Academic journal article Attention, Perception and Psychophysics

Inhibition of Return in Static but Not Necessarily in Dynamic Search

Academic journal article Attention, Perception and Psychophysics

Inhibition of Return in Static but Not Necessarily in Dynamic Search

Article excerpt

If and when search involves the serial inspection of items by covert or overt attention, its efficiency would be enhanced by a mechanism that would discourage reinspections of items or regions of the display that had already been examined. Klein (1988, 2000; Klein & Dukewich, 2006) proposed that inhibition of return (IOR) might be such a mechanism. The present experiments explored this proposal by combining a dynamic search task (Horowitz & Wolfe, 1998, 2003) with a probe-detection task. IOR was observed when search was most efficient (static and slower dynamic search). IOR was not observed when search performance was less efficient (fast dynamic search).These findings are consistent with the "foraging facilitator" proposal of IOR and are unpredicted by theories of search that assume parallel accumulation of information across the array (plus noise) as a general explanation for the effect of set size upon search performance.

Inhibition of return (IOR) was first discovered by Posner and Cohen (1984; for a review, see Klein, 2000). Most scholars have explored IOR using the model task pioneered by Posner in which targets requiring simple responses are preceded by nonpredictive cues. Typically, at short cue- target onset asynchrony (CTOA), response times (RTs) to targets appearing at a previously cued location are shorter than for targets appearing at an uncued location. However, at long CTOA, the pattern is reversed and RTs to targets presented at a cued location are longer than those to targets presented at an uncued location. In their seminal study, Posner and Cohen proposed that the function of IOR is to encourage orienting toward novel items.

Visual search is a widely used paradigm in attention research. In a typical search task, participants are asked to search for a target among a group of distractors. The target may be present or absent in the array of items, and the participant makes a present/absent decision. When the search task is easy, the target "pops out" of the array, and RT is relatively unaffected by the number of items presented (set size). In contrast, when the search task is more difficult, RT increases, usually linearly, with the size of the array, and the slope of the RT 3 set size function on target-absent trials is steeper than that on target-present trials. To avoid a theoretical commitment-and although recognizing that using two labels implies a dichotomy when a continuum of difficulty might provide a more accurate characterization-we will refer to these two types of search tasks as "easy" and "difficult," respectively. As for the prototypical pattern of results with difficult search tasks, two broad classes of theory have been advanced to explain the effect of set size on RT. One class of theory (e.g., Broadbent, 1987; Ratcliff, 1978; Townsend, 1974) asserts that information is processed or accumulated in "parallel" across all the items in the array, and a response is issued once enough evidence has been accumulated for either the "present" or "absent" decision. A second class of theory (Treisman & Gelade, 1980; Wolfe, 1994; Wolfe, Cave, & Franzel, 1989) asserts that individual items in the array or regions containing a small number of items are inspected serially by an attentional operator to determine whether the item is a target, or whether the region contains one. The linear relationship between RT and set size, and the finding that the slopes for target-absent trials are often approximately twice those of target-present trials, are straightforward consequences of this serial, selfterminating search strategy. (We will refer to these two classes of theory on how difficult search is accomplished as parallel and serial, respectively.) Although serial theories implicitly assume that, once rejected as distractors in difficult search tasks, items are not (or not likely to be) reinspected, originally no explicit mechanism was described that would prevent or discourage such reinspections (but see Koch & Ullman, 1985). …

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.