PERSONALITY FACTORS IN MARITAL COMPATIBILITY: II. · 2020. 4. 2. · PERSONALITY FACTORS IN MARITAL...

PERSONALITY FACTORS IN MARITALCOMPATIBILITY: II.

From the Department of Psychology, Stanford University

LEWIS M . TERMAN AND PAUL BUTTENWIESER

RELATION OF MARITAL HAPPINESS TO HUSBAND-WIFE

RESEMBLANCES IN TRAIT-SCORES

Is the mere fact of resemblance or difference between spouses inthe traits measured favorable or unfavorable to marital compatibility?There is a popular opinion that unlike or even directly opposed typesof personality are attracted to each other and that such matings arefavorable to marital happiness. However, investigations of maritalselection have usually revealed positive rather than negative correla-tions between spouses in such traits as height or weight, intelligence,education, and social-economic status. Less information exists asto the direction of marital selection, if any occurs, in the case ofqualities related to personality.

We have computed husband-wife correlations on the thirteen per-sonality variables and on schooling for 341 married couples, and for109 divorced couples. Our interest, of course, is not primarily todiscover the extent to which marital selection in these traits has takenplace, important as this is, but to find out whether the selection thatoccurs is in any way related to marital success. We have accord-ingly divided the married subjects into two groups, one composedof 126 couples with highest happiness scores, the other composed of215 couples with the lowest scores. Table 16 gives the husband-wifecorrelations for these groups separately and for the divorced group.

It is evident from this table that the characteristics of personalitymeasured by these tests have played little part in marital selectionin any of the groups. The moderately high correlations for schoolingare about of the order of magnitude found in other investigations,but these too fail to show significant differences between the groups.Of the thirteen test variables, in only one (interest score for "Y. M.C. A. Director") is the correlation for the most happily marriedgroup significantly higher than for both of the other groups. The

267

268 L. M. TERMAN AND P. BUTTENWEISER

TABLE 16HU9BAND-WLFE CORRELATIONS ON PERSONALITY VARIABLES AND ON

SCHOOLING

(Unmatched groups)

Married couples,happiness scores8 or 9 (high)

N = 126

Married couples.happiness scores

1 to 7 (low)N=21S

Divorced couples

N=109

C. p. A.ChemistLawyerLife insurance salesmanTeacherY. M. C. A.Office clerkInterest maturityM. F.Bl-NB2-SB3-IB4-DSchooling

.2O2±.O58

.125±:.0S9

.424-±.O49

.289±.05S

.242±.O57

.358±.O53

.410±.050

.300±.05S

.029±.060

.108±.060.074±.060.020±.06O.241 ±.057.446±.O48

.037±.046

.076±.046

.167±.04S

.244±.O43

.295±.O42

.133±.O45

.294±.O42

.256±.O43

.O8S±.O46

.217±.O44

.121±.O45

.165 ±.045

.289±.O42

.487±.O35

.O26±.O65

.146±.O63

.345±.O57

.304±.058

.231±.O61

.115±.O64

.226±.O61

.322±.O58—.012±.065

.044±.065

.059±.064

.072±.064

.O16±.O65

.485±.O49

correlations for C. P. A. show a similar difference but smaller. Inthe case of "dominance," both married groups yield significantlyhigher correlations than the divorced group. In "lawyer" interestboth the divorced group and the more happily married group show-higher correlations than the less happily married, a finding so in-consistent that one is inclined to impute it to chance. The onlyother group difference approaching statistical significance is that be-tween the less happily married and the divorced in "neurotic ten-dency," the former showing the greater resemblance.

Of the 39 test correlations for the three groups, all but one arepositive, the exception being —.012±.065. Despite the smallness ofmost of the correlations, the evidence points consistently to theconclusion that such selection as takes place in these traits is in thedirection of "like" rather than "unlike" matings, unless indeed theslight tendency of husbands and wives to resemble one another isthe result of their having lived together. If the latter explanationwere valid, divorced couples should yield lower correlations than themarried groups, which is not the case to any very significant extent.

There is some evidence of marital selection in the case of the fol-lowing variables: "Lawyer," "Life Insurance Salesman," "Teacher,"

PERSONALITY FACTORS IN MARRIAGE: II 269

"Office Clerk," and "Interest Maturity." Since "interest maturity"is largely a function of age, and since the husband-wife correlationfor age is high, the correlations for "interest maturity" are to someextent spuriously high.

Table 17 gives the correlations between each spouse's trait scores

TABLE 17CORRELATION OF TRAIT SCORES WITH MARITAL HAPPINESS SCORES, SEPARATELY

FOR HUSBANDS AND WIVES(Married group, 341 couples)

Wife's Husband'sHusband's trait happiness happiness Wife's trait

scores on: score score scores on:

Husband's Wife'shappiness happiness

score scoreC. P. A.ChemistLawyerLife insurance

salesmanTeacherY. M. C. A.Office clerkInterest maturityM. F.Bl-NB2-SB3-IB4-DSchooling

.071

.084

.014

.123

.198

.086

.020

.137

.085

.130

.019

.156

.044J15

.070

.130

.086

—.098.125

—.007—.088

.147- . 0 3 1—222

.042—3H

.137

.099

C. p. A.ChemistLawyerLife insurance

salesmanTeacherY. M. C. A.Office clerkInterest maturityM. F.Bl-NB2-SB3-IB4-DSchooling

.025

.069

.062

.019

.029

.012

.074

.041

.047

.140

.059

.123

.124

.016

.008

.073

.053

.009

.021- . 0 2 8—.019

.020

.031—.144—.112—.142

.072

.042

and the happiness scores of husband and of wife taken separately.The data in this table have a more specific bearing on the issueunder consideration than those of the preceding table.

The probable errors of the correlations in Table 17 are low, rang-ing from .034 to .037 and giving statistical significance to a numberof the coefficients. Correlations of .10 or more in the table havebeen italicized as being at least suggestive of a trend. Only threeare as high as .20: husband's "teacher" interest with wife's happiness(+.198) ; husband's "neurotic tendency" with his own happiness(—.222); husband's "introversion" with his own happiness (—.244).These reduce to two, as "introversion" and "neurotic tendency" arepractically one trait. Of the 28 correlations of husband's trait scores,only twelve are as high as .10; of the wife's, only six. No traitscore of the wife correlates more than .144 with the happiness score


of either the hushand or wife. "Teacher" interest on the part ofthe hushand has the distinction of correlating more highly with thewife's happiness than any other trait in the list. The hushand's"neurotic tendency" (or "introversion") has less effect upon hiswife's happiness than upon his own! This trait in the wife affectsahout equally her own happiness and her hushand's, hut neither tcany considerahle extent.

Contrary to popular opinion, the wife's "dominance" score corre-lates positively instead of negatively with her hushand's happiness."Dominance" in the hushand correlates positively with his ownhappiness hut practically zero with his wife's. The hushand's "inter-est maturity" correlates positively, though only slightly, with hoth hiswife's happiness and his own, hut the wife's score in this trait isnot significantly related to either her hushand's happiness or her own.The wife's "self sufficiency" score correlates negatively with her ownhappiness hut not significantly with her hushand's. The wife's school-ing has no hearing on marital happiness, hut hoth correlations ofhushand's schooling w'ith happiness score are positive and approachsignificance.

Only two of the correlations are high enough to have appreciahlevalue in predicting marital happiness: the positive correlation hetweenhushand's "teacher" interest and his wife's happiness, and the negativecorrelation hetween hushand's "neurotic tendency" (or "introver-sion") and his own happiness. Prediction could douhtless he improvedto some extent hy the multiple correlation technique, hut the zero-order coefficients are so low as hardly to justify the lahor that thiswould involve.

One possihility remains of finding predictive values in the scoresof these variahles; namely, from an examination of the difference-scores hetween hushand and wife. Even though the ahsolute scoreof neither hushand nor wife on a given variable is correlated signifi-cantly with the happiness score of either, it is conceivahle that thedifference hetween their scores might he, especially if account weretaken of the direction of the difference.

Tahle 18 gives the results of such a treatment of the data. Theraw scores of hoth hushands and wives were transmuted into z-scores,or units of deviation from their respective means. The differenceshetween a hushand's z-score and his wife's z-score, for each trait,were then coded from 1 to 17, a low score indicating that the hus-


TABLE 18CORRELATION OF DIFFERENCE-SCORES IN THE TRAITS WITH MARITAL

HAPPINESS

(341 married couples)

Husband-wifedifference on

C. P. A.Chemi9tLawyerLife insurance salesmanTeacherY. M. C. A.Office clerkInterest maturityMental masculinityBl-NB2-SB3-IB4-DSchooling

Correlation withmarital happiness

—.060—.004—.018

.101—.161—.047

.036—.104

.108

.015—.070

.056

.005

.088

band's z-score is higher than his wife's in a given trait, and that thewife's is lower than her husband's. Such a score signifies that the hus-band, relative to husbands in general, stands higher in the trait inquestion than does his wife relative to wives in general. It does notnecessarily indicate which spouse has the higher raw trait-score. Themedian code-score, 9, signifies that husband and wife have the samez-score in the trait. This might be anything from low to high;e.g., both husband and wife might be extreme introverts or extremeextroverts. The diflerence-scores were then correlated with the aver-age happiness score of husband and wife. In Table 18 positivecorrelation indicates association between high happiness score andwife-superiority to husband in the trait; negative correlation indicatesassociation between high happiness and husband-superiority in thetrait.

Figures 1, 2, and 3 present graphically the low association betweenmarital happiness and diflerence-scores in three of the Bernreutervariables: "neurotic tendency," "self sufficiency," and "dominance."

Only four of the correlations in Table 18 are as high as .10 andonly one ("teacher" interest) is above this figure. Husband-superior-ity to wife in "teacher" interest and "interest maturity" is associatedwith greater marital happiness, as is also wife-superiority (husband-


A j . \

High

Idiv—

Divorcee/

/ i 3 4 S « 7 S 9 11 1/ " '3 If 'S 14

Difference Score

FIGURE 1DISTRIBUTIONS OF HUSBAND-WIFE DIFFERENCE SCORES IN "NEURIWIC T E N D -

ENCY" FOR H, L, AND D GROUPS

(100 couples in each group)

f

f

/

•f—

y

-4f

1 AA/\

~^

V

\A

11

VV/ziA

\ 1

rced^ . .

\\"\

\

/ je 3 .4 e a 7 s 9 /o // /z /3 M is' /6

Difference Score

FIGURE 2DISTRIBUTIONS OF HUSBAND-WIFE DIFFERENCE SCORES IN "SELF SUFFICIENCY'-

FOR H, L, AND D GROUPS

(100 couples in each group)


It

ttt

Itn

to

g

7fS

s£/

V-

// •

'1//

• ' / /

1/V

IAfi

//

-

/ //

/I1

11

\w/ \

i

llf f

t

\\\\

/ •

V\

\

Divorced

,

\\

\

//

'̂ .

\\\ ^

VA\

\\\/8 M /S /f

DifTfrenct Score

FIGURE 3DISTRIBUTIONS OF HUSBAND-WIFE DIFFERENCE SCORES IN "DOMINANCE" FOR

H, L, AND D GROUPS

(100 couple9 in each group)

inferiority) in "life insurance salesman" interest and "mental mas-culinity." The only possibility of improving the predictive value ofthe trait scores would have to lie in the direction of improving theratings on marital happiness. That this is open to improvementmay be freely conceded, but in view of the near-negative results fromcomparison of the scores of the most happily married with those ofthe divorced group (p. 283 iif.), it seems unlikely that a better criterionof happinfess would improve prediction to any considerable extent.

Pending the improvement of current instruments for measuringpersonality, and pending especially the identification and psychologicaldescription of the most important personality variables, it seems neces-sary to look for an alternative approach to our problem. One suchapproach is suggested by the analysis that has been made of theresponses of the H, L, and D groups to the 545 individual items ofthe personality tests used. It is conceivable that a method of pre-diction could be worked out which would be based on the associationfound between marital happiness and the responses to empiricallyselected items. The following section deals with this possibility.

2 7 4 L. M. TERMAN AND P. BUTTENWEISER

HUSBAND-WIFE CORRELATIONS ON THE INDIVIDUAL

ITEMS OF THE STRONG AND BERNREUTER TESTS

Husband-wife correlations (tetrachoric) were computed for eachof the 545 test items, separately for the H, L, and D matched groups.The point of interest was not the absolute amount of husband-wifeagreement or disagreement in the three groups, for a given item, butthe relative amount of correlation yielded by the three groups. Thesearch was for items which would show a significant difference be-tween the correlation for the H group on the one hand and for theL and D groups on the other. The correlation for the H groupmight be either positive or negative, so long as the coefficient differedsufficiently from either or both of those for the L and D groups(preferably, of course, from both).

These tetrachoric coefficients of marital resemblance were com-puted by means of Chesire, SafEr, and Thurstone's diagrams (2).Values calculated from these diagrams agree to two decimal placeswith those computed by means of Pearson's and Lee's tables. Sincethree answers are possible in the case of each item (Yes, ?, No forBernreuter; Like, Indifferent, Dislike for Strong), the distribution ofhusbands' answers by wives' answers yields a 3-by-3 table. Tocompute the tetrachoric r, such a nine-fold table must be reduced tofour-fold form. This reduction may be effected in two ways, (1) bythrowing the " I " and " D " answers into a single category as againstthe "L" answers, and (2) by lumping the "L" and " I " answers asagainst the "D". For the responses of each of the H, L, and Dgroups, tetrachoric r's were computed separately for each of the tworesponse-divisions, and the average of the two r's so found was takenas the value for the item in question.

Table 19 presents all the correlations which showed large enoughdifference between the H group and either the L or D group tosuggest that it was probably not due to chance factors. This is farfrom saying that the L and D divergences from H in the table arecertainly significant. We have not computed the exact degree ofsignificance of the differences for the reason that the determinationof the probable error of a tetrachoric correlation is beyond the boundsof what is practicable in the case of so large a number of coefficients.̂

P.E., = j .674S V(l-r2) [l_(sin-l r)2] 1 y/p q p' q'***• 1 90° J 7"7VN

(Kelley, 1, p. 258.)

PERSONALITY FACTORS IN MARRIAGE : 11 275

B

B

B

B

B

BB

1

4

10

13

17

2636

B 38

B 43

B 47

B 48

B 50B 52

B 53

B 54

B 55

TABLE 19HUSBAND-WIFE CORRELATIONS ON INDIVIDUAL TEST ITEMS FOR

H, L, AND D MATCHED GROUPS

(B=Bernreuter, S=Strong)

Agreementfavors

Great- Aver- happinessest age or un-

H L D Diff. Diff. happiness

Uncomfortabie to bedifferent or uncon-ventionai? .06 —.19 —.

Ever crossed street toavoid meeting aperson? —.10

Easily discouragedwhen opinions ofothers differ fromyour own ? .64

Athletics interest youmore than intellectualaffairs? .42

Affected by praise orblame of manypeople? —.16

Frequently feel grouchy .32Ever solicited funds

for a cause ? .49Conversation more

helpfui in formuiatingideas than reading? .27

Like to bear responsi-biiities alone ? .34

Want someone withyou when you receivebad news? .31 —.16

Bothered to have peopiewatch you at work?

Avoid arguments ?Prefer to do your plan-

ning alone? —.08 —.05Find teiiing others of

your own personalgood news is thegreatest part of theenjoyment of it? —.14 —.

Often feei ionesomewith other peopie? .31

Thrifty and carefulabout making loans? .04

.19

.28

.15

.34

.18

.03

.40

.16

.16

—.21

.25

.25

.04

.17

.40

.20

—.07

.12

.27

.38

.49

.38

.34

.35

.29

.34

.22

.26

.365

.44

.23

.335

.135

.19

.225

.20

H

U

H

H

U

H

H

H

.14 .47 .32

.32 .40 .215

H

.09

.49—.19—.39

—.05—.40

.28

.89.21.885

HH

u

.05

.15

.28

.13

—.12

.36

.27

.43

.32

.18

.295

.28

U

H

U


TABLE 19 {continued)

B

B

B

B

B

B

BB

B

B

B

B

B

B

B

B

60

67

70

75

78

79

8082

84

88

92

95

97

98

100

101

Ever rewrite lettersbefore mailing them ?

People ever come toyou for advice ?

Like attention fromacquaintances whenill?

Play your best evenwhen opponentsuperior?

When in low spiritstry to find someoneto cheer you up ?

Understand problembetter by studying italone than by dis-cussing it?

Lack self confidence?Willing to take

chance alone indoubtful situation?

Usually avoid askingadvice ?

Rather stand whenlate at meetingthan take frontseat?

Ever argue a pointwith respected olderperson ?

"Have it out" with aperson who spreaduntrue rumors aboutyou?

People more stimu-lating to you thananything else?

Prefer a play to adance?

Prefer to be alone attimes of emotionalstress ?

Usually prefer towork with others?

H

—.12

.31

.41

.22

—.08

.00

.38

.40

.22

.18

.38

.30

.13

.81

.50

.21

L

.15

.05

—.12

—.06

.00

.40—.02

.12

.12

—.04

.15

.14

—.11

.35

.03

.11

D

.16

?

.09

—.14

.19

.04

.04

—.07

—.10

—.25

—.05

—.02

—.14

.16

.30

—.06

Great-est

Diff.

.28

?

.53

.36

.27

.40

.40

.47

.32

.43

.43

.32

.27

.65

.47

.27

Agreementfavors

Aver- happinessageDiff.

.275

?

.425

.32

.175

.22

.37

.375

.21

.325

.33

.24

.255

.555

.335

.185

or un-happiness

U

H

H

H

U

UH

H

H

H

H

H

H

H

H

H


TABLE 19 (continued)

B

B

B

B

B

B

B

B

SS

ss-sssssssssssssssss

102

104

112

115

121

122

123

124

6161722

29434550515354

666974767779

868993

100

Usually work betterwhen praised?

Feelings alternate be-tween happiness andsadness withoutreason?

Prefer making hurrieddecisions alone?

Often in a state ofexcitement ?

Like to be with otherpeople a great deal?

Can be optimistic whenothers depressed?

Does discipline makeyou discontented?

Are you usually con-sidered indifferentto opposite sex?

AstronomerBookkeeperBuilding contractorCertified public

accountantDentistInterior decoratorInventorLandscape gardenerLawyer, criminalLibrarianLife insurance

salesmanOrchestra conductorPhysicianPrivate secretaryRancherReal estate salesmanReporter, sporting

pageSecret service manSocial workerSurgeonWorker in Y. M. C.

A., etc.

H

.24

—.08

.20

.20

.25

—.17

—.24

.45—.13

.06—.02

.04

.37

.29—.16—.14

.38—.26

.63—.13

.19

.22

.46

.13

.24—.05

.36

.23

.39

L

—.15

.13

- . 1 8

—.22

—.01

—.12

.01

—.08.06.30.14

.25

.09

.13

.00

.30.13.28

—.10.24

—.04.02.11

—.04

.23

.17—.06

.15

.09

D

—.11

.28

—.13

—.01

—.04

.17

.16

.04

.36

.23

.23

.22—.38—.12

.29

.25—.04

.17

—.14.13

—.03—.15

.19—.24

—.14.39.27

—.20

.20

Great-est

Diff.

.39

.36

.38

.42

.29

.34

.40

.53

.49

.24

.25

.21

.75

.41

.45

.44.42.54

.77

.37

.23

.37

.57

.37

.37

.44

.42

.43

.30

Agreementfavors


.37

.285

.355

.315

.275

.195

.325

.47

.34

.205

.205

.195

.515

.285

.305

.415

.335

.485

.75

.315

.225

.285

.31

.27

.195

.33

.255

.255

.245

or un-happiness

H

U

H

H

H

U

u

HU

uuuHHUuHU

HUHHHH

HUHH

H



ss

ssssssssssssssssssssssssssss

ssss

111116

120121122123124126

127134135136139148157174187197

199206

208212217

221229

234239240241

245

247248250

251

BridgePlaying a musical

instrumentChopping woodAmusement parksPicnicsExcursionsSmokers (party)Conventions (meet-

ings)Full-dress affairsMusical comedySymphony concertsPet canariesSporting pages"Popular Mechanics"ArithmeticLiteraturePublic SpeakingRepairing electrical

wiringOperating machineryInterviewing prospects

in sellingMaking a speechTeaching adultsMeeting new situa-

tionsDoing research workDisplaying merchan-

dise in a storeRegular hours of workSaving moneyContributing to charityRaising money for

charityLooking at a collec-

tion of antiquefurniture

Conservative peopleEnergetic peoplePeople who borrow

thingsQuick-tempered people

H

.66

.40

.16.39.39.31.33

.39

.37

.57

.45

.49

.51

.47

.01

.34—.17

—.02.18

.26—.02

.50

.32.09

.17

.25.04.52

.50

.44

.38

.53

.23

.34

L

.53

—.06—.32

.19.20.09.25

.31

.23

.52

.34

.02

.14

.52.40

—.30.25

—.30—.44

.02.10.08

—.03.33

—.12—.10

.19

.16

.18

—.02.23.02

.16—.11

D

.23

.30—.02

.03.08.28

—.18

.14

.09

.38—.13

.26

.21

.10

.12

.40—.03

—.06.16

—.07—.33—.23

—.06—.23

—.06—.10

.32

.10

.10

.12—.08

.18

—.33.03

Great-est

DiflF.

.43

.46

.48

.36

.31

.22.51

.25

.28

.19

.58

.47

.37

.37

.39

.64

.42

.28

.62

.33.31.73

.38

.32

.29

.35

.28

.42

.40

.46

.46

.51

.56

.45

Agreementfavors


.28

.28

.33

.28

.25

.125

.295

.165

.21

.12

.345

.35

.335

.16

.25

.29

.28

.16

.32

.285.095.575

.365.04

.26

.35

.215.39

.36

.39

.305

.43

.315

.38

or un-happiness

H

HHHHHH

HHHHHHHU?U

HH

H?H

H?

HHUH

H

HHH

HH



H L D

Great-est

Diff.

Agreementfavors

Aver- happinessage or un-Diff. happiness

S 253 PessimistsS 261 Talkative peopleS 263 Irreligious peopleS 264 People who have done

you favorsS 266 Gruff menS 273 Very old peopleS 293 TeetotalersS 296 Men who use perfumeS 299 Develop theory of a

new machine, e.g.,auto

S 302 Determine cost ofoperation of machine —

S 305 Sell machineS 306 Prepare advertising

for machineS 310 Steadiness and perma-

nence of workS 311 Opportunity for promo-

tionS 315 Opportunity to understand

how superior expectswork to he done

S 321 Thomas A. EdisonS 325 J. J. Pershing, soldierS 327 Booth Tarkington,

authorS 328 John Wanamaker,

merchantS 329 (To be) president of

a societyS 339 Street car conductor

vs. motormanS 353 Definite salary vs.

commissionS 357 Selling article quoted

10% above cost vs.10% below cost —

S 362 Great variety vs. sim-ilarity of work

S 365 Technical vs. supervisoryresponsihility

S 371 Reading book vs.going to movies

.29

.06

.07

.40

.48

.44

.48

.42

.12

.30

.06

.17

.37—.02

.41—.02

—.30.22.38

—.11.13.31

—.17—.11

.59

.36

.45

.51

.35

.46

.65

.53

.38

.32

.29

.37

.23

.295

.36

.485

HUU

HHHHH

.16 —.21 .12 .37 .205 H

.22

.30

.24

.27

.21

.42

.44

.44

.10

.48

.25

.40

.43

.07

.35

.29

.30

—.06.04

.16

—.05

.10

.14

.03

.00

.27

.15

.13

—.05

—.10

.08

—.23

.20

.19

.14

.03

—.02

.00

.00

—.02.21.08

.34

.30

—.30

—.06

.03

.30

.08

—.04

—.08

.36.27

.26

.32

.21

.44

.41

.44

.24

.33

.55

.46

.53

.37

.58

.33

.38

.26.265

.17

.295

.16

.36

.32

.40

.205

.255

.335

.455

.465

.26

.425

.21

.245

UH

H

H

H

HHH

U

H

H

H

H

U

H

H

H



s

ss

s

sssss

372

373

377

385

386

387

400402

412

Belonging to many vs.few societies

Few intin:iate friends vs.many acquaintances

Liking jealous peoplevs. liking conceitedpeople

Usually liven up thegroup on a dullday

Am quite sure ofmyself

Accept just criticismwithout getting sore •

Plan my work in detailPut drive into the

organizationLoan money to ac-

quaintances

H

.29

.16

.30

.15

.30

—.26.24

—.09

.30

L

.13

—.16

.04

—.20

.01

—.05.11

.02

—.10

D

—.03

—.13

—.02

—.22

—.16

.53—.11

.15

—.10

Great-est

Diff.

.32

.32

.32

.37

.46

.79.35

.24

.40

Agreementfavors


.24

.305

.29

.36

.375

.50

.24

.175

.40

or un-happine8s

H

H

H

H

H

UH

U

H

It is known, however, that this is considerably larger than the prob-able error of the Pearson r.

It will be noted that Table 19 gives the largest divergence ofL or D group from the H group, and also the average of the L andD divergences. The average divergence is the best single indicationof the value of the item, although similarity of L and D divergenceis also desirable. In the last column of the table H indicates thathusband-wife agreement in response tends to be associated with mari-tal happiness, U that it tends to be associated with marital unhap-piness or divorce (or both).

Table 19 presents an interesting contrast to the data on total scoresin the fact that so many of the individual items are associated signifi-cantly with marital compatibility. The extreme L or D divergencefrom the H correlation is 50 or higher for 22 of the items, and40 or more for 78 items. Even the average of the L and D diver-gences is 40 or more for 15 items, and 30 or more for 56 of them.The probabilities are appreciably against an L-D average divergenceof -30 or more being due to chance. Average L-D divergence in


the range of 20 to 30 carries at least a suggestion of validity for anitem if the L and D coefScients are consistent.

The reader will doubtless find many surprises in Table 19 bothwith respect to degree of association with happiness and the direction(i.e., whether spouse agreement is favorable or unfavorable). Con-sider first the items showing the greatest degree of association. Thehighest ranking item is B 50, re avoiding arguments. It appearsthat among the 545 items the greatest single danger to marital hap-piness is for one spouse to like and the other to dislike to argue.This sounds reasonable enough, but why the second highest itemshould be interest in the occupation of life insurance salesman issomewhat baffling, unless the explanation lies in the particular typeof social extroversion which this kind of salesmanship requires. Itis reasonable that agreement on preference between attending a playand attending a dance (B 98) should be associated with maritalcompatibility, for the preference doubtless connotes a good deal withrespect to one's total personality. Similarly whether one would likeor dislike teaching adults (S 212) might very well token a greatdeal for one's personality and interests. The choice between thework of street-car conductor and street-car motorman (S 339) prob-ably hinges on the relative interest in people and in mechanical things,and likeness between spouses on this point could easily be conduciveto happiness. The same could be said for preference between varietyand sameness of work (S 362). It is notorious that interest in theopposite sex (B 124) frequently brings marital complications whenone spouse has it and the other does not. Why S 29, interest inthe occupation of dentist, or S 296, like or dislike of men who useperfume, should give the results they do we are unable to say. Twoitems yielding large differences such that agreement of spouses isnegatively related to happiness are interest in the occupation oflibrarian (S 53) and ability to accept criticism (S 387). The typicallibrarian probably tends to be more than ordinarily bookish, seclusive,and non-social, traits in which mating of likes could conceivably beless favorable to happiness than mating of unlikes. We are not ableto suggest a plausible explanation of the fact that spouse agreementon ability or inability to accept criticism without getting "sore"(S 387) should be unfavorable to happiness; one might perhaps evenhave expected the reverse. Whatever the reason, the evidence is

2 8 2 L. M. TERMAN AND P. BUTTENWEISER

Strong that for one spouse to be able to accept criticism when theother cannot is more favorable than agreement.

It seems reasonable that spouse agreement should be favorable (ordisagreement unfavorable) on such items as wanting attention whenill (B 70), willingness to take a chance alone (B 82), liking forpet canaries (S 136), liking for regular hours (S 234), active interestin charities (S 240), interest in looking at antique furniture (S 245),attitude toward teetotalers (S 293) (disagreement a fertile sourceof discord), or lending money to acquaintances (S 412). We leaveit to the reader to explain why it is unfavorable for spouses to dis-agree on their liking for Pershing (S 325) and on their tolerancefor pessimists (S 253). Agreement is unfavorable on crossing thestreet to avoid meeting someone (B 4) . Response to this item maybe influenced by a tendency to harbor personal dislikes, the presenceof which in both spouses would be likely to breed trouble. Whyagreement of spouses on interest in landscape gardening (S 50) shouldalso be unfavorable is not clear, unless this interest is associated withwhat is commonly called tbe "artistic temperament" and its tradi-tional restiveness under the yoke of marriage.

There are a few puzzling disagreements between items that appearto be closely similar. For example, B 17, on being much affectedby the praise or blame of many people, would seem to be much likeB 102, working better when praised. However, agreement on thelatter is favorable to happiness and agreement on the former un-favorable. Similarly with S 412 on loaning money to friends, andB 55 on being thrifty and careful in making loans; with the formeragreement is favorable and with the latter unfavorable. However,these two items are not so nearly identical as they appear to be, forone may be thrifty and careful in making loans and yet at timesmake them. Even so, it is not easy to see why likeness betweenspouses in regard to care and thriftiness in making loans should beassociated with unhappiness. Again it seems strange that spouseagreement should be unfavorable on "irreligious people" (S 263),when it was favorable in the case of S 293, "teetotalers."

Other items on which spouse likeness is unfavorable are B 60,rewriting letters (perhaps indicating indecision or over-scrupulosity);;B 104, alternation of feelings without cause (heavily weighted byits author for neurotic tendency); B 123, discipline causing dis-content (indocility or rebellious attitude) ; S 86, "secret service man"'


(sleuthing interest); and S 66, "orchestra conductor" (artistic tem-perament).

We forego further speculation as to possible interpretations ofthe husband-wife correlations on the individual test items until theseitems have been submitted to new and much larger groups and thecorrelations have been confirmed. Items whose validity can not besubstantiated will have to be discarded, however plausible they mayseem, and those which stand the test of repeated trial will have tobe retained whether we can explain their behavior or not.

Table 19 offers valuable evidence of the merits of the data in thefact that the L and D divergences from the H group are usuallyin the same direction. Taking the absolute figures as they stand,the direction is the same for 78 of the 130 items; if we ignoredirectional differences of 10 or less, the direction is the "same" for96 items. One could not assume that the direction of divergencewould be the same for every item, even if the data were perfectlyreliable and valid; there are no doubt genuine psychological differencesbetween L and D groups which would prevent this. However, inview of the known similarities of such groups it speaks well for thedata that the found correlations agree on the whole so well.

Before closing this section it will be well to remind the readerthat here as elsewhere in this study one must guard against takingthe fact of association of a given datum with marital happiness asproof of a causal relationship, or of the direction of causal relation-ship if such exists. In the case of many of the items it seems reason-able to believe that high correlation between husband and wiferesponses in the H group is the result rather than the cause of thehappiness; spouses who are happily mated may develop similarity oflikes, dislikes, and attitudes simply because they are happy. Con-versely, two spouses may find themselves unhappy in their marriagefor any one of a hundred very special or even trivial reasons, and,being unhappy, may each tend to develop likes, dislikes, and attitudesthat are antagonistic to those possessed by the other.

TRAIT CHARACTERISTICS OF THE H , L , AND D GROUPS AS

INDICATED BY MEAN SCORES IN THE PERSONALITY TESTS

Do the happily married tend to differ from the unhappily marriedor divorced groups of the same sex in the personality traits repre-sented in our thirteen variables? Perhaps most people would expect


to find certain differences, as, for example, that the happily marriedwould tend to be less "neurotic," less "introverted," more mature intheir interests, and more interested in social-service and "uplift"activities. Some would expect the H women to be less masculine-minded than the L or D, and the D women to be more "self-sufficient" than the L or H. There is one circumstance, however,which tends to reduce the expected differences—the fact that a sub-ject's marital happiness is in part a function of the spouse's personality.

The means of our six groups on the thirteen variables are given inTable 20. The scores in question are z-scores derived separatelyfor the men and women. They accordingly do not reveal sex dif-

TABLE 20MEAN Z-SCORES OF H , L, AND D GROUPS IN THE THIRTEEN VARIABLES

(100 of each sex in each group)

C.P.A."u

Chemist'M

Lawyer'u

Life insurance salesman

Teacher'M

Y.M.C.A. worker"M

Office clerk'u

Interest maturity

Masculinity-femininity"M

Bl-N"M

B2-S

M

B3-I"M

B4-D

H

5.31.19

4.88.20

5.23.23

5.11.18

5.18.18

5.03.21

5.01.21

5.46.20

4.67.20

4.70.19

4.88.20

4.67.18

5.13.19

HusbandsL

4.82.20

4.79.20

4.93.20

5.16.17

4.61.21

4.83.21

4.97.21

4.71.20

5.14.19

5.39.21

4.98.21

5.39.20

4.79.20

D5.09.18

4.67.20

5.27.21

5.40.19

4.10.20

4.39.19

4.65.22

4.96.18

4.29.18

5.40.18

4.95.20

5.57.19

4.99.18

H

5.00.20

5.16.20

5.22.20

5.00.18

5.21.19

5.08.21

4.99.20

5.39.17

4.71.24

4.66.21

4.63.19

4.66.20

5.17.21

WivesL

4.90.21

4.63.20

5.01.19

4.97.18

5.01.21

5.04.20

4.96.20

4.95.19

4.56.21

5.33.21

4.86.21

5.25.20

4.81.21

D5.27.21

5.37.20

5.29.22

4.51.19

4.30.19

4.40.20

4.42.23

5.03.18

5.45.20

4.64.20

5.79.21

4.95.20

5.42.21


ferences but do make possible inter-comparisons among the threegroups of each sex.

In Table 21 , which gives the critical ratios of the differences

TABLE 21CRITICAL RATIOS OF THE DIFFERENCES BETWEEN THE MEANS GIVEN IN TABLE 20

C.P.A.ChemistLawyerLife insurance

salesmanTeacherY.M.C.A.

workerOfBce

clerkInterest

maturityivieniai

masculinityBl-NB2-SB3-IB4-D

H—L

4-1.78-fO.32-1-0.99

—0.20+2.03

H-0.68

-f0.13

+2.93

—1.67—2.47—0.34—2.66-f 1.24

HusbandsH—D

-1-0.84-fO.76—0.13

—1.13+4.00

+2.28

-1-1.19

-1-1.84

-t-1.41—2.70—0.25—3.45-1-0.53

L—D

—1.00+0.43- 1 . 1 8

—0.95-f-1.76

-i-1.20

-M.06

—0.92

+3.24—0.04+0.10—0.66—0.76

H—L

+0.35+ 1.90+0.76

+0.12+0.71

+0.14

+0.10

+ 1.74

+0.47—2.26—0.81—2.08+ 1.21

WivesH—D

—.095—0.75—0.23

+ 1.83+3.38

+2.34

+ 1.86

+ 1.44

—2.39+0.07—4.16—1.01—0.85

L—D

—1.27—2.67—0.95

+ 1.74+2.53

+2.25

+ 1.77

—0.30

—3.08+2.36—3.14+ 1.05—2.07

recorded in Table 20, a plus sign means a higher mean score forthe first member of the pair concerned, a minus sign a lower score.In each case the pairs concerned are indicated in the column heading.Critical ratios of 2.00 or more are in italics.

The scarcity of large differences between like-sex groups is suchas to offer little support for the belief that the happily married indi-vidual is typically different from the L or D individual of the samesex in the majority of personality traits measured by the tests used.Some of the differences, however, are significant and interesting. Letus scrutinize these differences with reference to their plausibility, tak-ing them in the order given in the tables.

1. Certified Public Accountant. No significant difference, al-though H husbands tend to have slightly more of this type of interestthan L husbands ( C R 1.78).

2. Chemist. No significant difference for men. Among thewomen, however, the L group is markedly below the D group ( C R2.67) and probably below the H group ( C R 1.90).


3. Lawyer. No significant difference.4. Life Insurance Salesman. There is a suggestion that H and

L women have more of this type of interest than the D (CR's 1.83and 1.74, respectively).

5. Teacher. Marked differences in the case of both sexes. TheH men have significantly more of this type of interest than eitherthe L or D (CR's 2.03 and 4.00, respectively), and the L probablymore than the D (CR 1.76). Among the women, both H and Lgroups rate markedly above the D (CR's 3.38 and 2.53).

6. Y. M. C. A. Worker. The differences here are in line withthose for teacher, although somewhat smaller. H men have moreof this interest than D men (CR 2.28), and both H and L womenhave more than the D (CR's 2.34 and 2.25).

7. Office Clerk. No clearly significant differences, although atendency is noticeable for both H and L women to have more ofthis type of interest than D women (CR's 1.86 and 1.77).

8. Interest Maturity. H men rate markedly higher than L(CR 2.93) and probably higher than D (CR 1.84). H womentend to rate somewhat higher than either the L or D (CR's 1.74and 1.44).

9. Mental Masculinity. The outstanding difference among menis the markedly high rating of the L group as compared to the D(CR 3.24), the H group being about half way between the two.For the women the outstanding fact is the lower masculinity of theH as compared with either the L or D (CR's 2.39 and 3.08).

10. Bl-N, "Neurotic Tendency." The H men are distinctlyless "neurotic" as a group than the L (CR 2.47) and than the D(CR 2.70). The L women are the most "neurotic," being separ-ated from the H by a CR of 2.26 and from the D by a CR of2.36, the means of the H and D being almost identical.

11. B2-S, "Self Sufficiency." We find here no significant dif-ferences for the men, but large differences for the women. Amongwomen the striking fact is the high "self sufficiency" of the D group,which exceeds the H group by a CR of 4.16 and the L group by aCR of 3.14.

12. B3-I, "Introversion." As would be expected from the closecorrelation between Bl-N and B3-I, the critical ratios agree closelywith those already given for "neurotic tendency." The H men are"less introverted" than the L (CR 2.66) and than the D (CR 3.45).


H women are less "introverted" than the L (CR 2.08) and a trifleless than the D (CR 1.01).

13. B4-D,"Dominance." Only one difference is very significant;L women are less "dominant" than the D (CR 2.07).

One of the most striking facts in Table 21 is that whereas H mendiffer from the L and D in much the same way, this is not true ofthe women. That is, L and D men are much alike, differing sig-nificantly from one another only in mental masculinity (the lowbeing considerably more masculine). Among the women, however,the L and D groups present the greatest contrast. In fact, L and Dwomen are much less alike than the H and L women. Putting itin another way, it is the D woman who stands out from the otherfemale groups, and the H man who stands out from the other malegroups.

Compared with L or D men, H men have more uplift interests,greater interest maturity, and are less "neurotic" (or "introverted").In M-F score they are about midway between the more masculine-minded L group and the more feminine-minded D group.

The H women differ significantly from the L women only inbeing less "neurotic" (or "introverted"), although there is strongsuggestion that their interests are more mature and more like thoseof workers in the physical sciences. As compared with D women,H women are much less "self sufficient," less masculine-minded, andfar more interested in uplift activities. Their moderate excess ofscore over D for the occupation of ofBce clerk suggests greater docilityand greater interest in routine; their higher score for the occupationof life insurance salesman is probably connected with their greatersocial and uplift interests.

D women differ from L women in all the ways in which theydiffer from the H, and to much the same degree. In addition, theydiffer from the L in being more "dominant," less "neurotic," andmore interested in science.

SUMMARY

The Bernreuter Personality Inventory and the Strong VocationalInterest Test were administered to 345 married couples and to 116divorced couples. The married couples supplied anonymous datawhich yielded marital happiness ratings. The two tests were scored


for thirteen variables, including seven characteristic types of voca-tional interest, and "interest maturity," "masculinity-femininity,"neurotic tendency, "self sufficiency," "introversion," and "dominance."A comparative study was made of the 100 most happily marriedcouples, the 100 least happily married couples, and 100 divorcedcouples with respect to mean scores in the thirteen variables, degreeof husband-wife resemblance in the variables, and the direction ofhusband-wife differences. All of these comparisons based upon totalscores showed low or negligible correlations with marital happiness.However, statistical treatment of the responses of the subjects tothe 545 items of the tests showed that more than a quarter of theitems taken singly appear to have appreciable validity as indicatorsof marital compatibility. These items and others suggested by themwill be tried out further with larger populations. Incidentally nocorrelation was found between marital happiness and age at mar-riage, age difference between spouses, number of offspring, or spouse-parent attachments and conflicts.

REFERENCES

1. KELLEY, T . L. Statistical method. New York: Macmillan, 1923. Pp.xi+390.

2. CHESIRE, L., SAFFIR, M., & THURSTONE, L. L. Computing diagrams forthe tetrachoric correlation coefficient. Chicago: Univ. Chicaco Book-store, 1933. Pp. 57.

Stanford UniversityCalifornia

LES FACTEURS DE PERSONNALITfi DANS LA COMPATIBILITYMARITALE

(Resume)On a fait subir I'lnventaire de Personnaliti Bernreuter et le Test de

rinteret Professionnel Strong a 34S couples d'epoux et a U6 couples divorces.Les couples d'epoux ont fourni des donnees anonymes qui ont ciAe desevaluations du bonheur marital. On a mis les deux tests en ^chelle pourtreize variables, y compris sept types caracteristiques de l'interet profes-sionnel, et "la maturite de l'interet," "la masculinite-femininite," "la tendancea la nevrose," "la suffisance," "l'introversion," et "la dominance." On afait une itude comparative des 100 couples d'epoux les plus heureux, des100 couples d'epoux les moins heureux, et de 100 couples divorces i l'^garddes resultats moyens des treize variables, du degre de ressemblance entrele man et la femme dans les variables, et de la direction des differences


entre le raari et la femme. Toute ces comparaisons basees sur les resultatstotaux ont montre des correlations peu elevies ou negligeables avec le bon-heur marital. Cependant, le traitement statistique des reponses des sujetsaux 545 points des tests a montre que plus que le quart des points prisindividuellement semble avoir une valeur appreciable comme indicateurde Ia compatibility maritale. On fera subir ces points et d'autres qu'ilssuggJrent k plus de sujets. A propos, on n'a trouve aucune correlationentre le bonheur marital et Page au mariage, la difference entre leB agesdes epoux, le nombre des enfant, ou les attachments et Ies conflits entreepoux et parent.

TERMAN ET BUTTENWIESER

PERS6NLICHKEITSFAKT0REN BEI DER EHELICHEN VERTRAG-LICHKEIT

(Referat)Das Bernreuter Personlichkeitsinventar und der Strong Berufseignungs-

test wurden 345 Ehepaaren und 116 geschiedenen Paaren gegeben. DieEhepaare lieferten namenlose Angaben, woraus Auskunfte uber das Eheg-luck gesammenit vfurden. Die Zvcei Tests wurden auf Grund drei Vari-ablen verwertet, welche sieben charakteristische Arten von Berufsinteres-sen, "Interessenreife," "Mannlichkeit-Weiblichkeit," "neurotischer Neigung,""Selbststandigkeit," "Introversion," und "Herrschsucht" einschlossen. Einvergleichendes Studium der 100 glCicklichsten Ehepaare, der 100 am wenig-sten glucklichen Ehepaare, und 100 geschiedener Paare in bezug auf dieDurchschnittswerte der dreizehn Variablen, den Grad der Mann-Frauahn-lichkeit der Variablen, und die Richtung der Mann-Frauunterschiede wurdegemacht. Alle diese Vergleiche zeigten niedrige oder unbedeutende Kor-relationen mit Ehegluck. Die statistische Behandlung der Angaben der Vpn.uber die 545 Einzelheiten des Tests bewies aber, dass mehr als ein Viertelder Einzelheiten, einzeln genommen, eine betrachtliche Gultigkeit als Merk-male der ehelichen Vertraglichkeit zu besitzen scheint Diese Einzelheitenund andere, die durch diese entstanden sind, werden spater bei einer gros-seren Gruppe weiter ausgefuhrt. Nebenbei fand sich keine Korrelationzwischen Ehegluck und dem Alter zur Zeit der Heirat, dem Altersunter-schied zwischen den Eheleuten, der Anzahl der Nachkommenschaft, oderden Gatten-Elternbeziehungen und Konflikten.

TERMAN UND BUTTENWIESER

PERSONALITY FACTORS IN MARITAL COMPATIBILITY: II. · 2020. 4. 2. · PERSONALITY FACTORS IN MARITAL...

Documents

Transcript of PERSONALITY FACTORS IN MARITAL COMPATIBILITY: II. · 2020. 4. 2. · PERSONALITY FACTORS IN MARITAL...