On the Validity of Cognitive Interpretations of Scores From Alternative Concept-Mapping Techniques
Titel:
On the Validity of Cognitive Interpretations of Scores From Alternative Concept-Mapping Techniques
Auteur:
Ruiz-Primo, Maria Araceli Shavelson, Richard J. Li, Min Schultz, Susan E.
Verschenen in:
Educational assessment
Paginering:
Jaargang 7 (2001) nr. 2 pagina's 99-141
Jaar:
2001-05-01
Inhoud:
The emergence of alternative forms of achievement assessment and the corresponding claims that they measure "higher order thinking" rouse the need to examine their cognitive validity. In this article, we provide a framework for examining cognitive validity claims that includes conceptual and empirical analyses and use it to evaluate the validity of a "connected understanding" interpretation of 3 concept-mapping techniques: (a) construct-a-map from scratch, in which students constructed a map using concepts provided; (b) fill-in-the-nodes, in which students filled in a 12-blank-node skeleton map with concepts provided; and (c) fill-in-the-lines, in which students filled in a 12-blank-line skeleton map with a description of the relation provided for each pair of connected concepts. The first technique imposes little structure on the students (low-directedness), whereas the other 2 techniques are much more structured (high-directedness). The framework focuses on the analysis of the mapping tasks' intended demands (conceptual analysis), and the tasks' correspondence with inferred cognitive activities and performance scores (empirical analyses). To infer cognitive activities, we examined respondents' (teachers, expert students, and novice students) concurrent and retrospective verbalizations in performing the mapping tasks and compared the directedness of the mapping tasks, the characteristics of verbalization, and the scores obtained across techniques. We concluded that the framework allowed us to determine that (a) the 3 mapping techniques provided different pictures of students' knowledge, and (b) inferred cognitive activities across mapping techniques differed in relation to the directedness of the task. The low-directed technique provided students with more opportunities to reveal their conceptual understanding (explanations and errors) than did the high-directed techniques.