Would you prefer your coefficients with a little bias, or rather with a lot of variance?

in European Journal of Psychological Assessment (2018), 34

Executive functions as moderators of the worked example effect on statistical reasoning. A replication study

Poster (2018)

Students' exploration strategies in computer-simulated complex problem environments: A latent class approach

in Computers and Education (2018), 126

Complex problem solving (CPS) is considered an important educational achievement indicator. Previous research has indicated that CPS performance depends to a substantial extent on the way students explore ... [more ▼]

Complex problem solving (CPS) is considered an important educational achievement indicator. Previous research has indicated that CPS performance depends to a substantial extent on the way students explore problem environments. In this study, we investigated qualitative differences in the way students interact with such environments. In a sample of N = 2226 Hungarian students in Grades 6 to 8, we applied a latent class approach to investigate the use of the principle of isolated variation as an exploration strategy across six CPS tasks that were developed within the MicroDYN approach. Six qualitatively different class profiles emerged: proficient explorers, intermediate explorers, low-performing explorers, rapid learners, emerging explorers, and nonpersisting explorers. We further validated the profiles by comparing the latent classes with regard to students' overall CPS performance and additional indicators of task exploration. In analyzing age-related and gender differences on a cross-sectional level, we found only a small progression toward better performing class profiles from Grade 6 to Grade 8 (e.g., 14.6% of students in Grade 6 were proficient explorers vs. 24.6% in Grade 8; 27.1% of students in Grade 6 were low-performing explorers vs. 25.8% in Grade 8), and there were no substantial gender differences. This study contributes to the understanding of how students interact with complex problems and is the first to address whether variations in these behaviors indicate qualitatively different levels of strategic behavior. We discuss the theoretical underpinnings and potential of identifying class profiles of students' exploration behavior in the field of educational psychology. [less ▲]

Computer-based assessments in PISA. Challenges and potentials for research and policy

Speeches/Talks (2017)

Komplexes Problemlösen im Kontext internationaler Large-Scale Assessments

Speeches/Talks (2017)

Intrinsische Motive umfassend und ökonomisch messen. Entwicklung und Validierung von 16 Kurzskalen für den Einsatz in der Motivationsforschung.

Scientific Conference (2017, September)

Dimensional and social comparisons effects on domain-specific self-concepts and interests: An examination of the generalized I/E model in first- and third-graders

Scientific Conference (2017, September)

Selektives Enkodieren in figuralen Matrizen.

Scientific Conference (2017, September)

Problem solving in PIAAC. Fundamentals and vision.

Speeches/Talks (2017)

Relationship satisfaction: Establishing measurement and structural invariance across men with gay and straight identity

Poster (2017, September)

The Relationship Assessment Scale (RAS; Hendrick, 1988) is a widely employed, seven-item measure of relationship satisfaction. However, research on its measurement invariance across people differing in ... [more ▼]

The Relationship Assessment Scale (RAS; Hendrick, 1988) is a widely employed, seven-item measure of relationship satisfaction. However, research on its measurement invariance across people differing in their sexual orientation identity has yet to be conducted. Consequently, it is still an open question whether the RAS can be used for comparative research across gay and straight people. This study starts filling this gap in examining RAS' measurement invariance across gay and straight men. To this end, we drew on a sample of 644 German men (half gay/straight), who responded to the German-version RAS (Hassebrauck, 1991), which has been extensively validated and found to be equivalent to its English counterpart. A stepwise multigroup confirmatory factor analysis was used to examine configural, metric, and scalar invariance as well as the structural invariance of latent factor variances and means across groups. Configural, metric, and partial scalar measurement invariance (by freeing one item intercept) could be established, indicating that the RAS enables comparative research targeting variances and means. Further, structural invariance testing revealed non-invariant variances but invariant latent means across groups, indicating that gay and straight men's reported relationship satisfaction is equally high albeit differently distributed. [less ▲]