Robert M. DeKeyser
Professor Emeritus, School of Languages, Literatures, and Cultures
Professor Emeritus, Second Language Acquisition
Professor Emeritus of Second Language Acquisition Robert DeKeyser is originally from the Flemish part of Belgium. After completing his BA at the University of Leuven, his MA and PhD at Stanford University, and a short stint with the Belgian National Science Foundation, he taught in the Linguistics Department at the University of Pittsburgh for 17 years. He has published in Applied Psycholinguistics, Studies in Second Language Acquisition, Language Learning, Language Testing, the Modern Language Journal, and Bilingualism, among others. He edited Practice in a Second Language (Cambridge University Press, 2007) and is author of many chapters in various handbooks. He served as editor of Language Learning from 2005 to 2010, as co-editor of the book series Studies in Bilingualism (Benjamins) from 2010 to 2013, and as associate editor of Bilingualism, Language and Cognition from 2013 to 2017. Robert DeKeyser's research interests concern primarily cognitive aspects of second language acquisition, from implicit and explicit learning mechanisms, automatization processes, and age differences in learning, to more applied concerns such as aptitude-treatment interaction, error correction, and the effects of study abroad.
Publications
Scrutinizing LLAMA D as a measure of implicit learning aptitude
In this study, we tested the hypothesis that researchers’ variable test instructions are the source of the inconsistent results.
Author/Lead: Takehiro Iizuka, Robert M. DeKeyserSince Gisela Granena’s influential work, LLAMA D v2, a sound recognition subtest of LLAMA aptitude tests, has been used as a measure of implicit learning aptitude in second language acquisition research. The validity of this test, however, is little known and the results of studies with this instrument have been somewhat inconsistent. In this study, we tested the hypothesis that researchers’ variable test instructions are the source of the inconsistent results. One hundred fourteen English monolinguals were randomly assigned to take LLAMA D v2 under one of three test instruction conditions. They also completed two implicit aptitude tests, three explicit aptitude tests, and a sound discrimination test. The results showed that, regardless of the type of test instructions, LLAMA D scores did not align with implicit aptitude test scores, indicating no clear evidence of the test being implicit. On the contrary, LLAMA D scores were negatively associated with scores on one implicit aptitude test, the Serial Reaction Time (SRT) task, but only in the condition where the instructions drew participants’ focal attention to the stimuli. This negative association was interpreted as focal attention working against learning in the SRT task. Implicit learning aptitude may be the degree to which one is able to process input without focal attention.
Read More about Scrutinizing LLAMA D as a measure of implicit learning aptitude