Birth order differences in normal personality traits: Perspectives from within and outside the family

The hypotheses of Sulloway (1996) regarding birth order differences in five-factor model personality traits were tested in a sample of 231 college students with the NEO Five-Factor Inventory (NEO-FFI; Costa & McCrae, 1992). Data were collected from three sources (self, peer, and parent) to systematically evaluate previous observations that birth order differences are more commonly found when ratings are obtained from family members than from observers outside the family (Ernst & Angst, 1983). Using a between-family design, students were selected only from families with two or three full biological siblings and no half-siblings, step-siblings, or adopted siblings. Firstborn (n = 103) and laterborn (n = 128) students were compared using NEO-FFI ratings by the self, by a same-sex college peer, and by a biological parent. No birth order differences were found for any of the five NEO-FFI scores using any of the three rating sources. Effect sizes (Cohen’s d) were less than .20 for all comparisons. Suggestions for new approaches to the study of birth order differences are considered, including the use of within-family designs and more contextualized personality variables.