Applying an exemplar model to the artificial-grammar task: String completion and performance on individual items

Abstract

Jamieson and Mewhort (2009a) demonstrated that performance in the artificial-grammar task could be understood using an exemplar model of memory. We reinforce the position by testing the model against data for individual test items both in a standard artificial-grammar experiment and in a string-completion variant of the standard procedure. We argue that retrieval is sensitive to structure in memory. The work ties performance in the artificial-grammar task to principles of explicit memory.

Keywords:

• Artificial grammar
• Exemplar theory
• Implicit learning
• Minerva 2
• String completion

Acknowledgments

The research was supported by grants from the Natural Sciences and Engineering Research Council of Canada to both authors. We thank Phil Audette and Brian Hauri for their help in collecting the data. We thank Beth Johns and Matthew Kelly for comments on a draft of the paper. We thank Zoltan Dienes for providing us with his ranks and discussion of his data. Codes for the model and ample output are available from the first author.

Notes

¹ We report correlations here as descriptive statistics. Due to the small number of observations across which they are computed, any inferential analysis is questionable.

² We thank Professor Dienes for providing these ranks to us.

³ Since these experiments were conducted, the Toronto Stock Exchange (TSX) has appeared regularly in the daily news. The regular discussion around the TSX has caused participants to treat the trigram TSX differently than they had in the past.

⁴ Redington and Chater (1996) discussed the problem in the JOG procedure as the fixed-effect-fallacy. In a proper JOG experiment, each item would be equally likely to serve as a training or test item. We did not rotate items in our procedure because we were attempting to discover rather than ameliorate item effects.