Verbal reports as data in text comprehension research: An introduction

The current debate about the use of verbal reports in studying psychological processes refers back to the turn of the Century when the methodology of introspective reports in research on mental processes, whichhad been developed to a highly sophisticated level by the Würzburg school, enjoyed a legitimacy equal to the Wundtian experiments. The advent of behaviorism, with its justified criticism of the vague methodology, removed from the agenda of psychology for several decades not only the introspective methods but also the whole issue of mental processes. Apart from the relatively isolated field of the psychology of thinking, where a kind of introspection, the 'thinkingaloud' method, survived (Patrick, 1935; Maier, 1931; Duncker, 1945), the interest in verbal reports äs data was revitalized only with contemporary cognitive psychology which conceives äs genuine matter of psychology the study of restoring the mind and its processes. Although in recent years several forms of verbal report methodology have evolved and been used in various fields of cognitive psychology, there are still doubts äs to whether these methods are eligible when studying cognitive processes. In this introductory chapter, we shall first briefly review the most influential paper arguing against verbal report methods (Nisbett and Wilson, 1977). We will then present the framework offered by Ericsson and Simon (1980,1984) which puts serious epistemological constraints on such methods. Nevertheless, it legitimizes the verbal report methodology, if properly used. Next, we shall deal with verbal report methods in text research, and emphasize particularly the thinking aloud method (which is apparently the most widely used procedure) and literary text understanding, i.e., special comprehension processes which are very difficult to explore and where verbal