Bayesian statistics for evaluation research: an introduction
In: Contemporary evaluation research 8
10 Ergebnisse
Sortierung:
In: Contemporary evaluation research 8
In: Social science quarterly, Band 70, Heft 2, S. 534-535
ISSN: 0038-4941
In: American behavioral scientist: ABS, Band 30, Heft 6, S. 661-676
ISSN: 1552-3381
In: American behavioral scientist: ABS, Band 30, Heft 6, S. 661
ISSN: 0002-7642
In: Knowledge, Band 5, Heft 1, S. 56-83
The purpose of this article is to present the basic ideas of Bayesian statistics and consider their relevancefor the topic of utilization ofevaluation research. It is argued that whether utilization is viewed in terms of modifying thinking concerning social programs or in terms of affecting specific decisions, the Bayesian approach, with its requirement of coherence in inference and decision making, can be useful in analyzing data and expressing results, in guiding appropriate utilization, and in conceptualizing aspects of the utilization process.
In: The Journal of social psychology, Band 89, Heft 1, S. 35-45
ISSN: 1940-1183
In: Evaluation and Program Planning, Band 8, Heft 2, S. 161-169
In: Evaluation and program planning: an international journal, Band 8, Heft 2, S. 161-169
ISSN: 0149-7189
In: Knowledge, Band 1, Heft 4, S. 477-498
In: Decision sciences, Band 6, Heft 4, S. 739-751
ISSN: 1540-5915
The paper presents an extension of decision theory to the analysis of social power. The power of a person, A, over another person, B, is viewed in terms of the effect A has on B's decision. The analysis is based on the idea that B's decision regarding the performance of alternative behaviors is a function of 1) B's utility for the consequences of the behaviors and 2) B's subjective probabilities that the behaviors will lead to these consequences. In these terms, A's power over B lies in A's ability to mediate various consequences for B, contingent upon B's compliance or noncompliance.Subjects were asked to consider eight situations in which hypothetical individuals had to make a choice between two courses of action. In each situation another person (A) was attempting to induce the hypothetical individual (B) to choose one of the alternatives, while various situational factors were influencing B to choose the other alternative. The subjects were asked to consider B's utilities and subjective probabilities in each situation and to indicate whether or not B should comply with A and to make ratings of A's power. The decision theory analysis did well in predicting whether or not subjects would indicate that B should comply with A. Also, subjects generally were able to correctly specify whether A or the situational factors had more influence over B's decision. Finally, the subjects' ratings of A's power in the eight situations were highly related to the decision theoretic measure of power.