ABSTRACT
Significance, as contrasted to confidence, is given to the testing of hypotheses. The admixing of the concepts of confidence and significance has become so prevalent in the psychological literature that one typically reads statements, in the reports of psychological research, indicating that certain experimental results were significant at, say, the 5 percent “level of confidence.” It may be that the confusion arises from the fact that one can utilize a confidence interval as a significance test and in doing so may hastily, but incorrectly, conclude that there is no difference between the two concepts. Confidence, a concept customarily reserved for discussions of interval estimation, is the faith which one is willing to place in a statement that an interval established by a sampling process actually contains or bounds a parameter of interest. Several different book reviewers who have commented on the confusion that exists in the psychological literature regarding the statistical concepts of confidence and significance.
