Search

An inductive inference problem will often be structured so that the target variable (hypotheses) are logically distant from the observable events (data). In this situation it may be difficult or impossible to assess the probabilistic connection between them, but it may be possible to decompose the problem through the use of intermediate or explanatory variables. That is, it will often be possible to assess the likelihood of the observed data given some intermediate variable, and the likelihood of that intermediate variable given another, and so on, until the hypotheses of interest are reached. Inferences which incorporate one or more intermediate variables are called hierarchical, cascaded, or multistage inferences. The present paper presents a normative model for the solution of the general hierarchical inference problem. The formulation begins with a formal description of the hierarchical inference tree, including a discussion of various simplifying conditional independence assumptions. The solution is first derived for three special case models of differing structure, and then the algorithm for the general solution is given for two cases: one in which the conditional independence assumptions have been made and one in which they have not. [ABSTRACT FROM AUTHOR]

In many cases Bayes' theorem is an appropriate algorithm for the aggregation of probabilistic evidence. As with other statistical procedures, there are restrictions that must be taken into account. In the present paper we shall comment on several approaches that have been devoted to one of these restrictions; the incorporation of uncertainty about the true state of a datum. A datum is a variable which can be partitioned into equivalence classes. These classes represent the possible data states which will be also called events. In Bayes' theorem an event is an item of information which will be used for revising the opinion about the relative likelihood of hypotheses. In any specific situation only one of the possible events will be the true event. An event may come from a source whose reporting or observational accuracy is not perfect. An example may illustrate the issue. A medical doctor wants to come to a diagnosis. To achieve this he considers several data. One datum might be the result of a medical test, which has three possible states: positive, negative, inconclusive. If the doctor reports the state of the datum to be positive, he may be wrong by whatever reasons. That is, the report of an event must not necessarily coincide with the actual or true event of the datum under consideration. This kind of uncertainty about the true event in any specific situation is a characteristic of the source. In most cases it will reduce the diagnostic impact of an event. Whenever the report of an event and the true event coincide imperfectly, measures of source inaccuracy must be incorporated into Bayes' theorem. The task can be considered as two stage probabilistic induction. The first step is induction from the reported to the actual event, the second is induction from the actual event to the creditation of hypotheses. This is the reason why Gettys and Willke (1969) speak about "cascaded inference." [ABSTRACT FROM AUTHOR]

This paper is a secondary analysis of the data from two risk taking gambling experiments to test the value of risk hypothesis. The value of risk hypothesis is said to be true if the value of taking a risk is positive and if the level of risk at maximum expected value shifts towards greater risk for groups compared with individuals. Based upon the task, a method for imputing the value of risk is presented and then used to test the hypothesis. Data support the assumption of a positive value for risk taking but do not support a shift in maximum expected value risk. Therefore, the data do not support; the value of risk hypothesis. [ABSTRACT FROM AUTHOR]

The paper develops a model of organizational participative decision making (PDM) which emphasizes equilibrium and social change. Using attitudes as a mediating variable, the analysis explores organizational homeostatic reactions to an experimental PDM program. Various mechanisms are described by which PDM may contribute to motive satisfaction of managers and subordinates and to organizational goal attainment. Using this approach, the effects of several mediating variables on PDM are entertained. The PDM literature is summarized briefly and criticized in depth. Research in nonorganizational settings is not relevant, for the central issues in organizational PDM are ignored. Although less at fault, observational studies of organizations offer insufficient support for the PDM hypothesis. Experimental studies in organizations could provide the best support, but the data are mixed (yet suggestive). It is suggested that future research focus on mediating factors in PDM. [ABSTRACT FROM AUTHOR]