Maternal sensitivity, typically measured as patterns of mother's response to infant signals has been identified as an important determinant of the normally developing child's social and cognitive competence. This project will study the sources of variation in maternal response to infant signals using Signal Detection Theory. Studies to date which have identified maternal sensitivity as an important variable have left unanswered the degree to which variation in maternal response is due to differences in sensitivity at the sensory system level or with the response (choice making) system. In Experiment I we will assess the separate associations of maternal sensitivity and response bias measures with measures of mother-infant interaction and infant developmental outcome. We will also measure the association of infant visual attention with mothers' sensitivity and response bias levels in response to infant social signals. Mothers will visit the laboratory twice when their infants are 5 months old. First, a Signal Detection task will be used to measure maternal sensory sensitivity and response bias to variations in the cry signal independently of each other. Second, a simulated child-care task will be used to measure mothers' perception of control over stopping an audiotaped infant cry. These measures will be correlated with visual attentiveness of the infant at 6 months, with feeding patterns at 9 months, with Bayley scores at 18 months and compliancy behavior at 24 months. In Experiment II we again apply Signal Detection Theory to study the independent effects of sensitivity and response bias in predicting maternal response to infant temperament. Mothers will visit the laboratory twice when their infants are 5 months old. First, the simulated child-care task will be used to measure mothers' perception of control. Second, mothers will participate in the Signal Detection task. In the instructions for this task, mothers will receive an experimental manipulation, labeling the cry as either from an easy or a difficult infant. By experimentally manipulating (i.e., labeling) temperament we avoid the confounding of infant and maternal measures (e.g., questionnaire assessments) usually present when studying the effects of infant temperament on maternal response.
