The Class Activities Questionnaire (CAQ) was used to compare a Canadian high school participating in a research project employing an information retrieval system with another not involved in the project. This paper evaluates the validity of the CAQ as a measurement instrument for this study. The CAQ was administered to 1831 students, 806 Science and 1025 English students, from the two high schools in Ottawa. A principal components factor analysis with a varimax rotation confirmed the earlier results of Steele, the developer of the instrument, that the cognitive and behavioral dimensions of instruction can be assessed and described using the CAQ. Although additional items are recommended to improve the measurement characteristics of the CAQ, results indicate that it is sufficiently reliable and valid for experimental use. Caution is advised against its use for routine classroom purposes. (Author/LR)