Need for space: The key distance effect depends on spatial stimulus configurations

In numerous psychological experiments, participants classify stimuli by pressing response keys. According to Lakens, Schneider, Jostmann, and Schubert (2011), classification performance is affected by physical distance between response keys--indicating a cognitive tendency to represent categories in spatial code. However, previous evidence for a key distance effect (KDE) from a color-naming Stroop task is inconclusive as to whether: (a) key separation automatically leads to an internal spatial representation of non-spatial stimulus characteristics in participants, or if the KDE rather depends on physical spatial characteristics of the stimulus configuration; (b) the KDE attenuates the Stroop interference effect. We therefore first adopted the original Stroop task in Experiment 1, confirming that wider key distance facilitated responses, but did not modulate the Stroop effect as was previously found. In Experiments 2 and 3 we controlled potential mediator variables in the original design. When we did not display instructions about stimulus-response mappings, thereby removing the unintended spatial context from the Stroop stimuli, no KDE emerged. Presenting the instructions at a central position in Experiment 4 confirmed that key separation alone is not sufficient for a KDE, but correspondence between spatial configurations of stimuli and responses is also necessary. Evidence indicates that the KDE on Stroop performance is due to known mechanisms of stimulus-response compatibility and response discriminability. The KDE does, however, not demonstrate a general disposition to represent any stimulus in spatial code.