Anticipatory response control in motor sequence learning: evidence from stimulus-response compatibility

Three experiments using a serial four-choice reaction-time (RT) task explored the interaction of sequence learning and stimulus-based response conflict. In Experiment 1, the spatial stimulus-response (S-R) mapping was manipulated between participants. Incompatible S-R mappings produced much higher RTs than the compatible mapping, but sequence learning decreased this S-R compatibility effect. In Experiment 2, the spatial stimulus feature was made task-irrelevant by assigning responses to symbols that were presented at unpredictable locations. The data indicated a Simon effect (i.e., increased RT when irrelevant stimulus location is spatially incompatible with response location) that was reduced by sequence learning. However, this effect was observed only in participants who developed an explicit sequence representation. Experiment 3 replicated this learning-based modulation of the Simon effect using explicit sequence-learning instructions. Taken together, the data support the notion that explicit sequence learning can lead to motor 'chunking', so that pre-planned response sequences are shielded from conflicting stimulus information.     

Unintentional and intentional learning of noncorresponding stimulus-response associations in the Simon task

The Simon effect is a robust phenomenon that persists after extensive practice. However, several studies using a transfer paradigm have shown that the Simon effect is eliminated after practicing a location-relevant task with an incompatible spatial mapping. The present study examined whether this transfer effect is a result of implicit, procedural knowledge developed through repeated execution of noncorresponding responses in the practice session or a consequence of explicitly learning and reinstating a noncorresponding mapping rule. Results from two experiments show that, although a small part of the transfer effect may be due to residual activation of noncorresponding S-R associations from the prior task, the larger and more stable part is likely due to response-selection strategies performed intentionally in the practice task.     

Control over location-based response activation in the Simon task: behavioral and electrophysiological evidence

In 4 Simon experiments the authors examined control over 2 routes of sensorimotor processing: response priming in the unconditional route and response selection via the conditional route. The Simon effect diminished as the frequency of noncorresponding trials increased. Location-based response priming was observed only when the stimulus followed a corresponding event but not after a noncorresponding trial. Therefore, the unconditional route appears to be suppressed whenever the task context indicates priming as potentially disadvantageous. Moreover, the task-irrelevant stimulus location was used for response selection as a function of correspondence probability. Although exact repetitions of stimulus-response sequences caused a marked speed-up of responses, this 3rd mechanism is independent of unconditional route suppression and frequency-based adjustments in the conditional route.     

Evidence for gating of direct response activation in the Simon task

TheSimon effect denotes faster responses when the task-irrelevant stimulus position corresponds to the response position than when it does not. Accounts of this effect assume that stimulus position automatically activates a spatially corresponding response while the correct response is being computed. Yet the Simon effect has been found to be reduced after noncorresponding trials. Some authors have interpreted these sequential modulations of the Simon effect as evidence for a mechanism gating positionbased response activation. Alternatively, sequential modulations have been explained in terms of feature-integration processes, which depend upon the fact that different sequences of spatial-correspondence conditions covary with different degrees of feature overlap between subsequent trials. The present study investigates whether sequential modulations of the Simon effect can occur when feature overlap in the different conditions is the same. Therefore, a Simon task with four stimulus positions and two response positions was used. Sequential modulations of the Simon effect were found in trial sequences with constant amounts of feature overlap between trials. Although the feature-integration account cannot explain this result, it is consistent with the idea of a gating (i.e., cognitive control) mechanism.     

Sequential modulations of logical-recoding operations in the Simon task

The Simon effect consists of faster responses to the color (or another nonspatial feature) of spatially corresponding stimuli than to spatially noncorresponding stimuli. Recently, several studies observed the Simon effect after corresponding predecessor trials, but not after noncorresponding predecessor trials. To explain these sequential modulations, Stürmer et al. (2002) proposed a mechanism modulating the ability of stimulus position to automatically activate a response. The present study investigated which events are effectively triggering this mechanism in a variant of the Simon task, in which both stimuli and responses varied in color (participants wore colored gloves) as well as in horizontal position. In the same-color task (e.g., green stimulus-green response), a normal Simon effect showed up after corresponding trials, but no effect occurred after noncorresponding trials. In the alternate-color task (e.g., green stimulus-red hand), no effect occurred after spatially corresponding trials, whereas an inverted Simon effect was found after noncorresponding trials. Additional analyses showed that repetition (or alternation) effects did not affect the results. The results are discussed in terms of a conflict-monitoring account (Stürmer et al., 2002), and in terms of a feature-integration account (Hommel et al., 2002).     

A feature-integration account of sequential effects in the Simon task

Recent studies have shown that the effects of irrelevant spatial stimulus-response (S-R) correspondence (i.e., the Simon effect) occur only after trials in which the stimulus and response locations corresponded. This has been attributed to the gating of irrelevant information or the suppression of an automatic S-R route after experiencing a noncorresponding trial—a challenge to the widespread assumption of direct, intentionally unmediated links between spatial stimulus and response codes. However, trial sequences in a Simon task are likely to produce effects of stimulus- and response-feature integration that may mimic the sequential dependencies of Simon effects. Four experiments confirmed that Simon effects are eliminated if the preceding trial involved a noncorresponding S-R pair. However, this was true even when the preceding response did not depend on the preceding stimulus or if the preceding trial required no response at all. These findings rule out gating/suppression accounts that attribute sequential dependencies to response selection difficulties. Moreover, they are consistent with a feature-integration approach and demonstrate that accounting for the sequential dependencies of Simon effects does not require the assumption of information gating or response suppression.     

Demonstration of nondeclarative sequence learning in mice: development of an animal analog of the human serial reaction time task

In this paper, we demonstrate nondeclarative sequence learning in mice using an animal analog of the human serial reaction time task (SRT) that uses a within-group comparison of behavior in response to a repeating sequence versus a random sequence. Ten female B6CBA mice performed eleven 96-trial sessions containing 24 repetitions of a 4-trial sequence. During the 12th session, the repeating sequence was replaced with the random sequence halfway through the session. Reaction time (RT) to respond to an illuminated nose-poke was recorded, and performance was compared at the halfway point in each session to test for any change in behavior. For learning effect, RTs decreased over the no-switch repeating-sequence sessions. For interference effect, behavior did not change appreciably at the halfway point during the last repeating-sequence session. However, RTs deteriorated significantly after the switch from repeating to random sequences halfway through session 12. The mice demonstrated a robust interference effect when switched from repeating to random sequences. This pattern of behavior in humans performing the SRT is interpreted as evidence of nondeclarative sequence learning. The similarity between the human and mouse SRTs will enable more direct comparisons of mouse-human nondeclarative memory behavior and will provide a useful behavioral end-point in mouse-models of basal ganglia dysfunction.     

Spatial interference and response control in sequence learning: the role of explicit knowledge

In several sequence learning studies it has been suggested that response control shifts from the stimuli to some internal representation (i.e., motor program) through the learning process. The main questions addressed in this paper are whether this control shift is related to explicit knowledge and whether the formation of these internal representations depends on the stimulus attributes. In one experiment we compared the learning of a response sequence triggered by either spatial location or location symbol (left-right) by using a serial response task (SRT). Symbols were presented at either a centered or random location. The results showed that in the symbolic conditions the shift of response control correlated with the emergence of explicit knowledge. Only participants with complete explicit knowledge seemed to learn the sequence structure beyond probabilistic information (response time "RT" did not depend on the frequency of the response). Moreover, these participants were able to overcome, when needed, spatial interference (RT was the same for both spatially corresponding and non-corresponding trials). However, when spatial location was relevant, RT was always faster, especially for more frequent responses. These results suggest that the relevant stimulus dimension (location or symbol) seems to engage different sequence learning mechanisms.     

Effects of saccades and response type on the Simon effect: If you look at the stimulus,the Simon effect may be gone

The Simon effect has most often been investigated with key-press responses and eye fixation. In the present study, we asked how the type of eye movement and the type of manual response affect response selection in a Simon task. We investigated three eye movement instructions (spontaneous, saccade, and fixation) while participants performed goal-directed (i.e., reaching) or symbolic (i.e., finger-lift) responses. Initially, no oculomotor constraints were imposed, and a Simon effect was present for both response types. Next, eye movements were constrained. Participants had to either make a saccade toward the stimulus or maintain gaze fixed in the screen centre. While a congruency effect was always observed in reaching responses, it disappeared in finger-lift responses. We suggest that the redirection of saccades from the stimulus to the correct response location in noncorresponding trials contributes to the Simon effect. Because of eye–hand coupling, this occurred in a mandatory manner with reaching responses but not with finger-lift responses. Thus, the Simon effect with key-presses disappears when participants do what they typically do—look at the stimulus.     

Effect of locus of warning tone on auditory choice reaction time

Forty Ss pressed a left- or right-hand key depending on the ear in which they heard a 500-Hz stimulus tone. Half of the Ss were instructed to press the key on the same side as the ear stimulated (corresponding condition), while the other half pressed the key on the opposite side (noncorresponding condition). A 200-Hz warning tone preceded the stimulus tone by either 200 or 400 msec. The warning tone was presented to the left ear, the right ear, or to both ears in a predetermined random sequence. The locus of the warning tone affected RT on noncorresponding trials but not on corresponding trials. The effect consisted of a significant slowing of information processing on trials where the warning tone was contralateral to the response. Results were explained in terms of an initial tendency to respond toward the source of the warning tone.     

Response conflict determines sequential effects in serial response time tasks with short response-stimulus intervals

In serial choice reaction time (RT) tasks, performance in each trial critically depends on the sequence of preceding events. In this study, the authors specifically examined the mechanism underlying RT sequence effects at short response-stimulus intervals (RSIs), in which performance is impaired in the current trial N if events alternate rather than repeat from trial N-2 to trial N-1. Different accounts of this RT pattern in terms of perceptual noise, response-selection monitoring, and response conflict were tested in 4 experiments. Second-order RT costs were caused by the response sequence rather than the stimulus sequence. Manipulation of stimulus contrast, stimulus classification difficulty, and set-level compatibility did not modulate the response-related second-order RT effect, whereas this effect increased when spatially incompatible responses were demanded. These findings support a response conflict account of higher order sequential effects in short-RSI situations.     

Response strategies and the Simon effect

The study investigated whether the Simon effect, and its facilitation and interference components, shows up in reaction time (RT) or in movement time (MT), depending on the response strategy. Experiment 1 replicated a study by Hietanen and Rämä. Subjects had to press one of two lateralised keys in response to one of two stimuli. The stimuli were presented in the center (neutral condition) or to the left or right side (corresponding or non-corresponding conditions). To press the response key, a reaching movement was necessary, and both RT and MT were recorded. One group of subjects showed an RT facilitation effect and an MT interference effect. Another group of subjects showed both MT facilitation and MT interference effects. It was hypothesized that the two groups used different response strategies. In Exps. 2 and 3, the subjects were explicitly instructed to use the two strategies that were hypothesized for Exp. 1. The results showed that whether facilitation and interference manifest themselves in RT or MT depends on the response strategy adopted by the subjects.     

Secondary-task effects on sequence learning

With a repeated sequence of stimuli, performance in a serial reaction-time task improves more than with a random sequence. The difference has been taken as a measure of implicit sequence learning. Implicit sequence learning is impaired when a secondary task is added to the serial RT task. In the first experiment, secondary-task effects on different types of sequences were studied to test the hypothesis that the learning of unique sequences (where each sequence element has a unique relation to the following one) is not impaired by the secondary task, while the learning of ambiguous sequences is. The sequences were random up to a certain order of sequential dependencies, where they became deterministic. Contrary to the hypothesis, secondary-task effects on the learning of unique sequences were as strong or stronger than such effects on the learning of ambiguous sequences. In the second experiment a hybrid sequence (with unique as well as ambiguous transitions) was used with different secondary tasks. A visuo-spatial and a verbal memory task did not interfere with the learning of the sequence, but interference was observed with an auditory go/no-go task in which high- and low-pitched tones were presented after each manual response and a foot pedal had to be pressed in response to high-pitched tones. Thus, interference seems to be specific to certain secondary tasks and may be related to memory processes (but most likely not to visuo-spatial and verbal memory) or to the organization of sequences, consistent with previous suggestions.     

Effect of sequence length on the execution of familiar keying sequences: lasting segmentation and preparation?

The author assessed the mechanisms underlying skilled production of keying sequences in the discrete sequence-production task by examining the effect of sequence length on mean element execution rate (i.e., the rate effect). To that end, participants (N = 9) practiced fixed movement sequences consisting of 2, 4, and 6 key presses for a total of 588 trials per sequence. In the subsequent test phase, the sequences were executed with and without a verbal short-term memory task in both simple and choice reaction time (RT) paradigms. The rate effect was obtained in the discrete sequence-production task-including the typical quadratic increase in sequence execution time (SET, which excludes RT) with sequence length. The rate effect resulted primarily from 6-key sequences that included 1 or 2 relatively slow elements at individually different serial positions. Slowing of the depression of the 2nd response key (R2) in the 2-key sequence reduced the rate effect in the memory task condition, and faster execution of the 1st few elements in each sequence amplified the rate effect in simple RT. Last, the time to respond to random cues increased with position, suggesting that the mechanisms that underlie the rate effect in new sequences and in familiar sequences are different. The data were in line with the notion that coding of longer keying sequences involves motor chunks for the individual sequence segments and information on how those motor chunks are to be concatenated.     

The impact of anatomical and spatial distance between responses on response conflict

Different features of objects can be associated with different responses, so that their concurrent presence results in conflict. The Simon effect is a prominent example of this type of response conflict. In two experiments, we ask whether it is modulated by the anatomical or spatial relation between responses. Predictions were derived from an extended variant of the leaky, competing accumulator (LCA) model proposed by Usher and McClelland (Psychological Review, 108, 550–592, 2001). The relation between responses was represented by the lateral-inhibition parameter of the model. For the anatomical distance between responses the expectations were largely confirmed, but not for spatial distance. First, the Simon effect was stronger when responses were performed with two fingers of the same hand than with different hands. Second, the Simon effect was larger only for responses with different hands at short reaction times and disappeared at long ones, whereas for responses with fingers of the same hand, the Simon effect was essentially the same for shorter and longer reaction times. This difference resulted in smaller variability of reaction times in noncorresponding than in corresponding conditions. The dependence of decision processes, as modelled by the LCA model, on the anatomical relation between responses supports the broad hypothesis that the accumulation of evidence on the state of the world is intricately linked with the activation of response codes, that is, the selection of the appropriate actions.     

Memory for Movement in Blind Children

The author assessed the mechanisms underlying skilled production of keying sequences in the discrete sequence-production task by examining the effect of sequence length on mean element execution rate (i.e., the rate effect). To that end, participants (N = 9) practiced fixed movement sequences consisting of 2, 4, and 6 key presses for a total of 588 trials per sequence. In the subsequent test phase, the sequences were executed with and without a verbal short-term memory task in both simple and choice reaction time (RT) paradigms. The rate effect was obtained in the discrete sequence-production task—including the typical quadratic increase in sequence execution time (SET, which excludes RT) with sequence length. The rate effect resulted primarily from 6-key sequences that included 1 or 2 relatively slow elements at individually different serial positions. Slowing of the depression of the 2nd response key (R₂) in the 2-key sequence reduced the rate effect in the memory task condition, and faster execution of the 1st few elements in each sequence amplified the rate effect in simple RT. Last, the time to respond to random cues increased with position, suggesting that the mechanisms that underlie the rate effect in new sequences and in familiar sequences are different. The data were in line with the notion that coding of longer keying sequences involves motor chunks for the individual sequence segments and information on how those motor chunks are to be concatenated.     

Approach and avoidance movements are unaffected by cognitive conflict: A comparison of the Simon effect and stimulus–response compatibility

Participants in this study reached from central fixation to a lateral position that either contained or was opposite to the stimulus. Cognitive conflict was induced when the stimulus and response directions did not correspond. In the Simon task, the response direction was cued by the color of the lateral stimulus, and corresponding and noncorresponding trials varied randomly in the same block of trials, resulting in high uncertainty and long reaction times (RTs). In the stimulus-response compatibility (SRC) task, participants reached toward or away from the stimulus in separate blocks of trials, resulting in low uncertainty and short RTs. In the SRC task, cognitive conflict in noncorresponding trials slowed down RTs but hardly affected reach trajectories. In the Simon task, both RTs and reach trajectories were strongly influenced by stimulus-response correspondence. Despite the overall longer RTs in the Simon task, reaches were less direct and deviated toward the stimulus in noncorresponding trials. Thus, cognitive conflict was resolved before movement initiation in the SRC task, whereas it leaked into movement execution in the Simon task. Current theories of the Simon effect, such as the gating of response activation or response code decay, are inconsistent with our results. We propose that the SRC task was decomposed as approaching and avoiding the stimulus, which is sustained by stereotyped visuomotor routines. With complex stimulus-response relationships (Simon task), responses had to be coded as leftward and rightward, with more uncertainty about how to execute the action. This uncertainty permitted cognitive conflict to leak into the movement execution.     

Sequential analysis of a Simon task--evidence for an attention-shift account

 We investigated the attention-shift hypothesis of the Simon effect by analysing the effect of repeating relevant colour or irrelevant location of the stimulus in four serial reaction time tasks. In Experiment 1 with short response-stimulus intervals (RSI), we assume that there is no time to engage attention at the fixation cross before the onset of a new stimulus. In agreement with the hypothesis, Experiment 1 reveals no Simon effect when the stimulus location is repeated. In Experiment 2 with long RSI, we observe a Simon effect for location repetitions and alternations. In Experiment 3 with long RSI, we hinder the disengagement of attention by displaying the stimulus after response execution. As expected, the Simon effect is reduced for location repetitions. In Experiment 4 with stimuli additionally presented at the fixation cross, responses are faster if the attention shift towards the centrally presented stimulus corresponds with the location of the required response. Additionally, we argue that binding of the stimulus features into an object or event file better explains the so-called blocking of the automatic response-priming route after a noncorresponding trial. Received: 2 February 2000 / Accepted: 10 November 2000     

Analyzing distributional properties of interference effects across modalities: chances and challenges

In research investigating Stroop or Simon effects, data are typically analyzed at the level of mean response time (RT), with results showing faster responses for compatible than for incompatible trials. However, this analysis provides only limited information as it glosses over the shape of the RT distributions and how they may differ across tasks and experimental conditions. These limitations have encouraged the analysis of RT distributions using delta plots. In the present review, we aim to bring together research on distributional properties of auditory and visual interference effects. Extending previous reviews on distributional properties of the Simon effect, we additionally review studies reporting distributional analyses of Stroop effects. We show that distributional analyses of sequential effects (i.e., taking into account congruency of the previous trial) capture important similarities and differences of interference effects across tasks (Simon, Stroop) as well as across sensory modalities, despite some challenges associated to this approach.     

Response inhibition and adaptations to response conflict in 6- to 8-year-old children: Evidence from the Simon effect

Several studies have shown that the Simon effect, which is the advantage of spatial correspondence between stimulus and response locations when the stimulus location is task-irrelevant, decreases with increasing response times and is affected by preceding-trial correspondence. These modulations suggest the existence of control mechanisms that adapt our behavior to current goals by responding to the conflict experienced within a trial and by preventing the recurrence of a conflict in the subsequent trial. The aim of the present study was to assess whether these control mechanisms, which are well consolidated in adults and in children older than 8 years of age, are present in children between 6 and 8 years old. To this end, we tested 32 first-grade (6–7 years) and 34 second-grade (7–8 years) children on a Simon task in which correspondence sequence was manipulated on a trial-by-trial basis. The Simon effect was larger for first- than for second-graders and decreased with increasing response times only in second-graders. Crucially, for both groups, the effect was reduced when the preceding trial was noncorresponding, and the reductions were comparable for the two groups, indicating that trial-by-trial control mechanisms are already present in first-grade children and may be dissociated from within-trial control adjustments.