Cracks, Gaps, and Holes in Validation Practice as Evidenced From a Validation Synthesis of the English Version of the Rosenberg Self-Esteem Scale

Abstract

Abstract: We conducted a validation synthesis of the English version of the Rosenberg Self-Esteem Scale. Twenty-eight studies met our criteria for inclusion. Using the 2014 Standards for Educational and Psychological Testing ( AERA, APA, & NCME, 2014 ) as a frame of reference, we examined the sources of validity and reliability evidence gathered; identified trends, gaps, and issues with this evidence; and made recommendations to guide future validation practice. None of the studies directly cited any edition of the Standards. Validity evidence has relied predominantly on internal structure, followed by relations to other variables, and little to no other sources of evidence. Reliability evidence was reported in most studies but consisted exclusively of internal consistency estimates. Increased awareness and use of the Standards and better training in measurement could help improve validation practice. Recommendations to researchers include becoming more familiar with modern validity theory, paying more attention to sources of validity evidence that address score meaning, presenting clear and explicit a priori expectations of the evidence needed to support the validity of intended inferences, and evaluating obtained evidence with a more critical eye to determine the degree to which theoretical rationales and the evidence support the meaning of test scores for a given use, sample, and context.