Estimating cognitive structure from qualitative data

Date of publication: October 26, 1987


The paper describes a method which estimates consumers' cognitive structure with regard to a certain product class from qualitative data, l.e. data In the form of raw text. The raw text is first translated to a series of cognitive categories by means of computer-assisted content analysis. A matrix of proximities between these cognitive categories is then computed, using the distances between these categories In the text as input. Various ways to analyse this matrix are described. Method and results are illustrated using three sets of data: Focus groups, depth Interviews, and free continuous word association, all on the same product (cameras). The three data sets are compared In terras of reliability and convergent validity, showing that depth interviews are preferable to the other two data collection methods. Various marketing applications of the method are discussed, including product innovation and/or variation, copy development, and campaign evaluation.

  • PDF
  • This could also be of interest