Dynamisch testen van inductief redeneren: Een studie naar de betrouwbaarheid en validiteit van een nieuw meetinstrument

This thesis evaluated the reliability and validity of a newly developed dynamic test intended to measure inductive reasoning in children in the second grade of Dutch primary education. Participants were recruited through convenience sampling across primary schools in the Netherlands. In total, 85 children (mean age was 7,9 years) from six schools participated in the study. Reliability was assessed using item response theory, specifically the two-parameter logistic (2PL) model. Convergent and divergent validity were examined by correlating scores on the different components of the dynamic test with scores on the Raven’s Standard Progressive Matrices and a Cito reading comprehension test, respectively. These instruments are well-established measures with strong psychometric properties, supporting their use as external criteria for validity evaluation. Results indicated that the reliability of most components of the dynamic test was satisfactory for children with higher ability levels,...Show moreThis thesis evaluated the reliability and validity of a newly developed dynamic test intended to measure inductive reasoning in children in the second grade of Dutch primary education. Participants were recruited through convenience sampling across primary schools in the Netherlands. In total, 85 children (mean age was 7,9 years) from six schools participated in the study. Reliability was assessed using item response theory, specifically the two-parameter logistic (2PL) model. Convergent and divergent validity were examined by correlating scores on the different components of the dynamic test with scores on the Raven’s Standard Progressive Matrices and a Cito reading comprehension test, respectively. These instruments are well-established measures with strong psychometric properties, supporting their use as external criteria for validity evaluation. Results indicated that the reliability of most components of the dynamic test was satisfactory for children with higher ability levels, but insufficient for children with average or lower ability levels. Furthermore, while convergent validity was supported, strong correlations were found between reading comprehension and performance on several components of the dynamic test, which challenges its divergent validity. In particular, the training component – intended to distinguish dynamic testing from conventional static testing – showed a strong correlation with reading comprehension (r = .50). This suggests that the training component may rely not only on inductive reasoning but also on reading comprehension skills. Suggestions for further development of the dynamic test are provided, along with implications for future research. Specifically, future test development should focus on designing a training component that is less dependent on reading skills. Additionally, future research should employ a study design that allows for a more thorough evaluation of reliability and validity, as certain aspects of the current design limited these analyses.Show less

Leiden University Student Repository

Documents

In Collections

Dynamisch testen van inductief redeneren: Een studie naar de betrouwbaarheid en validiteit van een nieuw meetinstrument