Validation of new tablet-based problem-solving tasks in primary school students.

Problem-solving is an important skill that is associated with reasoning abilities, action control and academic success. Nevertheless, empirical evidence on cognitive correlates of problem-solving performance in childhood is limited. Appropriate assessment tools are scarce and existing analog tasks require extensive coding. Thus, we developed and validated new tablet-based versions of existing analog tasks assessing technical problem-solving with gear construction tasks. To validate these tasks, 215 children (6-8 years) performed the problem-solving tasks in both modalities (analog, digital). To investigate whether performances in both modalities were correlated with other cognitive abilities, participants performed three additional tasks assessing language, reasoning and problem-solving. Structural equation modelling showed that performance was substantially correlated across modalities and also correlated with language, reasoning and another problem-solving task, showing the convergent validity of the digital tasks. We also found scalar measurement invariance across task modalities indicating that both task versions can be used interchangeably. We conclude that both versions (analog and digital) draw on similar cognitive resources and abilities. The analog tasks were thus successfully transferred to a digital platform. The new tasks offer the immense benefits of digital data collection, provide a valid measuring tool advancing problem-solving research in childhood and facilitate the application in the field, e.g., in the classroom.