Academic journal article Memory & Cognition

Implicit Sequence Learning Is Represented by Stimulus-Response Rules

Academic journal article Memory & Cognition

Implicit Sequence Learning Is Represented by Stimulus-Response Rules

Article excerpt

For nearly two decades, researchers have investigated spatial sequence learning in an attempt to identify what specifically is learned during sequential tasks (e.g., stimulus order, response order, etc.). Despite extensive research, controversy remains concerning the information-processing locus of this learning effect. There are three main theories concerning the nature of spatial sequence learning, corresponding to the perceptual, motor, or response selection (i.e., central mechanisms underlying the association between stimulus and response pairs) processes required for successful task performance. The present data investigate this controversy and support the theory that stimulus-response (S-R) rules are critical for sequence learning. The results from two experiments demonstrate that sequence learning is disrupted only when the S-R rules for the task are altered. When the S-R rules remain constant or involve only a minor transformation, significant sequence learning occurs. These data implicate spatial response selection as a likely mechanism mediating spatial sequential learning.

Learning sequences of motor behaviors based on spatial stimuli (e.g., parallel-parking a car, playing the piano, dancing the tango, etc.) is an essential part of our daily lives. It is, therefore, not surprising that the study of sequence learning has been of considerable scientific interest for decades (for reviews, see Cleeremans, Destrebecqz, & Boyer, 1998; Clegg, DiGirolamo, & Keele, 1998). Despite the abundance of research, fundamental questions remain about the cognitive processes involved. One unresolved issue involves the information-processing locus of implicit sequence learning. Some theories propose that people learn associations between sequential stimuli, so the learning occurs in relatively early stimulus-encoding processes (stimulus- based theory; e.g., Clegg, 2005; A. Cohen, Ivry, & Keele, 1990; Howard, Mutter, & Howard, 1992; Keele, Jennings, Jones, Caulton, & Cohen, 1995; Mayr, 1996; Stadler, 1989; Verwey & Clegg, 2005). Other theories propose that people learn the sequence of rules associating stimuli and responses, so the learning occurs within central response selection processes (stimulus-response [S-R] rule theory; see, e.g., Deroost & Soetens, 2006; Hazeltine, 2002; Schumacher & Schwarb, 2009; Schwarb & Schumacher, 2009; Willingham, Nissen, & Bullemer, 1989). Finally, other theories propose that people learn associations between sequential responses, so the learning occurs in relatively late response execution processes (response-based rule theory; e.g., Bischoff-Grethe, Goedert, Willingham, & Grafton, 2004; Koch & Hoffmann, 2000; Willingham, 1999; Willingham, Wells, Farrell, & Stemwedel, 2000). Here we focus specifically on the S-R rule theory of sequence learning, since we believe that it provides a unifying explanation of the sequence-learning literature.

Since its introduction, the serial reaction time (SRT) task has become a paradigmatic procedure for studying implicit spatial sequence learning (Nissen & Bullemer, 1987). In the SRT task, participants respond to a target presented in one of several (typically 3-6) possible locations on each trial. Although participants are not told that a sequence exists, the targets are, in fact, presented in a specific sequential order (typically 6-12 positions long) that may be repeated several times throughout a given block of trials. Reaction times (RTs) decrease with practice, presumably (at least in part) because participants benefit from knowledge of the sequence.

Because other factors of general skill learning may also speed RT with practice, many SRT studies measure sequence learning with a transfer effect procedure (see, e.g., A. Cohen et al., 1990; Keele et al., 1995; Willingham et al., 1989). The transfer effect is measured by comparing the mean RTs on a block of random trials (termed the transfer block) with the mean RTs from the preceding and succeeding sequenced-trial blocks. …

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.