Interactive problem solving revisited: Replicating the effects of interactivity using matchstick algebra problems.

Abstract

The concept of embodied cognition has driven researchers to the creation of novel experimental procedures that would make it possible to evaluate the impact of motor activity on cognitive processes. Among those isthe integration of motor activity into the process of problem solving [Weller, Villejoubert, and Vallée-Tourangeau, 2011]. The 'interactive' condition within the Weller et al.[2011] study enabledthe participants to interact with physical objects (sticks) while problem-solving. This experimental manipulation substantially increased the success rate for most matchstick algebra problems. Given the differential source of difficulty in these problems [Knoblich et al., 1999], we hesitate to suggest that there is a single (or general) type of motor activity that could explain the uniform effect of interactivity on the success rate.In order to verify these results and to evaluate the effect of interactive problem solving, we carried out a conceptual replication of Weller et al. [2011] study. Three experimental variants were applied: the first variant aimed to evaluate the interactivity effect on the solving, the second variant attempted to vary the amount of motor activity produced by the solver, while the third variant aimedto maximize interactivity through the application of virtual reality technologies. In addition to the success rate, we considered the number of epistemic and pragmatic actions [Kirsh, Maglio, 1994] visible in video-recordings of the individual experimental instances. An increase in epistemic actions could serve as an indicator of a transfer of a part of the cognitive functions onto the material environment. However, our results failedto replicate the above mentioned effects [Weller et al., 2011]. Despite a substantially higher number of epistemic actions in the second experiment, the success ratere mained unaffected. One possible explanation is that the mere possibility of physical manipulation does not automatically makethe solving process interactive. We discuss the challenges and outlook of the interactive problem solving, as well asits role in the studies of embodied cognition.