|
Article Excerpt
This longitudinal study investigated the relative contributions of infant temperament, maternal sensitivity, and psychosocial risk to individual differences in preschool children's cognitive development. It also examined specific moderating effects between predictors as well as the specific mediating role of maternal sensitivity in the relation between psychosocial risk and children's cognitive development. A mixed sample comprising 27 low-risk (i.e., adult mothers) and 62 high-risk (i.e., adolescent mothers) mother-child dyads was evaluated at home on five occasions. At 6 and 10 months, infants' cognitive development was assessed using the Bayley Scales of Infant Development (BSID) (Bayley, 1993). At 15 and 18 months, mothers completed Goldsmith's (1996) Toddler Behavior Assessment Questionnaire (TBAQ), and observers completed Pederson and Moran's (1995) Maternal Behavior Q-Sort. At 36 months, preschoolers' cognitive development was reassessed using the BSID. Results showed that controlling for infant mental development scores, all three classes of variables contributed to differences in cognitive functioning. Psychosocial risk moderated the relations between interest persistence and preschooler cognitive development, and the interaction between risk and infant anger proneness tended toward significance (p = .07). Maternal sensitivity significantly, but partially, mediated the relation between psychosocial risk and cognitive development. Implications of these results for current understanding of the processes underlying socioemotional influences on cognitive development are discussed.
**********
Over the past two decades, there has been a growing body of research on the socioemotional predictors of the quality of children's cognitive functioning (Lewis, 1993; Robinson & Acevedo, 2001; van IJzendoorn, Dijkstra, & Bus, 1995). Infant emotionality or temperament (Halpern, Garcia Coll, Meyer & Bendersky, 2001), maternal behavior and psychological characteristics (Miller, Miceli, Whitman & Borkowski, 1996; van Bakel & Riksen-Walraven, 2002), and more distal ecological characteristics such as poverty and familial risk (Liaw & Brooks-Gunn, 1994), among others, have all been posited to contribute to cognitive development at the preschool and school-age periods. However, while evidence has been accumulating regarding the potential relation between each of these classes of variables and child cognitive functioning, little work has been conducted to assess the relative contributions of each within a more general model, as would be suggested by most theoretical approaches to child development (Belsky, 1984; Bronfenbrenner, 1979; Hinde, 1987; Sroufe, 1996). Consequently, few specific developmental process hypotheses have been explicitly examined. This is the general objective of the current study. Specifically, we will investigate the relative contributions of infant temperament, maternal sensitivity, and psychosocial risk in accounting for individual differences in cognitive development at the preschool age. Furthermore, specific interactions between (a) psychosocial risk and infant temperament characteristics and (b) maternal behavior and temperament in predicting preschool cognitive development will also be examined. Finally, the specific mediating effect of maternal sensitivity in the relation between psychosocial risk and cognitive development will be addressed. All hypotheses are addressed in a longitudinal design.
Predictors of Preschooler Cognitive Development
The major theoretical models of child development postulate that infant characteristics, parent-child interaction, and other more distal factors all contribute to child development across time (Belsky, 1984; Bronfenbrenner, 1979; Hinde, 1987; Sroufe, 1996). This particular study is set within these general models of development. Following these models, the socioemotional predictors of preschool children's cognitive development can be categorized in three groups: (1) infants' own emotional and behavioral characteristics such as temperamental differences in self-regulation capacities and attention (Fox, 1994; Kagan, 1997), (2) maternal interactive behavior (Bornstein & Tamis-LeMonda, 1997), and (3) broader ecological characteristics such as maternal IQ and education, maternal age, and family income and status.
Infants' temperamental and emotional characteristics. At least two hypotheses link infant temperamental characteristics with cognitive development. The first hypothesis suggests that individual differences in child irritability, activity levels, and attentional processes influence information-processing capacities and thus affect children's ability to learn from their environments and the cognitive development that follows. While there have been a few studies showing potential relations between temperament and information processing in infancy (Johnson, Posner, & Rothbart, 1991; Keenan, 2002; Lemelin, Tarabulsy, & Provost, 2002), studies that have addressed this issue from a longitudinal perspective have reported more varied and sometimes contradictory findings. For example, Matheny (1989) reported significant correlations between several dimensions of temperament assessed by both questionnaires and observations, such as activity level, adaptability, distractibility, and attention persistence and children's mental development from 1 to 12 years of age. In another study involving 208 pairs of twins, DiLalla, Thompson, Plomin, Phillips, Fagan, Haith, et al. (1990) reported predictive relations between two dimensions of temperament assessed at 9 months of age (e.g., activity level and task orientation) and measures of cognitive development from 1 to 3 years of age. However, in comparing the results of these two studies, there appeared to be some contradiction in that the relation between activity level and IQ was negative in the Matheny study, while the relation between the same two variables was positive in the DiLalla et al. study.
This variability in findings has led certain authors to suggest a second major hypothesis linking temperament to cognitive functioning. Here, it is proposed that the relation between these two variables is moderated by other aspects of the developmental ecology (Lewis, 1993; Robinson & Acevedo, 2001). For example, in a recent study, Halpern et al. (2001) reported significant negative correlations between difficultness assessed by mothers at 8 months and mental development during infancy, at 12 and 18 months, in both an appropriate-for-gestational-age (AGA) and a small-for-gestational-age (SGA) group of infants. However, in the SGA group, but not in the AGA group, the quality of mother-infant interaction during a free-play session moderated the relation between observed temperament and 12-month mental development, suggesting that the contribution of temperament to cognitive development was moderated by both biological and social risk factors.
The suggestion that infant characteristics are moderated by other ecological factors has received support from other studies. Maziade, Cete, Boutin, Bernier, and Thivierge (1987) reported strong positive correlations between difficult temperament and later IQ, but only for children of middle and upper socioeconomic status. In another recent study, Karrass and Braungart-Rieker (2004) indicated that maternal reports of infant distress to novelty at 4 months were positively related to IQ at 36 months but that later assessments of distress to novelty, at 12 months, were positively related only in the presence of insecure mother-infant attachment. Such interaction effects between temperamental characteristics of infants and other aspects of the developmental ecology have been reported in other studies, mostly with regard to mother-infant interaction (Miceli, Whitman, Borkowski, Braungart-Rieker, & Mitchell, 1998; Robinson & Acevedo, 2001), or psychosocial risk, usually defined as a function of socioeconomic factors (Andersson & Sommerfelt, 1999).
From a theoretical perspective, such interactions may be justified in different ways. For example, Maziade and colleagues (1987) have suggested that highly irritable infants may elicit more frequent mother-infant interaction than less irritable infants. To the degree that elicited maternal behavior in high-risk environments may be less appropriate for favorable cognitive development and that elicited behavior in low-risk environments may be helpful for cognitive development, one would expect an interaction between temperamental features of the child and maternal interactive sensitivity in predicting cognitive development. Thus, the proposed association between temperamental difficulties and lower levels of cognitive functioning in the preschool years may be moderated as a function of the quality of maternal behavior or the more distal ecological characteristics of the environment to which it is associated.
However, this pattern does not emanate clearly from the reviewed research results. Variations in findings may be explained in part by differences in how temperamental or emotional characteristics of infants are conceptualized among researchers in a general sense (see Goldsmith, Buss, Plomin, Rothbart, Thomas, Chess, et al., 1987) and with reference to the specific dimensions of temperament that are examined. Results sometimes concern the notion of "difficult" temperament (Maziade et al., 1987), distress to novelty (Karrass & Braungart-Rieker, 2004), activity level and task orientation (DiLalla et al., 1990), and others. Procedures used to assess such aspects of infants vary greatly, showing low or moderate convergence (Rothbart & Bates, 1998). Moreover, the frequent use of global categories such as positive and negative emotionality, as well as the frequent lack of concern for the ecological context in which measures are taken, may also help explain discrepancies in results from different research groups (but see Karrass & Braungart-Rieker, 2004).
In spite of the fact that in the current literature the relations between infant emotional and temperamental characteristics and cognitive development have not always been demonstrated with robustness (the reported correlations usually ranging from weak to moderate), the data nevertheless suggest that these infant characteristics may contribute, perhaps in interaction with other variables, to a part of the variance in cognitive functioning. What appears to be important in addressing this issue is to distinguish between different facets of temperament and to examine them as a function of different aspects of the developmental ecology. In the present study, we consider four different aspects of temperament as defined by Goldsmith (1996): activity level, social fearfulness, anger proneness, and infant interest. The prediction of cognitive functioning based on these four dimensions of temperament will be examined within the context of two aspects of the developmental ecology: maternal interactive behavior and psychosocial risk.
Maternal interactive behavior. Results from studies aiming at predicting cognitive development in infants and preschool children from measures of the quality of the mother-infant relationship or maternal behavior are more convergent than studies predicting cognitive development from temperamental features of the child. Generally, it has been found that...
|
