Reliability generalization of the Multigroup Ethnic Identity Measure-Revised (MEIM-R).

[Correction Notice: An Erratum for this article was reported in Vol 63(5) of <i>Journal of Counseling Psychology</i> (see record 2016-33161-001). The name of author Erika Feinauer was misspelled as Erika Feinhauer. All versions of this article have been corrected.] Individuals' strength of ethnic identity has been linked with multiple positive indicators, including academic achievement and overall psychological well-being. The measure researchers use most often to assess ethnic identity, the Multigroup Ethnic Identity Measure (MEIM), underwent substantial revision in 2007. To inform scholars investigating ethnic identity, we performed a reliability generalization analysis on data from the revised version (MEIM-R) and compared it with data from the original MEIM. Random-effects weighted models evaluated internal consistency coefficients (Cronbach's alpha). Reliability coefficients for the MEIM-R averaged α = .88 across 37 samples, a statistically significant increase over the average of α = .84 for the MEIM across 75 studies. Reliability coefficients for the MEIM-R did not differ across study and participant characteristics such as sample gender and ethnic composition. However, consistently lower reliability coefficients averaging α = .81 were found among participants with low levels of education, suggesting that greater attention to data reliability is warranted when evaluating the ethnic identity of individuals such as middle-school students. Future research will be needed to ascertain whether data with other measures of aspects of personal identity (e.g., racial identity, gender identity) also differ as a function of participant level of education and associated cognitive or maturation processes. (PsycINFO Database Record