Application of Skill Theory to compare scientific reasoning of young children in different tasks

In order to study the development of scientific reasoning in children, it is necessary and also challenging to compare performances of children over different tasks and/or situations. Previous research indicates that children's performance is highly influenced by the context, which stresses the need for a task-independent measure. In this paper, skill theory (Fischer, 1980) is used as a developmental ruler. Various types of analyses of the cognitive content and structure of scientific explanations in young children are presented, which show how a coding scheme based on skill theory can provide a more thorough and multifaceted view of the child's cognitive development than standard measurements or instruments. We also demonstrate how skill theory can be used in the context-specific nature of cognitive performances and fluctuations in performance level. Skill theory promises to be a good system as a less task-specific performance measure, both in research contexts and in educational settings. Where: Netherlands journal of Psychology, Volume 67, 9-19 Application of skill theory to compare scientific reasoning of young children in different tasks Netherlands journal of Psychology | Skill theory as a performance measure 9 Science education for pre-schoolers has been highly promoted in the last decade (e.g. Eshach & Fried, 2005). Studies on science education often focus on scientific reasoning. The term 'scientific reasoning' can refer to two separate types of reasoning (Crowley & Galco, 2001; Zimmerman, 2000). The first type concerns the experimentation skills of children (e.g. Klahr & Nigam, 2004; Schauble, 1996; see also the review of Zimmerman, 2000). The second form, which this article focuses on, refers to the conceptual (change in) understanding of more specific scientific phenomena. Intervention studies have focused on both kinds of reasoning. In many cases, researchers use highly specific performance measures. In a previous study, we argued that the level of scientific reasoning and the effectiveness of an intervention (aimed at enhancing children's level of scientific reasoning) cannot be measured by studying one single task or task domain, since a task can be seen as an important contributo