25.05.2009

J.J.M. (Rob) Schoonen

Psycholinguistic perspectives on assessment in applied-linguistic research

 

In most empirical research on language learning and language use, whether conducted to test theories of SLA or to investigate educational progress, it is the language user or learner who is at the focus of interest. In order to reach their research goals, researchers have to assess language user’s linguistic skills and language knowledge. Therefore, language testing, including the operationalization of the constructs we seek to understand, is part of almost all research in applied linguistics.

The empirical literature shows that in language testing “knowledge” has been assessed more frequently than “skills”, and end products (of receptive or productive language use) more frequently than the underlying processes. There might be good reasons for this ‘bias’, but I would like to claim that for a good understanding of language proficiency and language learning both knowledge and skill dimensions of language proficiency need our attention and that it would thus be beneficial for research to include more psycholinguistically-oriented measures in our assessments.

In this talk, I will present results we obtained in past and current research in which we used some psycholinguistically-oriented measures, in addition to more traditional knowledge measures. The psycholinguistic measures pertain to the fluency with which language users are able to perform certain linguistic tasks.

The first study (the NELSON project) was in the domain of Dutch L1 and EFL reading and writing. In this study we combined ‘traditional’ knowledge tests with psycholinguistic, time-sensitive tests using RT measures. The second study (the WiSP project) has a similar design, but focuses on (Dutch L2) speaking proficiency. Speaking is more time constraint and thus might be more sensitive to fluent accessibility of linguistic knowledge, or lack thereof. Finally, I will briefly outline an ongoing study that we have recently started, exploring a processing dimension of lexical skills in primary school children, i.e., accessibility of lexical knowledge, in addition to the well-known dimensions of breadth and depth of lexical knowledge.

To conclude we will discuss how this psycholinguistic approach compares to more task-oriented approaches, to what extent these kinds of measures can contribute to diagnostic language testing and, finally, how this approach relates to recent developments in validity theory.