SOFTWARE STYLE AND INTERACTION AROUND THE MICROCOMPUTER

A study is reported in which pairs of users (both teacher pupil and pu pil-pupil pairs) collaboratively used each of eight standard, off-the- shelf items of educational software, four related to mathematics and f our to English language teaching. Initially, the software was also cla ssified as being either open (high in user control) or closed (lower i n user control), giving two programs in each category within each subj ect domain. The subjects' interactions (verbal and gestural) with each other as they used the programs were video-recorded and coded, and th e resulting data factor-analysed for each program. Cluster analysis of the factor solutions yielded groups of software, which corresponded i n part with the original open-closed classification, but which also qu alified it in the case of some of the programs. Two major clusters, ea ch of three programs, emerged. One cluster was comprised of three clos ed programs, whilst the other was comprised of two open programs plus one program that we had initially classified as closed. The remaining two programs were separate outliers. It was concluded that software st yle does indeed affect the patterns of interactions among pairs of use rs; for example, the group of programs that were more closed were asso ciated with greater amounts of explaining whilst the group of programs that were more open were associated with greater amounts of assessing and confirming (i.e. taking stock of progress). The results raise the issue of the frequency and nature of the prompts and feedback given b y programs and the possibility of designing software to shape the user s' interactions in specific ways.