Communicative language testing: Integrating a listening comprehension section and communicative features in a computer -based Spanish test

This dissertation was motivated by an observed incongruity between the methodology employed in current second language teaching—the communicative approach—and the pedagogical orientation found in the majority of Spanish placement tests—the form-focused methods that characterize traditional language teaching. Two main aspects were identified as particularly needing attention: the general absence of a listening comprehension measure and the persisting presence of features from pre-communicative days. The first goal of this study was to determine whether adding a listening comprehension measure to a test of vocabulary, grammar, and reading comprehension has an effect on its accuracy to assess Spanish proficiency. The second goal was to determine whether there are any significant differences in the performance of a subset of items written in accordance to communicative guidelines (experimental items) and another subset of items that follow a more form-based approach (control items), taken from existing exams. A review of the literature on the communicative approach to language teaching and testing from the 1970s until present provided the necessary guidelines for the writing of the experimental items and the selection of the control items. An analysis of the most salient features found in Spanish placement exams, as well as a review of the latest research on computer-based language testing are also included. The research instrument was a computer-based, multiple-choice, self-scoring test, specifically designed for this study. The test was administered during the 2001 Spring semester to a total of 233 students enrolled in elementary and intermediate Spanish language classes at two universities. The analyses of the data confirmed that there is a significant effect associated with the methodology reflected in the items on a test. The listening section was found to accurately discriminate among levels of skill. The study also demonstrated that it is possible to develop a test that: (1) has the advantages of delivery by computer, (2) has features that are in accordance with the communicative approach employed in second language instruction, and (3) is a valid measure of Spanish ability. More importantly, the test demonstrated the feasibility of including a component for the assessment of listening skills