Affect Dysregulation

|

DPP16~1! 440

1026

02013004

11:29 am

REVISED PROOF

Development and Psychopathology 16 ~2004!, 43–68 Copyright © 2004 Cambridge University Press Printed in the United States of America DOI: 10.10170S0954579404040404

Affect dysregulation in the mother–child relationship in the toddler years: Antecedents and consequences

NICHD EARLY CHILD CARE RESEARCH NETWORK

Abstract The purpose of this study was to examine child, maternal, and family antecedents of children’s early affect dysregulation within the mother–child relationship and later cognitive and socioemotional correlates of affect dysregulation. Children’s affect dysregulation at 24 and 36 months was defined in the context of mother–child interactions in semistructured play and toy cleanup. Dyads were classified as dysregulated at each age based on high negative affect. Affect dysregulation was associated with less maternal sensitivity and stimulation, more maternal depressive symptoms, and lower family income over the first 36 months of life. Children with early negative mood, lower Bayley Mental Development Index scores and insecure-avoidant ~15 months! or insecure-resistant attachment classifications ~36 months! were more likely to be in an affect-dysregulated group. Controlling for family and child variables, affect-dysregulated children had more problematic cognitive, social, and behavioral outcomes at 54 months, kindergarten, and first grade. The findings are discussed in terms of the early role played by parents in assisting children with affect regulation, the reciprocal nature of parent–child interactions, and the contribution of affect regulation to children’s later cognitive, social, and behavioral competence.

The first 3 years of life are an important time in the development of a child’s emotional and behavioral self-regulation ~Thompson, 1998!. Although the literature often refers to selfregulation as a characteristic of the child, schol-

ars increasingly argue that self-regulation in infancy is best conceptualized as a quality of the infant–caregiver relationship, rather than as a characteristic of the infant alone. Sroufe ~1996! noted that the role of the caregiver in

This study is directed by a Steering Committee and supported by NICHD through a cooperative agreement ~U10!, which calls for scientific collaboration between the grantees and the NICHD staff. Participating investigators, listed in alphabetical order, are Virginia Allhusen, University of California, Irvine; Jay Belsky, University of London; Cathryn L. Booth, University of Washington; Robert Bradley, University of Arkansas, Little Rock; Celia A. Brownell, University of Pittsburgh; Margaret Burchinal, University of North Carolina, Chapel Hill; Susan B. Campbell, University of Pittsburgh; K. Alison Clarke–Stewart, University of California, Irvine; Martha Cox, University of North Carolina, Chapel Hill; Sarah L. Friedman, NICHD, Bethesda, Maryland; Kathyrn Hirsh–Pasek, Temple University; Aletha Huston, University of Texas, Austin; Elizabeth Jaeger, Temple University; Jean F. Kelly, University of Washington; Bonnie Knoke, Research Triangle Institute, Research Triangle, North Carolina; Nancy Marshall,

Wellesley College; Kathleen McCartney, Harvard University; Marion O’Brien, University of Kansas; Margaret Tresch Owen, University of Texas, Dallas; Chris Payne, University of North Carolina, Greensboro; Deborah Phillips, National Research Council, Washington, DC; Robert Pianta, University of Virginia; Wendy Robeson, Wellesley College; Susan Spieker, University of Washington; Deborah Lowe Vandell, University of Wisconsin, Madison; and Marsha Weinraub, Temple University. The authors express appreciation to the study coordinators at each site who supervised the data collection, the research assistants who collected the data, and especially the families and child care providers who welcomed the authors into their homes and workplaces and cooperated willingly with repeated requests for information. Address correspondence and reprint requests to: NICHD Early Child Care Research Network, NICHD, 6100 Executive Blvd., 4A01, Rockville, MD 20852.

43

|

|

DPP16~1! 440

2026

02013004

11:29 am

REVISED PROOF

44 the infant’s regulation of emotion is so fundamental that researchers have used terms such as “co-regulation” ~Fogel, 1993! and “mutual regulation” ~Tronick, 1989!. These terms describe the process whereby the child’s selfregulation arises out of the patterns of arousal and affect modulation repeatedly experienced in early caregiving relationships. Both the goodness of fit model of Thomas and Chess ~1977! and attachment theory ~Ainsworth, Blehar, Waters, & Wall, 1978; Bowlby, 1968; Carlson & Sroufe, 1995! emphasize the importance of the child’s relationship with the primary caregiver as central to understanding the developmental processes leading from early affective arousal and attentional control to later functioning. Thomas and Chess ~1977! and Sameroff and Chandler ~1975! argued that the caregiver–child relationship emerges as a function of ongoing transactions in which each member of the dyad changes the other’s behaviors and expectations in an ongoing feedback loop. The quality of the dyadic relationship over time, including children’s mood, compliance with maternal requests, and ability to use affect to communicate security needs, appears to be an important component of individual differences in children’s ability to regulate negative affect and may have long-term implications. During the first 3 years of life, children are navigating several successive stage-salient tasks related to self-regulation. These include establishing physiological regulation, synchronizing dyadic interaction, forming attachments to caregivers, learning to manage impulses, establishing autonomy, and internalizing norms and standards for social interaction ~Cicchetti, Toth, Bush, & Gillespie, 1988; Kopp, 1982, 1989; Sroufe, 1979!. Most children negotiate these developmental tasks relatively smoothly, although some are seen as showing brief periods of disruptive or noncompliant behavior ~Campbell, 2002!. However, for a small group of children these difficult transitions may escalate into negative, coercive exchanges with caregivers that are seen as a potential precursor of later behavior problems ~Bates, Pettit, & Dodge, 1995; Campbell, 2002; Campbell, Shaw, & Gilliom, 2000!. A common theme in most conceptual-

NICHD Early Child Care Research Network

izations of children’s early problem behavior is whether and in what contexts they are unable to control negative affect ~Thompson, 1994, 1998!, and thus, for example, show extreme levels of anger and defiance, or affect dysregulation, in interactions with their caregivers. Children’s inability to control negative affect in early interactions with their caregivers may forecast continuing difficulties with affective regulation across multiple contexts. Emotions and emotion regulation are increasingly conceptualized as functional in all of the child’s transactions with the world ~Saarni, Mumme, & Campos, 1998!, including the physical world, and development of emotion regulation entails coordinating and regulating affective responding across multiple domains ~Dodge, 1989!. Hence, problems in early affect regulation may disrupt optimal cognitive development. Furthermore, as children enter school, both their adjustment and their achievement depend in part on their ability to manage the emotional and regulatory demands of the classroom. This may include coping with boredom, frustration, or anxiety, as well as dealing with the interpersonal requirements of the classroom setting, including impulse control and regulation of interpersonal affect. Because the roots of emotional and interpersonal regulation lie in the caregiver–child dyad ~Carlson & Sroufe, 1995; Kopp, 1989; Sroufe, 1996!, affect dysregulation in this context may be a precursor of both socioemotional and cognitive outcomes. In the current report we use measures collected as part of the NICHD Study of Early Child Care to identify aspects of affect dysregulation that take into account the child’s behavior in relation to mother across different interactive contexts. These contexts include minimally stressful semistructured play sessions and a potentially more stressful but shorter clean up after play episodes. We define children’s affect dysregulation in the context of the mother–child relationship at 24 and 36 months with these measures and then examine the developmental antecedents and sequelae of varying degrees of affect dysregulation. More specifically, we examine how well earlier aspects of the mother–child relationship and family context predict affect dys-

|

|

DPP16~1! 440

3026

02013004

11:29 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers regulation in the second and third years, and how well affect dysregulation in the mother– child relationship predicts later child functioning in the social and cognitive domains. Most existing work addresses either antecedents or sequelae of affect regulation, but seldom both. And few studies have operationalized affect regulation dyadically despite the strong theoretical grounds for doing so. We adopted a relationship-based approach in defining affect dysregulation because it appears to capture a potentially important and developmentally salient aspect of the child’s emerging self-control. Children’s negative affect when it occurs in the context of the caregiver–child relationship may be an especially important marker of potential problems with affect regulation. This is because the caregiver is usually expected to be a source of support in situations in which children have difficulty managing their own negative emotions. Strong or persistent negative affect in the context of the caregiver–child relationship may reflect a lack of mutual cooperation, harmony, or support as toddlers and preschoolers grapple with stage-salient issues such as establishing autonomy and learning to cooperate with others. Affect dysregulation in the caregiver– child relationship may partly reflect the quality of maternal scaffolding and limit setting during a developmental period when young children still depend on adults to help them control arousal and negative emotion, despite their desire to become independent. Thus, caregivers of young children need to strike a balance in responding to their toddlers’ competing demands for autonomy and support. Some caregivers may be better able to help their young children learn to use and manage potent affects, whereas others may be less skilled. The children of less skillful caregivers may be more angry and frustrated and use affect less effectively within the relationship. From this perspective, then, young children’s ability to regulate negative emotion may be seen as emerging from the early caregiver–child relationship, evident in child mood and behavior as well as qualities of the caregiver–child relationship, including attachment security. Interest in young children’s affect regulation derives from several perspectives. One is

45

a temperament perspective and the search for individual difference parameters that appear to predict later socioemotional functioning and to differentiate between well-adjusted and poorly adjusted children ~Bates, Maslin, & Frankel, 1985; Kochanska, 1995; Rothbart & Bates, 1998!. In general, studies of stable, biologically based individual differences in infants and toddlers have emphasized reactivity and self-regulation that reflect variations in irritability, impulsivity, resistance to control, fearfulness, and attentional processes ~Rothbart & Bates, 1998!. A growing body of evidence indicates that high levels of negative affect, impulsivity, and irritability in infancy and toddlerhood predict later behavior problems ~Campbell et al., 2000; Keenan, Shaw, Delliquadri, Giovannelli, & Walsh, 1998; Sanson, Oberklaid, Pedlow, & Prior, 1991!. However, this relationship is not always clear or direct ~e.g., Bates, Pettit, Dodge, & Ridge, 1998; Belsky, Hsieh & Crnic, 1998!. Not all difficult infants evidence behavior problems in the preschool years, and there are often complex interactions between child characteristics and aspects of parental behavior that together predict which difficult children will and will not show problem behavior at later ages. Prediction from early temperamental characteristics to later functioning requires complex transactional models that take into account the child’s reactivity and regulation as well as parenting behavior. Several recent studies provide support for this transactional perspective by demonstrating that temperamental characteristics of the child and features of maternal parenting style interact to predict outcomes. Belsky, Woodworth, and Crnic ~1996! characterized toddler boys and their families as having difficulties with the “terrible twos” on the basis of both children’s defiance in response to parental control attempts and parents’ use of negative control. In roughly 20% of the sample, parent– child dyads engaged in escalating negative control attempts across the second year. Mothers, fathers, and caregivers rated boys in this group higher on externalizing symptoms at age 3. Follow-up of this sample revealed an interaction between child characteristics and maternal behavior in predicting later behavior

|

DPP16~1! 440

4026

02013004

11:29 am

REVISED PROOF

46 problems ~Belsky et al., 1998!. Maternal negative affect and intrusiveness in the second and third years of life were more strongly predictive of later behavior problems among irritable than nonirritable toddlers. Toddlers who were more positive in mood were less susceptible to the influence of negative maternal behavior. Shaw, Keenan, and Vondra ~1994! also showed that negative and irritable children with less flexible or sensitive mothers were more likely to have difficulties controlling their behavior and affect than irritable toddlers with more responsive and supportive mothers. Conversely, other research suggests that positive parenting interacts with children’s irritability or fearfulness to predict positive outcomes ~Pettit & Bates, 1989; Kochanska, 1995!. A growing body of work is thus emerging to support the idea that child characteristics interact with aspects of parenting to predict early signs of behavior problems, especially externalizing problems reflecting poor emotional control, or to protect irritable children from potentially negative outcomes ~e.g., Bates et al., 1998; Kochanska, 1995!. Attachment theorists also consider the quality of the attachment relationship as having implications for how well children learn to manage negative affect ~Carlson & Sroufe, 1995!. The experience of a satisfying, reciprocal relationship that meets the child’s needs for comfort and security leads the child to value the relationship and become more amenable to socialization efforts. For example, Kochanska ~1995, 1997! suggested that caregiver–child mutuality and cooperation, partly reflected in secure attachment, may render some children more susceptible to the internalization of parental values and expectations, as well as more willing to comply with maternal requests. Thus, children should be better able to manage arousal and negative affect in the context of a secure attachment relationship. In contrast, insecurely attached children, who presumably have less harmonious relationships with their mothers, may be less easily socialized and more likely to engage in negative behavior with their mothers even in relatively undemanding situations that require cooperation and shared goals. Studies of individual differences in children’s affect regulation skills, especially com-

NICHD Early Child Care Research Network

pliance and defiance characteristic of the terrible twos, have tended to emphasize child characteristics ~Braungart–Rieker & Stifter, 1996; Calkins, 1994! that are then examined as a function of different parenting strategies. Three recent studies, however, have specifically examined dyadic regulation of affect as a predictor of early behavior problems ~Denham, Workman, Cole, Weissbrod, Kendziora, & Zahn–Waxler, 2000; Leadbeater, Bishop, & Raver, 1996; Martin, Clements, & Crnic, 2001!. In all three studies, when parent–child dyads were characterized by negativity, anger, and conflict when children were 2 years old, the children were more likely to exhibit disruptive, externalizing behavior problems later as preschoolers. In summary, affect regulation is a construct that generally encompasses behaviors and strategies that children use to control and communicate affect and arousal, especially negative affect ~Calkins, 1994; Kopp, 1989; Thompson, 1994, 1998!. Affect regulation is increasingly conceptualized as an important early emerging skill with implications for both social and cognitive development ~Calkins, 1994!. This is because both constructive social interactions and the ability to learn from the environment are likely to be facilitated when young children can modulate attention, arousal, and negative affect to engage in rewarding and enriching social and cognitive activities. Conversely, children who are highly aroused and unable to modulate negative emotions such as anger and frustration are likely to engage in fewer positive social interactions with either adults or peers. They also may have a more difficult time focusing attention in ways that enhance the acquisition of cognitive and language skills ~Rothbart & Bates, 1998; Rothbart, Posner, & Hershey, 1995!. Based on this review of the importance of the caregiver–child relationship context for children’s emerging emotion regulation, we defined children’s affect dysregulation on the basis of child behavior observed in two ageappropriate, interactional contexts with mother at 24 and 36 months. These ages were chosen because they represent a period when children’s relationship with mother has stabilized into a predictable dyadic partnership and when

|

|

DPP16~1! 440

5026

02013004

11:29 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers the challenge of balancing needs for autonomy and dependence is at its peak. We selected situations that differed in the degree of challenge to the dyad. However, they both rely on dyadic cooperation, mutuality, and a sense of shared goals. The situations included an undemanding, semistructured play interaction and a potentially more emotionally charged compliance task in which mothers requested that children clean up a set of appealing toys with which they had been playing. At each age, we defined affect dysregulation in the mother– child relationship on the basis of the child’s observed negative affect during play with mother, or child negative affect and defiance during cleanup. Because the mother–child play situation was undemanding and positive for most dyads, child negative affect in this setting was considered an important indicator of poor control of emotion directed at mother, and possibly an indicator of a more generally negative mother–child relationship. During toy cleanup, although some degree of resistance may be normative ~Crockenberg & Litman, 1990; NICHD Early Child Care Research Network, 1998!, high levels of defiance and negative affect in this context also signal a poorly regulated mother–child relationship in which conflict over disparate goals overshadows mutual cooperation and harmony. We expected that most children would show no signs of affect dysregulation in the mother–child relationship at either age. When affect dysregulation was observed, we expected that it would be associated with earlier and concurrent aspects of child behavior and family context. In particular, we expected that children who evidenced no affect dysregulation would have families that could more easily support the developmental needs of young children and that this would be reflected in greater demographic and psychosocial resources ~maternal education, financial resources, maternal mood, and maternal language skills! and more sensitive and stimulating maternal caregiving. These children would be more likely to have been characterized as less difficult in infancy, more advanced cognitively, and securely attached at 15 months. We also examined children’s cognitive and social outcomes at 54 months, kindergarten,

47

and first grade as a function of earlier affect dysregulation. We expected that children who showed no affect dysregulation would function most competently on outcome measures. Although our primary focus on the antecedents of affect dysregulation in the mother– child relationship was on child and family characteristics, we also examined whether the quantity or quality of child care experienced by children accounted for any associations between affect dysregulation and other child0 family variables, including children’s outcomes at 54 months and school age. Method Participants Participants in the NICHD Study of Early Child Care were recruited from hospitals near the following locations throughout 1991: Little Rock, Arkansas; Irvine, California; Lawrence, Kansas; Boston, Massachusetts; Philadelphia, Pennsylvania; Pittsburgh, Pennsylvania; Charlottesville, Virginia; Morganton, North Carolina; Seattle, Washington; and Madison, Wisconsin. Potential participants were selected from among 8,986 mothers giving birth during selected 24-hr sampling periods to determine their eligibility for the study. Subjects were excluded from the sample if ~a! the mother was under 18, ~b! the mother was not conversant in English, ~c! the family planned to move, ~d! the child was hospitalized for more than 7 days after birth or had obvious disabilities, ~e! the mother had a known or an acknowledged substance-abuse problem, or ~f ! the family lived at considerable distance from the site or in a location that posed a danger to home visitors. Of those contacted in the initial sampling periods, 5,265 met the eligibility criteria for the study and agreed to be contacted after their return home from the hospital. A subset of this group was selected in accordance with a conditionally random sampling plan. When the infants were 1 month old, a total of 1,364 families with healthy newborns were enrolled in the study. The recruited families came from a wide range of socioeconomic and sociocultural backgrounds, and included 24%

|

DPP16~1! 440

6026

02013004

11:29 am

REVISED PROOF

48 ethnic-minority children, 11% low-education mothers ~less than high school education!, and 14% single-parent mothers. ~Note that these percentages are not mutually exclusive.! The recruited families did not differ significantly from other families eligible to participate on major demographic variables, except that the mothers in the study were slightly ~4%! more likely to report that they planned to be employed compared with the nonparticipating mothers. Of the mothers recruited, 53% planned to work full time ~.30 hr0week!, 23% part time ~10–30 hr0week!, and 24% did not intend to be employed during the child’s first year. The sample was not designed to be nationally representative. However, we compared the demographic characteristics of the sample with those of people living in the same census tracts and nationally. The sample was similar to families in the census tract records and the nation as a whole on key demographic variables ~household income and ethnicity!. Overview of data collection Data for this report were collected from the time the child was 1 month of age through first grade. When the child was 1 month of age, basic demographic information on the child and family was gathered and mothers completed questionnaires during a home visit. Data on child-care usage were collected through phone calls at 3-month intervals, up to age 3 years and at 4-month intervals thereafter, and in face to face contacts with mothers at 1, 6, 15, 24, 36, and 54 months and Grade 1. Data were gathered in the laboratory and in the home on children’s behavioral functioning, mother’s psychological functioning, mother–child interaction, and the quality of the home environment when children were 6, 15, 24, 36, and 54 months of age, and in the spring of Grade 1. Teachers also completed questionnaires about the child in kindergarten and Grade 1. Overview of measures This presentation of measures is organized to reflect the ways in which the variables func-

NICHD Early Child Care Research Network

tioned in the analyses. We first present measures used to create affect-dysregulation variables at 24 and 36 months, and then describe how different affect-dysregulation groups were constructed. Next we discuss the child and family variables used to describe the characteristics of the affect-dysregulation groups at 24 and 36 months, including variables that will be used as covariates in the analyses of subsequent child social and cognitive outcomes at 54 months, kindergarten, and Grade 1. Finally, we describe measures used to assess social and cognitive outcomes at 54 months, kindergarten, and Grade 1. Affect dysregulation Affect dysregulation was assessed during two laboratory playroom observations of mother– child interaction, meant to index aspects of their relationship at 24 and 36 months: a toy cleanup activity and a semistructured mother– child play interaction. Lab cleanup. At the end of a 15-min toy play period which opened the lab sessions at 24 and 36 months, the visit coordinator handed the mother containers for the toys and instructed her to have the child participate in picking up the toys; no other directions were given to mother ~see NICHD Early Child Care Research Network, 1998, for details!. Clean up lasted no more than 5 min. Child behaviors coded on a 5-point scale from 1 5 not at all characteristic to 5 5 very characteristic were: compliance, assertive noncompliance, defiant noncompliance, passive noncompliance, positive affect, and negative affect. Only the defiant noncompliance and negative affect scores were used. A rating of $3 on both of these scales at one age point was required for a child to be assigned to the affect-dysregulation group at that age. The intraclass correlation ~Winer, 1971! of intercoder reliabilities of the defiant noncompliant and negative affect scales were .71 and .81 at 24 months, and .83 and .81 at 36 months. Table 1 presents descriptive statistics for the selected lab cleanup ratings. Means, and standard deviations for the defiant noncompliance and negative affect ratings are presented,

|

|

DPP16~1! 440

7026

02013004

11:29 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers

49

Table 1. Measures of dyadic affect dysregulation Child Age 24 Months

36 Months

Lab Clean Up Defiance M ~SD! Low ~ratings of 1–2! n High ~ratings of 3–5! n Negative affect M ~SD! Low ~ratings of 1–2! n High ~ratings of 3–5! n

1.13 ~0.51! 989 ~97%! 29 ~3%! 1.15 ~0.51! 994 ~98%! 24 ~2%!

1.16 ~0.54! 979 ~97%! 31 ~3%! 1.20 ~0.56! 971 ~96%! 39 ~4%!

Mother–Child Interaction 24-month negative affect M ~SD! Low ~ratings of 1–2! n High ~ratings of 3– 4! n 36-month negative affect M ~SD! Low ~ratings of 1–3! n High ~rating of 4–7! n

1.43 ~0.70! 926 ~91%! 95 ~9%!

1.66 ~1.08! 953 ~94%! 62 ~6%!

Note: The criteria for high negative affect in the mother–child interactions differ over age because the ratings of interactions could range from 1 to 4 at 24 months and from 1 to 7 at 36 months.

along with the proportion of children displaying low and high levels of negativity. Relatively few children showed high levels of both defiance and negative affect. As shown in Table 2, only 17 children were high in both defiance and negative affect during toy cleanup at 24 months, whereas 7 children showed high levels of negative affect alone and 12 children showed high levels of defiance alone. Only 21 children showed high levels of both defiance and negative affect during toy cleanup at 36 months. Mother–child interaction. Mother–child interaction was videotaped in semistructured 15min play observations at both ages ~see NICHD Early Child Care Research Network, 1999 for details!. At 24 months child negative mood was rated on a 4-point scale.At 36 months, child negativity, which paralleled the 24-month rating of child negative mood, was rated on a 7-point scale. Winer ~1971! intraclass correlations, assessing interrater reliability for this scale, were .88 and .85 for 24 and 36 months, respectively. Children were assigned to the affectdysregulation group if they displayed exten-

sive negative affect in interaction with their mother. At 24 months, extensive negative affect was indicated by scores of 3– 4, and at 36 months by scores of 4–7. Table 1 shows the mean and standard deviation of negative affect at each of the two ages and the proportion of children classified as displaying extensive negativity during interactions with their mothers. As shown in this table, 95 children met this criterion at 24 months and 62 at 36 months. Children were classified as dysregulated if they showed high levels of negative affect during interaction with the mother or if they showed high levels of both defiance and negative affect during the toy cleanup. The bottom half of Table 2 displays the numbers of children meeting these criteria at both ages. As the table shows, a total of 105 children were classified as dysregulated at 24 months because they displayed high levels of negative affect during both interaction and cleanup ~n 5 6!, in interaction alone ~n 5 89!, or during cleanup alone ~n 5 10!. At 36 months, 77 children were classified as dysregulated.

|

DPP16~1! 440

8026

02013004

11:29 am

REVISED PROOF

50

NICHD Early Child Care Research Network

Table 2. Number of children displaying affect dysregulation During Lab Cleanup 24 Months

36 Months

Negative Affect

Negative Affect

Defiance

No

Yes

No

Yes

No ~n! Yes ~n!

982 12

7 17

961 10

18 21

Mother–Child Interactions 24 Months

36 Months

Negative Affect

Negative Affect

Lab Cleanup

No

Yes

No

Yes

No ~n) Yes ~n)

911 10

89 6

925 18

57 2

Note: The children who were high on both Negative Affect and Defiance were classified as dysregulated during lab cleanup; children considered dysregulated by either criterion were classified as dysregulated.

Family and child characteristics The following variables were used as family and child predictors or correlates of affect dysregulation in univariate analyses ~Table 3!: maternal education, vocabulary, personality, and depressive symptoms; partner in the home; income; caregiving quality; child gender and ethnicity ~Black, Hispanic, other, or White!; 6-month negative mood; 6-month temperament; 15-month cognitive functioning; 15-, 24-, and 36-month attachment; and amount and quality of child care. Variables significantly associated with affect dysregulation at the univariate level were included in a single multivariate analysis predicting to dysregulation status ~Table 4! and were also covariates in the analyses predicting child outcomes ~Table 5!. ~Exceptions were child variables from assessments after age 6 months, to avoid overcontrolling for child outcomes; and child care quality, because it was not available for the full sample. Site was a covariate, as it was complexly associated with both predictor and

outcome variables; child gender was added as a covariate because of its association with the outcome measures.! Maternal education was represented by five levels as determined at 1 month: 1 5 less than high school, 2 5 high school graduate, 3 5 some college, 4 5 college graduate, and 5 5 graduate education. Maternal vocabulary was assessed with the Peabody Picture Vocabulary Test— Revised ~PPVT-R; Dunn & Dunn, 1981!. The PPVT assesses verbal intelligence operationalized as receptive knowledge of vocabulary. Scores are standardized with a mean of 100 and standard deviation of 15. Scores are moderately correlated with standard measures of verbal intelligence such as the WAIS Vocabulary subscale ~Weschler, 1981!. Maternal personality was assessed using three scales of the NEO Personality Inventory ~NEO PI; Costa & McCrae, 1985! obtained at 1 month. The three NEO scales were neuroticism, the extent to which the mother indicated she is anxious, hostile, and depressed ~reversed!; extraversion, the extent to which she is sociable, fun-loving, and optimistic; and agreeableness, the extent to which she is trusting, helpful, and forgiving. Thus, higher scores indicated better adjustment. Maternal depressive symptoms were measured using the Center for Epidemiological Studies Depression Scale ~CES-D; Radloff, 1977! administered at each home visit ~1, 6, 15, 24, 36, and 54 months, and Grade 1!. Cronbach’s alpha for the CES-D across the data collection points ranged from .88 to .91. The depression scores had moderately strong correlations over time ~.46 to .58!. Partner in the home was reported by mothers at each interview, and the proportion of time a partner was in residence was computed for each family from 6 to 54 months. Income was the income0needs ratio computed from maternal interview items collected at each home visit. Family income, excluding AFDC, was divided by the appropriate poverty threshold, determined by the year in which the income was earned and by household size. This variable was averaged across 0 –54 months to create an average income0needs ratio.

|

|

DPP16~1! 440

9026

02013004

11:29 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers

51

Table 3. Comparisons on selected family and child variables for affect-dysregulated groups Dysregulated 24 Months

Dysregulated 36 Months

No N 5 917

No N 5 942

Yes N 5 106

Yes N 5 81

Maternal Characteristics Maternal education a Maternal receptive vocabulary b,c Maternal agreeableness at 6 months b Maternal extraversion at 6 months Maternal neuroticism at 6 months d Maternal depression ~CESD score! c,d 6 months 15 months 24 months 36 months

14.6 100.3 46.7 42.6 29.5 8.6 8.8 9.1 8.9

~2.4! ~18.3! ~5.2! ~5.8! ~7.0!

13.7 93.7 44.8 41.3 31.3

~2.4! ~18.5! ~4.7! ~5.9! ~7.1!

14.5 100.2 46.6 42.5 29.6

~7.9! ~8.0! ~8.3! ~8.1!

11.3 10.0 12.0 12.0

~9.5! ~8.7! ~10.3! ~10.7!

8.6 8.8 9.1 8.9

~2.4! ~18.2! ~5.1! ~5.8! ~7.0!

13.9 92.2 45.4 41.8 30.6

~2.6! ~19.6! ~5.9! ~5.7! ~7.4!

~7.8! ~8.0! ~8.2! ~8.1!

11.5 10.3 11.9 12.6

~10.5! ~9.2! ~11.6! ~11.3!

Family Characteristics Partner 6–54 months d,e Income0poverty threshold a 6 months 15 months 24 months 36 months Maternal caregiving a, f,g 6 months 15 months 24 months 36 months

0.87 ~0.28! 3.9 3.9 3.9 3.8

~3.2! ~3.3! ~3.0! ~3.1!

0.09 ~0.80! 0.10 ~0.78! 0.14 ~0.74! 0.10 ~0.77!

0.75 ~0.38! 3.1 3.0 2.9 2.8

~2.4! ~2.4! ~2.4! ~2.6!

20.31 ~0.90! 20.28 ~0.86! 20.68 ~0.99! 20.48 ~1.07!

0.87 ~0.28! 3.9 3.8 3.8 3.7

~3.1! ~3.2! ~2.9! ~3.0!

0.07 ~0.79! 0.08 ~0.78! 0.09 ~0.79! 0.11 ~0.77!

0.76 ~0.39! 3.2 3.4 3.7 3.3

~3.3! ~3.1! ~3.8! ~3.8!

20.27 ~1.01! 20.20 ~0.96! 20.37 ~0.95! 20.76 ~1.06!

Child Characteristics Male gender ~%! Child ethnicity ~%! Black d,e Hispanic Other White d,e Negative mood SI 6 months b,d Carey temperament 6 months Bayley MDI 15 months a,e Attachment SS 15 months ~%! Aa Ba C D0U Q-security 24 months b,c SS 36 months ~%! A B C b,c D0U b Hours of care0week 6 months 15 months 24 months 36 months 54 months

50

57

51

47

10 6 4 80 1.41 ~0.68! 3.16 ~0.40! 110.2 ~13.6!

20 7 6 68 1.64 ~0.87! 3.21 ~0.44! 100.5 ~13.9!

11 6 4 80 1.41 ~0.69! 3.16 ~0.39! 109.6 ~13.9!

21 6 3 70 1.63 ~0.82! 3.19 ~0.50! 104.7 ~14.1!

12 64 9 15 0.31 ~0.20!

31 45 9 14 0.21 ~0.20!

13 63 9 15 0.30 ~0.20!

19 53 8 19 0.21 ~0.20!

5 64 16 15

4 48 25 23

5 64 16 16

4 46 28 23

23.5 25.1 25.4 27.0 30.6

~21.4! ~21.7! ~21.0! ~21.3! ~18.8!

23.4 25.4 28.2 27.5 33.7

~21.1! ~18.6! ~21.0! ~21.3! ~20.4!

23.8 25.4 25.7 27.2 30.9

~21.6! ~21.4! ~21.1! ~21.2! ~18.8!

19.8 22.7 25.5 24.5 31.3

~17.9! ~20.6! ~20.1! ~20.2! ~21.8!

|

DPP16~1! 440

10026

02013004

11:30 am

|

REVISED PROOF

52

NICHD Early Child Care Research Network

Table 3 ~cont.! Dysregulated 24 Months

Dysregulated 36 Months

No N 5 917

No N 5 942

Yes N 5 81

3.0 ~0.6! 3.0 ~0.6! 2.8 ~0.6! 2.8 ~0.5! 3.0 ~0.6!

3.0 ~0.5! 2.8 ~0.5! 2.6 ~0.6! 2.6 ~0.5! 2.9 ~0.6!

Yes N 5 106

Child Characteristics ~cont.! Quality ~ORCE composite rating! c,d 6 months 15 months 24 months 36 months 54 months

3.0 ~0.6! 3.0 ~0.6! 2.8 ~0.6! 2.8 ~0.5! 3.0 ~0.6!

3.0 ~0.5! 2.9 ~0.6! 2.7 ~0.6! 2.7 ~0.6! 2.9 ~0.6!

Note: N, maximum number listed; sample sizes vary depending on the outcome examined. The site is included as a covariate in all analyses. All values except percentages are means ~standard deviations!. a 24-month dysregulated groups different at p , .001. b 24-month dysregulated groups different at p , .01. c 36-month dysregulated groups different at p , .01. d 24-month dysregulated groups different at p , .05. e 36-month dysregulated groups different at p , .05. f 36-month dysregulated groups different at p , .001. g Mean levels of maternal caregiving decrease over time for both 24- and 36-month dysregulated groups ~ p , .001!.

Mother–child interaction during play at home at 6 months, and in a laboratory playroom at 15, and 24 months, was rated on 4-point global rating scales developed for this study; 7-point rating scales were used for ratings of 36 and 54 month–month laboratory interactions. A maternal sensitivity composite composed of the sum of sensitivity to nondistress, instrusiveness ~reverse scored!, and positive regard was formed at 6, 15, 24, 36, and 54 months ~see NICHD Early Child Care Research Network, 1999, for a discussion of the adjustment to convert all ratings to the same scale!. Intercoder reliability was calculated as the intraclass correlation ~Winer, 1971!. For the maternal sensitivity composites the intraclass correlations were .87, .83, .84, .84, and .84 for 6, 15, 24, 36, and 54 months, respectively. Cronbach’s alpha for the composites ranged from .70 to .84. The Home Observation for Measurement of the Environment ~HOME; Caldwell & Bradley, 1984! assesses overall quality of the physical and social resources available to a child within the home. The HOME consists of both direct observation and a semistructured interview with the mother. The focus is on the child as a recipient of stimulation from objects, events, and interactions occurring in the family surroundings. The total score is computed

as the sum of 45 items at 6 and 15 months, and as the sum of 55 items at 36 and 54 months, with higher values denoting higher levels of maternal responsiveness, child stimulation, and support to the child. All observers attended centralized training sessions before collecting the data and were required to maintain reliability by matching a master coder on 90% of items. The Infant0Toddler version was administered at 6 and 15 months, and the Early Childhood version was administered at 36 and 54 months. Cronbach’s alpha for the total score was .76 at 6 months, .80 at 15 months, .87 at 36 months, and .82 at 54 months. Correlations across time range from .48 to .70 ~ p , .0001!. A composite measure of quality of maternal caregiving was created at 6, 15, 24, 36, and 54 months from the mean of the standardized maternal sensitivity composite plus the standardized HOME total score. Infant negative mood was the 6-month 4-point child negative mood rating from videotaped mother–child interaction in the home ~see NICHD Early Child Care Research Network, 1999, for details!. The intraclass correlation for this scale was .69, indicating adequate interrater reliability. Infant temperament was based on fifty-five 6-point items from the Infant Temperament

|

DPP16~1! 440

11026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers

53

Table 4. Logistic regressions predicting affect-dysregulation group from child and family characteristics

Maternal Education Vocabulary Agreeableness Neuroticism Depression 6–24036 months Partner in home 6–24036 months Income0poverty 6–24036 months Maternal caregiving 6–24036 months Child ethnicity Black–White Hispanic–White Other–White Negative mood 6 months Bayley MDI 15 months Attachment 15 months A–B C–B D–B Attachment security 24 months Attachment 36 months A–B C–B D–B Site

24-Month Dys. ~n 5 919!

36-Month Dys. ~n 5 904!

0.05 ~0.07! 20.00 ~0.01! 20.02 ~0.03! 0.00 ~0.02! 20.01 ~0.02! 20.07 ~0.33! 20.00 ~0.06! 20.89 ~0.25!*** ns 20.54 ~0.34! 0.27 ~0.40! 0.03 ~0.50! 0.24 ~0.15! 20.04 ~0.01!*** ** 0.76 ~0.23!** 20.20 ~0.32! 20.24 ~0.26! 21.10 ~0.62!

0.15 ~0.08! 20.01 ~0.01! 0.01 ~0.03! 20.02 ~0.03! 0.03 ~0.03! 0.22 ~0.37! 0.09 ~0.06! 20.94 ~0.29!** ns 0.18 ~0.43! 0.67 ~0.48! 21.05 ~0.81! 0.16 ~0.17! 20.01 ~0.01! ns 0.08 ~0.20! 20.15 ~0.38! 0.20 ~0.28! 20.77 ~0.72! ns 20.04 ~0.50! 0.42 ~0.28! 20.22 ~0.31! ns

*

Note: Dys., dysregulated. The coefficients for the site are not listed. All values are B and ~standard errors!. **p , .01. ***p , .001.

Questionnaire completed by mothers at 6 months ~Medoff–Cooper, Carey, & McDevitt, 1993!. The composite measure reflected difficult temperament and was created by calculating the mean of the nonmissing items with appropriate reversal of scales, so that numerically large scores consistently reflected a more “difficult” temperament. Cronbach’s alpha was .81. Higher scores reflect a more negative disposition: that is, a child seen as more intense, less positive in mood, and less adaptable to daily routines. Children’s cognitive functioning was assessed using the original Bayley Scales of Infant Development ~Bayley, 1969! at 15 months during lab visits. The Mental Development Index ~MDI! represents overall cognitive level. Scores are standardized so that the mean is 100 with a standard deviation of 15.

Infant–mother attachment security was assessed at 15, 24, and 36 months. At 15 months the Ainsworth et al. ~1978! Strange Situation procedure was conducted. Videotapes of all Strange Situations were coded in a central location and attachment security was evaluated by a team of three coders blind to child-care status and other information about the child and family. These coders double-coded 1,201 Strange Situation assessments using the standard attachment classifications of secure ~B!, insecure-avoidant ~A!, insecure-resistant ~C!, disorganized ~D!, and unclassifiable ~U!. The group viewed disagreements and a code was assigned by consensus. Across all coder pairs, before conferencing, agreement with the 5-category classification system was 83% ~k 5 .69!. In this report the D and U classifications are combined.

| DPP16~1! 440

Table 5. Affect dysregulation and child outcomes from 54 months to first grade 24-Month Dysregulation

d

F

54

875 99

G1

F

d

F

848 97

54

G0

G1

427.4 421.5

472.5 467.5

373.6 367.6

454.7 451.8

464.4 460.4

486.8 482.7

Math ~WJ Applied Problems! a 6.48*

0.25*

3.03

Not dys. Dys.

11.48*** 427.2 424.9

0.37***

0.22

472.6 467.4

54

Reading ~WJ Letter–Word Identification! a 3.31

0.17

2.21

Not dys. Dys.

4.06* 373.3 371.4

0.21*

1.18

455.0 449.3

Phonological Knowledge ~WJ Incomplete Words! a 4.76* 0.19* 2.32 Not dys. Dys. Impulsivity ~CPT errors of commission log transformed! 7.62* 20.25* 0.12 Not dys. Dys. Sustained attention ~CPT errors of omission log transformed! 0.61 20.07 0.02 Not dys. Dys. Mother rating of social skills Sample size Not dys. Dys.

10.27** 464.4 460.7

4.32* 1.91 2.19

1.29 1.52

1.94 2.01

0.77 0.82

2.78

881 101

.32**

.00

486.6 485.6

857 100

847 97

20.19

20.17

REVISED PROOF

872 73

11:30 am

897 77

02013004

Cognitive0academic outcomes Sample size Not dys. Dys.

G0

Time Interaction Adj. M

Main Effect

12026

Time Interaction Adj. M

Main Effect F

36-Month Dysregulation

0.56 1.92 2.18

1.31 1.45

1.94 2.11

0.77 0.85

0.64

895 78

879 78

871 72

|

| DPP16~1! 440

CPRS, closeness 3.95*

0.18*

0.82

2.17 4.61 4.54

4.52*

0.35

Not dys. Dys. CBCL, internalizing 1.89

0.12

0.30**

48.5 50.6

46.8 48.0

47.4 48.5

48.3 49.3

99.1 96.7

101.7 97.9

105.7 99.8

822 96

820 97

0.00

5.55*

1.20

588 66 0.04

0.04

20.20***

1.12

4.22 4.24

0

2.12

49.8 52.5

50.0 53.4

49.2 48.9

48.8 49.3

48.2 49.8

1.35

48.7 48.6

46.9 47.5

47.5 47.6

48.5 47.8

0.11

99.0 96.6

101.5 99.8

105.2 104.7

843 75

843 73

1.29

0.03

0.00 4.19 4.20

0.86 49.6 51.9

50.3 51.1

1.38

4.28 4.30

.70

51.3 52.8

602 52 0.04

4.18 4.21

20.10

0.12

4.76 4.71

20.09

0.00

4.22 4.23

4.28 4.29

0.63 49.8 51.1

50.0 50.8

50.4 50.5

49.2 49.4

48.8 49.1

48.3 48.6

103.5 102.0

103.8 100.9

104.2 99.3

REVISED PROOF

50.2 52.0

4.66 4.62

11:30 am

55

10.82** Not Dys. Dys. Teacher rating of social skills Sample size Not dys. Dys. TCRS, closeness 0.16 Not dys. Dys. TRF, externalizing 14.75*** Not dys. Dys. TRF–Internalizing 0.71 Not dys. Dys.

0.11 4.60 4.57

0.60 51.3 52.8

0.02

Not dys. Dys. SSRS, social skills

0.13

4.76 4.73

02013004

20.19*

4.67 4.61

13026

Not dys. Dys. CBCL, externalizing

0.00

SSRS–Social Skills 5.87* Not dys. Dys.

0.22*

0.04

5.49 103.6 101.0

103.9 101.0

104.0 101.0

.24*

1.03

Note: 54, 54 months; G0, kindergarten; G1, first grade; dys., dysregulated; WJ, Woodcock–Johnson. Covariates: site, mother’s education, PPVT vocabulary, caregiving ~maternal sensitivity 1 HomeTotal!, depressive symptoms, income, partner in the home; child gender, ethnicity, negative mood at 6 months, age at assessment, and all interactions with age at assessment. a Raush scores were used in the longitudinal analyses of Woodcock–Johnson subscales. *p , .05. **p , .01. ***p , .001.

|

|

DPP16~1! 440

14026

02013004

11:30 am

REVISED PROOF

56 Infant–mother attachment security was assessed at 24 months using the Attachment Q-set. Two-hour home visits were conducted during which trained observers made notes of child behavior that was naturally occurring or in response to semistructured situations introduced by the observer. After the visit the observers used the Attachment Q-Sort ~AQS! to describe security with mother ~Vaughn & Waters, 1990; Waters & Deane, 1985!. The resulting profile is correlated with the profile of a prototypically secure child, as determined by experts in the field, and the correlation is the child’s Q-security score. Research assistants were tested on five videotaped observations scored by experts Drs. Posada and Gao. The Posada–Gao consensus score was the “gold standard” for this test. Their reliability agreement across the five test tapes, before conferencing to consensus, was .77, with a security score difference of .04. Research assistants were certified to collect data if the median correlation of their Q-security scores with the gold standard Q-security scores was .55, and no correlation was below .40. Across the data collection period within-site reliability was assessed. Across all research assistants at all sites, this correlation was .73, with a security score difference of .06. A modified Strange Situation procedure based on recommendations by Cassidy and Marvin and the MacArthur Working Group on Attachment ~1992! was used to assess attachment classifications at 36 months see ~NICHD Early Child Care Research Network, 2001, for details!. The MacArthur coding system classifies preschoolers as secure ~B! or insecure-avoidant ~A!, insecure-resistant ~C!, and insecure-controlling0insecure other ~D!. A team of three coders conducted the coding. Reliability was calculated based on 867 randomly paired cases. Disagreements were discussed and a consensus code was assigned. Intercoder agreement ~before conferencing! on the 4-category A B C D classifications was 75.7% ~k 5 .58!. Child care variables Amount of child care was the sum of all hours reported by mothers in all nonmaternal care

NICHD Early Child Care Research Network

arrangements, scored as mean hours per week during the periods 1– 6 months, 7–15 months, 16–24 months, 25–36 months, and 37–54 months. Children who experienced no routine nonmaternal care received a score of zero. Quality of child care was assessed during two half-day child care visits at 6, 15, 24, and 36 months, and one half-day visit at 54 months, using observational methods developed for this study ~Observational Record of the Caregiving Environment or ORCE; see NICHD Early Child Care Network, 1996; in press!. Positive caregiving was generated at 6, 15, and 24 months by summing five qualitative ratings of the care received by the focus child, made at the end of four observation cycles: sensitivity0 responsiveness to child’s nondistress expressions, positive regard, stimulation of cognitive development, detachment ~reversed!, and flat affect ~reversed!. At 36 and 54 months, two additional ratings, fostering exploration and intrusiveness ~reversed!, were added to the composite. The Cronbach’s alphas were .89, .88, .86, .82, and .79 at 6, 15, 24, 36, and 54 months, respectively. Interobserver reliability estimates ~Pearson correlations! for positive caregiving ratings were .94, .86, .81, .80, and .94 ~videotapes!, and .90, .89, .89, .90, and .98 ~live! at each age. Child outcome measures The set of child outcome variables included in the analyses was derived from measures obtained during laboratory, home, and child care or school visits. The criteria used to determine the final set of outcomes included coverage of major developmental domains ~cognitive0 academic and social!, good psychometric properties, and representation of important constructs. We also sought to limit the set to a small number to permit us to construct a relatively coherent picture of the antecedents and sequelae of early affect dysregulation. The child outcomes were measured over time: some at 54 months and first grade and others at 54 months, kindergarten, and first grade. Cognitive0academic outcomes. Three subscales of the Woodcock Johnson Achievement Battery, Letter–Word Identification, Incom-

|

|

DPP16~1! 440

15026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers plete Words, and Applied Problems, to assess reading, phonological knowledge, and math, respectively, were administered during lab visits at 54 months and Grade 1. All subscales have a mean of 100 and standard deviation of 15. In addition, Raush scores are available for longitudinal analyses. Raush scores are not standardized by age, and so true developmental change can be charted by increasing scores. Scores on the three subtests were correlated in our sample at 54 months ~r 5 .38– .59, p , .001! and first grade ~r 5 .34–.56, p , .001!. Sustained attention and impulsivity were measured using the Continuous Performance Task ~CPT; Rosvold, Mirsky, Sarason, Bransome, & Beck, 1956!. This task was administered to each child individually toward the end of a 2.5-hr laboratory visit at the 54 months and first-grade assessments. At 54 months, with an experimenter in the room, the child was seated at a table in front of a 2-in. square screen and a red button. Dotmatrix pictures of 10 familiar objects ~e.g., butterfly, fish, flower! were generated by a computer and presented on the screen. The child was asked to press the button “as fast as you can” each time a target stimulus ~a chair! appeared on the screen. A total of 220 stimuli were presented in 22 blocks. The target stimulus was randomly presented within each block and appeared twice within each block for a total of 44 presentations, leaving 176 presentations of nontarget stimuli. Each stimulus appeared on the screen for 500 ms at 1500-ms interstimulus intervals. The child’s task was twofold: ~a! to press the red button as soon as he or she saw the image of the chair on the screen, and ~b! to refrain from pressing the button at the appearance of other nontarget stimuli. The task took 7 min and 20 s. A more challenging version of this task was used at first grade. Children were seated in front of a computer monitor and letter stimuli were randomly presented in 30 blocks, 10 stimuli per block. The target stimulus, the letter X, appeared twice in each block. The stimulus duration was 200 ms and the interstimulus interval was 1500 ms. The child was instructed to press the red button whenever an

57

X appeared on the screen and to refrain from pressing when any other letter appeared. Before administration, children’s letter knowledge was assessed. The task lasted 8 min and 30 s. Sustained attention was assessed by the number of errors of omission. Errors of omission occurred when children failed to press the button in response to the appearance of the target stimulus. Therefore, children with greater ability to sustain attention had fewer lapses in attention and made fewer errors of omission. Impulsivity was reflected in the number of errors of commission made by the child—that is, incorrect button-press responses to nontarget stimuli. Children with more incorrect responses to nontargets were considered more impulsive than were children with fewer incorrect responses. Measures of sustained attention and impulsivity derived from the CPT have adequate test–retest reliability ~r 5 .65–.74!, based on data from a sample of 138 nonreferred boys ~Halperin, Sharma, Greenblatt, & Schwartz, 1991!. Children’s performance on the CPT has high construct validity as a measure of attention ~Halperin et al., 1991!, and adequate predictive validity ~e.g., Barkley, 1994; Barkley, Brodzinsky, & DuPaul, 1992; Campbell, Pierce, March, Ewing, & Szumowski, 1994!. Performance on these two measures of attention is correlated at r 5 .24 and r 5 .48 ~ p , .0001! at 54 months and first grade in the current sample. Cross age correlations are r 5 .26 and .19 ~ p , .0001! for sustained attention and impulsivity, respectively. Social and behavioral outcomes. Mothers and teachers rated children’s social skills using their respective forms of the Social Skills scale from the Social Skills Rating System ~Gresham & Elliott, 1990!, rating how often a social behavior occurs on a 3-point scale ~0 5 never, 1 5 sometimes, 2 5 very often!. The teacher Social Skills Scale contains three subscales: Cooperation, Assertion, and Self-control. Cooperation includes behaviors such as paying attention to the teacher’s instruction and putting away work materials properly. Assertion includes initiating appropriate behaviors, such as starting conversations with peers, introduc-

|

DPP16~1! 440

16026

02013004

11:30 am

REVISED PROOF

58 ing oneself, and volunteering to help peers with classroom tasks. Self-control includes behaviors that are effective in conflict situations, such as responding to teasing or peer pressure appropriately, receiving criticisms well, and controlling temper. The maternal version includes an additional Responsibility subscale. Responsibility includes behaviors that demonstrate ability to communicate with adults and exhibiting regard for property or work. For the elementary-level forms, Cronbach’s a 5 .87 and .94 for the total mother and teacher Social Skills scales, respectively. The total score was used in these analyses. The validity ~content, criterion, and construct! of the SSRS is documented extensively in Gresham and Elliot ~1990!. The Child Behavior Checklist ~CBCL; Achenbach, 1991a! and Teacher Report Form ~TRF; Achenbach 1991b! are slightly different versions of a widely used measure to assess the social competence and problem behavior of children 4–18 years. This family of measures has been standardized on large samples of children in the United States and abroad. A series of behaviors ~about 100 items per version! are rated on 3-point scales from 0 ~not true of the child ! to 2 ~very true of the child !. Broad band scales of internalizing and externalizing problems are derived from a computerized scoring program. Mothers and teachers completed the CBCL and TRF at 54 months, kindergarten, and Grade 1. The Child–Parent Relationship Scale ~CPRS; Pianta, 1994! and Student–Teacher Relationship Scale ~STRS; Pianta, 1992! are 30-item, 5-point Likert-type scales designed to assess the adult’s perception of the quality of the relationship between the child and the adult. The Closeness scale was used in these analyses. The 54-month version had 11 items and the later versions had 8 items. The total score was the mean across items. Higher scores indicated the respondent reported a warmer, closer relationship with the child. The alpha values ranged from .64 to .69 for the parentcompleted measure and from .83 to .85 for the teacher-completed form. The CPRS and STRS were completed by caregivers or teachers and mothers, respectively, at 54 months, kindergarten, and Grade 1.

NICHD Early Child Care Research Network

Analysis and Results Of the 1,364 families enrolled in the study, 1,023 were included in the analyses in this report; 341 were excluded because of missing data on the dysregulation, covariates, or outcome variables. The included and excluded families do not differ significantly on child’s gender or dysregulation classification at 24 or 36 months. Mothers who were excluded and included in these analyses also did not differ on maternal depression scores at any age, or maternal neurotism or extroversion scores obtained at 1 month. The two groups were also comparable on income at 15, 24, 36, and 54 months and at Grade 1. However, families with missing outcome data were more likely to include a child who was Black, x 2 ~3, N 5 1,364! 5 15.7, p , .001, and a mother rated as less agreeable on the personality inventory, t ~1,270! 5 23.21, p , .001, with less maternal education, t ~1,361! 5 25.87, p , .001, and with lower maternal vocabulary scores on the PPVT, t ~1,165! 5 22.91, p , .01. The results are presented in three sections. First, children in the two affect-dysregulation groups at 24 and 36 months ~0 5 not dysregulated, 1 5 dysregulated ! are compared for concordance over time. Second, they are compared on relevant child and family characteristics, including some that were measured longitudinally from 6 to 36 months. Third, the two affect-dysregulation groups are compared on the child outcomes, collected longitudinally from 54 months to Grade 1, adjusting for a variety of child, family, and child care variables. Concordance in affect-dysregulation groups over time Children classified as dysregulated at 24 months were substantially more likely to be classified as dysregulated at 36 months, x 2 ~1, N 5 1,023! 5 44.75, p , .001, but there was no extensive overlap across age. As shown in Table 6, about 84% of the sample did not show affect dysregulation at either age. Of the 161 children classified as dysregulated at either age, 26 children were classified as dysregulated at both ages.

|

|

DPP16~1! 440

17026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers Table 6. Agreement between affectdyregulation classifications at 24 and 36 months 36-Month Dysregulation

24-Month Dysregulation

No

Yes

No Yes

862 ~84%! 80 ~8%!

55 ~5%! 26 ~3%!

Note: The difference in numbers between Table 2 and this table is because a few children have missing data on the lab cleanup or mother–child interaction and so meet criteria based only on one or the other. The values are the number of subjects. x 2 ~1, n 5 1,023! 5 44.75, p , .001.

Comparing affect-dysregulation groups on child and family characteristics First, we compared the dysregulated groups on selected family and child variables. Analyses of variance compared families on all measures collected at a single age, and repeated measures analyses compared families on all measures collected longitudinally. Table 3 lists the means and standard deviations or proportions for these variables. Both sets of analyses asked whether the two dysregulation groups ~0, 1! at 24 months and the two dysregulation groups at 36 months differed in analyses that adjusted for site. These analyses were conducted separately for each of the selected characteristics. Both effect sizes and test statistics are reported below when significant group differences were detected. Effect sizes were computed as the difference between the adjusted means for the two dysregulation groups divided by the pooled standard deviation, the square root of the residual error variance. The first set of analyses compared the maternal and family characteristics that were collected only once. Many of them differed according to the child’s dysregulation status. Mothers of children dysregulated at 24 months, but not 36 months, reported less education, d 5 .38; F ~1, 1,011! 5 12.8, p , .001. Maternal vocabulary scores on the PPVT were significantly lower among children classified as dysregulated at 24 months, d 5 .33;

59

F ~1, 1,006! 5 9.68, p , .01, and at 36 months, d 5 .35; F ~1, 1,006! 5 8.66, p , .01!. The mothers of children dysregulated at 24 months but not 36 months also reported poorer psychological adjustment, lower agreeableness, d 5 .34; F ~1, 993! 5 10.3, p , .01, and higher neuroticism, d 5 25, F ~1, 993! 5 5 .62, p , .05. The dysregulation groups did not differ on extraversion. In addition, dysregulated children were less likely to have mothers with partners in the household at 24 months, d 5 .39, F ~1, 1,011! 5 13.8, p , .001, and 36 months, d 5 .26, F ~1, 1,011! 5 4.78, p , .05, than those children who were not dysregulated. Significant differences among the dysregulation groups emerged on many of the child characteristics. No gender differences emerged. More African American and fewer White children were in the dysregulated groups at 24 months, x 2 ~3, N 5 1,023! 5 10.2, p 5 .017, and at 36 months, x 2 ~3, N 5 1,023! 5 8.58, p 5 .036!. Children tended to show more negative affect in interactions with their mothers when they were 6 months old if dysregulated at 24, d 5 .28, F ~1, 905! 5 7 .11, p , . 01, or 36 months, d 5 .24, F ~1, 905! 5 4.07, p 5 .04!, although there were no significant differences in mothers’ reports of temperament ratings at 6 months of age. Children also performed more poorly on the Bayley at 15 months than their nondysregulated peers if they were dysregulated at 24 months, d 5 .68, F ~1, 976! 5 39.9, p , .001, or at 36 months, d 5 .25, F ~1, 976! 5 4.07, p , .05. Finally, the dysregulated children also differed on attachment. Compared with nondysregulated children, children who were classified as dysregulated at 24 months were less likely than nondysregulated children to have been securely attached at 15 months and they were more likely to have been classified as insecureavoidant, x 2 ~3, N 5 996! 5 27.1, p , .001, based on analyses of individual cells using EXACON ~Bergman & El-Khouri, 1987, 1998!. However, children in the dysregulated group at 24 months were more likely than other children to be classified as insecure-resistant or insecure-controlling0insecure-other at 36 months, x 2 ~3, N 5 1,003! 5 12.1, p , .01, based on analyses of individual cells. In addition, children who were classified as dysreg-

|

DPP16~1! 440

18026

02013004

11:30 am

REVISED PROOF

60 ulated at 36 months were also more likely to be insecure-resistant at 36 months, x 2 ~3, N 5 1,003! 5 12.7, p , .01. Children classified as dysregulated at 24 months also had lower 24month Q-security scores, F ~1, 1,005! 5 14.83, p , .01, as did children classified as dysregulated at 36 months, F ~1, 1,005! 5 7.71, p , .01. Overall, therefore, children who were classified as dysregulated at either 24 or 36 months also were less likely to have secure attachments to their mothers both earlier and concurrently, although the specific patterns varied somewhat over time. Next, we asked whether children’s classifications as affect-dysregulated showed different patterns of associations over time on repeated assessments of maternal caregiving, maternal depression, family income, and amount and quality of child care experiences. Repeated measures analyses of variance were conducted in which the affect-dysregulation groups at 24 and 36 months were the predictors of interest and site was included as a covariate. Repeated assessments of each measure were collected across four waves of data collection from 6 to 36 months. Table 3 lists the means on the repeated assessments for children classified as dysregulated or not dysregulated at the two ages. The repeated measures analyses simultaneously tested whether the groups differed on the average level of a given outcome over time and whether they showed different patterns of change between 6 and 36 months on that outcome. A general linear mixed model approach was used so that we could include children who might be missing data at a particular age ~e.g., child care quality at ages at which the child was in exclusive maternal care!. Children classified as dysregulated differed from other children in four of the five repeated measures analyses. More maternal depressive symptoms were reported over time by mothers of children in the 24-month dysregulated group, d 5 .28, F ~1, 1,011! 5 11.4, p , .001, and the 36-month dysregulated group, d 5 . 27, F ~1, 1,101! 5 8.0, p , .01, but no differences in patterns of change over time were observed in either comparison. Lower family incomes were reported over time by mothers of children classified as dysregu-

NICHD Early Child Care Research Network

lated at 24 months, d 5 .32, F ~1, 1,011! 5 11.4, p , .001, but not at 36 months. Change over time in income was not related to either dysregulation group. Compared with the other children, dysregulated children at 24 months experienced less sensitive and stimulating maternal caregiving over time, d 5 .64, F ~1, 1,011! 5 51.3, p , .001, and their mothers became less sensitive and stimulating between 15 and 36 months, F ~3, 1,011! 5 20.1, p , .001. For example, the dyregulation groups became more discrepant, increasing from an effect size of 0.50 at 15 months to an effect size of 1.0 at 24 months. Similarly, dysregulated children at 36 months also experienced less sensitive and stimulating maternal caregiving over time, d 5 47; F ~1, 1,011! 5 22.1, p , . 001, and more decline in caregiving quality over time, F ~3, 1,011! 5 23.7, p , .001, than children who were not dysregulated at 36 months. Although the amount of nonmaternal care did not differ between the dysregulation groups at either age, children considered dysregulated at 24 months, d 5 .20; F ~1, 772! 5 4.93, p , .05, and at 36 months, d 5 .29; F ~3, 772! 5 7.88, p , .01, were observed to be in lower quality child care. Change in child care quality over time was not related to dysregulation group. Finally, we conducted logistic regression analyses to determine the independent contribution of each of the significant family and child characteristics, controlling for the effects of all other variables in the model, in predicting whether the child was classified as dysregulated at each age. As shown in Table 4, the logistic regression included maternal education, vocabulary, agreeableness, extraversion, and depressive symptoms; partner in the home; income; and caregiving quality; and child’s ethnicity ~Black, Hispanic, other, or White!, 6-month negative mood, 15-month cognitive functioning, 15- and 36-month attachment classifications, and 24-month Q-security. Child care quality was excluded because children not in child care had missing data. The repeated assessments were averaged from 6 to 24 months when predicting 24month dysregulation status, and from 6 and 36 months when predicting 36-month dysregulation status. Only three variables showed sig-

|

|

DPP16~1! 440

19026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers nificant independent prediction, and one of them emerged as a significant predictor of both 24- and 36-month dysregulation classifications. Children whose mothers were less sensitive and stimulating had children who were significantly more likely to be classified as dysregulated at 24 months ~B 5 2.89, SE 5 .25, p , .001! and 36 months ~B 5 2.94, SE 5 .29, p , .01!. In addition, children were significantly more likely to be classified as dysregulated at 24 months if they had lower Bayley MDI cognitive scores ~B 5 2.04, SE 5 .01, p , .001! or were classified as avoidant versus secure at 15 months ~B 5 .76, SE 5 .23, p , .001!.

Child outcomes We conducted repeated measures analyses of variance to determine the extent to which children classified as dysregulated at either 24 or 36 months exhibited more negative outcomes during preschool and the early elementary school years. The repeated measures analyses were conducted using a general linear mixed model approach ~Singer, 1998! in which repeated assessments of predictors as well as repeated assessments of outcomes were included. The models included site; maternal education, verbal ability, and concurrent measures of income, partner in the home, depressive symptoms, and caregiving; and the child’s ethnicity, sex, and negative mood rating at 6 months age and all interactions with age. The results of these analyses are reported in Table 5. This table lists the F statistics for main effect tests comparing the children dysregulated at 24 months with the other children in the first column and the associated effect size in the second column. The effect size is computed as the difference between the adjusted mean for the two groups divided by the estimated standard deviation ~square root of the estimated residual variance!. The third column lists the F statistics for the test of the Age 3 Dysregulation interaction and the next three columns list the adjusted means at each age for each group. The same values are listed in the 7th through 12th columns with respect to dysregulation status at 36 months.

61

Cognitive0academic outcomes. Children classified as dysregulated at 24 or 36 months scored higher on impulsivity and lower on all measures of academic achievement. As can be seen under the cognitive0academic outcomes in Table 5, children who displayed affect dysregulation at 24 months, F ~1, 978! 5 6.47, p , .05, and at 36 months, F ~1, 978! 5 11.48, p , .001, scored lower on a standardized math test on average over time but did not show patterns of gains over time different from those of other children. The children dysregulated at 24 months scored on average 2.3 points lower at 54 months and 5.2 points lower in first grade, resulting in an averaged over time effect size of d 5 .25. The children dysregulated at 36 months scored on average 5.9 points lower at 54 months and 5.0 points lower in first grade, resulting in an averaged over time effect size of d 5 .37. A similar pattern of results obtained for the other academic outcomes. Dysregulated children at 24 months scored lower than other children on phonological knowledge, d 5 .19; F ~1, 978! 5 4.76, p , .05, and higher on impulsivity, d 5 .25; F ~1, 967! 5 7.62, p , .01. Similarly, dysregulated children at 36 months scored lower than other children on reading, d 5 .21; F ~1, 978! 5 4.06, p , .05, and phonological knowledge, d 5 .32; F ~1, 978! 5 10.27, p , .01, and higher on impulsivity, d 5 .19; F ~1, 967! 5 4.32, p , .05. None of these analyses indicated that change over time on these cognitive measures was related to dysregulation status at either age. Social and behavioral outcomes: Mother report. Affect dysregulation at 24 but not 36 months was related to maternal ratings of social skills between 54 months and the end of first grade. Children in the dysregulation group at 24 months were rated as lower in social skills on three of the four mother-reported outcomes. Their mothers reported less closeness in the mother–child relationship, d 5 .18; F ~1, 981! 5 3.95, p , .05, more externalizing problems, d 5 2.19; F ~1, 981! 5 4.52, p , .05, and fewer social skills, d 5 .30; F ~1, 981! 5 10.82, p , .01, than the other children. In the analysis of mother-reported social skills, an Age3Dysregulation interaction, F ~1, 1,845!5

|

DPP16~1! 440

20026

02013004

11:30 am

REVISED PROOF

62 5.55, p , .05, emerged because mothers reported fewer gains over time in social skills for the dysregulated children than for the other children. Children in the dysregulated group at 36 months did not differ significantly from the other children on any of these maternal ratings of social skills. Social and behavioral outcomes: Teacher report. Teachers rated the social skills of children in the dysregulated groups at 24 or 36 months as more problematic on two of the four measures. Teachers reported that children in the dysregulated group at 24 months had more externalizing problems, d 5 2.20; F ~1, 953! 5 14.75, p , .001, and fewer social skills, d 5 .22; F ~1, 949! 5 5.87, p , .05, than the other children. Children in the dysregulated group at 36 months were also rated by their teacher as showing fewer social skills, d 5 .24, F ~1, 940! 5 5.49, p , .05, than the other children. Teachers reported no differences in the levels of internalizing symptoms, in their feelings of closeness to the study child, or in change over time on any measure for children who did or did not show affect dysregulation in interactions with their mothers at 24 or 36 months. Discussion In this study we have conceptualized the construct of children’s affect dysregulation in the early years at the dyadic level rather than at the individual child level. Our measures of affect dysregulation represent the child’s responses to mother during dyadic interactions that varied in their emotional demands. The child’s affect and affect regulation as we have assessed them are thus a product of the mother–child dyad, but the mother’s behavior, per se, was not included in the formation of the affect-dysregulation groups. Based on their negative affect with their mothers, we have characterized some children as showing “affect-dysregulation in the context of the mother–child relationship” in recognition of the central role of parents in assisting children with affect regulation in the first few years of life, and the bidirectional nature of parent–child interactions, including those which scaffold emerging affect regula-

NICHD Early Child Care Research Network

tion capacities in young children. Children characterized as affect-dysregulated in the current study displayed marked negative affect during low stress play with their mothers, and some children also exhibited high levels of defiance and other forms of negative affect when asked by their mothers to clean up toys after playing in the lab. In other words, as early as age 2 or 3, some mother–child dyads had developed relationships that appeared to be affectively negative, sometimes coercive, and relatively ineffective during times of low and moderate stress. About 16% of the sample of more than 1,000 children was identified as affectdysregulated at either 24 or 36 months. The children in affectively dysregulated relationships with their mothers differed in important ways from children not so identified. They came from families that were more likely to be African American, with fewer father figures and mothers who had less education, lower verbal ability, and poorer psychological adjustment. The children themselves had lower Bayley MDI scores at 15 months. In addition, their mothers were less sensitive and stimulating during both semistructured play and during naturally occurring interactions at home during the first years of these children’s lives. However, when all of these predictors were considered in a multivariate analysis, only the mother’s sensitivity and stimulation and the child’s cognitive development and insecureavoidant attachment classification at 15 months were associated with affect-dysregulation status at 24 or 36 months. Children classified as affect-dysregulated thus experienced and expressed negative affect in the context of parenting relationships that were less emotionally supportive overall, and they may have had fewer cognitive resources to assist in strategic emotion regulation and to elicit more positive caregiving. There were several associations between attachment, which was assessed in different ways at 15, 24, and 36 months, and affect dysregulation at both 24 and 36 months. Children who were insecure-avoidant at 15 months, but not insecure-resistant or insecure-disorganized children at that age, were more likely to be classified in the affect-dysregulated group at

|

|

DPP16~1! 440

21026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers 24 months. This finding is interesting because avoidance is often characterized by emotional overcontrol, or a minimization of negative affect when the attachment system is activated in the presence of the mother. However, some avoidant infants ~in the A2 subcategory! do show fussiness in the Strange Situation, and Ainsworth, Bell, and Stayton ~1972! showed that avoidant infants are angry with their mothers at home and ambivalent about contact with her, fussing and crying when they are put down. Thus, avoidant children who inhibit the expression of attachment behavior when their attachment systems are activated do behave in more negative ways with their mothers under some conditions. The results of this study also support the findings by others that avoidant infants are undercontrolled in other situations in later childhood ~Egeland, Pianta, & O’Brien, 1993; Munson, McMahon, & Spieker, 2001; Renken, Egeland, Marvinney, Mangelsdorf, & Sroufe, 1989!. The lack of association with disorganized attachment classifications is interesting. However, this classification is heterogeneous. We know little about how children whose Strange Situation behavior is classified as disorganized manage negative affect in interaction with the mother in less stressful situations. It is possible, for instance, that when their attachment systems are not activated they tend to inhibit the expression of negative affect in the presence of their mother ~i.e., a pattern opposite that observed in infants classified as avoidant!. A shift in the type of insecure attachment associated with affect dysregulation is already detectable by 36 months, when the association is with insecureresistant attachment for both 24- and 36month affect dysregulation groups. Crittenden ~1992! predicted such a shift to insecureresistance in the preschool years. Our findings are consistent with her speculation that more preschoolers will begin to use the more coercive strategies that are available to them as a result of their increasing cognitive and linguistic competence. Given the importance of affect regulation for the acquisition of competence in multiple domains ~Campos, Campos, & Barrett, 1989; Campos, Mumme, Kermoian, & Campos 1994; Dodge, 1989; Garber & Dodge, 1991; Kopp,

63

1989; Thompson, 1994!, it is perhaps unsurprising that the developmental outcomes in later preschool and the early school years were much less positive for children in the affectdysregulation groups. Children who were dysregulated as toddlers or young preschoolers were less cognitively and socially competent at 54 months, kindergarten, and Grade 1, and were more likely to be rated higher on behavior problems than were never dysregulated children. Notably, this was true even after controlling for earlier and concurrent family and child characteristics that were associated with affective dysregulation, including maternal education, vocabulary, depressive symptoms, partner in the home, income, and caregiving quality, and child ethnicity and 6-month negative mood. Thus, early affect dysregulation in the context of the mother–child relationship apparently is associated with continuing problems in both cognitive and social development through preschool and into the early school years. Mothers of affectively dysregulated children reported feeling less closeness with their children in later years, compared with mothers of children who were not dysregulated. In contrast, teachers did not report differential feelings of closeness across the groups of children although they did report differences in social skills and behavior problems. Perhaps this pattern of results reflects teacher professionalism. It will be interesting to test in future analyses whether observations of the quality of student–teacher interactions support or contradict teacher reports of the quality of the student–teacher relationship. The findings from this study converge with those of three smaller studies that also reported links between negativity and conflict in the mother–child relationship during the toddler years and subsequent behavior problems ~Denham et al., 2000; Leadbeater et al., 1996; Martin et al., 2001!. However, the large sample size and measurement strategies of the NICHD Study of Early Child Care have permitted us to control for a number of possible confounding factors and to extend the findings from previous research in several important ways: ~a! we included multiple measures of affect dysregulation in different contexts of

|

DPP16~1! 440

22026

02013004

11:30 am

REVISED PROOF

64 mother–child interaction that embodied different kinds of emotional demands or challenges; ~b! we identified several demographic, maternal, and family antecedents and concurrent correlates of affect dysregulation; ~c! our developmental outcome measures included not only behavior problems, but also more general measures of social and cognitive competence, and were obtained using multiple forms of assessment and multiple informants; and ~d! we followed the children longitudinally through preschool and into the school years as demands for affective and behavioral selfregulation steadily increased. The effects reported here are notable for their longevity and for their breadth. The convergence across very different assessments of competence as well as across informants makes us confident in concluding that affect dysregulation in the early years constitutes a significant risk for later social and cognitive development. What might account for these effects enduring into the preschool and early school years? Interestingly, dysregulated toddlers and preschoolers had not been identified by their mothers as more temperamentally difficult as infants than were children who exhibited no dysregulation. However, these children were more likely than their nondysregulated counterparts to become affectively negative with their mothers during play interactions as early as 6 months of age. We cannot rule out the possibility that negativity was characteristic of these children early on, which could explain their difficulties with negative emotions in the mother–child relationship at 24 months ~Mangelsdorf, Gunnar, Krestenbaum, Lang, & Andreas, 1990!. However, it is equally likely that the early negativity was a child response to the less sensitive and stimulating maternal caregiving that they received at every age. Just as Ainsworth et al. ~1978! concluded that duration of infant crying in the first quarter of life reflected differences in maternal responsiveness to crying, it is possible that some mothers induced or elicited negative affect in their infants, thereby creating an emotionally dysregulated relationship by either 24 or 36 months. Higher infant negativity with mothers was observed by 6 months for the children who would be in the dysregulated group at 24

NICHD Early Child Care Research Network

months. It emerged later for children classified as dysregulated at 36 months. For both groups, mothers of dysregulated toddlers were less sensitive and stimulating beginning in infancy. Thus, infants of such mothers experienced higher rates and0or intensities of negative affect in the very relationship on which they depend for reducing and managing their negative affect, and they would receive less sensitive responsiveness and scaffolding to help them learn to manage their negative emotions independently. Amplified during the emotionally difficult transitions of the toddler years, such dyads may be more likely to establish a pattern of mutual negativity and coercion ~Patterson, DeBaryshe, & Ramsey, 1989! that becomes progressively more difficult to escape. With less effective socialization of attention control ~Raver, 1996; Raver, Blackburn, Bancroft, & Torp, 1999! and compliance ~Kochanska, 1994; Londerville & Main, 1981! in such dyads, in addition to increasingly negative affect, children’s cognitive and achievementrelated outcomes as well as their social competence could be compromised; deficits could easily cumulate and further exacerbate the difficult mother–child relationship. As appealing as such an account may be, it fails to tell the whole story. Among mothers who were in the lowest quartile of maternal caregiving from 6 through 36 months, only 24% of their infants were in the affectdysregulated group at 24 months, and only 16% were in the affect-dysregulated group at 36 months. This suggests the still more complex possibility that the roots of affect dysregulation are truly dyadic—that is, that some mothers and infants are a poor affective fit for one another ~Crockenberg & Litman, 1990; Kopp, 1989; Mangelsdorf et al., 1990; Seifer, Schiller, Sameroff, Resnick, & Riordan, 1996!. The effects of such mismatches may not be readily apparent until the second or third year of life when the young child’s emotions are both quickly differentiating and intensifying ~Campos et al., 1994; Lewis, 1993!, when demands for self-regulation increase ~Kopp, 1989; Thompson, 1994!, and when autonomy bids emerge as issues for children and for their parents ~Belsky et al., 1998; Crockenberg & Litman, 1990!. Under these conditions, dyads that

|

|

DPP16~1! 440

23026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers have not effectively negotiated the affect regulation challenges of infancy may be particularly susceptible to affect dysregulation in the toddler and early preschool years. This susceptibility is perhaps especially likely to be expressed in adverse ecological conditions that place additional psychological stresses on families, and make it more difficult for mothers to support and manage their children’s changing emotional needs. Such a possibility is consistent with our findings that children were disproportionately more likely to be affectdysregulated when they were growing up in poverty with mothers who were poorly educated and depressed. There are several implications of the findings of this study for research on affect regulation. First, there are implications for conceptualizing affect regulation in the early years. Although several scholars have hypothesized that affect regulation occurs in the context of the parent–child dyad during infancy ~Kopp, 1989; Thompson, 1994; Tronick, 1989!, the current study is one of the first to operationalize early affect dysregulation in explicitly dyadic terms and is the largest and most complex. The findings that dysregulated toddlers and preschoolers differ with respect to both antecedent and concurrent correlates as well as social and cognitive sequelae several years later confirm the value of observing and studying early affect regulation in a dyadic context. Similarly, there are implications for predictive relationships among early child, maternal, and environmental characteristics and later developmental outcome. The findings of the current study suggest that it is perhaps not so much child or even maternal affective characteristics that are most predictive, but relationship qualities to which both partners contribute and which are best captured at the dyadic level. By observing child affect with the mother in undemanding play settings and more demanding compliance settings, we have gained a unique perspective on the extent to which the mother– child relationship serves as an effective or an ineffective regulator of the child’s negative emotions across contexts. This account suggests that future researchers should focus on dyadic measures of affective dysregulation be-

65

ginning in infancy, and not on assessments of infants or mothers alone. Despite the strength of our findings, an important feature of this study limits our conclusions. Although we have assessed children’s affect dysregulation in a dyadic context, we have not evaluated mothers’ contributions to the dyadic interactions. We do not know for which dyads the negative affect in interaction was mutual, thus reflecting a true coercive cycle, and for which dyads it was one-sided on the part of the child only. Thus, we are unable to examine the dyadic processes that contribute to affect dysregulation in mother–child relationships or to identify which particular qualities within the dyadic relationship are most critical to effective or ineffective affect regulation. The finding that affect dysregulation in the mother–child dyad at 24 months predicted more broadly to later outcomes than did 36month affect dysregulation further suggests the need for research to discern which features of mother–child interaction at age two are disrupted in affectively dysregulated dyads, and how they differ from disruptions at age 3. Perhaps the requirements for a good fit between mother and child vary somewhat at the two ages, so that some dyads are particularly at risk for dysregulation at particular developmental stages. This study suggests several additional avenues for future investigation. First, although we found a significant association between affect-dysregulation at 24 and 36 months, in fact most children were classified as affectively dysregulated at a single age. Either affect dysregulation trajectories are variable or our measurement is still imprecise. We did not find that using continuous measures of dysregulation made significant differences in our findings, but finer grained observations might make a difference. Second, although we assessed affect dysregulation in the toddler years, we have not examined later features of self-regulation such as resistance to temptation, delay of gratification, sustained attention, and affect regulation strategies that emerge during the preschool years. These may be important mediators of the longitudinal associations between early affect regulation and later cognitive and social development. Third, we have pri-

|

DPP16~1! 440

24026

02013004

11:30 am

REVISED PROOF

66 marily considered underregulated rather than overregulated emotion. We found few associations between early dysregulation and later internalizing behavior, perhaps because we did not look for early indicators of overregulation. It is interesting to speculate that children whose emotions are consistently checked in their relationships with mothers during the normative emotional upheavals of the toddler years may also be at risk for later deficits in functioning, perhaps especially in the socioemotional domain. In conclusion, we are encouraged by the intriguing and important findings of the current investigation and the potential value of the construct of affect dysregulation for understanding associations between emotions in the parent–child relationship and children’s

NICHD Early Child Care Research Network

social and cognitive development. These data also may have implications for prevention and intervention. For example, they suggest the importance of intervening early in poorly regulated mother–toddler relationships to help mothers learn to calm and scaffold toddlers who are upset and difficult to console, and to support sensitive responsiveness in mothers of young children who may be at risk for behavioral or learning problems. Because the sample is being followed into later primary grades, we will be able to identify longer term developmental trajectories for children dysregulated as toddlers and preschoolers, permitting us to determine both how stable and pervasive these effects are, and whether they are amplified or reduced by schooling experiences.

References Achenbach, T. M. ~1991a!. Manual for the Child Behavior Checklist04-18 and 1991 Profile. Burlington, VT: University of Vermont Department of Psychiatry. Achenbach, T. M. ~1991b!. Manual for the Teacher’s Report Form and 1991 Profile. Burlington, VT: University of Vermont Department of Psychiatry. Ainsworth, M. D. S., Bell, S. M., & Stayton, D. J. ~1972!. Individual differences in the development of some attachment behaviors. Merrill–Palmer Quarterly, 18, 123–143. Ainsworth, M. D. S., Blehar, M., Waters, E., & Wall, S. ~1978!. Patterns of attachment. Hillsdale, NJ: Erlbaum. Barkley, R. A. ~1994!. The assessment of attention in children. In G. Reid Lyon ~Ed.!, Frames of reference for the assessment of learning disabilities: New views on measurement issues ~pp. 69–102!. Baltimore, MD: Paul H. Brookes. Barkley, R. A., Grodzinsky, G. M., & DuPaul, G. J. ~1992!. Frontal lobe functions in attention deficit disorder with and without hyperactivity: A review and research report. Journal of Abnormal Child Psychology, 20, 163–188. Bates, J. E., Maslin, C. A., & Frankel, K. A. ~1985!. Attachment security, mother–child interaction, and temperament as predictors of behavior problem ratings at age three years. In I. Bretherton & E. Waters ~Eds.!, Growing points in attachment theory and research. Monographs of the Society for Research in Child Development, 50~Serial No. 209!, 167–193. Bates, J. E., Pettit, G. S., & Dodge, K. A. ~1995!. Family and child factors in stability and change in children’s aggressiveness in elementary school. In J. McCord ~Ed.!, Coercion and punishment in long-term perspectives ~pp. 124–138!. New York: Cambridge University Press. Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. ~1998!. The interaction of temperamental resistance to control and restrictive parenting in the development of

externalizing behavior. Developmental Psychology, 34, 982–995. Bayley, N. ~1969!. Bayley Scales of Infant Development. San Antonio, TX: Psychological Corporation. Belsky, J., Woodworth, S., & Crnic, K. ~1996!. Troubled family interaction during toddlerhood. Development and Psychopathology, 8, 477– 495. Belsky, J., Hsieh, K., & Crnic, K. ~1998!. Mothering, fathering, and infant negativity as antecedents of boys externalizing problems and inhibition at age 3 years: Differential susceptibility to rearing experience? Development and Psychopathology, 10, 301–320. Bergman, L. R., & El-Khouri, B. ~1987!. EXACON: A Fortran 77 program for the exact analysis of single cells in a contingency table. Educational and Psychological Measurement, 47, 155–161. Bergman, L. R., & El-Khouri, B. M. ~1998!. SLEIPNER: A statistical package for pattern-oriented analyses, V 2.0. Stockholm: Stockholm University, Department of Psychology. Bowlby, J. S. ~1968!. Attachment and loss: Vol. I. Attachment. New York: Basic Books. Braungart-Rieker, J. M., & Stifter, C. A. ~1996!. Infants’ responses to frustrating situations: Continuity and change in reactivity and regulation. Child Development, 67, 1767–1779. Caldwell, B. M., & Bradley, R. H. ~1984!. Home observation for the measurement of the environment. Little Rock, AR: University of Arkansas at Little Rock. Calkins, S. D. ~1994!. Origins and outcomes of individual differences in emotion regulation. In N. A. Fox ~Ed.!, The development of emotion regulation. Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59~Serial No. 240!, 53–72. Campbell, S. B. ~2002!. Behavior problems in preschool children: Developmental and clinical issues ~2nd edition!. New York: Guilford Press.

|

|

DPP16~1! 440

25026

02013004

11:30 am

|

REVISED PROOF

Affect dysregulation in the mother–child relationship in toddlers Campbell, S. B., Pierce, E. W., March, C. L., Ewing, L. J., & Szumowski, E. K. ~1994!. Hard-to-manage preschool boys: Symptomatic behavior across contexts and time. Child Development, 65, 836–851. Campbell, S. B., Shaw, D. S., & Gilliom, M. ~2000!. Early externalizing behavior problems: Toddlers and preschoolers at risk for later maladjustment. Development and Psychopathology, 12, 467– 488. Campos, J., Campos, R., & Barrett, K. ~1989!. Emergent themes in the study of emotional development and emotion regulation. Developmental Psychology, 25, 394– 402. Campos, J., Mumme, D., Kermoian, R., & Campos, R. ~1994!. A functionalist perspective on the nature of emotion. In N. Fox ~Ed.!, the development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59~203, Serial No. 240!. Carlson, E. A., & Sroufe, L. A. ~1995!. Contributions of attachment theory to developmental psychopathology. In D. Cicchetti & D. Cohen ~Eds.!, Manual of developmental psychopathology, Vol. I. Theory and methods ~pp. 581– 617!. New York: Wiley. Cassidy, J., Marvin, R. S., & The MacArthur Attachment Working Group. ~1992!. Attachment organization in preschool children: Procedures and coding manual. Unpublished coding manual, Pennsylvania State University. Cicchetti, D., Toth, S., Bush, M. A., & Gillespie, J. F. ~1988!. Stage salient issues: A transactional model of intervention. New Directions for Child Development, 39, 123–145. Costa, P., & McCrae, R. ~1985! The NEO Personality Inventory Manual. Odessa, FL: Psychological Assessment Resources. Crittenden, P. M. ~1992!. Quality of attachment in the preschool years. Development and Psychopathology, 4, 209–241. Crockenberg, S., & Litman, C. ~1990!. Autonomy as competence in 2-year-olds: Maternal correlates of child defiance, compliance, and self-assertion. Developmental Psychology, 26, 961–971. Denham, S., Workman, E., Cole, P, Weissbrod, C., Kendziora, K., & Zahn–Waxler, C. ~2000!. Prediction of externalizing behavior problems from early to middle childhood: The role of parental socialization and emotion expression. Development and Psychopathology, 12, 23– 45. Dodge, K. ~1989!. Coordinating responses to aversive stimuli: Introduction to special section on the development of emotion regulation. Developmental Psychology, 25, 339–342. Dunn, L. M., & Dunn, L. M. ~1981!. Peabody Picture Vocabulary Test—Revised. Circle Pines, MN: American Guidance Service. Egeland, B., Pianta, R., & O’Brien, M. A. ~1993!. Maternal intrusiveness in infancy and child maladaptation in early school years. Development and Psychopathology, 5, 359–370. Fogel, A. ~1993!. Developing through relationships. Chicago: University of Chicago Press. Garber, J., & Dodge, K. ~Eds.!. ~1991!. The development of emotion regulation and dysregulation. Cambridge: Cambridge University Press. Gresham, F., & Elliot, S. ~1990!. The social skills rating system. Circle Pines, MN: American Guidance Service.

67

Halperin, J. M., Sharma, V., Greenblatt, E., & Schwartz, S. T. ~1991!. Assessment of the continuous performance test: Reliability and validity in a nonreferred sample. Psychological Assessment: Journal of Consulting and Clinical Psychology, 3, 603– 608. Keenan, K., Shaw, D., Delliquadri, E., Giovannelli, J., Walsh, B. ~1998!. Evidence for the continuity of early problem behaviors: Application of a developmental model. Journal of Abnormal Child Psychology, 26, 441– 452. Kochanska, G. ~1994!. Beyond cognition: Expanding the search for the early roots of internalization and conscience. Developmental Psychology, 30, 20–22. Kochanska, G. ~1995!. Children’s temperament, mothers’ discipline, and security of attachment: Multiple pathways to emerging internalization. Child Development, 66, 597– 615. Kochanska, G. ~1997!. Mutually responsive orientation between mothers and their young children: Implications for early socialization. Child Development, 68, 94–112. Kopp, C. B. ~1982!. Antecedents of self-regulation: A developmental perspective. Developmental Psychology, 18, 199–214. Kopp, C. B. ~1989!. Regulation of distress and negative emotions: A developmental view. Developmental Psychology, 25, 343–354. Leadbeater, B., Bishop, S. & Raver, C. ~1996!. Quality of mother–toddler interactions, maternal depressive symptoms, and behavior problems in preschoolers of adolescent mothers. Developmental Psychology, 32, 280–288. Lewis, M. ~1993!. The emergence of human emotions. In M. Lewis & J. Havilland ~Eds.!, Handbook of emotions ~pp. 223–235!. New York: Guilford Press. Londerville, S., & Main, M. ~1981!. Security of attachment, compliance and maternal training methods in the second year of life. Developmental Psychology, 17, 289–299. Mangelsdorf, S., Gunnar, M., Krestenbaum, R., Lang, S., & Andreas, D. ~1990!. Infant proneness-to-distress temperament, maternal personality, and mother–infant attachment: Associations and goodness of fit. Child Development, 61, 820–831. Martin, S., Clements, M., & Crnic, K. ~2001!. Emotional regulation in mother–toddler dyads: Links to children’s internalizing and externalizing behavior problems. Poster presented at the biennial meeting of the Society for Research in Child Development, Minneapolis, MN. Medoff–Cooper, B., Carey, W. B., & McDevitt, S. C. ~1993!. Early Infancy Temperament Questionnaire. Journal of Developmental and Behavior Pediatrics, 14, 230–235. NICHD Early Child Care Research Network. ~1996!. Characteristics of infant child care: Factors contributing to positive caregiving. Early Childhood Research Quarterly, 11, 269–306. NICHD Early Child Care Research Network. ~1998!. Early child care and self-control, compliance, and problem behavior at 24 and 36 months. Child Development, 69, 1145–1170. NICHD Early Child Care Research Network. ~1999!. Child care and mother–child interaction in the first three years of life. Developmental Psychology, 35, 1399–1413.

|

DPP16~1! 440

26026

02013004

11:30 am

REVISED PROOF

68 NICHD Early Child Care Research Network. ~2001!. Child care and family predictors of preschool attachment and stability from infancy. Developmental Psychology, 37, 847–862. NICHD Early Child Care Research Network. ~in press!. The NICHD study of early child care: Contexts of development and developmental outcomes over the first seven years of life. In J. Brooks–Gunn & L. J. Berlin ~Eds.!, Young children’s education, health, and development: Profile and Synthesis Project report. Washington, DC: Department of Education. Patterson, G., Baryshe, B., & Ramsey, E. ~1989!. A developmental perspective on antisocial behavior. American Psychologist, 44, 329–335. Pettit, G. S., & Bates, J. E. ~1989!. Family interaction patterns and children’s behavior problems from infancy to age 4 years. Developmental Psychology, 25, 413– 420. Pianta, R. ~1992!. The Student–Teacher Relationship Scale. Charlottesville, VA: University of Virginia. Pianta, R. ~1994!. The Child–Parent Relationship Scale. Charlottesville, VA: University of Virginia. Radloff, L. ~1977!. The CES-D Scale: A self-report depression scale for research in the general population. Journal of Applied Psychological Measurement, 1, 385– 401. Raver, C. ~1996!. Success at catching and keeping toddler’s attention: An examination of joint attention among low-income mothers and their 2-year-olds. Early Development and Parenting, 5, 225–236. Raver, C., Blackburn, E., Bancroft, M., & Torp, N. ~1999!. Relations between effective emotional self-regulation, attentional control, and low-income preschoolers’ social competence with peers. Early Education and Development, 10, 333–350. Renken, B., Egeland, B., Marvinney, D., Mangelsdorf, S., & Sroufe, L. A. ~1989!. Early childhood antecedents of aggression and passive-withdrawal in early elementary school. Journal of Personality, 57, 257–281. Rosvold, H. E., Mirsky, A. F., Sarason, I., Bransome, E. D. Jr., & Beck, L. H. ~1956!. A continuous performance test of brain damage. Journal of Consulting Psychology, 20, 343–350. Rothbart, M. K., & Bates, J. E. ~1998!. Temperament. In W. Damon ~Ed.! & N. Eisenberg ~Vol. Ed.!, Handbook of child psychology: Vol. 3. Social, emotional, and personality development ~5th ed., pp. 105–176!. New York: Wiley. Rothbart, M. K., Posner, M. I., & Hershey, K. L. ~1995!. Temperament, attention, and developmental psychopathology. In D. Cicchetti & D. J. Cohen ~Eds.!, Manual of developmental psychopathology, Vol. 1: Theory and methods ~pp. 315–340!. New York: Wiley. Saarni, C., Mumme, D.L., & Campos, J. J. ~1998!. Emotional development: Action, cummunication, and un-

NICHD Early Child Care Research Network derstanding. In W. Damon ~Series Ed.! & N. Eisenberg ~Vol. Ed.!, Handbook of child psychology: Vol. 3. Social, emotional, and personality development ~5th ed., pp. 237–309!. New York: Wiley. Sameroff, A., & Chandler, M. ~1975!. Reproductive risk and the continuum of caretaking casualty. In F. Horowitz ~Ed.!, Review of child development research ~Vol. 4, pp. 187–244!. Chicago: University of Chicago Press. Sanson, A., Oberklaid, F., Pedlow, R., & Prior, M. ~1991!. Risk indicators: Assessment of infancy predictors of preschool behavioral maladjustment. Journal of Child Psychology and Psychiatry, 32, 609– 626. Seifer, R., Schiller, M., Sameroff, A., Resnick, S., & Riordan, K. ~1996!. Attachment, maternal sensitivity, and infant temperament during the first year of life. Developmental Psychology, 32, 12–25. Shaw, D. S., Keenan, K., & Vondra, J. I. ~1994!. Developmental precursors of externalizing behavior: Ages 1 to 3. Developmental Psychology, 30, 355–364. Singer, J. ~1998!. Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. Journal of Educational and Behavioral Statistics, 24, 323–355. Sroufe, L. A. ~1979!. The coherence of individual development: Early care, attachment, and subsequent developmental issues. American Psychologist, 34, 834–841. Sroufe, L. A. ~1996!. Emotional development. Cambridge: Cambridge University Press. Thomas, A., & Chess, S. ~1977!. Temperament and development. New York: Brunner0Mazel. Thompson, R. A. ~1994!. Emotion regulation: A theme in search of a definition. Monographs of the Society for Research in Child Development, 59~Serial No. 240!, 25–52. Thompson, R. A. ~1998!. Early sociopersonality development. In W. Damon ~Series Ed.! & N. Eisenberg ~Vol. Ed.!, Handbook of child psychology: Vol. 3. Social, emotional, and personality development ~pp. 25– 104!. New York: Wiley. Tronick, E. ~1989!. Emotions and emotional communication in infants. American Psychologist, 44, 112–119. Vaughn, B. E., & Waters, E. ~1990!. Attachment behavior at home and in the laboratory: Q-sort observations and Strange Situation classifications of one-yearolds. Child Development, 61, 1865–1973. Waters, E., & Deane, K. ~1985!. Defining and assessing individual differences in attachment relationships: Q-methodology and the organization of behavior in infancy and early childhood. Monographs for the Society for Research in Child Development, 41– 65. Weschler, D. ~1981!. Manual for the Weschler Adult Intelligence Scale—Revised. San Antonio, TX: Psychological Corporation. Winer, B. J. ~1971!. Statistical principles in experimental design ~2nd ed.!. New York: McGraw–Hill.

|