Information processing model of subjective estimates of the evolution of dynamic processes illustrated for anticipated future mortgage rates

In recent years, the importance of incorporating attribute uncertainty in models of spatial choice behaviour has been recognized in urban planning research. The majority of studies concerned with decision-making under uncertainty assume some a-priori probability distribution for discrete attribute levels or continuous attribute values. Consequently, it has been implicitly assumed that the decision maker perceives the uncertain attributes as reflected in the presumed discrete or continuous probability distributions. This assumption may, however, not be necessarily true. Capturing the shape of the probability distributions from the decision maker’s perspective likely increases the accuracy of models of decision-making under uncertainty. The aim of the current paper, therefore, is to develop an approach for measuring and modelling individuals’ subjective beliefs about uncertain attributes. The approach is illustrated using beliefs about future mortgage rates as an example. To understand the impact of trends in the data, we experimentally changed the trends in mortgage rates over 20 years with 5 years intervals and analysed the impact of such trends on subjective beliefs of anticipated future mortgages. More specifically, four patterns of the evolution of mortgage rates were created, i.e. monotonically increasing, monotonically decreasing, increasing for the first four intervals and then decreasing, and decreasing for the first four intervals and then increasing. Results suggest that the shape of the pattern (nature of the trend) significantly influences subjective probability assessments of future mortgage rates.Highlights Imputations of subjective probability distribution of uncertain environment is crucial for urban and transportation planningThe information trend on an uncertain event has significant impact on sits subjective probability envisaged by respondents.Social demographic variables are significant moderators for subjective mean and standard deviation of uncertain events.