Sander A. Los

On the origin of mixing costs: exploring information processing in pure and mixed blocks of trials

In a reaction task responding has often proved faster when the levels of the independent variable are presented isolated from each other, in pure blocks than when they are presented randomly intermixed, in mixed blocks. This difference in response time is denoted here as mixing costs. This paper presents a theoretical review of mixing costs with special emphasis on their origin. In Section 1, two views on the origin of mixing costs are delineated, which are subsequently elaborated in Sections 2 and 3. The strategic view asserts that subjects are less well prepared in mixed blocks than in pure blocks, due to greater uncertainty about the level of the independent variable to be presented on the forthcoming trial. The stimulus-driven view attributes mixing costs to more pronounced trial-to-trial adaptation of processing in mixed blocks than in pure blocks, due to greater intertial variability. Section 4 reviews mixing costs deriving from various areas of human performance, and concludes that the dominant strategic view in the literature is not warranted, and that stimulus-driven factors have been underestimated.

Intentional and unintentional contributions to nonspecific preparation during reaction time foreperiods.

The nonspecific preparation that follows a warning stimulus (WS) to speed responding to an impending imperative stimulus (IS) is generally viewed as a strategic, intentional process. An alternative view holds that WS acts as a conditioned stimulus that unintentionally elicits a tendency to respond at the moment of IS presentation as a result of a process of trace conditioning. These views were contrasted as explanatory frameworks for classical effects on reaction time of the duration and intertrial variability of the foreperiod, the interval between WS and IS. It is shown that the conditioning view accounts for the available data at least as well as the strategic view. In addition, the results of 3 experiments provide support for the conditioning view by showing that unintentional contributions to nonspecific preparation can be dissociated from intentional contributions.

The foreperiod effect revisited: conditioning as a basis for nonspecific preparation.

The foreperiod (FP) is the interval between a warning stimulus and the imperative stimulus. It is a classical finding that both the duration and the intertrial variability of FP considerably affects response time. These effects are invariably attributed to the participants state of nonspecific preparation at the moment the imperative stimulus is presented. In this article, we examined a proposal by Los, S. A. (1996) [On the origin of mixing costs: exploring information processing in pure and mixed blocks of trials. Acta Psychologica, 94, 145-188] that the real-time development of nonspecific preparation during FP relies on the same principle as trace conditioning. To this end, we adjusted the formal conditioning model developed by Machado, A. (1997) [Learning the temporal dynamics of behavior. Psychological Review, 104 (2), 241-265], and fitted this model to a representative data set we obtained from nine participants. Although the model accounted for only a moderate proportion of the variance, it accurately reproduced several key features of the data. We therefore concluded that the model is a promising first step toward a theory of nonspecific preparation.

Intentional and unintentional contributions to nonspecific preparation: Electrophysiological evidence.

The authors hypothesized that there are distinct intentional and unintentional influences on nonspecific preparation for a future event. In 2 experiments, participants responded to an imperative stimulus (S-sub-2) that was presented equiprobably either 400 ms or 1,200 ms after the offset of a warning stimulus (S-sub-1). During the S-sub-1-S-sub-2 interval, the authors measured the contingent negative variation (CNV), an event-related brain potential reflecting nonspecific preparation. S-sub-1 provided either no information or reliable information about the duration of the impending S-sub-1-S-sub-2 interval, thereby allowing an intentional influence on the state of preparation. The effect of S-sub-1 information on the CNV was approximately additive to the effect of the S-sub-1-S-sub-2 interval that was used on the preceding trial. This supports the view that the preceding S-sub-1-S-sub-2 interval contributes unintentionally to the state of nonspecific preparation guided by a process of trace conditioning.

Identifying stimuli of different perceptual categories in mixed blocks of trials: evidence for cost in switching between computational processes.

Responding to stimuli of different perceptual categories is usually faster when the categories are presented isolated from each other, in pure blocks, than when they are presented intermixed, in mixed blocks. According to criterion models, these perceptual mixing costs result from the use of a less conservative response criterion in pure than in mixed blocks. According to alternate processing models, mixing costs result from time-consuming switching in mixed blocks between different computational processes called on by the different perceptual categories. In 5 experiments, participants had to identify number stimuli of different categories. The results showed clear mixing costs whenever these categories differed in their assumed computational processing requirements but not when they differed on features that seemed trivial from a computational viewpoint. The results favor the alternate processing conception.

Reweighting sequential effects across different distributions of foreperiods: segregating elementary contributions to nonspecific preparation.

In the study of nonspecific preparation, the response time (RT) to an imperative stimulus is analyzed as a function of the foreperiod (FP), the interval between a warning stimulus and the imperative stimulus. When FP is varied within blocks of trials, a downward sloping FP—function is usually observed. The slope of this function depends on the distribution of FPs (the more negative the skewness, the steeper the slope) and on intertrial sequences of FP (the longer the FP on the preceding trial, the steeper the slope). Because these determinants are confounded, we examined whether FP—RT functions, observed under three different FP distributions (i.e., uniform, exponential, and peaked) can be predicted, one from the other, by reweighting sequential effects. It turned out that reweighting explained very little variance of the difference between the FP—RT functions, suggesting a dominant role of temporal orienting strategies.

Identifying stimuli of different perceptual categories in pure and mixed blocks of trials: evidence for stimulus-driven switch costs

Responding to stimuli of different perceptual categories is usually faster when the categories are presented isolated from each other, in pure blocks, than when they are presented randomly intermixed, in mixed blocks; a difference denoted as perceptual mixing costs. The present study examined the contribution of strategic and stimulus-driven factors to these costs. The first two experiments showed that perceptual mixing costs were not reduced when participants were informed at the start of each trial in mixed blocks about the impending category. Furthermore, Experiments 1 and 3 showed that mixing costs were concentrated on those trials of mixed blocks where the perceptual category was different from that of the preceding trial. These results support the view that perceptual mixing costs derive from stimulus-driven trial-by-trial adjustments in processing. Some general implications for processing models are discussed.

Sound speeds vision through preparation, not integration.

In manual choice reaction time (RT) tasks, people respond faster to a visual target stimulus when it is accompanied by a task-irrelevant tone than when it is presented alone. This intersensory facilitation effect is often attributed to multisensory integration, but here we show it to be a reflection of temporal preparation. According to this view, the more rapidly processed tone serves as a warning signal (S1), which initiates preparation for the more sluggish visual target (S2). To test this view, we varied the delay between S1 and S2 in conjunction with the modality of S1 (auditory or visual). For brief delays, responses to S2 were faster when S1 was auditory than when it was visual. Crucially, however, this intersensory facilitation effect disappeared after correction for the difference in S1-detection time, equating the effective preparation period. This shows that sound speeds response to a visual target only through preparation.

Outlines of a multiple trace theory of temporal preparation.

We outline a new multiple trace theory of temporal preparation (MTP), which accounts for behavior in reaction time (RT) tasks in which the participant is presented with a warning stimulus (S1) followed by a target stimulus (S2) that requires a speeded response. The theory assumes that during the foreperiod (FP; the S1–S2 interval) inhibition is applied to prevent premature response, while a wave of activation occurs upon the presentation of S2. On each trial, these actions are stored in a separate memory trace, which, jointly with earlier formed memory traces, starts contributing to preparation on subsequent trials. We show that MTP accounts for classic effects in temporal preparation, including mean RT–FP functions observed under a variety of FP distributions and asymmetric sequential effects. We discuss the advantages of MTP over other accounts of these effects (trace-conditioning and hazard-based explanations) and suggest a critical experiment to empirically distinguish among them.

Inhibition of return and nonspecific preparation: Separable inhibitory control mechanisms in space and time

I examined the relation between two inhibitory processes operating on spatial and temporal representations. In two experiments, participants had to detect a peripheral target that was presented after a variable interval following the onset of an uninformative peripheral cue. For the shortest cue-target interval, target detection was faster at the cued than at the uncued location, but this effect was reversed for the longer cue-target intervals. This finding has been taken to reflect a buildup of space-related inhibition over time, known asinhibition of return. Also, target detection was slower when the cue-target interval of the preceding trial was longer than that of the current trial than when this was not so. This sequential effect has been taken to reflect an intertrial carryover of time-related inhibition. Crucially, the spatial and temporal effects were additive in both experiments, suggesting a modular organization of the underlying inhibitory processes.