Saint

Bernard College (Alabama), took the LSAT in October 1974 and

December 1974 scoring 239 and 286 respectively.

Student M applied to eleven law schools, including

Deleware Law School, for the class entering in the fall of

1975. Student M was rejected by all these law schools.

Student M znrolled in the Evergreen LSAT Review course

prior to the October 1975 LSAT but did not take the LSAT until

December 1976. Student M scored 477 on the December 1976

LSAT. Student M applied only to Deleware Law School for the

class entering in the fall of 1977 and was accepted.

STUDENT N

Student N, a history major at Indiana University with

3 2.68 GPA, took the LSAT in Decembet 1972 and February 197A

scoring 391 and 444 respectively.

Student N applied to Chicago-Kent Col.,ege of Law, John

Marshall Law School, and Depaul University Law School for

admission to the class commencing in the fall of 1975.

Student N was rejected by all three schools.

Prior to the December 1976 LSAT, Student N was coached at

the Amity Review Course in Chicago. Student N scored 534

on the December 1976 LSAT and applied to 5 law schools for

admission to the class commencing in the fall of 1977.

167

18

Although rejected by Depaul and Chicago-Kent, Student

N was accepted at the John Marshall Law School.

STUDENT 0

Student 0, a political science major at Valparaiso

University (Indiana), with a 3.18 GPA, took the LSAT in October

1975 and December 1976 scoring 474 and 424 respectively.

Student 0 applied to Valparaiso University Law School and

Loyola University Law School of Chicago for the class commenc-

ing in the fall of 1976. Student 0 was rejected by both of

these schools.

Student 0 then enrolled at Test Prep in Chicago prior to

the December 1976 LSAT. Student 0 took the February 1977

LSAT scoring 517.

Student 0, although rejected t Valparaiso University Law

School and John Marshall Law School for the class entering in

the fall of 1977, was accepted at Lewis University Law

School, Loyola University Law School (New Orleans), and

Hemline University School of Law.

STUDENT P

Student P, a psychology major at University of California

at Los Angeles (UCLA), with a GPA of 3.30, took the December

1975 LSAT scoring 591.

Student P applied to UCLA Law School and Harvard Law

School and was not admitted by either, for the class com-

168

mencing in the fall of 1976. Student P then simultaneous-

lienrolled in both the Stanley H. Kaplan Educational Center

and the Amity Review Course in Los Angeles prior to the

December 1976 LSAT.

Student P scored 712 on the December 1976 LSAT and

applied to 10 law schools for admission to the class commenc-

ing in the fall of 1977. Although again rejected at UCLA and

Harvard, Student P was accepted at the University of San Diego

Law School and the University of California at Davis where

Student P enrolled.

STUDENT Q

Student Q, a biology major at UCLA with a GPA of 3.08,

.00k the LSAT in July 1974 and December 1974, scoring 438 and

479 respectively.

Student Q applied to 6 law schools for admission to the

class commencing in the fall of 1975. Student Q was rejected

by all 6 schools.

Student Q enrolled in the Law School Admissior. Test

Review Course (Evergreen) in Los Angeles prior to the October

1975 LSAT. Student Q scored 529 on the October 1975 LSAT.

Student Q then applied to 3 law schools for admission to

the class commencing in the fall of 1976. Although rejected

at Loyola University (Los An,e1,--1, Student Q was accepted at

both Southwestern University Law School, and Whittier Colleg,3

- Beverly School cf Law.

169

8

IX. ACTS AND PRACTICES OF COACHING SCHOOLS

All coaching schools make claims or have made claims

regarding their ability to improve scores on standardized

admission examinations.. Some claims are blatant, others

are subtle. However, they all have the same purpose -- to

convey to the student (or parents of students) that taking

a coaching course will increase an individual's score on an

examination that may well be the mwIt important examination

that person ever takes.

These claims appear in national newspapers of general

circulation, in college and high school newspapers, 66/ in

posters hanging on campus bulletin bo7-rds, in radio advertise-

ments,

mailed

places

in brochures distributed to students and parents,

circulars,

as bowling

in speeches, and in such other diverse

alley score sheets.67/

in

A representative sample of claims made by the coaching

schools, and the coaching school making each claim fo".dows:

thousands have earned increases of a hundred points or

more in their scores 68/

Stanley H. Kaplan Educational Center

66/ See generally, Category III File 772 3000, Vol. 111,

p. 51-61, for a sampling of newspaper advertisements.

67/ ''"egory III File 762 3033, Vol. IV, p.50.

68/ dory III File 762 3033, Vol'. I, p.l.

170

88

most effective course available

the resulting competence and confidence will allow you

to perform at your maximum potential 69/

gtanley H. Kaplan Educational Center

from the reports we have received we have found there

is an average increase of 50 to 100 points for testrepeaters 70/

Stanley H. Kaplan Educational Center

...achieved such dramatic results that many nationalmagazines have written articles about our success 71/

Stanley H. Kaplan Educational Center

we teach twice as many students as any other course in

the metropolitan New Yorkt,ew Jersey area 72/

John Sexton's LSAT Preparation Center

The LSAT Preparation Center is the only institutiondevoting its attention exclusively to the LSAT.73/

John Sexton's LSAT Preparation Center

the scores of many of our students have jumped 250 points

or more 74/

John Sexton's LSAT Preparation Center

69/ Category III File 762 3033, Vol. I, p.9.

70/ Category III File 762 3033, Vol. I of Investigational

Hearing Exhibits, Exhibit 5. This statement was made

in a speech by a representative of the Stanley H. Kaplan

Educational Centers, Ltd. to the Black American Law

Students Association of George Washington University.

71/ Category III File 762 3033, Vol. IV, p.42.

72/ Category III File 772 3000, Vol. XV, p. 21.

73/ Category III File 772 3000, John Sexton Physical

Exhibit C, p.4,

74/ Id. at 12.

171

189

the average score of its students greatly exceeds thenational. median 75/

John Sexton's LSAT Preparation Center

the average score of our students is one-third higherthan the national average 76/

John Sexton's LSAT Preparation Center

the median increase for our students who have previousscores is over 100 points 77/

John Sexton's LSAT Preparation Center

no other course can demonstrate better results thanours 78/

Evergreen LSAT Review Course

thousands of successful students prove we significantlyincrease scores 79/

Evergreen LSAT Review Course

since 1968 we've hei NI over 10,000 students gain admis-sion to law schoo.1

.rgreen LSAT Review Course

October LSAT average for those taking the course was630 81/

LSAT Review Course of Massachusetts

75/ Id. at 13.

76/ Id.

T7/ Id.

78/ Category III File 772 3000, Vol. XLIX, p.131.

79/ Id. at 132.

80/ Id. at 134.

81/ Category III File 772 3000, Vol. LIII, p.37.

172; 1) 0

114.1Gam

2 8 2.5

12.2

LMr

120

118

1.25 11111.4 1111-11.-

MICflOCOPY VA ioN I AR

Follow-ups show that our average student increases his

LSAT score over 80 points with some achieving a nearly

200 point increase, demonstrating the course's capacity

for enabling you to maximize your score within your

ability 82/

Law Board Review Course

...it will add crucial points to your test scoresTHE POINTS THAT MAY WELL GET YOU OVER THE THRESHOLD

OF SUCCESS 83/

Columbia Test Preparation Institute

the average increase of students who take our course is

80 points with some improving as much as 150 points 84/

Rutgers Review Course

...if an average improvement of 30 points results mere-

ly from having taken the test before, a still greater

score advantage would be expected to follow an effe::tive

instructional program... 85/

Amity LSAT Review

...the LSAT METHOD can help you make your first LSAT

score your best score.86/

LSAT Method

the Seminar equips the

taking skills essentialthe LSAT 87/

law school candidate test

for maximizing performance on

States ReviewWestern

82/ Category III File 772 3000, Vol. XXVII, p.115.

83/ Category III File 772 3000, Vol. XLVI, p.317.

84/ Category III File 772 3000, Vol. XXVI, p.16.

85/ Category TII File 772 3000, Vol. XXXII, p.117.

86/ Category III File 772 3000, Vol. LIV, p.42.

87/ Category III File 772 3000, Vol. XXXIV, p.99.

173

Even if the coaching schools made no advertising claims

their corporate names are implicit if not-explicit representa-

tions that they serve some affirmative benefit in preparation

for standardized admission examinations.

The use of the words "coach", "cram", "review", "educate",

"prepare', and words of similar import convey to the person

perceiving them that test takers will achieve some affirmative

benefit by attending the course. Thi.s is corroborated by con-

sumer testimony, the testimony of coaching school proprietors,

and the perceptions of guidance counsellors and educators.

The coaching schools' claims are not limited to effective-

ness in raising scores. The schools claim to have superior

curricula, 88/ small class sizes, 89/ experienced faculty whr

have scored high on standardized examinations, 90/ and exten-

sive classroom and homework materials that are continuously

updated to match the most recent standardized examinations.91/

These claims are of course all ancillary to the claim of

effectiveness in increasing score.

88/ Category III File 772 3000, Vol. XLIX, p.131.

89/ Category III File 772 3000, Vol. XXXII, p.99.

90/ Category III File 772 3000,Exhibit-C, p.8, 16-17.

John Sexton Physical

91/ Category III File 762 3033, Vol. I, p.l.

174