NCJ Number
96279
Journal
Missouri Journal of Dispute Resolution Volume: 1984 Dated: (1984) Pages: 77-86
Date Published
1984
Length
10 pages
Annotation
This paper develops the idea that the hypothesis-testing procedure used in social science research is analogous to the arbitral process.
Abstract
The opposing positions of the parties in an arbitration proceeding correspond to the null and alternative hypotheses in a social science research problem. Social scientists generate data to test hypotheses by defining variables and developing research models; they identify a set of data which permit valid conclusions about hypotheses. Similarly, the testimony and written evidence adduced in the arbitral forum produce data for arbitrators. Just as social scientists must test data to determine whether to accept or reject null hypotheses (which usually stand for the status quo), arbitrators must also evaluate data and situations (the testimony, evidence, and arguments) to resolve the submitted issue. The concluding step of a research project is relatively easy for social scientists; various statistical tests and tables provide a quantitatively objective basis on which to accept or reject null hypotheses and provide a confidence that the data and analyses correctly describe the true variation in the dependent variable. The arbitrator's job, however, is just beginning; although there are no statistical tables, there is a similar process of balancing risks. Just as social scientists must explain the significance of their findings, arbitrators' decisions must be succinct and clear so that their impacts are self-evident. Finally, as social science research conclusions must be based on empirical data, so arbitral decisions should reflect what has proven validity in conflict resolution. Sixteen references are included.