Compass Math Test Validity April 2011 Mike Orkin, Jo Ann Phillips, Hui Zhang Office of Institutional Research Peralta Community College District 1.

Slides:



Advertisements
Similar presentations
Analyzing Chi-Squares and Correlations
Advertisements

Revised Recommendations: System Agreement for the Use of Smarter Balanced 11 th Grade Assessment in Washington Higher Education Placement Process April.
CHAPTER TWELVE ANALYSING DATA I: QUANTITATIVE DATA ANALYSIS.
Developmental Education Assessment, Placement, and Progression Thomas Bailey Based on Research by Katherine Hughes, Shanna Jaggars, Judith Scott-Clayton.
STAAR EOC Elizabeth Lalor Senior Director of Academic Support
Evaluation of the Impacts of Math Course Placement Improvement Achieved through a Summer Bridge Program John R. Reisel, Leah Rineck, Marissa Jablonski,
1 Effective Use of Benchmark Test and Item Statistics and Considerations When Setting Performance Levels California Educational Research Association Anaheim,
Using a Regression Discontinuity Design to Estimate the Impact of Placement Decisions in the Los Angeles Community College District Tatiana Melguizo &
Gordon Associates Why Do We Do What We Do, And How Do We Know if it’s Working? By Ron Gordon, EdD.
GATEWAY COURSE SUCCESS Gateway not “gatekeeper” Principles of Reform Gateway not “gatekeeper” Principles of Reform.
MATH ASSESSMENT TEST OCMA May, HISTORY OF MAT Test originally developed in late 60’s.
Some Statistics Related to Placement Test Results and First Semester Math Grades at George Brown College.
Correlation: Relationship between Variables
Common core mathematics pathway overview
Does Remediation Help All Students
Melanie Formosa Jennifer McLaughlin Stephanie Keiles.
College of Engineering Freshman Admission Requirements.
MATH Placement Testing Results 2009 Data on New Cut Scores.
AN EVALUATION OF THE EIGHTH GRADE ALGEBRA PROGRAM IN GRAND BLANC COMMUNITY SCHOOLS 8 th Grade Algebra 1A.
1 Predicting Success in Math: The relationship between high school courses and remedial math in college Brandon Lagerquist Northshore School District Paul.
COMPASS National and Local Norming Sandra Bolt, M.S., Director Student Assessment Services South Seattle Community College February 2010.
The Impact of Postsecondary Remediation Using a Regression Discontinuity Approach: Addressing Endogenous Sorting and Noncompliance Juan Carlos Calcagno.
Using Demographic Data to Predict Students’ Achievement in DSPM Courses Daryl Stephens, ETSU TNADE October 27, 2005.
Integrated Mathematics The missions of Bristol Tennessee City Schools is to achieve college and career readiness through academic excellence.
10.3: Use and Abuse of Tests Carrying out a z-confidence interval calculation is simple. On a TI-83, it’s STAT-->TESTS-->7 Similarly, carrying out a z-signficance.
Why is college attendance a good idea? No High School Diploma $19,140 Annual Income High School Diploma $27,235 Annual Income Bachelor’s Degree $46,931.
Simmons Middle School. Simmons Math Department Colleen O’Connor ~ Seventh Grade Colleen O’Connor ~ Seventh Grade Jill Poetz ~ Seventh Grade Jill Poetz.
Academic Success First-time/Full-time Students Fall 2008, 2009 and 2010 Concurrently Enrolled in English and Select Courses.
Bivariate Regression Analysis The most useful means of discerning causality and significance of variables.
DEVELOPMENTAL MATH CHANGES AT BMCC JIM WHITTAKER, DIRECTOR OF INSTITUTIONAL RESEARCH.
ABLE + COLLEGE EQUALS SUCCESS OAACE April
ONEONTA HIGH SCHOOL 8 TH GRADE - OCTOBER 2010 PRESENTED ON DEC. 14, 2010 ACT EXPLORE RESULTS.
Math Locally W Stephen Wilson Leading Minds Conference Baltimore, Maryland April 24, 2008.
Using COMPASS to Find Your Way Dr. Beth Cipoletti John Ford Jessica George.
THE ACT-PLAN TEST IS AN IMPORTANT TEST FOR CHS SOPHOMORES!
Class of 2020 High Ability Mathematics Parent Night.
Testing Hypotheses about Differences among Several Means.
Setting a college-readiness cut score for the CRMT CRMT Working Group August 19, :30-3:30  Background  Pilot test summary  Current MPT-I placement.
MGS3100_04.ppt/Sep 29, 2015/Page 1 Georgia State University - Confidential MGS 3100 Business Analysis Regression Sep 29 and 30, 2015.
Perceptions of the Appropriateness of Student Placement in English, Math, Reading, and ESL Courses at Cañada College CAÑADA COLLEGE PLANNING, RESEARCH,
The Transformative Power of Remediation Steven Spurling City College of San Francisco.
Gifted Presentation Mike Nicholson, Senior Director of Research and Evaluation.
Using EXPLORE Results for Student Success. Gennine Brewer, MA Senior Consultant P-16 Assessment Services ACT Atlanta,
Sampling Design and Analysis MTH 494 Lecture-22 Ossam Chohan Assistant Professor CIIT Abbottabad.
1 Making Better Decisions on Placement in Mathematics Courses W. Robert Stephenson Iowa State University.
EXPLORE What is the EXPLORE? An assessment of academics as well as career interests. An assessment of academics as well as career interests. You will.
James C. Enochs High School Advanced Placement Information Meeting February 25, 2016.
Validity Evidence for the Orleans-Hanna Algebra Prognosis Test Jack B. Monpas-Huber, Ph.D. Director of Assessment & Student Information.
English Assessment Validity February 2012 Michael Orkin, Jo Ann Phillips, Hui Zhang, Sheryl Queen Office of Institutional Research Peralta Community College.
Examining the Enrollment and Persistence of Students with Discrepant High School Grades and Standardized Test Scores Anne Edmunds, Ed.D. Higher Education.
What is the Advanced Placement Program? The Advanced Placement Program are courses taught at local high schools, which allow students to participate in.
How to Apply to Utah State University. Deadlines Fall Semester: April 1 Spring Semester: October 1 Summer Semester: April 1.
Logistic Regression An Introduction. Uses Designed for survival analysis- binary response For predicting a chance, probability, proportion or percentage.
College Course Opportunities for High School Students Palm Harbor University High School 2016.
LESSON 5 - STATISTICS & RESEARCH STATISTICS – USE OF MATH TO ORGANIZE, SUMMARIZE, AND INTERPRET DATA.
HLC Academy on Student Persistence and Completion – A Presentation on Statistical Analyses of Illinois Tech Data May 24, 2016 Illinois Institute of Technology.
The Process The Results The Repository of Assessment Documents (ROAD) Project Sample Characteristics (“All” refers to all students enrolled in ENGL 1551)
Presenters: Conrad Perez-Math Faculty John Heathcote- Math Faculty Ryen Hirata-Outreach & Matriculation Coordinator.
Pre-Transfer Courses & Budget Cuts Mike Orkin Educational Services October, 2011 English, ESL, Math.
River Dell Math Placement 7th Grade into 8 th Grade.
Michelle Kershner (and Dan Kernler) Elgin Community College Elgin, IL
Lynn Hale – Graduation Interventionist
Ginni May, ASCCC Area A Representative
Becoming AB 705 Compliant Tammi Marshall, Cuyamaca College
Why do I have to take it? What score do I need? What is on ACT math?
PARCC RESULTS: PRESENTATION FAIRVIEW SCHOOL DISTRICT OCTOBER 2, 2018
Texas Success Initiative Assessment
Thompson Middle School
MGS 3100 Business Analysis Regression Feb 18, 2016
Presentation transcript:

Compass Math Test Validity April 2011 Mike Orkin, Jo Ann Phillips, Hui Zhang Office of Institutional Research Peralta Community College District 1

Compass Math Test Peralta students take assessment tests for placement into Math, ESL, and English courses. Course placement based on test scores is recommended not mandatory. A student may take courses without taking placement tests. 2

Compass Math Test Peralta uses the ACT Compass test for math placement. When students take Compass, they advance through “subtests” until they reach a “placement domain.” The higher the placement domain, the higher the difficulty level of the recommended courses. 3

Data Students taking Compass-designated math courses in F08, S09, F09, S10. Took Compass test immediately preceding semester when they took math course, or Never took Compass test prior to semester when they took math course. 4

Data Totals 12,657 student enrollments. 4,915 (39%) took Compass test. 7,742 (61%) did not. 5

Data by Placement Domain Courses Placement Domain Courses Compass No CompassTotal % Take Compass MPRAL 3,7492,6446,39359% MALGE 7555,2686,02313% MTRIG % MCOAL 2323,5903,8226% 6

Placement Domains and Key Courses MPRAL – Prealgebra  Arithmetic – Math 250 (0-32)  Pre-Algebra – Math 253 (33-39)  Elementary Algebra – Math 201 (40-100) 7

Placement Domains and Key Courses MALGE – Algebra.  Elementary Algebra – Math 201 (0-34)  Intermediate Algebra – Math 203 (35-49)  Statistics – Math 13 (50-100) 8

Placement Domains and Key Courses MTRIG – Trigonometry.  Trigonometry – Math 50 (0-45)  Pre Calculus – Math 1 (46-100) 9

Placement Domains and Key Courses MCOAL – College Algebra.  Statistics – Math 13 (0-34)  Pre Calculus – Math 1 (35-49)  Calculus – Math 3A (50-100) 10

Compliance Within Placement Domain Most complied with cutoff score assignment within their placement domain. 11

Compliance Within Placement Domain Placement DomainComplianceTotal % MPRALCOMPLY 3,26487% NONCOMPLY 48513% MPRAL Total 3,749 MALGECOMPLY 68290% NONCOMPLY 7310% MALGE Total 755 MTRIGCOMPLY 12570% NONCOMPLY 5430% MTRIG Total 179 MCOALCOMPLY 16973% NONCOMPLY 6327% MCOAL Total 232 Grand Total 4,915 12

Validity A test is “valid” if it measures what it’s supposed to measure. We looked at this in two ways: I.Are grades and success positively related to Compass scores? II.Do students who take Compass do “better” than students who don’t take Compass? 13

Validity Are grades positively related to Compass scores?  Analysis of Variance (ANOVA) determines if there are significant differences in average Compass scores across course grades.  On average, as grades increase, so do scores. 14

Validity ANOVA - Assessment test averages by grade. Placement Domain A - Test ave B - Test ave C - Test ave D - Test ave F/W- Test aveSig. MPRAL MALGE MTRIG MCOAL ALL

Validity Grade Counts Placement Domain A B C D FWTotal MPRAL ,1073,693 MALGE MTRIG MCOAL ALL 1, ,2974,854 16

Validity Do assessment test scores predict a student’s chance of success (C or better) in a course?  Logistic regression estimates the probability of success for a given assessment test score.  For all subtests, as scores get higher, the predicted chance of success increases, indicating predictive validity. 17

Validity 18

Validity 19

Validity 20

Validity 21

Validity 22

Perfect Success We have shown that the chance of success increases as Compass scores increase. However, if the assessment process works perfectly, there will be an appropriate course for every student, even those who get low Compass scores, and so everyone (or almost everyone) will succeed, regardless of their Compass score. 23

Perfect Success If everyone succeeds, the chance of success will be 100% no matter what the Compass score, and so the regression analysis will not be meaningful. In the same way, if there is an appropriate course for everyone, and if everyone succeeds, there will be no low grades, and, more generally, there won’t be a relationship between Compass scores and grades. 24

Perfect Success In other words, looking for a relationship between Compass scores and success can be clouded by too much success. It thus becomes useful to compare those who take the Compass test and those who don’t. 25

Compass v. no Compass Do students who take the Compass test do better than students who don’t take the test?  Assume that those who don’t take the Compass test essentially select a class at random. 26

Compass v. no Compass For non-Compass students, just by chance, some strong students will take lowest-level courses (too easy) and some weak students will take higher level courses (too difficult). If Compass is properly assessing students, only weaker students will take lowest level courses and stronger students will take higher level courses. 27

Compass v. no Compass Thus, students who take Compass should do worse than those who don’t at lowest subtest and better than those who don’t at higher levels. This is exactly what we observe in the data. 28

Compass v. no Compass Chi Square - Success – Compass Test v. No Test Placement Domain Courses Take Test % Success No Test % Success Sig. MPRAL 48.2%50.9%0.020 MALGE 65.0%57.5%0.000 MTRIG 74.9%56.9%0.000 MCOAL 82.8%60.8%

Compass v. no Compass Similar result when comparing grades of students who take Compass with those who don’t. 30

Compass v. no Compass ANOVA - Test v. No Test Course GPA Placement Domain Courses Test GPA No Test GPASig. MPRAL MALGE MTRIG MCOAL

Summary All four methods of analysis yield statistically significant results supporting Compass validity. 32

Summary On average, as grades increase, so do Compass scores. As Compass scores increase, the predicted chance of success increases. Students who take Compass do better than those who don’t in the higher subtests and worse in the lowest subtest, in both course success and grades. 33