Attitude survey data reduction using Chaid: An example in shopping centre market research

For response modelling, chi‐square automatic interaction detection (CHAID), it is claimed, is “easy to use, produces intuitively appealing results, and is statistically valid”. A market research project concerned with a proposed re‐development of a 100,000 sq. ft area of the largest city centre shopping centre in Europe afforded an ideal opportunity for assessing the practical value of CHAID. A survey of samples of existing shopping centre users sought responses on attractiveness and usage of various leisure facilities. Once the popular facilities had been identified, some appropriate segmentation of respondents showing a high level of interest in a particular facility was required. Initially, this was achieved by standard cross‐tabulation analysis with age, gender, frequency of visit, etc. and testing for significance. Subsequently, the same data was analysed using CHAID. The paper reports on the similarities and differences between results from the two approaches and focuses on the intuitive appeal and value of CHAID results to the shopping centre owner/managers.