26 August 2016
Testing Survey Data for Measurement Equivalence across Countries and Time
Determining whether people in certain countries, or at different time points score differently in measurements of interest, or whether constructs relate differently to each other across nations can indisputably assist in testing theories and advancing our sociological knowledge. However, meaningful comparisons require equivalent measurements of these constructs. This is especially true for subjective attributes such as values, attitudes, opinions, or behavior. In this course, we first discuss the meaning of the cross-group measurement equivalence, look at possible sources of nonequivalence, and suggest ways to prevent it. Next, we examine the social science methodological literature for ways to empirically test measurement equivalence and conduct such tests during the exercises with survey data using the software package Mplus. We focus on the multiple group confirmatory factor analysis (MGCFA) method. Furthermore, we consider what may be done when equivalence is not supported by the data and conclude with a discussion of recent developments such as the Bayesian estimation procedure to test for approximate invariance or multilevel structural equation modeling that can potentially be used to explain non-invariance. These methods offer exciting directions and solutions for future research in cross-group measurement equivalence assessment. Finally we discuss if time allows further issues, such as the newly developed alignment procedure or how multiple group structural equation modeling (rather than MGCFA) can be applied to cross-national data sets if the number of units (here countries) is too small for a multilevel analysis.
You can find the full syllabus of the course with complete information on the topics, literature, and day-to-day schedule on https://training.gesis.org.
Prof. Dr. Eldad Davidov is professor of sociology at the University of Zurich, Switzerland and vice-president of the European Survey Research Association;
Prof. Dr. Peter Schmidt is professor emeritus of social research methods at the University of Giess
Participants will find the course useful if:
- they are interested in conducting meaningful comparative research across countries, over time or across other social groups such as natives and immigrants;
- they are interested in understanding potential problems in comparative research, and in studying and applying the methodological prerequisites for meaningful comparisons across groups.
- Knowledge of and some experience with confirmatory factor analysis and structural equation modeling. This could be acquired in the course “Introduction to Structural Equation Modeling: Confirmatory Factor Analysis with Mplus” in week 1.
- We will use the software package Mplus. A short introduction to the use of the program will be provided during the first exercise. Knowledge of Mplus could also be acquired in the short course “Introduction to Data Analysis Using MPlus" in week 0.
- Preferably (not obligatory) some knowledge and experience with measurement error in survey research. This could be acquired in the course “Understanding and Modeling Measurement Error in Social Surveys” in week 2. This course will also use the software package Mplus. Furthermore, this course will touch upon the topic of equivalence on the last day.
By the end of the course participants will:
- know how to test for measurement invariance across groups using various types of multigroup confirmatory factor analyses with the software package Mplus;
- have gained practical experience in testing for measurement invariance and computing latent means for different groups (e.g., countries and time-points);
- know how to deal with non-invariant measurements.
- If time allows, learn to run multiple group structural equation models for comparative research to test complex theories after testing for measurement invariance.
- Certificate of attendance issued upon completion.
- 2 ECTS points via the University of Mannheim for regular attendance and satisfactory work on daily assignments (EUR 20).
- 4 ECTS points via the University of Mannheim for regular attendance and satisfactory work on daily assignments and for submitting a paper/report of about 5000 words to the lecturer(s) up to 4 weeks after the end of the summer school (EUR 50).
EUR 250: Student/PhD student rate.
EUR 350: Academic/non-profit rate.
Early bird discount: EUR 50 for applicants who book and pay by April 30.
The rates include the tuition fee, course materials, the academic program, access to library and IT facilities, coffee/tea, and a number of social activities.
10 DAAD scholarships are available via the Center for Doctoral Studies in Social and Behavioral Sciences (CDSS) at the University of Mannheim.