Processing irrelevant location information: Practice and transfer effects in choice-reaction tasks

Left or right keypresses to a relevant stimulus dimension are faster when the stimulus location, although irrelevant, corresponds with that of the response than when it does not. This phenomenon, called the Simon effect, persisted across 1,800 trials of practice, although its magnitude was reduced. Practice with the relevant stimulus dimension presented at a centered location had little influence on the magnitude of the Simon effect when irrelevant location was varied subsequently, and practice with location irrelevant prior to performing with location relevant slowed responses. After practice responding to stimulus location with an incompatible spatial mapping, the Simon effect was reversed (i.e., responses were slower when stimulus location corresponded with response location) when location was made irrelevant. When the response keys were labeled according to the relevant stimulus dimension (the Hedge and Marsh [1975] task variation), this reversal from practice with a spatially incompatible mapping was found for both the congruent and the incongruent relevant stimulus-response mappings. Thus, task-defined associations between stimulus location and response location affect performance when location is changed from relevant to irrelevant, apparently through producing automatic activation of the previously associated response.     

Correlations between spatial compatibility effects: are arrows more like locations or words?

Responses are more efficient when their spatial mappings with features of targets are compatible compared to when they are incompatible, even when those features are irrelevant to task performance. Currently, a debate exists as to whether spatial information conveyed by different stimulus modes leads to qualitatively different spatial representations. We investigated the relations between three of the most commonly used spatial stimulus modes-arrows, locations, and location words-using correlations of compatibility effects between each of these modes as well as compatibility effects at different segments of their response time distributions. Our results show that when spatial information is irrelevant to task performance (the Simon task), the compatibility effects elicited by arrows and words are more strongly related with each other than with those of locations. However, when spatial information is task relevant (the stimulus-response compatibility SRC task), the compatibility effects elicited by arrows and locations are more related, and both are less related to the effect elicited by words. We suggest that these changing relations between stimulus modes are strategically determined based on which spatial coding technique for arrows is most advantageous to task performance. Furthermore, the varying relations between these spatial compatibility effects indicate that the compatibility effect with one stimulus mode is not always predictive of the compatibility effect in another mode.     

Display-control arrangement correspondence and logical recoding in the Hedge and Marsh reversal of the Simon effect

When left and right keypresses are made to stimuli in left and right locations, and stimulus location is irrelevant to the task, responses are typically faster when stimulus location corresponds with response location than when it does not (the Simon effect). This effect reverses when the relevant stimulus-response mapping is incompatible, with responses being slower when stimulus and response locations correspond (the Hedge and Marsh reversal). Simon et al. (Acta Psychol. 47 (1981) 63) reported an exception to the Hedge and Marsh reversal for a situation in which the relevant stimulus dimension was the color of a centered visual stimulus and the irrelevant location information was left or right tone location. In contrast, similar experiments have found a reversal of the Simon effect for tone location when relevant visual locations were mapped incompatibly to responses. We conducted four experiments to investigate this discrepancy. Both results were replicated. With an incompatible mapping, irrelevant tone location showed a small reverse Simon effect when the relevant visual dimension was physical location but not when the color of a centered stimulus or the direction in which an arrow pointed conveyed the visual location information. The reversal occurred in a more standard Hedge and Marsh task in which the irrelevant dimension was location of the colored stimulus, but only when the response keys were visibly labeled. Several of the results suggest that display-control arrangement correspondence is the primary cause of the Hedge and Marsh reversal, with logical recoding playing only a secondary role.     

Perceptual and attentional factors in encoding irrelevant spatial information

Numerous studies found superior performance when the irrelevant location of a stimulus and response location were corresponding than when they were not corresponding (Simon effect), suggesting that stimulus location is processed in an obligatory manner. The present study compared Simon effects from the location of a relevant (i.e., to-be-attended) object to those from the location of an irrelevant (i.e., to-be-ignored) object. In four experiments, participants were presented with a rectangular frame and a square, with the relevant object in green or red color and the irrelevant object in gray or white color. Participants’ task was to respond with a lateral keypress to the color of the relevant object, and we varied spatial correspondence between the location of the relevant or the irrelevant object and the response, respectively. Results consistently showed larger Simon effects from the location of the relevant than from the irrelevant object, even when the irrelevant object was made very salient. These results suggest that location processing is largely confined to relevant (i.e., attended) objects, stressing the role of attention shifts for location encoding.     

Effects of stimulus–stimulus short-term memory associations in a Simon-like task

The Simon effect refers to faster responding when an irrelevant stimulus location corresponds with the response to a relevant stimulus attribute than when it does not. We investigated whether a memory-based Simon-like effect would occur when the irrelevant spatial attribute was associated with the stimulus during a prior task. In a first task, an association between colour and location was formed by requiring participants to count the occurrences of two colour stimuli, each of which was always presented in a left or right location. In a second task, the colour stimuli were presented centrally and mapped to left and right keypresses, with the mapping being inconsistent or consistent with the prior colour-location associations. A Simon-like effect was evident at the start of the second task, with performance being better when the established colour–position associations were consistent with the colour–response mapping than when they were not. This result indicates that stimulus–stimulus short-term memory associations formed during the first task transferred to the second task. For the remainder of the second task, the data showed a more conservative speed–accuracy criterion for the inconsistent condition than for the consistent condition, though a processing efficiency measure suggested that the prior stimulus–stimulus short-term associations may also continue to directly influence performance. Results suggest that simple declarative knowledge, as represented by stimulus–stimulus STM links, exerts less persistent transfer effects than procedural knowledge as provided by stimulus–response STM links.     

A memory-based, Simon-like, spatial congruence effect: Evidence for persisting spatial codes

The Simon effect refers to the finding that in a task where stimulus location is irrelevant, reaction time is faster when stimulus and response locations are congruent than when they are not. Dominant theories of the Simon effect have generally attributed this spatial congruence effect to a spatial code automatically generated upon stimulus presentation. A common assumption of these theories is that this spatial code decays in less than a few hundred milliseconds following stimulus onset. We report two working-memory experiments suggesting a reexamination of this assumption—a Simon-like spatial congruence effect persisted over a delay of as long as 2400 ms. We propose that, in addition to generating short-lived perceptual codes, spatial information may be coded in working memory as part of the context associated with stimulus events. When reactivated by cues from the original event, such information may influence response selection and produce spatial congruence effects (in this case, positive when participants made a “yes” response and negative when they made a “no” response).     

Mixing location-irrelevant and location-relevant trials: influence of stimulus mode on spatial compatibility effects

The performance advantage for spatially compatible mappings of physical locations to keypress responses, relative to incompatible mappings, is eliminated when stimulus color, rather than location, is relevant on half of the trials. In Experiment 1, we compared the effects of mixing for different stimulus modes (physical locations, arrow directions, and location words) to determine whether this elimination of the stimulus-response compatibility (SRC) effect would generalize to other stimulus modes. The SRC effect was unaffected when the location information was conveyed by arrows and was amplified when the location information was conveyed by words. In Experiment 2, we used vocal left-right responses instead of keypresses, and the SRC effects for all three stimulus modes were enhanced by mixing. In both experiments, for all stimulus modes, mixing reduced or reversed correspondence effects for trials on which the location information was irrelevant when the mapping for those trials on which it was relevant was incompatible. These findings suggest that when trial types are mixed, direct activation of the corresponding response, regardless of mapping, does not occur for physical locations mapped to keypresses. However, such activation does occur when stimuli or responses are verbal, apparently because performance is mediated in part by activation of a verbal name code for the stimulus.     

The compatibility of spatial and non-spatial relationships

This experiment reverses the procedure used by Hedge and Marsh (1975) who obtained two-choice RTs to a relevant stimulus attribute (colour) in the presence of an irrelevant attribute (location). On the basis of their finding that RT was faster when the colour and location of the correct response button were both the same, or both different from that of the stimulus, they concluded that performance is facilitated where the same logical process ‘same’ or ‘alternative’ can be applied to both attributes. The ‘Simon effect’ it was suggested may be explained by this process.Earlier work by the author had suggested performance is facilitated where the same recording rule applies to all the spatial relations in the task. Performing Hedge and Marsh's experiment with location as the relevant attribute failed to replicate their finding. This result limits the generality of their explanation to tasks where spatial relationships are the ‘irrelevant’ variable.     

Stimulus–response compatibility for mixed mappings and tasks with unique responses

For two stimulus locations mapped to two keypresses, reaction time is shorter when the mapping is compatible than when it is not (the stimulus–response compatibility, SRC, effect). A similar result, called the Simon effect, occurs when stimulus location is irrelevant, and colour is relevant. When compatibly mapped trials are intermixed with incompatibly mapped trials or Simon task trials, the compatibility effect is eliminated, and the Simon effect is influenced by the location mapping. In five experiments, we examined whether similar mixing effects occur when the two spatial mappings or location-relevant and location-irrelevant tasks use distinct keypresses on the left and right hands. Mixing had considerably less influence on the SRC and Simon effects than it does when the intermixed trial types or tasks share the same responses, even though response time was lengthened to a similar extent. Mixing two tasks for which stimulus location was irrelevant yielded no within-task Simon effect, but the effect was also absent when four stimuli were assigned to two responses on a single hand. The relative lack of influence of mixing on the SRC and Simon effects when the tasks have unique responses implies that suppression of direct activation of the corresponding response occurs primarily when the tasks share responses.     

Influence of irrelevant information on human performance: Effects of S-R association strength and relative timing

Six experiments examined effects of stimulus-response (S-R) association strength and relative timing on the magnitude of consistency effects for irrelevant information in Stroop-like tasks. Keypresses were made to two-dimensional stimuli (a colour or location word surrounded by a coloured rectangle or arrow), with the irrelevant information presented simultaneously with or prior to the relevant information. With simultaneous presentation, irrelevant information affected performance regardless of whether its S-R association was weak or strong, if the relevant S-R association was weak (e.g., colour word to keypress). However, a weak irrelevant S-R association (location word to keypress) had little effect when paired with a strong relevant S-R association (arrow direction to keypress), except when the stimulus onset asynchrony (SOA) between the irrelevant and relevant information was 300 ms. When the relevant information was colour, the effect of an irrelevant colour word persisted at a 500-ms SOA but that of an irrelevant physical colour did not, reflecting different decay functions for irrelevant verbal and non-verbal information. The persisting effect of an irrelevant colour word was reduced by articulatory suppression and eliminated at extended SOAs of 3 s. The results indicate that whether the consistency effect patterns are symmetric or asymmetric is determined by the relative strengths of the relevant and irrelevant S-R associations, as specified by the criteria of conceptual and mode similarity. The magnitude of the consistency effect is also a function of the temporal overlap of the resulting response activation, which is determined primarily by mode similarity.     

A Simon effect in memory retrieval: Evidence for the response-discrimination account

According to the traditional view, the effects of irrelevant stimulus location on the selection of a spatial response to a nonspatial stimulus feature (Simon effect) result from long-term associations between spatial stimulus codes and spatially corresponding response codes. According to an alternative view, the response-discrimination account, Simon effects arise from interactions between spatial stimulus codes and response labels in working memory (WM). The latter account predicts Simon effects when participants use spatial labels for response representation in WM, even when the actual responses have no spatial features (e.g., saying the word "plate"). The prediction was tested in an experiment, in which participants first encoded two words at different locations, and then responded to a stimulus by saying the word from the location indicated by stimulus color. The manipulation concerned the correspondence between irrelevant location of the colored stimulus and the retrieval cue for the vocal responses (i.e., word location in the encoding display). A Simon effect in memory retrieval was observed, supporting the response-discrimination account.     

S-R compatibility and response selection

In serial stage models, perception and action are usually thought to be linked to each other by an S-R translation mechanism. However, phenomena of S-R compatibility suggest a more direct relationship between perceptual and action domains. We discuss behavioral and psychophysiological evidence that irrelevant stimulus information automatically activates response codes, but then decays over time.

In a series of reaction time studies and electrophysiological experiments, we investigated both temporal and functional properties of the assumed automatic response activation process. We found that the amount of interference due to irrelevant spatial information depends upon how long its availability precedes that of the information relevant for response selection. This indicates that response activation decays rather quickly. If response-relevant and irrelevant spatial information are simultaneously available, electrophysiological measurements show that automatic activation of the spatially corresponding response rises soon after stimulus onset, but then dissipates and gets replaced by the activation of the response indicated by the relevant stimulus attribute.

We conclude that these findings do not support a pure translation account, but rather suggest the presence of two parallel and (at least partially) independent routes from perception to action: A direct route, allowing for automatic activation of response codes if stimulus and response features overlap, and an indirect route linking S and R codes in an arbitrary manner. Via the direct route responses may be primed independent of task-specific contingencies, while the correct response is selected via the indirect route. This use suggests that (a) the transmission of stimulus information to response stages does not (fully) depend on task relevance and that (b) different stimulus features can be transmitted asynchronously and independently from one another.     

The enhanced Simon effect for older adults is reduced when the irrelevant location information is conveyed by an accessory stimulus

The Simon effect, better performance when irrelevant stimulus location corresponds with the response location than when it does not, typically is larger for older than younger adults. However, Simon and Pouraghabagher [Simon, R. J., & Pouraghabagher, A. R. (1978). The effect of aging on the stages of processing in a choice reaction time task. Journal of Gerontology, 33, 553-561] found no age difference using an accessory-stimulus Simon task in which the relevant dimension was the color of a visual stimulus and the irrelevant dimension the location of a tone. Experiment 1 confirmed that older adults show a larger Simon effect than younger adults for the visual Simon task and that this age-related deficit is reduced or eliminated for the auditory-accessory task. Experiment 2 provided evidence suggesting that a small part of the age-related deficit in the visual Simon task is due to having to code the location of the relevant stimulus, but Experiment 3 showed that the majority of the deficit is due to the relevant and irrelevant information being conveyed by the same stimulus. Reaction-time distribution analyses show similar functions for younger and older adults, suggesting that the time course of activation is similar for both age groups.     

Spatial compatibility interference effects: a double dissociation between two measures

ABSTRACT

In spatial compatibility tasks, when the spatial location of a stimulus is irrelevant it nevertheless interferes when a response is required in a different spatial location. For example, response with a left key-press is slowed when the stimulus is presented to the right as compared to the left side of a computer screen. However, in some conditions this interference effect is not detected in reaction time (RT) measures. It is typically assumed that the lack of effect means the irrelevant spatial code was not analysed or that the information rapidly decayed before response. However, we show that even in conditions where there appears to be no spatial interference when measuring RTs, effects can nevertheless be detected after response when recording facial electromyography responses. This dissociation between two measures highlights the importance of diverging methods to investigate visuomotor processes as conclusions based on only one measure can be misleading.     

The relationship between stimulus processing and response selection in the Simon task: Evidence for a temporal overlap

Summary As is indicated by the Simon effect, choice reactions can be carried out faster when the response corresponds spatially to the stimulus, even if the stimulus location is irrelevant to the task. In Experiments 1–4 the relationships between the Simon effect and stimulus eccentricity, signal quality, and signal-background contrast are investigated. The Simon effect was found to interact with all of these factors, at least when manipulated blockwise. These results are at odds with previous results and are difficult to interpret from an additive-factor-method view. An alternative interpretation is suggested that attributes the results to the temporal relationship between the processing of the relevant stimulus information and stimulus location. The assumption is that a decrease in the Simon effect is caused by every experimental manipulation that markedly increases the temporal distance between the coding of the relevant stimulus information and that of the irrelevant stimulus location. This assumption was tested in Experiment 5 in a more direct way. The stimuli were built up on a screen over time, so that the temporal distance between the presence of location and identity information could be controlled experimentally. The results provide further support for a temporal-delay interpretation of interactions between irrelevant stimulus-response correspondence and factors that affect early stages of information processing.     

Age differences in response selection for pure and mixed stimulus-response mappings and tasks

Two experiments examined effects of mixed stimulus-response mappings and tasks for older and younger adults. In Experiment 1, participants performed two-choice spatial reaction tasks with blocks of pure and mixed compatible and incompatible mappings. In Experiment 2, a compatible or incompatible mapping was mixed with a Simon task for which the mapping of stimulus color to location was relevant and stimulus location was irrelevant. In both experiments, older adults showed larger mixing costs than younger adults and larger compatibility effects, with the differences particularly pronounced in Experiment 1 when location mappings were mixed. In mixed conditions, when stimulus location was relevant, older adults benefited more than younger adults from complete repetition of the task and stimulus from the preceding trial. When stimulus location was irrelevant, the benefit of complete repetition did not differ reliably between age groups. The results suggest that the age-related deficit associated with mixing mappings and tasks is primarily due to older adults having more difficulty separating task sets that activate conflicting response codes.     

Don't be afraid of irrelevant words: The emotional Stroop effect is confined to attended words

The emotional Stroop effect denotes slower responses to the colour of negative words (e.g., death) compared to neutral words (e.g., mug). Popular explanations assume a general power of negative words to capture visual attention. However, in the typical task, the irrelevant word stimulus and the relevant colour stimulus are perceptually integrated. We compared interference from negative words, which were part of the relevant visual object, to interference from negative words that were part of an irrelevant object, or occurred in the background, respectively. Results showed that only negative words in the relevant object delayed colour-naming responses, compared to neutral words. Negative words outside the relevant object failed to affect performance. This finding is at odds with the claim that negative words could capture spatial or object-based mechanisms of visual attention. However, the finding is consistent with the idea that negative words interfere with the allocation of dimensional attention to different features of an attended object.     

Task-defined associations are mode specific for selection of relevant dimension but mode independent for selection of location mapping

When compatible and incompatible mappings of a location-relevant task are mixed, or a location-relevant task is mixed with a task for which stimulus location is irrelevant, the benefit of the compatible mapping is eliminated for physical locations and enhanced for location words. Two experiments examined the influence of presenting the location information for the mixed conditions in different stimulus modes (physical location or word). Experiment 1 showed that the effects of mixing location-relevant and location-irrelevant tasks on the spatial compatibility and Simon effects are reduced when the location information is presented in different modes for the two tasks. Experiment 2 showed, in contrast, that the mode distinction had little influence on the effects of mixed compatible and incompatible mappings for location-relevant tasks: The compatibility effect was eliminated for physical locations and enhanced for words, as when there is no mode distinction. Thus, when location is relevant for one task and colour for the other, the task-defined associations of locations to responses are mode specific, but when location is relevant for both tasks, the associations are mode independent.     

Task-defined associations are mode specific for selection of relevant dimension but mode independent for selection of location mapping

When compatible and incompatible mappings of a location-relevant task are mixed, or a location-relevant task is mixed with a task for which stimulus location is irrelevant, the benefit of the compatible mapping is eliminated for physical locations and enhanced for location words. Two experiments examined the influence of presenting the location information for the mixed conditions in different stimulus modes (physical location or word). Experiment 1 showed that the effects of mixing location-relevant and location-irrelevant tasks on the spatial compatibility and Simon effects are reduced when the location information is presented in different modes for the two tasks. Experiment 2 showed, in contrast, that the mode distinction had little influence on the effects of mixed compatible and incompatible mappings for location-relevant tasks: The compatibility effect was eliminated for physical locations and enhanced for words, as when there is no mode distinction. Thus, when location is relevant for one task and colour for the other, the task-defined associations of locations to responses are mode specific, but when location is relevant for both tasks, the associations are mode independent.     

The influence of irrelevant stimulus changes on stimulus and response repetition effects

In this study the influence of irrelevant stimulus changes from one trial to another in a serial reaction time task was investigated. Two experiments were performed in which subjects were required to respond to stimulus colour. Four colours were mapped on two response keys, so that colour and response repetition effects could be dissociated. In Experiment 1, the irrelevant stimulus dimension was location and in Experiment 2 it was shape. Both experiments were performed with a short and a long response-stimulus interval (RSI)-condition. In both experiments, the irrelevant dimension influenced the response repetition effect but not the colour repetition effect. In the reaction times, a response alternation effect was observed only when the irrelevant location of the stimulus changed in the long RSI-condition. The error rates showed a response alternation benefit for both irrelevant dimensions, in the short and the long RSI-condition. The benefit for response alternations is explained in terms of a response bias towards change that is triggered by a changing stimulus feature. We assume that the response bias is stronger for location than for colour and that accuracy is more sensitive to this bias than response latencies.