The influence of representation control on task performance and learning

Abstract

Although infographics are often used in digital newspapers or magazines to communicate information, users frequently have difficulties identifying the relevant information. Moreover, the spatial organization of the information in the infographic and the users ́ requirements rarely match. One way to overcome these challenges might be to offer the users interactive control over the format and content of the representations. Representation control can be used to modify the infographics in order to externalize transformations that would usually have to be performed cognitively. This can enhance task performance, but it is unknown whether it facilitates or hinders learning of the underlying structure of the task. The objective of this dissertation was to examine the usage of representation control options and its effects on task performance and learning. In order to do so, I conducted three experimental studies with 650 participants in total. All studies revealed that users took the option to reorganize and reduce information in infographics in a strategic way. However, the availability of representation control improved only the performance in the task with a specific goal but not in the goal-free task. To test whether representation control also influences learning, I adapted a two- phase design with a practice and a testing phase in the third study. This experimental series revealed that practicing with representation control in general did not benefit learning. Participants who did not have representation control during testing could not benefit from their prior experience with representation control neither when solving the same task nor when doing a near transfer task. Comparing different forms of representation control revealed that the more automated types led to the best performance during the practice phase, but not during the testing phase. The results of my dissertation suggest that it is important to consider the type of task to be performed and to differentiate between task performance and learning when investigating representation control or deciding whether representation control should be offered to the user.