BACKGROUND: A holistic approach to performance assessment recognizes the theoretical complexity of multifaceted critical thinking (CT), a key objective of higher education. However, issues related to reliability, interpretation, and use arise with this approach. AIMS AND METHOD: Therefore, we take an analytic approach to scoring students' written responses on a performance assessment. We focus on the complementarity of holistic and analytic approaches and on whether theoretically developed analytical scoring rubrics can produce sub-scores that may measure the 'whole' performance in a holistic assessment. SAMPLE: We use data from the Wind Turbines performance assessment, developed in the iPAL project this study where 55 students at a German university participated. RESULTS: The (sub)scores generated from the scoring scheme empirically reproduced the theoretically assumed structure of CT, with valid and reliable scores in a three-dimensional model. The proposed interpretation of CT as assessed with a performance assessment and measured by the rating scheme was supported preliminarily. CONCLUSION: Our results support the complementarity of holistic and analytic approaches to assessing CT. When combined, they provide interpretable scores for a complex, multifaceted construct useful in diagnostic contexts.
BACKGROUND: A holistic approach to performance assessment recognizes the theoretical complexity of multifaceted critical thinking (CT), a key objective of higher education. However, issues related to reliability, interpretation, and use arise with this approach. AIMS AND METHOD: Therefore, we take an analytic approach to scoring students' written responses on a performance assessment. We focus on the complementarity of holistic and analytic approaches and on whether theoretically developed analytical scoring rubrics can produce sub-scores that may measure the 'whole' performance in a holistic assessment. SAMPLE: We use data from the Wind Turbines performance assessment, developed in the iPAL project this study where 55 students at a German university participated. RESULTS: The (sub)scores generated from the scoring scheme empirically reproduced the theoretically assumed structure of CT, with valid and reliable scores in a three-dimensional model. The proposed interpretation of CT as assessed with a performance assessment and measured by the rating scheme was supported preliminarily. CONCLUSION: Our results support the complementarity of holistic and analytic approaches to assessing CT. When combined, they provide interpretable scores for a complex, multifaceted construct useful in diagnostic contexts.