Academic journal article Perception and Psychophysics

Constant versus Variable Response Signal Delays in Speed-Accuracy Trade-Offs: Effects of Advance Preparation for Processing Time

Academic journal article Perception and Psychophysics

Constant versus Variable Response Signal Delays in Speed-Accuracy Trade-Offs: Effects of Advance Preparation for Processing Time

Article excerpt

In two experiments, we used response signals (RSs) to control processing time and trace out speed-accuracy trade-off (SAT) functions in a difficult perceptual discrimination task. Each experiment compared performance in blocks of trials with constant and, hence, temporally predictable RS lags against performance in blocks with variable, unpredictable RS lags. In both experiments, essentially equivalent SAT functions were observed with constant and variable RS lags. We conclude that there is little effect of advance preparation for a given processing time, suggesting that the discrimination mechanisms underlying SAT functions are driven solely by bottom-up information processing in perceptual discrimination tasks.

A fundamental characteristic of human performance is the ability to trade speed for accuracy (Pachella, 1974). In virtually all perceptual, cognitive, and motor tasks, people can choose to respond relatively quickly and produce responses with relatively low accuracy, or they can choose to respond more slowly and achieve greater accuracy. Many researchers have studied these speed-accuracy trade-offs (SATs) in order to characterize the inherent flexibility of the mechanisms underlying human performance.

Two experimental procedures that have often been used to study SATs are the deadline procedure and the response signal (RS) procedure. In the deadline procedure, participants are told in advance of each trial to respond within a certain prespecified time after stimulus onset-that is, to respond quickly enough to beat a deadline (e.g., Green & Luce, 1973; lien, Ruthruff, Remington, & Johnston, 2005; Link & Tindall, 1971; Pachella, Fisher, & Karsh, 1968; Ratcliff & Rouder, 2000; for reviews, see Luce, 1986; Pachella, 1974; Wood & Jennings, 1976). People are remarkably good at generating responses just before the deadline has elapsed (Pachella & Pew, 1968), and their response accuracy increases monotonically with the length of the deadline. Thus, the deadline procedure allows researchers to manipulate participants' processing time and trace out empirical SAT functions like the one shown in Figure 1. This procedure can even be used in experiments in which the prespecified deadline varies randomly from trial to trial (e.g., Gopher, Armony, & Greenshpan, 2000; Kleinsorge, 2001; Link, 1971; but see Strayer & Kramer, 1994b).

An alternative to the deadline procedure is the RS procedure, in which participants are given an explicit signal at the moment at which they must respond (e.g., Corbett & Wickelgren, 1978; Dosher, 1976, 1982; Dosher, Han, & Lu, 2004; McElree & Carrasco, 1999; Ratcliff & McKoon, 1989; Reed, 1973, 1976; Wickelgren & Corbett, 1977). Reed (1973) introduced this procedure with a recognition memory experiment in which a single probe letter was presented in each trial for an old/new judgment. The probe letter was displayed for 0.5, 1, 2, 4, or 8 sec, and the participants were told to respond immediately at the offset of the probe letter. Letter offset thus served as the RS in this experiment, although auditory RSs have been more common in subsequent experiments with visual test stimuli (e.g., Dosher, 1976). The experimenter-controlled time from the onset of the test stimulus to the RS is known as the RS lag. After some practice, participants can fairly consistently respond within approximately 200-250 msec of the RS, so the RS lag effectively controls the amount of time used to perform the task on a given trial. Like the deadline procedure, then, the RS procedure allows researchers to manipulate participants' processing time, thereby tracing out empirical SAT functions of the form shown m Figure 1.

Although both the deadline and the RS procedures are widely used for tracing out empirical SAT functions, there are several clear and potentially important differences between them. For example, the RS procedure necessarily involves the presentation not only of the main task-relevant stimulus, but also of an auxiliary stimulus telling the participant when to respond. …

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.