ERIC Number: ED388689
Record Type: RIE
Publication Date: 1995-Apr
Heresies of the New Unified Notion of Test Validity.
By focusing on "appropriateness" and "adequacy" of inference and action, unified validity may be misused in rejecting valid test outcomes. The notion of levels of validity is challenged, the necessity of assumption is argued, and experience is proposed as the basis of validity. "Consequential validity" is interpreted as an optional predictive validity, a tangential validity that depends on organizational or political prerogative. Measurement validity is distinguished from test validity, which usually has more importance. Test validities such as content and predictive validities are perceived as demonstrable in contrast to construct measurement, which can never be proven. The claim that all validity is construct validity is challenged. The same claim can be made for any type of validity, and not all valid tests require constructs. Tests with valid constructs may not provide adequate predictive and content validity, whereas selection tests may be adequate. It is argued that test purpose is a more important validity issue than credibility of construct. "Structural item validity" is suggested as an alternative descriptor for items free of random and systematic bias; systematic bias does not constitute measurement of an additional construct. Noting the futility of attempting to demonstrate "construct validity," it is suggested that the term be renamed "construct feasibility." (Contains 11 references.) (Author/SLD)
Publication Type: Opinion Papers; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Note: Paper presented at the Annual Meeting of the National Council on Measurement in Education (San Francisco, CA, April 19-21, 1995).