Big-GENDER Project

Big Data Meta-Analyses of Gender Differences in Students' Achievement and Achievement Motivation Based on Large-Scale Assessments

An essential requirement for any scientific and political discourse on gender differences in school (and beyond) is a reliable body of empirical knowledge on the nature, size, variability, and moderating factors of these differences. This knowledge is highly relevant for at least three reasons: It can be used (a) to learn about gender differences before university entry as plausible antecedents of still existing gender gaps in academic fields, (b) to provide scientific evidence that can help dispel the persistent stereotypes (e.g., that only boys can excel in mathematics) that may discourage girls from pursuing careers in science, technology, engineering, and mathematics (STEM), and (c) to identify target points for evidence-based decision making in educational policy (e.g., boys from families with low socioeconomic status [SES]). The main goal of this meta-analytic big data project is therefore to provide highly robust and widely generalizable knowledge on cross-national gender differences in students’ achievement and achievement motivation (concerning means and variances). To this end, we will meta-analyze individual student data from 999 representative student samples from 112 different countries/economic regions (total N > 4 million) participating in 24 cycles of international large-scale assessments covering the period from 1995 to 2015: the Trends in International Mathematics and Science Study (TIMSS; Grades 4, 8, and 12), the Progress in International Reading Literacy Study (PIRLS; Grade 4), and the Programme for International Student Assessment (PISA; 15-year-olds). This project will be the first to quantitatively synthesize this wealth of data with meta-analytic methods. Specifically, we will conduct three domain-specific meta-analyses to examine gender differences (concerning mean levels and variability) in achievement and achievement motivation in mathematics, science, and reading, respectively. We will study students’ age and SES, the selectivity of the sample (e.g., the bottom 10% or the top 5% of the achievement distributions), sociocultural indicators of gender equality, and historical changes as moderators of gender differences. Further, we will conduct one meta-analysis to examine gender differences in achievement and motivational profiles in multiple domains among three groups of top-performing students who belong to the top 5% in mathematics, science, or reading in their respective countries. To sum up, our project will provide novel insights into cross-national, temporal, and age-related trends concerning gender differences in the general student population and among top-performing students, as well as on the complex interactions between gender, the selectivity of the sample, SES, and sociocultural indicators of gender equality. BIG-GENDER  is funded by the German Research Foundation (DFG).

Research Associate: Dr. Lena Keller (University Potsdam)

Duration: 2020 - 2023

Funding: The German Research Foundation (DFG)

Cooperation partner: Prof. Dr. Martin Brunner (initial inquirer; University Potsdam)

go to the top

BIS⁺ Project

Intelligence research in adolescence and adulthood

The BIS⁺ project focus on intelligence as maximal and typical behavior in adolescents and adults. Intelligence as a maximal behavior will be assessed following prominent models of intelligence structure research. Specifically, we will apply the Berlin Intelligence Structure Model (BIS; Jäger, 1982, 1986) and the Carroll-Horn-Cattell Model (CHC; Schneider & McGrew, 2018). Intelligence as a typical behavior describes the individual tendency to engage in and find pleasure in cognitively demanding activities (Cacioppo & Petty, 1982), also referred to as Need for Cognition (NFC). In this project, the factors of the BIS model will be theoretically transformed into CHC factors and this transformation will be empirically tested. In addition, hypotheses on differentiation effects of intelligence will be investigated. These assume that the intelligence structure shifts with age and/or ability. In the present project, this is tested with respect to the influence of the g-factor on specific cognitive performances. Finally, the relationship and interaction of intelligence as maximal behavior and NFC will be investigated. Previous findings indicate only weak to moderate correlations between the two constructs. According to investment theories, however, tighter relationships should emerge over time and in higher-performing groups. In addition, the question of how both constructs interact in predicting academic and occupational performance is explored.

Research Associate: Julian Urban M.Sc., Julian Preuß M.Sc.

Funding: Hogrefe Verlag

Project manager: Prof. Dr. Franzis Preckel, Dr. Vsevolod Scherrer, Dr. Moritz Breit

Project partner: Dr. Clemens Lechner, gesis Leibniz Institut für Sozialwissenschaften; Prof. em. Dr. Heinz Holling

go to the top

DGIT Project

Differentiation in German Intelligence Tests

Intelligence tests are among the most important and most widely used psychometric test instruments. In addition to scores for general intelligence, many of the tests also provide scores for more specific cognitive abilities, which can be used to create ability profiles. In the scoring and interpretation of the tests, it is usually assumed that both the measurement accuracy of the individual scales and the validity of the ability profiles are given across the entire spectrum of intelligence - independent of the characteristics of the person tested. However, according to differentiation hypotheses of intelligence, a subject's age and ability systematically influence the relationship between different test scales and their measurement accuracy. Empirical evidence on these hypotheses is weak due to a small number of methodologically adequate studies. Thus, it is largely unclear which differentiation effects actually exist and to what extent they can be generalized across age groups or test content. With large-scale studies based on German-language intelligence structure tests, we therefore aim to generate comprehensive empirical evidence. The results of this project will contribute significantly to nuanced, evidence-based recommendations for the construction, use, and interpretation of intelligence tests.

Research Associate: Dr. Moritz Breit

Funding: Deutsche Forschungsgemeinschaft (DFG)

Project manager: Dr. Moritz Breit, Prof. Dr. Franzis Preckel

Project partner: Prof. Dr. Franzis Preckel, Prof. Dr. Martin Brunner

Cooperation partner: Prof. Dr. Martin Arendasy, Prof. Dr. D.A. Briley, Prof. Dr. Monika Daseking, Dr. Jan-Philipp Freudenstein, Prof. Dr. Ulf Kieschke, Dr. Ludwig Kreuzpointner, Dr. Peter Melchers, Dr. Dylan Molenaar, Prof. Dr. Christoph Perleth, Dr. Stefan Schipolowski, Prof. Dr. Ulrich Schroeders, Nadine Schuchart, Dr. Franz Pauls, Prof. Dr. Gabriele Ricken, Dr. Franziska Walter, Prof. Dr. Oliver Wilhelm

Running time: april 2023 - march 2026

go to the top