Mechanisms underlying transfer of task-defined rules across feature dimensions

Transfer of an implied incompatible spatial mapping to a Simon task

When location words left and right are presented in left and right locations and mapped to left and right keypress responses in the Hedge and Marsh (1975) task (Arend & Wandmacher, 1987), a compatible mapping of words to responses yields a benefit for stimulus-response location correspondence (sometimes called the Simon effect), whereas an incompatible mapping yields a benefit for noncorrespondence (called the Hedge and Marsh reversal). Experiment 1 replicated the correspondence benefit and its reversal by using Chinese location words [symbol: see text] (left) and [symbol: see text] (right) in the Hedge and Marsh task. Experiments 2 and 3 examined whether the tendency to respond with the noncorresponding response when the mapping is incompatible transfers to the task version in which the mapping is compatible, and Experiment 4 examined whether transfer similarly occurs from the compatible mapping to the task version with incompatible mapping. Transfer of the incompatible relation was apparent in a lack of correspondence benefit when the mapping was changed to compatible, but transfer of the compatible relation to the incompatible mapping did not occur. The results suggest that an association between noncorresponding stimulus-response locations is acquired when the word-response mapping is incompatible, even though this relation is only implicit, regardless of whether through misapplication of a logical recoding rule or spatial representations shared by the locations and words. These associations then continue to affect processing of location when the mapping is compatible.

Role of stimulus and response feature overlap in between-task logical recoding

The reversal logical recoding rule (i.e., "respond opposite") induced by an incompatible task (e.g., a task requiring to respond to red or green stimuli by pressing a key of the alternative colour compared to that of the stimulus) can be transferred to another task when the two tasks are combined in a task-switching paradigm. When the task to which the rule is transferred is a Simon task, this causes the disappearance of the typical advantage for responses that spatially correspond to the stimulus, or even results in an advantage for spatially noncorresponding responses. The present study aimed at investigating whether the transferred rule is independent of the specific stimulus and response dimensions for which it has been created. Previous studies suggest that when a Simon task is coupled with a colour incompatible task, the Simon effect may disappear or reverse even when stimuli in the two tasks, apart from being both visual and appearing on the same computer screen, have no other features in common. Results of the present study corroborate the hypothesis that feature overlap between stimuli is not necessary for the between-task transfer of the logical rule. However, an overlap between the representations of responses appears to be crucial. No modulation of the Simon effect was observed when the Simon task required bimanual responses while the colour-compatibility task required either vocal responses or responses executed with the two feet. In contrast, we did observe such a modulation when the discriminative response dimension and the effectors/response device were the same in the two tasks, even though these two tasks provided for different stimuli.

Influence of short incompatible practice on the Simon effect: transfer along the vertical dimension and across vertical and horizontal dimensions

In spatial compatibility and Simon tasks, the response is faster when stimulus and response locations are on the same side than when they are on opposite sides. It has been shown that a spatial incompatible practice leads to a subsequent modulation of the Simon effect along the horizontal dimension. It has also been reported that this modulation occurs both along and across vertical and horizontal dimensions, but only after intensive incompatible training (600 trials). In this work, we show that this modulatory effect can be obtained with a smaller number of incompatible trials, changing the spatial arrangement of the vertical response keys to obtain a stronger dimensional overlap between the spatial codes of stimuli and response keys. The results of Experiment 1 showed that 80 incompatible vertical trials abolished the Simon effect in the same dimension. Experiment 2 showed that a modulation of the vertical Simon effect could be obtained after 80 horizontal incompatible trials. Experiment 3 explored whether the transfer effect can also occur in a horizontal Simon task after a brief vertical spatial incompatibility task, and results were similar to the previous experiments. In conclusion, we suggest that the spatial arrangement between response key and stimulus locations may be critical to establish the short-term memory links that enable the transfer of learning between brief incompatible practices and the Simon effects, both along the vertical dimension and across vertical and horizontal dimensions.

Transfer of learning in choice reactions: The roles of stimulus type, response mode, and set-level compatibility

The Simon effect refers to the advantage of responding to spatially compatible stimuli. This effect can be eliminated or even reversed to favor spatially incompatible stimuli after participants practice a choice-reaction task with spatially incompatible mappings (e.g., pressing left and right keys to stimuli on the right and left, respectively). This transfer of incompatible spatial associations has been observed under conditions in which responses were made manually (e.g., keypresses, moving a joystick). The present study used vocal responses to reveal the primary determinants of the transfer effect, dissociating the influences of stimulus type, response mode, and their interaction (set-level compatibility). The results suggest that contextual match between the practice and transfer tasks with respect to stimulus type and response mode determined transfer of incompatible associations to the Simon task, and stimulus type determined the efficiency of acquiring new associations. However, there was little evidence that set-level compatibility plays any major role in either acquisition or transfer of spatial associations.