Join the conversation: todaysmeet.com/SB AC Assessment Update: June 2015.

Join the conversation:

todaysmeet.com/

SBAC

Assessment Update:June 2015

Reconceptualizing assessment without at the same time

reconceptualizing instructionwill have little benefit.

Pullin, 2008

Required State Data for Each of Eight State Priority Areas

February 2015 Draft LCAP Evaluation Rubric

Implementation Timeline for New State Assessments and Accountability

English Language Artsand Mathematics

English Language Development

Science and History / Social Studies

https://goo.gl/7SlA20

172015 Post-Test Workshop: Reporting Summative Assessment Results

Agenda

• Principles of Scoring• Understanding the Reports• Using the Online Reporting System• Overview of the Reporting Timeline• Interpreting, Using, and Communicating Results

Principles of ScoringPrinciples of Scoring


Computer Adaptive Testing:Philosophy

“Computer adaptive testing (CAT) holds the potential for more customized assessment with test questions that are tailored to the students’ ability levels, and identification of students’ skills and weaknesses using fewer questions and requiring less testing time.”

Shorr, P. W. (2002, Spring). A look at tools for assessment and accountability. Administrator Magazine.


How Does a CAT Work?• Each student is administered a set of test questions that

is appropriately challenging.• The student’s performance on the test questions

determines if subsequent questions are harder or easier.• The test adapts to the student item-by-item and not in

stages.• Fewer test questions are needed as compared to a fixed

form to obtain precise estimates of students’ ability.• The test continues until the test content outlined in the

blueprint is covered.


How Does a CAT Work?Example: A Student of Average Ability

Medium

Med-Low

Low

Very Low

Med-High

High

Very High

Test QuestionsAnswers (R/W)

1 2 3 4 5 6 7 8 9 10

R R W R W W W W R R


Computer Adaptive Testing:Behind the Scenes

• Requires a large pool of test questions statistically calibrated on a common scale with ability estimates, e.g., from the Field Test

• Uses an algorithm to select questions based on a student’s responses, to score responses, and to iteratively estimate the student’s performance

• Final scale scores are based on item pattern scoring


Scoring the CAT• As a student progresses through the test, his or her pattern of

responses is tracked and revised estimates of the student’s ability are calculated.

• Successive test questions are selected to increase the precision about the level of achievement given the current estimate of his or her ability.

• Resulting scores from the CAT portion of the test are based the specific test questions selected as a result of the student’s responses, but NOT the sum of the number answered correctly.

• The test question pools for a particular grade level are designed to include enhanced pool of test questions that are more or less difficult for that grade but still matching the test blueprint for that grade.


Human Scored Items in the CAT

• Some items administered on the Smarter Balanced adaptive test component require human scoring of items

• The adaptive algorithm will select these items based on performance on prior items.

• Since these items cannot be scored in real time by a human, performance on these items will not impact later item selection.


Performance Tasks (PTs)

• In all Smarter Balanced tests, a PT and a set of stimuli on a given topic are administered as well as the CAT.

• PTs are administered at the classroom/group level so they are not targeted to students’ specific ability level.

• The items associated with the PTs may be scored by machine or by human raters.


Final Scoring• For each student, the responses from the PT and CAT portions are

merged for final scoring.• Resulting ability estimates are based on the specific test questions that

a student answered, not the total number of items answered correctly.• Higher ability estimates are associated with test takers who correctly

answer difficult and more discriminating items.• Lower ability estimates are associated with test takers who correctly

answer easier and less discriminating items.• Two students will have the same ability estimate if they have the same set

of test questions with the same responses.• It is possible for students to have the same ability estimate through

different response patterns• This type of scoring is called “Item Pattern Scoring.”


Final Scoring: Contribution of CAT and PT Sections

Number of Items defined by Test Blueprints

ELA/Literacy Mathematics

Grade CAT PT CAT PT

3–5 38–41 5–6 31–34 2–6

6–8 37–42 5–6 30–34 2–6

11 39–41 5–6 33–36 2–6


Final Scoring: Contribution of CAT and PT Sections

(cont.)• Based on the test blueprint, the CAT section is emphasized

because there are more CAT items/points than PT items/points.

• Claims with more items/points are emphasized.– Mathematics: Concepts and Procedures Problem

Solving/Modeling and Data Analysis Communicating Reasoning

– ELA: Reading Writing Speaking/Listening Research• Because scores are based on pattern scoring, groups of

items that are more difficult and discriminating will have a larger contribution on final scores.

• Therefore there is no specific weight associated with either PT or CAT Sections


Final Scoring: Mapping

• After estimating the student’s overall ability, it is mapped onto the reporting scale through a linear transformation.

• Mathematics:– Scaled Score = 2514.9 + 79.3 *

• ELA:– Scaled Score = 2508.2 + 85.8 *

• Limited by grade level lowest and highest obtainable scaled score


Properties of the Reporting Scale

• Scores are on a vertical scale.– Expressed on a single continuum for a content area– Allows users to describe student growth over time

across grade levels• Scale score range

– ELA/Literacy: 2114–2795– Mathematics: 2189–2862

• For each grade level and content area, there is a separate scale score range.


Smarter Balanced Scale Score Ranges by Grade Level

Grade Subject Min Max Subject Min Max

3 ELA 2114 2623 Mathematics 2189 2621

4 ELA 2131 2663 Mathematics 2204 2659

5 ELA 2201 2701 Mathematics 2219 2700

6 ELA 2210 2724 Mathematics 2235 2748

7 ELA 2258 2745 Mathematics 2250 2778

8 ELA 2288 2769 Mathematics 2265 2802

11 ELA 2299 2795 Mathematics 2280 2862


Achievement Levels

• Achievement level classifications based on overall scores– Level 1—Standard Not Met– Level 2—Standard Nearly Met– Level 3—Standard Met– Level 4—Standard Exceeded


Achievement Levels by Grade


Smarter Balanced Scale Score Ranges for ELA/Literacy

Grade Level 1 Level 2 Level 3 Level 4

3 2114–2366 2367–2431 2432–2489 2490–2623

4 2131–2415 2416–2472 2473–2532 2533–2663

5 2201–2441 2442–2501 2502–2581 2582–2701

6 2210–2456 2457–2530 2531–2617 2618–2724

7 2258–2478 2479–2551 2552–2648 2649–2745

8 2288–2486 2487–2566 2567–2667 2668–2769

11 2299–2492 2493–2582 2583–2681 2682–2795


Achievement Levels by Grade


Smarter Balanced Scale Score Ranges for Mathematics

Grade Level 1 Level 2 Level 3 Level 4

3 2189–2380 2381–2435 2436–2500 2501–2621

4 2204–2410 2411–2484 2485–2548 2549–2659

5 2219–2454 2455–2527 2528–2578 2579–2700

6 2235–2472 2473–2551 2552–2609 2610–2748

7 2250–2483 2484–2566 2567–2634 2635–2778

8 2265–2503 2504–2585 2586–2652 2653–2802

11 2280–2542 2543–2627 2628–2717 2718–2862


Measurement Precision:Error Bands

• For each scale score estimated for a student, there is measurement error associated with each score. An error band is a useful tool that describes the measurement error associated with a reported scale score.

• Error bands are used to construct an interval estimate corresponding to a student’s true ability/proficiency for a particular content area with a certain level of confidence.

• The error bands used to construct interval estimates were based on one standard error of measurement. – If the same test is given to student multiple times,

about 68 percent of the time, the student will score within this band.


Achievement Levels for Claims• Achievement Levels for claims are very similar to subscores.

They provide supplemental information regarding a student’s strengths or weaknesses.

• No achievement level setting occurred for claims.• Only three achievement levels for claims were

developed since there are fewer items within each claim.• Achievement levels for claims are based on the distance a

student’s performance on the claim is from the Level 3 proficiency cut.

• A student must complete all items within a claim to receive an estimate of his or performance on a claim.


Achievement Levels for Claims (2)

• A student’s ability, along with the corresponding standard error, are estimated for each claim.

• The student’s ability estimate for the claim is compared to the Level 3 proficiency cut .

• Differences between and greater than 1.5 standard errors of the claim would indicate a strength or weakness.

( )C( )P

C P


Achievement Levels for Claims (3)1.5C SE

C P1.5C SE

CP

At/Near Standard

Below Standard


Achievement Levels for Claims (4)

Above Standard

P C

1.5C SE

Understanding the Understanding the ReportsReports


Available ReportsSecure Location

Preliminary student test results ORS

Preliminary and partial aggregate test results ORS

Student Score Reports (ISRs) TOMS

Final student data TOMS

Public Location

Smarter Balanced ELA and mathematics CDE CAASPP Web page

CST/CMA/CAPA for Science and STS for RLA

CDE CAASPP Web page


Secure ReportingReport LEA School* Parent

Preliminary Student Data ORS

ORS

Preliminary Aggregate Data ORS

ORS

Final Student Score Reports (ISRs)pdf/paper

TOMS†/Paper††

TOMS†/Paper

Paper

Final Student Data File TOMS

* Access to ORS will be granted to CAASPP Test Site Coordinators in August.† PDFs of the Student Score Reports will be available in TOMS.†† LEAs must forward or mail the copy of the CAASPP Student Score Report to each student’s parent/guardian within 20 working days of its delivery to the LEA.


Preliminary Test Results: Student and Aggregate

• Through the Online Reporting System (ORS)• Available approximately three to four weeks after

student completes both parts—CAT and PT—of a content area

• Added daily

Use Caution: The results are partial and may not be a good representation of the school or district’s final aggregate results. The results are preliminary; the processing of appeals may result in score changes.


Student Score Reports (ISR): Overview

• One-page report– Double-sided:

All Smarter Balanced CAPA for Science

– Single-sided: CST/CMA for Science (Grade 10) STS for RLA

• Student’s final CAASPP test results• Reports progress toward the state’s academic content standards• Indicates areas of focus to:

– Help students’ achievement– Improve educational programs

• LEA distributes to parents/guardians


Student Score Reports: Shipments to LEAs

• Two copies of each student’s Student Score Report– One for the parent/guardian– One for the school site

• 2015 LEA CAASPP Reports Shipment Letter• 2015 School CAASPP Reports Shipment Letter

Note: Per California Code of Regulations, Title 5, Section 863, LEAs must forward one copy to parent/guardian within 20 business days. Schools may file the copy they receive, or they may give it to the student’s current teacher or counselor. If the LEA receives the reports after the last day of instruction, the LEA must mail the pupil results to the parent or guardian at their last known address. If the report is non-deliverable, the LEA must make the report available to the parent or guardian during the next school year.


Test Results Reported on the Student Score Reports

Grade

Smarter Balanced ELA and

mathematics

CST, CMA, or CAPA Science

3

4

5

6

7

8

10

11

Grade STS RLA

2

3

4

5

6

7

8

9

10

11

For students who took Smarter Balanced ELA and mathematics, CST, CMA or CAPA for Science

For students who took STS RLA


Elements of the Student Score Report

11

22

33

44

55

66 77

88

Front Page Back Page


11

Elements of the Student Score ReportFront Page


22



33



44



55

Elements of the Student Score ReportBack Page


66



77



Elements of the Student Score Report: Science Grades 5, 8, & 10 only

88

Back Page


Elements of the Student Score Report: Early Assessment Program Grade 11 only

88

Back Page


Student Score Reports (cont.)

• A guide explaining the elements of student score reports will be available electronically on the caaspp.org reporting Web page.


Final Student Data File

• Downloadable file in CSV format– Data layout to be released soon on caaspp.org

• Includes test results for all students tested in the LEA

• Available within four weeks after the end of an LEA’s test administration window in TOMS

• Additional training planned


Public Web Reporting Site

• Available on the CDE Web site through DataQuest• Planned release in mid-August• Access two testing programs through one Web site

– Smarter Balanced ELA and mathematics– CST/CMA/CAPA Science and STS RLA

• Additional training planned

Reporting TimelineReporting Timeline


Timeline for Preliminary Results, Student Score Reports, and Final Student Data File

Appeals

Additional preliminary results received

Rescores


Timeline for Public Reporting on DataQuest

TOM TORLAKSONState Superintendent of Public Instruction

Communications Toolkit (Cont.)

• Sample parent and guardian letter to accompany the Individual Student Report

• Reading Your Student Report, in multiple languages, to help parents and guardians read and interpret the Individual Student Report

• Documents that include released questions that exemplify items in the Smarter Balanced assessments to help parents/guardians understand the achievement levels

• Short video to help parents/guardians understand the Individual Student Report

http://www.cde.ca.gov/ta/tg/ca/communicationskit.asp

65

TOM TORLAKSONState Superintendent of Public Instruction

66

Questions?

Join the conversation: todaysmeet.com/SB AC Assessment Update: June 2015.

Documents

Transcript of Join the conversation: todaysmeet.com/SB AC Assessment Update: June 2015.