Journal of Environmental Psychology (1988), 8, 45-55 ASPECTS OF HOUSING PREFERENCE: REVISITING A CROSS-CULTURAL STUDY WITH THE HINDSIGHT OF IMPROVED DATA ANALYSIS MARK GROVES* and ROSS THORNE Department of Architecture, University of Sydney, Sydney, NSW 2006, Australia Abstract Since the late 1960s, social scientists have been confronted with the problem of attempting to ascertain how people interacted with the real world of buildings and outdoor spaces. With the benefit of the results of nearly 20 years of research, this paper reconsiders the concept of environmental preferences with particular reference to recent work on the prototypical attributes of housing environments. The theore- tical re-evaluation of Canter and Thorne's (1972) studio, combined with PINDIS data analyses, indicated a consistent clustering of stimuli by the two respondent groups (Scots and Australian architectural students). The slight differences in the networks of association between clusters were attributed to group differences in environmental experience. The discussion highlights the limitations of preference paradigms that use an analysis of prototypical environmental attributes as their base. Rather than provide a cookbook approach to design, research has a re- sponsibility to provide designers with an understanding of process and to equip them with appropriate techniques to incorporate process in the design of place. Introduction The year 1967 saw the formation of the Building Performance Research Unit in Glasgow; 1969 saw the publication of the first journal devoted to environment- behaviour research and the first inter-disciplinary conference of the newly formed Environmental Design Research Association. This conference, like its annual successors, brought together members of the design and social science professions, frequently to present joint research. Social scientists were now being confronted with the problem of attempting to ascertain how people interacted with the real world of buildings and outdoor spaces. What were people's attitudes to their working and living environments? How satisfied were they with these environments? How did people of different cultures respond to the same environments? During the intervening years, from the late 1960s until the present, this kind of research has become commonplace; methods of analysing data are considerably more sophistic- ated, and the theoretical interpretation of findings now has the benefit of the results of nearly twenty years of research. With all this in mind it was decided to return to the data of a project and carry out a re-analysis at a number of levels. This study of house types, conducted by Canter and Thorne (1972) and cited elsewhere (e.g. Ittelson et al., 1974; Becker, 1977; Nasar, 1984), was reported as showing particular preferences by the two different groups surveyed. In this paper: (i) the original data will be reviewed to assess whether the interpretation of 'preference', as originally stated, requires revision; * To whom correspondence and requests for reprints should be sent. 0272 4944/87/010045 + 11 $03.00/0 ~ 1988 Academic Press Limited