Academic journal article Psychonomic Bulletin & Review

Sequential Adjustments before and after Partial Errors

Academic journal article Psychonomic Bulletin & Review

Sequential Adjustments before and after Partial Errors

Article excerpt

In choice reaction time tasks, subjects speed up before making an error, but slow down afterward to prevent the occurrence of a new error. In some trials, the correct response is preceded by an incorrect electromyographic (EMG) activation too small to reach the response threshold. In this article, we show that these incorrect EMG activations give rise to the same sequential effects as overt errors: Before a trial containing an incorrect EMG activation, subjects speed up, whereas after that trial, they slow down. These activations reflect errors that have been detected, inhibited, and corrected in time. As such, they index the involvement of online executive control.

It has been proposed that performance is controlled by a supervisory system (Botvinick, Braver, Barch, Carter, & Cohen, 2001; Rabbitt, 1966) that monitors information processing. This supervisory system can trigger strategic adjustments of behavior to optimize task performance. In reaction time (RT) tasks, Rabbitt reported "post-error slowing"; after making an error, subjects become more cautious, in order to prevent new errors. On average, the RTs of correct trials following an error increase while error rates decrease (Laming, 1979b). Conversely, subjects are less cautious on trials preceding an error: The mean RTs of such trials are shorter than the mean RTs of correct responses preceding a correct response (Smith & Brewer, 1995). Henceforth, this effect will be called preerror speeding. Post-error slowing and pre-error speeding can be considered sequential effects.

On certain correct trials in between-hand choice RT tasks, the electromyographic (EMG) activity of the agonist involved in the required response is preceded by EMG activity of the agonist involved in the alternative, incorrect response; there is general agreement that these trials reflect small incorrect EMG activations at a level too small to reach the response threshold (Coles, Gratton, Bashore, Eriksen, & Donchin, 1985; Eriksen, Coles, Morris, & O'Hara, 1985; Smid, Mulder, & Mulder, 1990). As such, we will henceforth call these activations partial errors. Correct trials containing a partial error cannot be considered purely correct, so they will here be termed partialerror trials. Correct trials not preceded by a partial error will be named pure-correct trials.

Since it is well established that post-error slowing follows and pre-error speeding precedes full errors, one can wonder whether partial errors, which do not result in an overt failure, produce comparable sequential effects. In other words, is the executive control system able to function in a graded manner, so that it can trigger sequential adjustments proportional to the size of the dysfunction encountered in information-processing operations, or does it function instead in an all-or-none manner? This question makes sense in the context of contemporary models of executive control, in particular the conflict loop theory (Botvinick et al., 2001), whose core idea is that a conflictmonitoring module measures online the degree of conflict. Conflict is defined as the sum of the products of the activations of the possible responses weighted by the negative connection strengths between them. If only one response is activated, there is no conflict, but as soon as at least two responses are activated, conflict occurs. According to this model, when an error is committed, both responses are activated during a brief period following the incorrect response, leading to a large amount of conflict. Therefore, according to Botvinick et al., "periods of high conflict lead to a reduction in response priming resulting in slower but more accurate responding," explaining the post-error slowing. In a partial-error trial, the two responses are also activated in the same trial, so conflict does occur, but less so than in a full-blown error trial. Nevertheless, the period of moderate conflict should lead to adjustment of the RT on the next trial, but less so than after error. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.