How to Analyze Large-Scale Assessments Data from Matrix Booklet Sampling Design: Focus on Psychometrics behind and Hands-on Analysis Using Actual Sample Data

This course available at no cost 

The goal of this course is to provide researchers who are interested in large-scale assessments data from matrix booklet sampling design (e.g. NAEP, TIMSS, PISA, or PIRLS) with the practical knowledge and tools to analyze such data.  This course will provide participants with training on the AM analysis tool as well as the psychometrics behind and sampling design of typical large-scale assessments. AM is a free statistical software package developed for the analysis of large- scale assessment data with complex sampling and the matrix booklet sampling design.

Using the publicly-released NAEP mini-sample data, instructors will also provide participants with data analysis strategies, including the marginal maximum likelihood approach to scale score estimation, use of appropriate sampling weights, and appropriate variance estimation. 

This course includes hands-on exercises that require the use of a web browser.

    Topics convered in this course include:

  • Welcome - Emmanuel Sikali 
  • Psychometric Models Used in Large Scale Assessment - Emmanuel Sikali
  • Sampling Concepts for NAEP Analysts - Emmanuel Sikali
  • Direct Estimation using NAEP Data with AM (SAS or SPSS) - Emmanuel Sikali 
  • Data Access and Funding - Emmanuel Sikali


  • Emmanuel Sikali, National Center for Education Statistics
  • Andrew Kolstad, National Center for Education Statistics
  • Young Yee Kim, American Institutes for Research


Course No. 105-2013


VRLC Course
Recorded 04/30/2013
Recorded 04/30/2013