; Distinguishing the contributions of implicit and explicit processes to performance of the weather prediction task
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Distinguishing the contributions of implicit and explicit processes to performance of the weather prediction task

VIEWS: 19 PAGES: 14

Examinations of the cognitive neuroscience of category learning frequently rely on probabilistic classification-learning tasks-namely, the weather prediction task (WPT)-to study the neural mechanisms of implicit learning. Accumulating evidence suggests that the task also depends on explicit-learning processes. The present investigation manipulated the WPT to assess the specific contributions of implicit- and explicit-learning processes to performance, with a particular focus on how the contributions of these processes change as the task progresses. In Experiment 1, a manipulation designed to disrupt implicit-learning processes had no effect on classification accuracy or the distribution of individual response strategies. In Experiment 2, by contrast, a manipulation designed to disrupt explicit-learning processes substantially reduced classification accuracy and reduced the number of participants who relied on a correct response strategy. The present findings suggest that WPT learning is not an effective tool for investigating nondeclarative learning processes. [PUBLICATION ABSTRACT]

More Info
  • pg 1
									Memory & Cognition
2009, 37 (2), 210-222
doi:10.3758/MC.37.2.210




                          Distinguishing the contributions of implicit
                            and explicit processes to performance
                                of the weather prediction task
                                                           AmAndA L. Price
                                           Elizabethtown College, Elizabethtown, Pennsylvania

                Examinations of the cognitive neuroscience of category learning frequently rely on probabilistic classification-
             learning tasks—namely, the weather prediction task (WPT)—to study the neural mechanisms of implicit learn-
             ing. Accumulating evidence suggests that the task also depends on explicit-learning processes. The present
             investigation manipulated the WPT to assess the specific contributions of implicit- and explicit-learning pro-
             cesses to performance, with a particular focus on how the contributions of these processes change as the task
             progresses. In Experiment 1, a manipulation designed to disrupt implicit-learning processes had no effect on
             classification accuracy or the distribution of individual response strategies. In Experiment 2, by contrast, a
             manipulation designed to disrupt explicit-learning processes substantially reduced classification accuracy and
             reduced the number of participants who relied on a correct response strategy. The present findings suggest that
             WPT learning is not an effective tool for investigating nondeclarative learning processes.



   Responding to environmental events often demands                     drack et al., 2001). For this reason, the WPT has been used
rapid categorization of both familiar and novel stimuli. The            extensively to examine the neural mechanisms that sup-
learning required for such categorization readily occurs                port nondeclarative learning (Aron, Gluck, & Poldrack,
under a variety of circumstances, and it is generally be-               2005; Foerde, Knowlton, & Poldrack, 2006; Poldrack
lieved that such learning is supported by multiple memory               et al., 2001; Poldrack, Prabakharan, Seger, & Gabrieli,
systems (Ashby, Alfonso-Reese, Turken, & Waldron, 1998;                 1999; Shohamy et al., 2004) and to assess the integrity
Ashby & Valentin, 2005; Maddox & Ashby, 2004; Poldrack                  of nondeclarative learning in various neuropsychological
et al., 2001; Poldrack & Packard, 2003; Smith, Patalano, &              populations (Beninger et al., 2003; Eldridge, Masterman,
Jonides, 1998; but see Lagnado, Newell, Kahan, & Shanks,                & Knowlton, 2002; Foerde et al., 2008; Hopkins, Myers,
2006, and Nosofsky & Johansen, 2000, for discussions of                 Shohamy, Grossman, & Gluck, 2004; Kéri, Szlobodnyik,
a unitary system of categorization). At least one of these              Benedek, Janka, & Gádoros, 2002; Knowlton et al., 1996;
systems supports declarative, explicit category learning,1              Knowlton et al., 1994).
which requires active memorization and/or hypothesis                       Although the WPT is commonly considered a measure
testing, whereas at least one other system supports non-                of nondeclarative learning, neuropsychological evidence
declarative, implicit category learning, which develops in              suggests that the task may rely on declarative-learning
a gradual and nonconscious manner.                                      processes. Patients with focal damage of the medial
   A number of cognitive, neuropsychological, and neuro-                temporal lobe (MTL), who experience a selective defi-
imaging studies have supported the existence of multi-                  cit of declarative memory, exhibit minimal learning even
ple category-learning systems (for relevant reviews, see                with considerable practice (Hopkins et al., 2004; Meeter,
Ashby & Valentin, 2005; Shohamy, Myers, Kalanithi,                      Myers, Shohamy, Hopkins, & Gluck, 2006). Further-
& Gluck, 2008). In these investigations, one of the most                more, neuropsychological research with healthy older
commonly used paradigms is the weather prediction                       adults and patients with Parkinson’s disease indicates that
task (WPT), a probabilistic classification task designed                WPT learning is greater among those with higher levels
to assess gradual, incremental learning over many trials                of executive function (Knowlton et al., 1996; Price, 2005).
(Knowlton, Squire, & Gluck, 1994). On this task, partici-               Executive function is positively associated with explicit
pants must categorize a set of visually presented cues that             category learning but has a negative or null relationship
are probabilistically related to one of two outcomes and                with implicit category learning (DeCaro, Thomas, &
then receive feedback on the accuracy of their response.                Beilock, 2008; Maddox & Filoteo, 2001), suggesting that
The probabilistic cue–outcome relationships are thought                 WPT performance depends on explicit processes. Behav-
to disrupt declarative-learning processes (Knowlton,                    ioral evidence provides further support for this possibil-
Mangels, & Squire, 1996; Knowlton et al., 1994; Pol-                    ity. Concurrent performance of an attention-demanding


                                                       A. L. Price, pricea@etown.edu


© 2009 The Psychonomic Society, Inc.                                210
                                                                                         WPT Learning Processes            211

secondary task, which has previously been shown to dis-           delivery of feedback triggers dopamine signals that alter
rupt explicit, but not implicit, category learning (Jiménez       synaptic connections associated with the immediately pre-
& Vázquez, 2005; Waldron & Ashby, 2001; Zeithamova                ceding stimulus and associated response (Arbuthnott, Ing-
& Maddox, 2007), impairs performance on the WPT                   ham, & Wickens, 2000; Hollerman & Schultz, 1997; Kerr
(Foerde et al., 2006; Foerde, Poldrack, & Knowlton, 2007;         & Wickens, 2001; Schultz, 1992). If feedback is delayed,
Newell, Lagnado, & Shanks, 2007). Furthermore, WPT                the feedback-mediated dopamine release will modify in-
performance is positively associated with participants’           appropriate synapses, thereby disrupting implicit learn-
awareness of the underlying structure (Foerde et al., 2007;       ing. Explicit category learning is unaffected by a delay
Gluck, Shohamy, & Myers, 2002; Knowlton et al., 1994;             in feedback, presumably because the relevant rule can be
Lagnado et al., 2006; Newell et al., 2007; Price, 2005;           verbalized and maintained in working memory until fee
								
To top