Infant temperament and subject loss due to crying during operant conditioning

Infants who failed to complete a 2-day operant-conditioning task were compared with a stratified random sample of those who did on measures of infant temperament and several demographic characteristics. A discriminant-function analysis revealed that female infants who cried differed from female infants who did not cry on measures of duration of orienting and latency to approach sudden or novel stimuli. Reliable prediction of crying and noncrying could not, however, be made for males. No sex differences emerged in the incidence of crying or in the number of sessions completed. Partially successful females (i.e., those completing 1 of the 2 sessions) could reliably be discriminated from those who cried during the first session on measures of age at testing and maternal ratings of smiling behavior. The results of this study suggest that, as with habituation studies, subject loss in operant-conditioning studies is influenced by individual differences among the infants which may or may not adversely affect external validity.