Cognitive processes at work in CALL

Abstract

The seminal work by Ericsson and Simon established verbal reports as a genuine way to get a glimpse into the “black box” of the workings of the human mind, and it is now recognised as a method of value in the pursuit of new knowledge relating to, for instance, learner strategies, cognitive strategies, human-computer interaction, and functionality of software and hardware. The present paper presents a think-aloud study (concurrent verbal report) which was conducted with university students of English as a Foreign Language, who were using a CALL program to learn English syntax. The students were asked to verbalise their thoughts while conducting the task of parsing sentences and building tree diagrams on the computer. The computer program Visual Interactive Syntax learning (VISL) is operational in fifteen different languages ranging from English, French, and German to Arabic, Danish, Dutch, Latin, Greek, and Japanese, and allows the students to learn and work with syntax in their own speed. VISL contains a large pre-tagged corpus but also allows for free text input. The tree diagrams build on a strict division of form and function, and this is supported by the colour scheme of the interface. The purpose of the think-aloud (TA) study was to gain direct access—as far as that is possible—to the mental processes and strategies which the students employ in their use of the software in order to better understand the learning strategies they made use of in these processes. The study and the recordings took place in two phases: an initial recording at the beginning of the experiment and a later, second recording after a 10-week period of weekly sessions of working with the software. This study was intended to supply data that could supplement and illuminate quantitative data which were measured in a pre-test/post-test formula, in a design which included a control group. In the initial recordings there were some critical remarks relating to the software but these were almost completely absent from the second recording. An interesting aspect is that despite the fact that the content of this learning process—English syntax—continued to create problems for many of the participating students, this was not projected into dissatisfaction with the software; quite the contrary. The think-aloud protocols were analysed for information on strategies and human-computer interaction. The new information from this study is that it appears that the computer is able to further the qualities in the students which are desirable in solving the learning task. At the outset of the experiment the low-achieving students had very few cognitive strategies, but a high rate of affective strategies. During the time of the training these students developed the more appropriate strategies for the task. In other words, the computer (and/or the software) helped students develop a better strategy for learning, which leads to the question of whether teaching of strategy or awareness-raising exercises related to appropriate strategies would lead to higher learning outcomes if initiated from the beginning of the instructional efforts.