ABSTRACT
The purpose of this letter is to present commentary on the internal consistency estimates (α; Cronbach, 1951) obtained in the study by Hauck, de Lara, Pereira and Ruschel (2012) towards the validation of the five Big Five Personality factors in Brazil (Hutz et al., 1998). This new information may have a critical impact on the psychometric conclusions drawn by the authors, since measurement error should always be taken into account when evaluating the equivalence between test forms. The authors state that the levels of internal consistency obtained from the reliability coefficients varied from reasonable to satisfactory in magnitude; however, they did not put any emphasis on the fact that the variance in magnitude of these coefficients suggests that they cannot be considered homogeneous, nor that different amounts of measurement error are present in subscale scores. The present critique involves two types of measurement precision: of the scores of the five reduced scales examined, and the precision of the α coefficient.
KEYWORDS:
Reliability; Big Five; measurement error; alpha coefficient