The Hedge and Marsh Task, also sometimes called the Reverse-Simon Task, is a choice reaction time task where there is dimensional overlap between the irrelevant stimulus and the response and between the relevant stimulus and the response, but there is no dimensional overlap between the two stimulus dimensions. This is accomplished by having two separate but correlated response dimensions. In the dimensional overlap taxonomy, it is considered a Type 5 task.
Hedge and Marsh (1975) examined a variation of the standard Simon task in which subject pressed a left key or a right key in response to the color of a stimulus (red or green), while the stimulus color appeared on the left or right side of the screen; in addition, however, they color-coded the response keys so that the response set not only had dimensional overlap with the irrelevant stimulus set (due to left-right position) but also with the relevant stimulus set (due to color). In addition to S-R consistency, therefore, the instructions given to subjects could be either congruent (e.g. “press the key with the same color as the stimulus”) or incongruent (“press the key with the opposite color from the stimulus”). There is no dimensional overlap between the relevant stimulus set (color) and the irrelevant stimulus set (location), and therefore there is no S-S consistency.
Because the Hedge and Marsh task has two types of overlap, there are two compatibility effects that can be obsereved. First, the experiment showed a standard mapping effect: overall reaction times were much faster wth compatible (i.e congruent) mapping, i.e. when each color stimulus was mapped to a correspondingly colored response (e.g. “press the key with the same color as the stimulus”) than when the mapping was incompatible, that is incongruent (e.g. “press the key with the opposite color from the stimulus”).
The effect of irrelevant stimulus-response consistency, however, was unusual. Hedge and Marsh found the surprising result that the irrelevant S-R consistency effect reverses when the mapping instructions are incompatible. That is, when the instruction were to press the key with the same color as the stimulus, the subjects responded faster when the stimulus position was consistent with the response position (the normal S-R consistency effect); however, when the instructions were to press the key with the opposite color from the stimulus, subjects responded faster when the stimulus position was inconsistent with the response key position (a reversed S-R consistency effect). This effect of the irrelevant stimulus in a Hedge and Marsh task is called a Hedge and Marsh effect or a reverse-Simon effect.
The reverse-Simon effect was met with some initial skepticism. Simon, Sly and Vilapakkam (1981) suggested that the result might just be an artifact of the way that the display and the response keys were arranged: if the left key is green and the right key is red, then a display that shows a green patch on the left side or a red patch on the right side matches the arrangement of the keys. On the other hand, a display that shows a red patch on the left and a green patch on the right does not. They called this display-control correspondence, suggesting that people reach faster when the displays and controls correspond than when they do not. This would give the appearance of a reverse irrelevant S-R consistency effect.
However, Kornblum and Stevens (1999) replicated the conditions for the Hedge and Marsh (1975) task with auditory stimuli and verbal responses. Stimuli were the words “boot” and “gate” spoken at either a high pitch or a low pitch. Subjects had to respond to either the pitch (making the word irrelevant) or the word (making the pitch irrelevant), by making one of two responses: saying the word “boot” at a high pitch or the word “gate” at a low pitch. This task has the same structure and dimensional overlap as the standard Hedge and Marsh task, making it a Type 5 task in the Dimensional Overlap Taxonomy. However, there was naturally no display and there were no controls in the experiment, and therefore there could be no display-control correspondence. Despite the fact that this task had no display or controls, both the regular S-R consistency effect (for compatible mappings) and the reversed S-R consistency effect (for incompatible mappings) were obtained. (See also DeJong, Liang, & Lauber, 1994, for another attempt to remove display-control correspondence from the task.)
Despite initial suspicions, the reverse S-R consistency effect has also been replicated under a wide variety of conditions and manipulations (e.g. Berbner, 1979; DeJong, Liang & Lauber, 1994; Hasbroucq & Guiard, 1991; Lu & Proctor, 1994; Smith & Brebner, 1983; Zhang, 1999). A number of explanations have been suggested to account for this reversal of the S-R consistency effect in the Hedge & Marsh task, although it is still a matter of much debate.
It should be pointed out that a number of studies have incorrectly claimed that they replicate the Hedge and Marsh task in the auditory domain (e.g. Arend & Wandmacher, 1987; Proctor & Pick, 1999;Ragot & Guiard, 1992). In these tasks, the relevant stimuli, irrelevant stimuli, and responses are all characterized by spatial location. For example, the word “left” or “right” might be spoken into the left or the right ear, requiring a response of a left or right key press. In these tasks, all three dimensions overlap with one another, including the two stimulus dimensions, which are both spatial. As a result, these are Type 8 tasks in the dimensional overlap taxonomy and should be considered variants of the Stroop task, not variants of the Hedge and Marsh task.
NOTE: Sometimes you will see researchers use the term “Simon task” to refer to this task. However, in the Dimensional Overlap taxonomy, the standard Simon task is not a Type 5 task: it is a Type 3 task. Although the standard Simon task does have an irrelevant stimulus dimension (usually location) that overlaps with the response, it does not have any overlap between the relevant stimulus dimension and the response. In a traditional Simon task, the mapping between stimulus and response is completely arbitrary. The dimensional overlap model therefore contends that the cognitive processing in the Simon task and the Hedge and Marsh task are fundamentally different, because of this difference in relevant stimulus-response overlap.