Elsevier

Cognition

Volume 111, Issue 2, May 2009, Pages 275-279
Cognition

Brief article
Post-error slowing: An orienting account

https://doi.org/10.1016/j.cognition.2009.02.002Get rights and content

Abstract

It is generally assumed that slowing after errors is a cognitive control effect reflecting more careful response strategies after errors. However, clinical data are not compatible with this explanation. We therefore consider two alternative explanations, one referring to the possibility of a persisting underlying problem and one on the basis of the low frequency of errors (orienting account). This latter hypothesis argues that infrequent events orient attention away from the task. Support for the orienting account was obtained in two experiments. Using a new experimental procedure, Experiment 1 demonstrated post-error slowing after infrequent errors and post-correct slowing after infrequent correct trials. In Experiment 2, slowing was observed following infrequent irrelevant tones replacing the feedback signals.

Introduction

Cognitive control is responsible for adjusting our information processing network to context demands and goal settings. Empirically, behavioural adaptation effects are taken as a reflection of cognitive control processes. Perhaps one of the most replicable effects is the observation that responses are slower after an error than after a correct trial. Cognitive control theories attribute post-error slowing to adaptive control mechanisms that induce more careful behaviour to reduce the probability of error commission. Conflict monitoring theory (Botvinick, Braver, Barch, Carter, & Cohen, 2001), for instance, explains post-error slowing in terms of a decrease in baseline response activation after errors which is functionally equivalent to increasing the response threshold. As a result, post-error trials are predicted to be slower and more accurate. Conflict monitoring theory adequately simulated the data by Laming (1968) who indeed observed this pattern. Consequently, post-error slowing is now widely accepted as a cognitive control effect, and is used as a marker for cognitive control in clinical studies (e.g., Bogte et al., 2007, Kerns et al., 2005, Sergeant and van der Meere, 1988).

Although the combination of post-error slowing and accuracy increase has been reported (Laming, 1968), an overview of the literature suggests that increased accuracy after errors is usually not observed (e.g., Hajcak and Simons, 2008, Hajcak et al., 2003, Rabbitt and Rodgers, 1977). Hence, other explanations need to be considered. Gehring, Goss, Coles, Meyer, and Donchin (1993) suggested that post-error slowing could be caused by the persistence of the malfunctioning process that led to an error on the previous trial, leading to a correlation in task efficiency across trials. This account does not only predict post-error slowing, but also a post-error accuracy decrease.

In the present paper, we propose that post-error slowing is caused by the relative infrequency of errors which causes attentional capture. This was already hinted at by Burns, 1965, Rabbitt and Phillips, 1967, pp. 38): “Burns himself preferred to suggest that the occurrence of an error was followed by an orienting response which inhibited rather than facilitated subsequent responses”. In line with this, Barcelo, Escera, Corral, and Periáñez (2006) reported slowing after infrequent events (oddballs) and interpreted this in terms of a time-consuming orientation to the oddball and a reorientation to the task. We refer to this hypothesis as the orienting account.

The orienting account makes two unique predictions. First, when errors are more frequent than correct trials, correct trials should elicit the orienting response and slowing should be observed after infrequent correct trials. On the basis of a persisting problem and the cognitive control hypothesis, one should always predict post-error slowing irrespective of the relative frequencies of errors and correct responses. Second, if the orienting response causes the slowing after errors, it is also predicted that orienting towards completely irrelevant unexpected signals should slow down subsequent responding.

Both predictions were tested in the following experiments. In Experiment 1 we manipulated the error rates by means of an adaptive program. We predict post-error slowing when errors are infrequent and post-correct slowing when correct trials are infrequent. In Experiment 2, we replace the feedback signal by an irrelevant high or low tone. We predict slowing when an infrequent tone follows the response.

Section snippets

Participants

Sixteen students (15 female; average age of 18 years and 8 months) of Ghent University participated in turn for course credits.

Procedure

Stimuli were 0.4° by 0.4° colored squares presented centrally on a white background. The brightness of the colors was adjusted in order to keep every participant’s performance to a prespecified level (35%, 55% or 75% accuracy). Colors are described according to the HSV color model with three parameters: hue (0–360), saturation (0–100) and value (0–100). The four colors

Experiment 2

To further investigate the influence of expectancy on slowing, we designed a second experiment where an irrelevant signal substitutes the feedback signal. If post-error slowing is caused by an orienting response, one would also expect slowing after an infrequent irrelevant signal. Indirectly, this was already suggested in Barcelo et al. (2006) where occasionally (26 times in a block of 140 trials) a novel unique sound was presented. The slowing after these novel sounds is in line with our

General discussion

In Experiment 1, it was demonstrated that post-error correct RT is modulated by the frequency of errors. Post-error slowing was observed when errors were infrequent, but when errors were frequent, slowing was observed after correct trials. This cannot be explained by mechanisms of adaptive cognitive control or by the persistence of an underlying problem that caused the error. The hypothesis that infrequent events slow down task-relevant processing was further confirmed in Experiment 2 where

References (22)

  • E.A. Crone et al.

    Heart rate and skin conductance analysis of antecedents and consequences of decision making

    Psychophysiology

    (2004)
  • Cited by (427)

    View all citing articles on Scopus
    View full text