ERIC Number: ED408853
Record Type: RIE
Publication Date: 1997-Mar
Implications for Teaching Graduate Students Correct Terminology for Discussing Validity and Reliability Based on a Content Analysis of Three Social Science Measurement Journals.
Daniel, Larry G.; Witta, E. Lea
Although reliability and validity are characteristics of test data, social scientists often attribute reliability and validity erroneously to the tests themselves. To determine the extent to which this problem exists, 150 reliability and validity studies selected from 3 prominent social science measurement journals over a 3-year period were analyzed for common errors in terminology and categorized according to methodology types used in assessing reliability and validity. Results indicate over 50 percent of the articles contained more than one inappropriate statement concerning reliability or validity. It is suggested that professional journal reviewers and editors could improve research practice by catching and correcting a larger percentage of these errors. In the educational research classroom, it is recommended that teachers emphasize that reliability and validity are properties of data, model correct language about score characteristics while discussing reliability and validity in the presence of their students, and correct students' inappropriate use of language. Study data are appended. Contains 12 references. (MSE)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual Meeting of the American Educational Research Association (Chicago, IL, March 24-28, 1997).