{"pk":25471,"title":"Memory constraints affect statistical learning;\nstatistical learning affects memory constraints","subtitle":null,"abstract":"We present evidence that successful chunk formation during a\nstatistical learning task depends on how well the perceiver is\nable to parse the information that is presented between\nsuccessive presentations of the to-be-learned chunk. First, we\nshow that learners acquire a chunk better when the\nsurrounding information is also chunk-able in a visual\nstatistical learning task. We tested three process models of\nchunk formation, TRACX, PARSER, and MDLChunker, on\nour two different experimental conditions, and found that only\nPARSER and MDLChunker matched the observed result.\nThese two models share the common principle of a memory\ncapacity that is expanded as a result of learning. Though\nimplemented in very different ways, both models effectively\nremember more individual items (the atomic components of a\nsequence) as additional chunks are formed. The ability to\nremember more information directly impacts learning in the\nmodels, suggesting that there is a positive-feedback loop in\nchunk learning.","language":"eng","license":{"name":"","short_name":"","text":null,"url":""},"keywords":[{"word":"statistical learning; chunking; memory"}],"section":"Papers","is_remote":true,"remote_url":"https://escholarship.org/uc/item/5wm2t4q9","frozenauthors":[{"first_name":"Joshua","middle_name":"R","last_name":"de Leeuw","name_suffix":"","institution":"Indiana","department":""},{"first_name":"Robert","middle_name":"L","last_name":"Goldstone","name_suffix":"","institution":"Indiana","department":""}],"date_submitted":null,"date_accepted":null,"date_published":"2015-01-02T02:00:00+08:00","render_galley":null,"galleys":[{"label":"PDF","type":"pdf","path":"https://journalpub.escholarship.org/cognitivesciencesociety/article/25471/galley/15095/download/"}]}