Магистратура
2026/2027
Углубленные методы в психометрике и анализе данных
Статус:
Курс по выбору (Обучение и оценивание как наука)
Кто читает:
Департамент образовательных программ
Где читается:
Институт образования
Когда читается:
2-й курс, 2, 3 модуль
Охват аудитории:
для всех кампусов НИУ ВШЭ
Язык:
английский
Кредиты:
6
Контактные часы:
64
Course Syllabus
Abstract
This course aims to introduce advanced statistical methods and statistical models which are used in psychometrics and data analysis of psychological, sociological and educational data. Students will learn different approaches to latent variables analysis, such as Confirmatory Factor Analysis (CFA) and some models within Item Response Theory (IRT) framework such as bifactor IRT models etc. Topics of structural equation modelling, analysis of mediated and moderated relations between latent variables will be also introduced. Course continues with discussion of generalized linear models and extension of this models – generalized linear mixed effects models (GLMM). Different types of GLMM, their assumptions and application in social data analysis will be reviewed. At the end of the course some models with discrete latent variables (Latent Class Analysis, Cognitive Diagnostic Models) will be discussed. During the course students learn to select, apply and discuss the results of statistical models appropriate for addressing a given research problem.Prerequisites:1) Basic knowledge of statistics (especially regression analysis and factor analysis)2) Basic knowledge of Item Response Theory (recommended, but now required)3) Experience of working with base R.
Learning Objectives
- To familiarize students with Generalized Linear Mixed Models (GLMM) when the individuals are clustered (e.g., students belonging to different schools), introduce the idea of fixed-effects and random-effects terms in models.
- To familiarize students with Confirmatory Factor Analysis (CFA) Structural Equation Modeling (SEM) paradigm, to demonstrate the concepts of moderation and mediation in SEM.
- To familiarize students with special chapters of individual differences modeling in social sciences: Measurement Invariance (MI) and Differential Item Functioning (DIF) analysis, Latent Class Analysis, Latent Profile Analysis, Cognitive Diagnostic Modeling.
- To illustrate students how they can use R software to perform GLMM, CFA and SEM analysis.
Expected Learning Outcomes
- Calibrates, selects, improves model quality, and interprets CFA models
- Calibrates, re-norms and interprets parameters of IRT models in CFA parametrization
- Performs measurement invariance analyses for different types of data and compares the fit of nested models
- Conducts path analysis, including mediation and moderation
- Calibrates and interprets alternative CFA specifications (e.g., bifactor, second-order, multidimensional)
- Conducts latent class and latent profile analysis for identifying unobserved subgroups
- Conducts cognitive diagnostic modeling (CDM)
- Specifies and interprets fixed and random effects in multilevel models
- Calibrates Rasch models and explanatory IRT models within the multilevel framework
Course Contents
- Multilevel Modeling: Fixed and Random Effects
- Explains IRT as a special case of multilevel modeling. Explanatory IRT models.
- Confirmatory Factor Analysis (CFA) and Structural Equation Modeling (SEM)
- Relations between IRT and CFA
- Alternative CFA specifications
- Measurement invariance (MI)
- Path analysis and SEM
- Categorical and ordinal latent variables
Interim Assessment
- 2026/2027 3rd module0.24 * Written homework + 0.24 * Final Test + 0.12 * R Exercises + 0.4 * Exam
Bibliography
Recommended Core Bibliography
- Applied latent class analysis, , 2002
- Hierarchical linear models : applications and data analysis methods, Raudenbush, S. W., 2002
- Multilevel analysis : techniques and applications, Hox, J., 2002
Recommended Additional Bibliography
- Handbook of structural equation modeling, , 2012
- Principles and practice of structural equation modeling, Kline, R. B., 2011