Often research in judgment and decision making requires comparison of multiple competing models. Researchers invoke global measures such as the rate of correct predictions or the sum of squared (or absolute) deviations of the various models as part of this evaluation process. Reliance on such measures hides the (often very high) level of agreement between the predictions of the various models and does not highlight properly the relative performance of the competing models in those critical cases where they make distinct predictions. To address this important problem we propose the use of pair-wise comparisons of models to produce more informative and targeted comparisons of their performance, and we illustrate this procedure with data from two recently published papers. We use Multidimensional Scaling of these comparisons to map the competing models. We also demonstrate how intransitive cycles of pair-wise model performance can signal that certain models perform better for a given subset of decision problems.

Broadleaf weeds apparently have patchy distributions within a field while POST control decisions are made assuming a regular spatial distribution. As a result, yield loss from weed competition may be overestimated, possibly leading to mistakes in choosing the optimal control treatment. Data on distribution of broadleaf weeds in 14 soybean fields were used in simulation experiments to investigate the potential for improving decision making with information about weed patchiness. The feasibility of modeling weed distribution in individual fields was also examined. Overall, the cost of assuming a regular distribution when making POST decisions was found to be low. Errors that occurred most often involved recommending more intensive control than was actually required, although in a few cases less intensive control was recommended. Error in the yield loss estimated for the uncontrolled population did not indicate the potential for a mistake in decision making for a field. Accurately modeling distribution of weeds within fields may be difficult as a result of correlations between distributions of individual species within a field and variation in distributions between fields.

This research examines the predictability of a profit maximization model, an expected value-variance utility maximization (E-V) model, and two versions of the target-MOTAD model for modeling risky agricultural production decisions. Model solutions were translated into expected value and variance of farm income for analysis. Direct comparison and chi-square analysis of actual and predicted expected income distributions were used in the analyses. It was concluded that the utility maximization and cash-cost target-MOTAD models predicted distributions of farm income better than the variable-cost target-MOTAD and profit maximization models.

The accelerated aging of the Canadian population is a recognized fact and leads to an increasing number of seniors with cognitive impairments (Curateur public du Québec, 2010a). This has a definite impact on health professionals who have to assess their competency to live independently and manage their finances. This decision, which has important consequences for the person, must be based on an objective and rigorous assessment. The purpose of this paper is to analyse the available tools, both in the scientific literature and in clinical settings, to better document the various components to assess seniors’ competency to live independently and manage their finances. The goal is to help practitioners who work with older people with cognitive impairments to accurately assess their ability to manage themselves and their property. A review of the relevant literature and training available, as well as three group consultations, showed that there is no consensus about the tools used to assess the capacity to take care of oneself and one’s property. Additional studies are thus needed to fill the gap in knowledge about specific tools used to assess competency.