Purposive preferences for multi-attributed alternatives: A study of choice behaviour using personal construct theory in conjunction with decision theory

Abstract

The thesis is based on the notion that a person's behaviour is largely a result of the interplay between his beliefs and values.A model is described which ccmbines Personal Construct Theo~J (as a means of describing beliefs) and Multi-Attributed Utility Theory (as a means of describing values) in order to predict purposive choice behaviour.The model is applied to choice of records, books, clothes and role behaviours and is found to predict choices with a high degree of accuracy.Prediction using personal constructs is shown to be superior to that using supplied dimensions.Furthermore, construct weights elicited by a lottery tech~que are shown generally to be purposespecific and to give better predictions of behaviour than UIUt weights.'J.J 1 !-iii-"When a man starts to learn, he is never clear about his objectives.His purpose is faulty; his intent is vague.He hopes for rewards that will never materialize, for he knows nothing of the hardships of learning ••• •Learning is never what one expects.Every step of learning is a new task."(Castaneda, 1970, p.84)For Denise, who helps me through the "hardships".I~~--iv-GENERAL OVERVIEW Chapter 1 is an introductory chapter in which Personal Construct Theory and Multi-Attributed Utility Theory are described.The advantages and disadvantages of each theory are outlined and a synthesis is proposed which also incorporates the notion of purpose.The chapter concludes with a summary and statement of aims.Chapter 2 contains a detailed description of the methodology used in the study, and includes various methodological considerations.The question of validation is raised and the chapter concludes with a summary of the method.Chapter J documents the use of the method with seven subjects.The areas of application include choices among records, books and clothes.The results are discussed in relation to their theoretical and methodological implications.Two variations on the model are also described.One involves the comparison of predictions using constructs and semantic differential scales; the other describes an attempt to indicate the model's applicability in recurrent situations.Chapter 4 is written in two parts.The first part discusses the problem of ranking and weighting dimensions.The question of unit weighting schemes is discussed and the data from Chapter J are analysed in relation to this question.In the second part, applications of the model to the field of personality are discussed and the chapter ends with a summary of the major points thus far.Chapter 5 documents the use of the method to study the sentencing behaviour of nine magistrates.The question of validation is again -vdiscussed, with reference to the theory of generalizability.Chapter 6 contains the perceived implications of the results and some speculations concerning future use of the model.The studies ,~ described in Chapter 1 are reconsidered in terms of the model, and reformulations of the concepts of 'role' and 'attention' are offered.Chapter 7 contains a summary of the study and conclusions to be drawn from the results.The implications of the results for practical applications are presented, and the chapter concludes with some speculations concerning a theory of values.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.