GESIS Leibniz Institute for the Social Sciences: Go to homepage

Psychometrics - Data Analysis of Items and Scales

Psychometrics refers to the methodology of measuring individual psychological characteristics (e.g. traits, attitudes, competencies and behaviours). It offers basic methods for the development and evalaution of measurement instruments as well as general procedures for controlling the quality of measurements. We consult and support you in the psychometric evaluation of existing data sets on measurement instruments and in the development of new (short) scales based on the data you have collected. In our consultation, we stress the selection of individually suitable statistical models (e.g. latent variable models such as structural equation models or item response theory models), especially with regard to the development and evaluation of your measurement instruments as well as their evaluation and further data-analytical use. 

Our Consulting Services for Psychometric Analysis, Validation and Testing

Psychometric item analysis for testing measurement instruments

We support you in calculating psychometric indicators for the quality of individual items and scales on the basis of your collected data. In addition, we consult you individually on suitable quantitative data analysis methods for item analysis, scale shortening and permissible scale value calculations (e.g. through exploratory and confirmatory factor analysis, factor loadings, ant-colony optimisation, etc.).

Psychometric testing of measurement invariance

We consult and support you in the quantitative analysis of measurement equivalence or measurement invariance between groups (e.g. gender, age, education, languages) as well as longitudinal measurement invariance (as a prerequisite for latent change measurement), commonly known in item response theory as Differential Item Functioning (DIF).

Psychometric validation of measurement instruments

We consult you on the application of suitable quantitative methods to check your data for reliability, validity, dimensionality and model fit (structural equation models/ item response theory). In addition, we also support you in carrying out these analyses and in interpreting the results, if desired.

Basic Knowledge

We have compiled a range of information and practice-oriented assistance for frequently requested advisory topics. These already answer a large number of questions that arise during the preparation and analysis of data. You can find them under Basic knowledge.

Consulting Fees

Our counselling as help for self-help is free of charge. We are happy to support you with more in-depth services, for which costs will be incurred. You can find more information in our cost overview (109 kB)

Raykov & Bluemke (2021): Examining Multidimensional Measuring Instruments for Proximity to Unidimensional Structure Using Latent Variable Modeling.