Bertling, Jonas Pablo: Measuring reasoning ability : applications of rule-based item generation. 20.2.2014
Inhalt
- General introduction
- Theoretical background
- Reasoning ability as a construct
- Cross-cultural validity of reasoning tests
- The term ``culture'' in cross-cultural studies
- The challenge of ``culture-fair'' testing
- Cognitive and cultural complexity as possible factors for bias
- Item difficulty modeling and rule-based automatic item generation
- Explanatory item response modeling
- Differential item and facet functioning
- The Figural Analogy Test (FAT): Item generation and construct validation
- Introduction
- Figural-spatial analogies as indicators of fluid intelligence
- Rule-based generation of figural analogy items
- Research questions
- Method
- Development of the new item-generative framework
- Specific hypotheses
- Sample
- Instruments and procedure
- Results
- Discussion
- The Number Series Test (NST): Item generation of parallel forms
- Introduction
- Number series items as indicators of reasoning ability
- An information processing model for number series
- Item difficulty modeling of number series items: Previous attempts and problems
- Research questions
- Method
- Results
- Discussion
- A cross-cultural investigation of the Latin Square Task
- Introduction
- The Latin Square Task
- Similarities to the popular number placement game SUDOKU
- Cross-cultural validity of the LST
- Research questions
- Method
- Results
- Psychometric properties of the LST for the two samples
- Item difficulty modeling for the two samples
- Differential item functioning analyses
- Differential facet functioning analyses
- Qualitative analyses of DIF in LSTs
- Discussion
- Epilogue
- Prediction of item difficulties by means of explanatory IRT models
- Understanding and enhancing construct validity
- Limitations
- References
- Study 1 – Additional Materials
- Study 2 – Additional Materials
- Study 3 – Additional Materials