Modelling Decisions of Woolproducers: Hierarchical Decision Models and Personal Construct Theory

Abstract

Models used to predict the supply of agricultural commodities that may be reliable for minor changes in key variables often have difficulties when major changes in the variables occur. A general aim of this research was to advance some explanations for the differences in forecasting performance between situations of major and minor changes in key variables. A major part of the study was to find a descriptive and predictive model of behaviour, which allowed these stimuli to be elicited, while minimising the level of contamination of the results from the theoretical assumptions of the model. An indepth appraisal was made of the various models of decision making based on utility theory. It was concluded they involve assumptions and behaviours that decision makers are unwilling or unable to comply with for some simple decisions, but more important, for complex decisions. An approach was chosen that incorporated Gladwin's (1977) hierarchical decision model and Kelly's (1955) personal construct theory. Personal construct theory provides an explanation for human behaviour and learning, and for selection of aspects in the hierarchical decision models that are used as the operational models of people's decision processes. The approach allows for the simplifying procedures people use in making their decisions. Theoretical bias is reduced because no assumptions are made about which factors should be included in the models. Models of production and marketing decisions based on this approach were developed and tested using information from wool producers in the Armidale Rural Lands Protection Board area of the New England Tableland in NSW. Most of the models achieved success rates of 80 per cent or better for individual decisions.