Long-term negative priming: Support for retrieval of prior attentional processes

Abstract

Negative priming reveals that participants respond slowly to a probe target that was a task-irrelevant distractor in the preceding prime display (e.g., Tipper, 1985) and is thought to reflect processes mediating short-term behaviour. However, since the first surprising reports that negative priming is found with meaningless stimuli across delays of 30 days (e.g., DeSchepper & Treisman, 1996), researchers have questioned the existence of long-term negative priming effects. Because long-term negative priming could indicate that task-irrelevant information leaves a memory trace that impacts performance over time, such a finding is of immense theoretical importance. Indeed, the current research finds support for the existence of long-term negative priming as well as its generality across different stimuli and conditions. The authors propose that the initial processes that prevent response to irrelevant stimuli may be stored in memory, where retrieval of these processes can mediate behaviour over time.

Acknowledgments

The research reported here was made possible through financial support to Sarah Grison by a PhD Studentship from Booklands Booksellers, the Overseas Research Student Scheme, and the Biotechnology and Biological Sciences Research Council (BBSRC Grant 5/S1355).

Notes

Although not the focus of the current research, it is important to note that models have been developed to explain negative priming effects based on a variety of mechanisms, such as attentional inhibition (e.g., Tipper, 1985), response blocking (e.g., Tipper & Cranston, 1985), episodic retrieval (e.g., Neill et al., 1992), code coordination (e.g., Lowe, 1979), feature mismatch (e.g., MacDonald, Joordens, & Seergobin, 1999), perceptual mismatch (e.g., Park & Kanwisher, 1994), transfer-inappropriate processes (e.g., Neill & Mathis, 1998), dual inhibition and memory mechanisms (e.g., Kane, May, Hasher, Rahhal, & Stoltzfus, 1997), perceptual load (e.g., Lavie & Fox, 2000), and temporal discriminability (e.g., Milliken, Joordens, Merikle, & Seiffert, 1998).

In an effort to design a task that would minimize context changes between the prime and the probe, and thus perhaps be more likely to obtain long-term negative priming effects, the research described here incidentally employed a switch in task between the prime and probe. Importantly, a variety of research has found short-term negative priming effects when there is a task switch between prime and probe trials. For example, Tipper and his colleagues have consistently found short-term negative priming when the ignored prime distractor and subsequent probe target are physically identical, but there is a change in selection criteria between the prime and probe (e.g., Milliken et al., 1994; Tipper & Cranston, 1985; Tipper, Weaver, & Houghton, 1994; Tipper, Weaver, & Milliken, 1995). Furthermore, research indicates that short-term negative priming can be obtained when the prime and probe tasks differ (e.g., Healy & Burt, 2003; Kahan, 2000; Loach & Mari-Beffa, 2003; Milliken & Joordens, 1996; Milliken et al., 1998; Milliken, Lupiáñez, Debner, & Abello, 1999; Milliken & Rock, 1997; Neill, Terry, & Valdes, 1994; Ortells, Fox, Noguera, & Abad, 2003; Ortells & Tudela, 1996; Wood & Milliken, 1998), when the stimuli change between the prime and the probe display (e.g., Allport et al., 1985; Buchner et al., 2003; Dalrymple-Alford & Budayr, 1966; Driver & Baylis, 1993; Lowe, 1979; Neill, 1977; Tipper & Driver, 1988), and when the required response is altered between the trials (e.g., Chiappe & MacLeod, 1995; Tipper et al., 1988). Finally, although there is very little research investigating the existence of long-term negative priming, one unpublished study found these effects with word stimuli over a lag of 100 displays and a delay of one week when participants performed a naming task in the prime and a lexical decision task in the probe (E. Neumann & Russell, 2000).

* p < .05; # p < .07.