Issues in Assessment Design, Vertical Alignment, and Data Management : Working with Growth Models Pete Goldschmidt UCLA Graduate School of Education &

Slides:



Advertisements
Similar presentations
Using Growth Models to improve quality of school accountability systems October 22, 2010.
Advertisements

Hierarchical Linear Modeling: An Introduction & Applications in Organizational Research Michael C. Rodriguez.
The Research Consumer Evaluates Measurement Reliability and Validity
1 COMM 301: Empirical Research in Communication Kwan M Lee Lect4_1.
Chapter 4 – Reliability Observed Scores and True Scores Error
MEASURING COLLEGE VALUE-ADDED: A DELICATE INSTRUMENT Richard J. Shavelson SK Partners & Stanford University AERA Ben Domingue University of Colorado Boulder.
1 SSS II Lecture 1: Correlation and Regression Graduate School 2008/2009 Social Science Statistics II Gwilym Pryce
Regression Analysis Once a linear relationship is defined, the independent variable can be used to forecast the dependent variable. Y ^ = bo + bX bo is.
UCLA Graduate School of Education & Information Studies National Center for Research on Evaluation, Standards, and Student Testing Moving Beyond Status:
Using Growth Models to Monitor School Performance Over Time: Comparing NCE, Scale and Scores on NRTs and SBTs American Educational Research Association.
Evaluating Pretest to Posttest Score Differences in CAP Science and Social Studies Assessments: How Much Growth is Enough? February 2014 Dale Whittington,
Longitudinal Experiments Larry V. Hedges Northwestern University Prepared for the IES Summer Research Training Institute July 28, 2010.
Communicating through Data Displays October 10, 2006 © 2006 Public Consulting Group, Inc.
The Lumina Center Grantseeking Workshop Series Presents Outcomes & Evaluations April 20, 2006.
Using Growth Models for Accountability Pete Goldschmidt, Ph.D. Assistant Professor California State University Northridge Senior Researcher National Center.
Using Hierarchical Growth Models to Monitor School Performance: The effects of the model, metric and time on the validity of inferences THE 34TH ANNUAL.
Treatment Effects: What works for Whom? Spyros Konstantopoulos Michigan State University.
Pengujian Parameter Koefisien Korelasi Pertemuan 04 Matakuliah: I0174 – Analisis Regresi Tahun: Ganjil 2007/2008.
Pertemua 19 Regresi Linier
Today Concepts underlying inferential statistics
Chapter 7 Correlational Research Gay, Mills, and Airasian
Chapter 7 Forecasting with Simple Regression
Classroom Assessment A Practical Guide for Educators by Craig A
Richard M. Jacobs, OSA, Ph.D.
Slide 1 Testing Multivariate Assumptions The multivariate statistical techniques which we will cover in this class require one or more the following assumptions.
Introduction to Regression Analysis, Chapter 13,
S-005 Types of research in education. Types of research A wide variety of approaches: –Theoretical studies –Summaries of studies Reviews of the literature.
Relationships Among Variables
Multiple Linear Regression A method for analyzing the effects of several predictor variables concurrently. - Simultaneously - Stepwise Minimizing the squared.
Analysis of Clustered and Longitudinal Data
Statistics for Managers Using Microsoft Excel, 4e © 2004 Prentice-Hall, Inc. Chap 12-1 Chapter 12 Simple Linear Regression Statistics for Managers Using.
Including a detailed description of the Colorado Growth Model 1.
NCLB AND VALUE-ADDED APPROACHES ECS State Leader Forum on Educational Accountability June 4, 2004 Stanley Rabinowitz, Ph.D. WestEd
Assessment Group for Provincial Assessments, June Kadriye Ercikan University of British Columbia.
Inferences about School Quality using opportunity to learn data: The effect of ignoring classrooms. Felipe Martinez CRESST/UCLA CCSSO Large Scale Assessment.
Human Capital Policies in Education: Further Research on Teachers and Principals 5 rd Annual CALDER Conference January 27 th, 2012.
Philosophy of IR Evaluation Ellen Voorhees. NIST Evaluation: How well does system meet information need? System evaluation: how good are document rankings?
The Impact of Including Predictors and Using Various Hierarchical Linear Models on Evaluating School Effectiveness in Mathematics Nicole Traxel & Cindy.
Slide 1 Estimating Performance Below the National Level Applying Simulation Methods to TIMSS Fourth Annual IES Research Conference Dan Sherman, Ph.D. American.
© 2003 Prentice-Hall, Inc.Chap 13-1 Basic Business Statistics (9 th Edition) Chapter 13 Simple Linear Regression.
Instruction, Teacher Evaluation and Value-Added Student Learning Minneapolis Public Schools November,
Measuring Complex Achievement
Extension to Multiple Regression. Simple regression With simple regression, we have a single predictor and outcome, and in general things are straightforward.
Project on Educator Effectiveness & Quality Chancellor Summit September 27, 2011 Cynthia Osborne, Ph.D.
Assumptions of value-added models for estimating school effects sean f reardon stephen w raudenbush april, 2008.
UCLA Graduate School of Education & Information Studies National Center for Research on Evaluation, Standards, and Student Testing Practical Considerations.
Chapter Thirteen Measurement Winston Jackson and Norine Verberg Methods: Doing Social Research, 4e.
Managerial Economics Demand Estimation & Forecasting.
Chapter 16 Data Analysis: Testing for Associations.
Lecture 4 Introduction to Multiple Regression
UCLA Graduate School of Education & Information Studies National Center for Research on Evaluation, Standards, and Student Testing Growth Models for Monitoring.
Impediments to the estimation of teacher value added Steven Rivkin Jun Ishii April 2008.
American Educational Research Association Annual Meeting AERA San Diego, CA - April 13-17, 2009 Denise Huang Examining the Relationship between LA's BEST.
Research Methodology and Methods of Social Inquiry Nov 8, 2011 Assessing Measurement Reliability & Validity.
T tests comparing two means t tests comparing two means.
Applied Quantitative Analysis and Practices LECTURE#30 By Dr. Osman Sadiq Paracha.
Education 793 Class Notes Inference and Hypothesis Testing Using the Normal Distribution 8 October 2003.
Reliability a measure is reliable if it gives the same information every time it is used. reliability is assessed by a number – typically a correlation.
Handout Six: Sample Size, Effect Size, Power, and Assumptions of ANOVA EPSE 592 Experimental Designs and Analysis in Educational Research Instructor: Dr.
Applied Quantitative Analysis and Practices LECTURE#17 By Dr. Osman Sadiq Paracha.
UCLA Graduate School of Education & Information Studies National Center for Research on Evaluation, Standards, and Student Testing United States Growth.
Stats Methods at IC Lecture 3: Regression.
Module II Lecture 1: Multiple Regression
Measuring College Value-Added: A Delicate Instrument
C3 Student Growth Percentiles: An Introduction for Consumers of the Data MSTC, February 19, 2015.
Chapter 14 Introduction to Multiple Regression
Comparability by Design
Dr. Robert H. Meyer Research Professor and Director
Dan Goldhaber1,2, Vanessa Quince2, and Roddy Theobald1
AACC Mini Conference June 8-9, 2011
Presentation transcript:

Issues in Assessment Design, Vertical Alignment, and Data Management : Working with Growth Models Pete Goldschmidt UCLA Graduate School of Education & Information Studies Center for the Study of Evaluation National Center for Research on Evaluation, Standards, and Student Testing CCSSE Value Added Brain Trust November 15, 16, Washington DC

Data The metric matters depending on which question we are interested in addressing. Questions concerning absolute growth require a vertically equated scale score. Relative questions are less sensitive to the metric. ranking schools comparison of various performance measures comparison of schools based on growth comparison of achievement gaps (static or longitudinal) value added comparisons

Types of Longitudinal Models Longitudinal Growth Panel Models Longitudinal School Productivity Models Longitudinal Program Evaluation Models

Keep track of students’ achievement form one grade to the next e.g., collect achievement scores at Grades 2, 3, 4, and 5 for students in a school Focus on students’ developmental processes What do students’ growth trajectories look like? Longitudinal Growth Panel Models (LGPM)

LGPM: At level 2 we add student characteristics for both the slope and intercept. At Level 3 we add school characteristics for everything assumed to vary across schools. Advantages of LGPM: Direct measure of growth. Do not need complete outcome data for each student. Growth model estimates are robust to sample sizes down to 30 per school (grade). Growth model estimates are robust to missingness (given missingness is not systematic – unless can model systematic part).

Longitudinal Growth models Advantages of multiple time points: Avoid spurious negative correlation between pre-test and gains In fact evidence suggest that as occasions are added to the mode; the correlation between initial status and growth (in absolute value) decreases. Generate more precise estimates of change as add occasions.

Multiple occasions allow for a more accurate portrayal of change over time.

Generate more reliable estimates with additional occasions The higher the reliability, the greater the ability to detect true differences among schools.

Use Growth Models to Examine specific dimensions of school quality Can examine how specific aspects of schools vary among schools and estimate value added for these specific indicators of school quality. E.G. Can examine the gap in achievement growth rates between title1 students and non-title1 students at each school. These will vary normally around a district or state average learning gap

Mean =  20, standard deviation var(U 20 ) =  20 1/2

Conclusions Smaller VA estimates indicate that the model is able to account for more unexplained variation among schools. May be desirable to look at the variation in specific aspects – e.g. performance gaps or learning gaps, and use these to examine school quality and performance. Longitudinal modeling can become quite complex (although simple models may fit the data empirically (despite theoretical assumptions). Hence, the cookie cutter approach to VA modeling may not be a good idea because someone needs to decide how well the model fits etc., which questions can be addressed, what assumptions are underlying the model, and what happens when the assumptions are violated.