Mothers’ Interactions With Temperamentally Frustrated Infants


Download Mothers’ Interactions With Temperamentally Frustrated Infants


Preview text

ARTICLE
MOTHERS’ INTERACTIONS WITH TEMPERAMENTALLY FRUSTRATED INFANTS
SUSAN D. CALKINS, ANNE HUNGERFORD, AND SUSAN E. DEDMON University of North Carolina at Greensboro
ABSTRACT: A sample of 162 six-month-old infants was selected from a larger sample of 346 infants on the basis of mothers’ report of their infants’ temperament and a laboratory assessment of temperament. Infants were classified as easily frustrated or less frustrated and observed in several types of interactions with their mothers in the laboratory. Mothers completed several measures that indicated their level of parenting stress, psychological functioning, and marital adjustment. Maternal behavior with infants was coded along the dimensions of sensitivity, intrusiveness, and physical stimulation. Results indicated that maternal intrusiveness was related to infant temperament and that maternal physical stimulation was predicted by an interaction of infant temperament and mothers’ perceived parenting stress. Implications of these findings for mother– infant interaction and subsequent child adjustment are discussed.
RESUMEN: Un grupo muestra de 162 infantes de seis meses de nacidos fue seleccionado de una muestra aun mayor de 346 infantes, sobre la base de lo que las madres reportaron acerca del temperamento de sus infantes y una evaluacio´n de laboratorio del temperamento. A los infantes se les dividio´ entre los que se frustraban fa´cilmente y los menos frustrados, y se les observo´ en varios tipos de interacciones con sus madres en el laboratorio. Las madres completaron varias encuestas que indicaron su nivel de tensio´n en cuanto al proceso de crianza, el funcionamiento sicolo´gico, as´ı como la adaptacio´n marital. La conducta materna con los infantes fue codificada a lo largo de las dimensiones de susceptibilidad, entremetimiento, y estimulacio´n f´ısica. Los resultados indicaron que el entremetimiento materno estaba relacionado con el temperamento del infante; adema´s, que una interaccio´n entre el temperamento del infante y la tensio´n acerca de la crianza, tal como era percibida por la madre, predec´ıan el nivel de estimulacio´n f´ısica de la madre. Se discuten las implicaciones de estos resultados en cuanto a la interaccio´n entre madre e infante y las subsecuentes adaptaciones del nin˜o.
RE´ SUME´ : Un e´chantillon de 162 be´be´s de six mois fut se´lectionne´ a` partir d’un e´chantillon plus grand de 346 be´be´s, sur la base des rapports fait par les me`res du tempe´rament de leur be´be´ et d’une e´valuation laboratoire du tempe´rament. Les be´be´s ont e´te´ classifie´s comme e´tant facilement frustre´s ou moins frustre´s et observe´s dans plusieurs types d’interactions avec leurs me`res dans le laboratoire. Les me`res comple´te`rent plusieurs mesures qui indique`rent leur niveau de stress de parentage, le fonctionnement psychol-
This research was supported by a National Institute of Mental Health FIRST Award (MH 55584) and a Research Council Grant and Summer Excellence Award from the University of North Carolina at Greensboro to Susan D. Calkins. The authors thank Martie Skinner for statistical assistance and Amy Clark, Mary Johnson, Cynthia Smith, Lauren Davis, Laura Lomax, and Natalie Weeks for their help in subject recruitment and data collection and coding. The authors also thank the families who generously gave their time to participate in the study. Direct correspondence to: Susan D. Calkins, Department of Psychology, P.O. Box 26164, University of North Carolina at Greensboro, Greensboro, NC 27402-6164; e-mail: [email protected]
INFANT MENTAL HEALTH JOURNAL, Vol. 25(3), 219– 239 (2004) ᮊ 2004 Michigan Association for Infant Mental Health Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/imhj.20002
219

220 ● S.D. Calkins et al.
ogique et l’ajustement marital. Le comportement maternel avec les be´be´s fut code´ selon les dimensions de sensibilite´, d’intrusion et de stimulation physique. Les re´sultats indique`rent que l’intrusion maternelle e´tait lie´e au tempe´rament du be´be´ et que la stimulation physique maternelle e´tait pre´dite par une interaction du tempe´rament du be´be´ et du stress perc¸u de parentage des me`res. Des implications de ces re´sultats pour l’interaction me`re-be´be´ et l’ajustement de l’enfant subse´quent sont discute´es.
ZUSAMMENFASSUNG: Eine Untersuchungsgruppe bestehend aus 162 sechsmonatigen Sa¨uglingen wurde aus einer gro¨ßeren Gruppe von 346 Kleinkindern auf Grund der mu¨tterlichen Einscha¨tzungen und einer Laboruntersuchung des Temperaments des Kleinkinds ausgewa¨hlt. Die Kleinkinder wurden in eine frustrierte, oder weniger frustrierte Gruppe klassifiziert und bei verschiedenen Arten der Interaktion mit ihren Mu¨ttern im Labor beobachtet. Die Mu¨tter fu¨llten einige Bo¨gen aus, die das Ausmaß des elterlichen Stress, der psychologischen Fa¨higkeiten und der Anpassung an das Elternsein maßen. Das mu¨tterliche Verhalten mit ihren Kleinkindern wurde nach den Dimensionen der Einfu¨hlsamkeit, der Einmischung und der ko¨rperlichen Stimulation kodiert. Die Ergebnisse weisen darauf hin, dass die mu¨tterliche Einmischung in einer Korrelation zum Temperament des Kindes stand, und dass die ko¨rperliche Stimulierung voraussagbar war durch das Zusammenspiel zwischen dem Temperament des Kindes und der mu¨tterlichen Wahrnehmung ihres elterlichen Stress. Die Bedeutung dieser Ergebnisse fu¨r die Mutter–Kind Interaktion und die darauf folgende Anpassung des Kindes werden diskutiert.
*** Despite many years of research on the topic of early parent – child relationships and the implications for child adjustment, questions remain about the predictors and consequences of these early interactions. One broad question often asked about early relationships concerns the relative contribution of parent factors and child factors to these relationships. This question often is addressed empirically by looking at the relations between parenting behavior and attachment security and by examining relations between infant temperament and attachment security. Although parent – child relationships are usually operationalized in terms of the security of attachment (Ainsworth, Blehar, Waters, & Wall, 1978), interactions between parents and infants, and the discrete behaviors that make up such interactions, are clearly important as well (Seifer & Schiller, 1995; Thompson, 1998). Although parenting factors often are conceptualized as parenting behaviors, other characteristics of the parent such as personality or psychological adjustment also may influence the parent – child relationship (Belsky, 1984). Thus, specific questions remain to be addressed concerning the factors that may predict the nature and quality of parent – child relationships early in development, and whether these factors may influence broader relationship dimensions such as attachment security and behavioral adjustment later in development. The purpose of the present study was to examine associations

Frustrated Infants ● 221
between a specific infant temperament characteristic (frustration) and maternal behavior. In addition, interactions between maternal functioning and infant temperament were examined as potential predictors of parenting quality.
In studies of maternal interaction during infancy, much attention has been focused on the role of maternal sensitivity versus intrusiveness during early infancy. Several studies indicate that sensitivity to an infant’s signals contributes to mutual regulation, which in turn assists the child in gaining mastery over his own behavior (Gianino & Tronick, 1988; Shaw, Keenan, & Vondra, 1994). Alternatively, intrusive parenting during the first few months of life may disrupt the process of mutual regulation and interfere with the development of self-regulatory skills. For example, Egeland, Pianta, and O’Brien (1993) found that intrusive caregiving in infancy was related to behavior problems in preschool and, in a follow-up assessment, observed the same relation with behavior problems in early elementary school. The assumption of this and other attachment research is that a child who is exposed to sensitive and responsive caregiving early in life develops a sense of the caregiver as available in times of stress. Thus, children who have established secure attachment relationships with their caregivers develop ways of understanding and organizing their emotional responses and experiences that enable them to cope effectively (Bretherton, 1985; Cassidy, 1994; Egeland et al., 1993; Lewis, Feiring, McGuffog, & Jaskir, 1984).
Another relevant dimension of maternal behavior in infancy is physical contact and stimulation. Because few studies have examined physical contact separately from measures of sensitivity or intrusiveness, its relation to the quality of mother – infant interaction and child functioning is less clear. However, van den Boom and Hoeksma (1994) found that mothers of highly irritable infants provided less physical contact and provided lower levels of “effective stimulation,” which included effective physical stimulation, than did mothers of less irritable infants. In addition, an intervention study involving low-income women and their infants experimentally manipulated physical contact by providing the experimental group with soft baby carriers and the control group with infant seats (Anisfeld, Casper, Nozyce, & Cunningham, 1990). Mothers in the intervention group initiated more vocalizations and showed more contingent responses to their infants’ vocalizations at 31⁄2 months relative to mothers in the control group. Infants in the experimental group were more likely to be securely attached at 13 months than infants in the control group, even after controlling for differences in maternal responsiveness in the two groups.
There is a substantial theoretical and empirical literature supporting the notion that sensitive caregiving and subsequent attachment security play an important role in childhood social and emotional adjustment. However, the questions of whether characteristics of the infant influence the nature of these early interactions, the likelihood that an infant will receive sensitive caregiving, and the development of a secure attachment relationship have been debated vigorously for the last several years (for a review, see Thompson, 1998). Characteristics of the infant that are most often explored are temperamental dimensions such as proneness-to-distress (Calkins & Fox, 1992; Mangelsdorf, Gunnar, Kestenbaum, Lang, & Andreas, 1990; Seifer, Schiller, Sameroff, Resnick, & Riordan, 1996). Because temperament is viewed as a relatively stable, biologically based individual difference in the infant’s behavioral style or emotional tendencies (Goldsmith et al., 1987), the assumption is that any relation between child temperament and parenting behavior or attachment relationship is caused, at least in part, by these early appearing child behaviors (Mangelsdorf et al., 1990; Seifer et al., 1996). Presumably, infant temperament that is characterized by frequent or intense bids for maternal intervention might lead to decreased maternal sensitivity and, consequently, less secure attachment over time (Calkins & Fox, 1992; Mangelsdorf et al., 1990).
Crockenberg’s (1986) review of the work examining relations between infant temperament

222 ● S.D. Calkins et al.
and maternal behavior found no consensus, though, on whether difficult temperament was related to less sensitive maternal responding. Indeed, some studies indicated the opposite relation. Crockenberg concluded that this inconsistency could be attributed to, among other things, the distribution of the temperament trait in a given sample and suggested that decreased sensitivity might be reliably observed in a sample with more extreme temperaments. Or, she argued, the negative relation might be observable only when other circumstances, such as low social support or stress, are present (e.g., Crockenberg, 1981). Studies conducted since Crockenberg’s review continue to vary on a number of dimensions including demographic characteristics, whether the sample is selected on the basis of extreme temperament, and whether potential moderators such as social support, stress, marital satisfaction, or maternal depression are assessed. This research also indicated inconsistent relations between maternal behavior and infant temperament (Mangelsdorf et al., 1990; Owens, Shaw, & Vondra, 1998; Pauli-Pott, Mertesacker, Bade, Bauer, & Beckmann, 2000; Seifer et al., 1996; van den Boom, 1994; van den Boom & Hoeksma, 1994). Thompson (1998) noted that more research is needed to understand the complex transactional relations that likely exist between infant and parent characteristics.
In addition to infant characteristics, maternal adjustment and functioning also have been associated with individual differences in maternal behavior. Psychosocial factors that interfere with parents’ awareness of or willingness to respond to infant signals are likely to be related to less sensitive behavior (Belsky, 1984; Donovan, Leavitt, & Walsh, 1998; Fish, Stifter, & Belsky, 1993). For example, higher levels of stress may make it more difficult to engage in effective parenting behaviors (Campbell, Pierce, March, & Ewing, 1991; Pianta & Egeland, 1990; Pianta, Sroufe, & Egeland, 1989).
Another cluster of parent factors that has been associated with early emerging familyinteraction difficulties is the psychological functioning of parents (Alpern & Lyons-Ruth, 1993; Cummings, 1995; Donovan et al., 1998; Seifer et al., 1996). The link between parent psychopathology and child adjustment is believed to be mediated through either the attachment relationship or parenting behavior (Cummings & Davies, 1994; Downey & Coyne, 1990). For example, Alpern and Lyons-Ruth (1993) specifically looked at the timing of depression and found that chronic depression, across the infant and preschool period in particular, affected the likelihood that the child would develop aggressive behavior problems. These researchers point to the interpersonal component of depression and suggest that negative parent mood increases the child’s level of negativity, an affective response that feeds a cycle of parent – child hostility. A related factor that may influence both the mother’s psychological state and her capacity to interact sensitively with her infant is the kind of marital relationship she is experiencing. Relationships characterized by a lack of affection and cohesion can create stress or hostility that may diminish sensitivity to the infant’s signals (Fish et al., 1993).
This overview of factors that might affect maternal behavior leads to several conclusions. First, the emphasis in the infancy work has focused largely on predicting attachment, even though attachment theory explicitly states that sensitive caregiving leads to secure attachment and thus could plausibly be a mediator of a variety of factors, including infant temperament, which might affect attachment security. Second, relatively few studies have examined the relation between infant temperament and maternal behavior in the months prior to the formation of the attachment bond. Third, in studies that do examine infant temperament, there is a lack of specificity concerning which temperament dimensions might affect maternal behavior and why. Fourth, limited data exist that examine both parent factors and infant temperament together as predictors of maternal behavior.
Of particular interest in the present investigation was the role of a temperament dimension that might be especially challenging for caregivers — infant frustration. Frustration has been

Frustrated Infants ● 223
observed in early infancy (Calkins & Fox, 1992; Stifter & Fox, 1990), and by 6 months of age, anger displays are consistently observed. Infants who are easily frustrated likely display frequent bouts of negativity and bids for parental intervention, making it more challenging for caregivers to respond sensitively. In addition, they also may be constrained in the development of appropriate regulatory behaviors that could serve to alleviate their distress (Calkins, 1994; Calkins & Fox, 1992; Calkins & Johnson, 1998), potentially leading to later difficulties with impulsivity, aggression, and other types of externalizing behavior problems. Calkins reported that among a sample of toddlers, mothers of easily frustrated children tended to do things for their children rather than allow the children to become frustrated (Calkins & Johnson, 1998). It is possible that intrusive caregiving in infancy precedes this kind of “preemptive interference” in toddlerhood.
Other studies have examined the dimension of frustration via maternal report of “distress to limitations” (Mangelsdorf et al., 1990; Seifer et al., 1996). In this study, we used both maternal report and a laboratory assessment to identify infants who were easily frustrated across contexts. In addition, we identified these infants from a larger community sample to generate a group of infants who were clearly different from other, less easily frustrated infants. Our assumption was that associations between maternal behavior and infant frustration would be more likely in such a sample (Crockenberg, 1986; van den Boom, 1994; van den Boom & Hoeksma, 1994). Finally, we collected maternal reports of parenting stress, marital adjustment, and psychopathology.
Our primary question concerned whether infant temperament was related to different types of maternal behavior. More specifically, we hypothesized that mothers of easily frustrated infants would show less sensitivity, more intrusiveness, and lower levels of physical stimulation relative to mothers of less frustrated infants. In addition, we addressed the role of maternal report of adjustment as a factor that might affect maternal behavior. Based on past research, we expected that mothers who reported higher levels of parenting stress, more psychiatric symptoms, and/or lower levels of marital satisfaction would show lower sensitivity, more intrusiveness, and less physical stimulation than mothers who reported fewer difficulties. In addition, we hypothesized that less optimal maternal functioning would exacerbate the effects of dealing with a challenging infant temperament. Thus, we expected that mothers of easily frustrated infants who reported more stress, lower marital satisfaction, or more psychiatric symptoms would be less sensitive, more intrusive, and provide less physical stimulation to their infants than mothers of easily frustrated infants who reported fewer difficulties in these domains.
METHOD
Participants
A total of 346 six-month-old infants and their mothers were participants in the larger study. Parents of this larger recruitment sample were contacted through local child-care centers, pediatricians’ offices, direct-mail marketing lists, and county health and human services facilities. After parents indicated that they were interested in participating in the study, they were contacted by telephone to schedule a laboratory assessment. At that time, they were mailed a temperament questionnaire [Infant Behavior Questionnaire (IBQ); Rothbart, 1981] and asked to complete the questionnaire and bring it with them to the laboratory assessment. During the laboratory assessment, the infants were assessed for frustrated temperament using a standard battery. On the basis of parents’ ratings on the IBQ and the laboratory assessment, 162 healthy,

224 ● S.D. Calkins et al.
full-term infants (79 males) and their mothers were selected for the follow-up study (Calkins, Dedmon, Gill, & Johnson, 2002). For the selected sample, average maternal age at the time of the 6-month visit was 29.22 years (SD ϭ 6.17). Families’ mean socioeconomic status, as measured by the Hollingshead (1975) index, was 35.69 (SD ϭ 11.37), and average maternal education was 14.22 years (SD ϭ 2.20). The majority of mothers were married rather than single (79.6 vs. 20.4%), and most participating families were European American (79.6%).
The 162 infants were selected based on both maternal report of distress to limitations and their behavior in a two-episode procedure (barrier task and arm restraint) designed to elicit frustration distress (LAB-TAB; Goldsmith & Rothbart, 1993). Although a third task (maternal prohibition) also was included in the assessment, a number of infants fretted prior to this task (n ϭ 30); therefore, this task was not used to select infants. Following from Rothbart and Derryberry (1981) and Stifter and Fox (1990), measures of frustration distress were operationalized as the latency to fuss (in seconds), the intensity of distress (scored every 10 s on a scale of 0 – 5, with 0 indicating no distress and 5 indicating a full-blown scream or cry), and the duration of fussing in seconds. Latency to fuss was subtracted from the total time of the task (120 s) so that higher scores indicated shorter latencies. A summary score of these distress measures across the two frustration tasks was created by standardizing and summing the measures for the entire sample (Cronbach’s ␣ ϭ .83, M ϭ Ϫ.02, range ϭ Ϫ9.02 – 11.00). High scores on the frustration distress index indicated that the child was very reactive to the frustrating tasks (short latency to cry or fuss, high-intensity fussing, long-duration fussing). For the entire sample (N ϭ 346), the correlation between the laboratory index of frustration and maternal report of distress to limits was modest, though significant, r ϭ .13, p ϭ .01.
To generate groups of infants who could be characterized as easily frustrated versus less frustrated, children scoring at or above the 50th percentile on both the laboratory index of frustration and maternal report of distress to limits were selected for membership in the easily frustrated group while children scoring below the 50th percentile on both the laboratory index and maternal report of distress to limits were selected for the less frustrated group. Of the larger sample, 77 infants met the criteria for the less frustrated group, and 85 met the criteria for the easily frustrated group. The remaining infants were not included in the final (selected) sample that is the focus of this report. For the selected sample, the correlation between the laboratory index of frustration and maternal report of distress to limits was r ϭ .58, p Ͻ .001. Analyses revealed that there was no relation between frustration group and sex of child (47 of 85 infants in the easily frustrated group were females compared with 36 of 77 infants in the less frustrated group), marital status, or child birth order. There was a relation between frustration group status and race, ␹2(1, n ϭ 162) ϭ 6.23, p ϭ .01. More children of African American mothers (n ϭ 33 of 85, 38.8%) were included in the easily frustrated group than in the less frustrated group (n ϭ 16 of 77, 20.8%). There also was a marginal difference between the less frustrated and easily frustrated groups in terms of SES, t (160) ϭ 1.93, p ϭ .06. Easily frustrated children had family Hollingshead SES scores that were lower (M ϭ 34.0) than did families of children in the less frustrated group (M ϭ 37.5). Socioeconomic status and race were not significantly correlated, r(162) ϭ Ϫ.12, n.s.
Interestingly, separate analyses of the two factors (race and SES) with the laboratory observation data and maternal ratings of frustration indicated that African American mothers tended to rate their infants as more frustrated (M ϭ 3.53) than did mothers of European American infants (M ϭ 3.08), t(153) ϭ Ϫ3.20, p Ͻ .01. However, mean levels of infant frustration during the laboratory assessment did not differ significantly for African American (M ϭ 2.52) and European American infants (M ϭ 1.93), t (160) ϭ Ϫ.74, n.s. Thus, African American infants were overrepresented in the easily frustrated group mainly because of maternal ratings of temperament. However, because the frustration groups included only those infants with

Frustrated Infants ● 225
congruent scores on both the laboratory assessment and maternal report of frustration, the correlations between the laboratory assessment of frustration and maternal report of frustration were high and not significantly different for European American infants, r(111) ϭ .60, p Ͻ .001, and African American infants, r(44) ϭ .57, p Ͻ .001 (Fisher’s r to z n.s., p ϭ .80). Because there were no a priori hypotheses regarding either SES or race, these factors were treated as covariates in all analyses.
Procedures
Mother – infant interaction. Although participants were assessed in a number of procedures, the focus of this investigation was on the maternal interaction tasks. These tasks were interspersed with tasks assessing attention, empathy, and the frustration tasks that were used to classify infants as easily frustrated and less frustrated. Mother – infant interaction was observed for a total of 12 to 14 min. The entire procedure was videotaped through a one-way mirror and typically lasted 45 min. Infants were assessed in each procedure only if they were calm and displaying no distress. It was not uncommon for breaks to be taken for feeding or diaper changing. To begin the assessment, infants were changed into a gender-neutral sleeper to minimize possible coding bias and then placed in a high chair that reclined at an angle to provide support for infants not capable of sitting upright. The five mother – infant interaction tasks of interest were conducted as follows:
● Peek-a-boo: While the infant was seated in the reclining high chair, the mother was instructed to play peek-a-boo with her child as she might at home. The mother was provided specific instructions about how to play only if she indicated that she did not know how. In such a case, she was given a demonstration by a research assistant. This task lasted 2 min.
● Puppet play: While the infant was seated in the reclining high chair, the mother was given a large Big Bird puppet and asked to play with the puppet with her child as she might at home. This task lasted 2 min.
● Soothing: Following the frustration task of Arm Restraint, the mother was asked to remove her infant from the infant seat and comfort him or her. This task lasted 1 min. However, if the infant did not soothe, the assessment would stop, and the mother would be given an opportunity to feed or comfort her infant until calm.
● Free play: Mothers and infants were placed on a quilt on the floor (with the infant seated in a supportive “boppy” cushion) with a set of blocks to play with. The mother was instructed to play with her infant as she might at home. This task lasted 3 min.
● Caregiving: As the last task, mothers were asked to move the infant to a changing table and were instructed to remove the neutral sleeper, change the infant’s diaper, wash hands and face, and change the child into his or her own clothes. This task varied in length from approximately 4 to 6 min.
Maternal self-report of parenting stress, psychological functioning, and marital adjustment. Following completion of the laboratory assessment, mothers were asked to take home, complete, and return a packet of the following questionnaires:
Symptom Checklist 90-Revised (SCL-90-R; Derogatis, 1986). The SCL-90-R contains 90 items reflective of commonly encountered adult psychopathology symptoms. Respondents rate each item in terms of how much distress it caused them over the past week, using a 0 (Not at

226 ● S.D. Calkins et al.
all) to 4 (Extremely) scale. The 90 items are then combined to yield a global index of total symptoms as well as nine subscale scores reflective of more specific types of psychopathology (e.g., anxiety). Reliabilities for the scale scores across larger samples of adults are adequate to excellent (Derogatis, 1986; Derogatis & Cleary, 1977). Of interest in the current study was the Total Symptom score, which was generated by the instrument’s computer-scoring program.
Parenting Stress Index-Short Form (PSI; Abidin, 1990). The PSI contains 36 items that assess various sources of stress and are rated on a 5-point scale. These ratings are then combined to generate four stress indices using the instrument’s computer-scoring program: Total Stress, Difficult Child Stress, Parent Domain Stress, and Parent – Child Relationship Stress. These subscales are moderately to highly intercorrelated (rs ϭ .42 – .82). We chose to use the Difficult Child Stress subscale because of its conceptual relevance to infant temperamental frustration. The PSI has shown high reliability and adequate validity (Abidin, 1997).
Dyadic Adjustment Scale (DAS; Spanier, 1976). This measure is designed to assess the quality of the relationship between partners. All mothers who were married or currently in a relationship completed the questionnaire. It consists of 32 items from which four subscale scores assessing various aspects of the relationship can be generated: Dyadic Satisfaction, Dyadic Cohesion, Dyadic Consensus, and Affection Expression. These subscales are all significantly intercorrelated (rs ϭ .21 – .67), and a total marital adjustment score was created by summing the four subscales. The DAS has demonstrated high reliability and adequate criterion and construct validity (Spanier, 1976).
A total of 121 mothers completed the entire questionnaire packet. Comparisons across all study measures of mothers who completed the questionnaires versus those who had incomplete or missing data indicated only one significant difference between the two groups. African American mothers were less likely to complete the questionnaires than European American mothers (p Ͻ .05). Descriptive statistics for the maternal-functioning measures are provided in Table 1.

Measures
Maternal behavior. Maternal and infant behavior was coded across the five tasks using a paper-and-pencil coding system adapted from Fish et al. (1993) and Diener, Mangelsdorf,

TABLE 1. Descriptive Statistics for Psychological Adjustment, Parenting Stress, and Marital Adjustment for Mothers of Easily Frustrated and Less Frustrated Infants

Easily Frustrated

Less Frustrated

Maternal Functioning

N

M

SD

N

M

SD

Psychological Adjustment SCL-90 Total Symptoms Score

71

51.82

11.07

67

53.25

8.78

Parenting Stress

PSI Difficult Child Stress Subscale*

73

24.30

7.25

66

18.83

5.08

Marital Adjustment DAS Total Score

67

106.56

20.25

65

109.54

18.23

* p Ͻ .001.

Frustrated Infants ● 227
McHale, and Frosch (2002). Maternal behavior was coded every 30 s along three dimensions using a scale of 1 (low) to 5 (high) for each dimension. Each dimension was coded separately across the entire task period. Three coders trained together on 10% of the selected sample to establish agreement on use of the coding system. Coders then separately scored at least an additional 10% of the selected sample to achieve reliability, operationalized as agreement within 1 scale point and calculated using the kappa statistic. Coders were reliable at .80 or above for each scale.
Sensitivity (␬ ϭ .86) was coded with respect to the extent to which the parent’s interactions were tuned to the baby’s behavior and included such behaviors as well-timed and synchronous responses, acknowledgment of infant’s affect, appropriate levels of stimulation, soothing, and picking up on the infant’s interests. Higher scores indicated greater sensitivity. Intrusiveness (␬ ϭ .85) referred to the extent to which the parent displayed overcontrolling behavior or was focused on her own agenda and ignored the baby’s cues. Intrusiveness included behaviors such as failing to modulate the pace or intensity of her behavior when the infant withdrew or turned away, appearing to force toys or self on infant, and intrusive physical interactions. Higher scores indicated greater intrusiveness. Physical Stimulation (␬ ϭ .82) referred to the extent to which the mother used physical contact such as holding, touching, and hugging/kissing as well as the extent of stimulating behaviors such as stroking, tickling, or playing with the infant’s limbs in an effort to increase the infant’s interest and arousal in the interaction. Higher scores indicated more physical stimulation. Descriptive statistics for maternal behavior are presented in Table 2.
To reduce the data, cross-episode correlations were examined. These correlations were in the modest to high range for all measures except physical stimulation. Individual measures across episodes were standardized and summed to create overall measures of sensitivity (␣ ϭ

TABLE 2. Descriptive Statistics for Maternal Behavior Across Five Mother–Infant Tasks for Easily Frustrated (n ϭ 85) and Less Frustrated (n ϭ 77) Infants

Maternal Behavior

Easily Frustrated

M

SD

Less Frustrated

M

SD

Total

Min

Max

Sensitivity Peek-a-boo Puppet Play Soothing Free Play Caregiving

3.69

.67

3.88

.56

1.50

5.00

3.66

.60

3.91

.61

2.00

5.00

4.02

.79

4.09

.73

1.50

5.00

3.71

.66

3.67

.68

2.00

5.00

3.70

.85

4.03

.72

1.33

5.00

Intrusiveness Peek-a-boo Puppet Play Soothing Free Play Caregiving

1.61

.79

1.46

.70

1.00

5.00

1.68

.69

1.53

.64

1.00

4.33

1.29

.49

1.24

.40

1.00

3.50

1.59

.60

1.58

.65

1.00

3.50

1.52

.81

1.21

.34

1.00

4.67

Physical Stimulation

Peek-a-boo

1.96

.98

2.14

1.04

1.00

5.00

Puppet Play

2.30

.78

2.38

.87

1.00

4.50

Soothing

2.04

1.04

2.56

1.31

1.00

5.00

Free Play

1.13

.21

1.25

.40

1.00

3.00

Caregiving

1.26

.39

1.37

.49

1.00

3.17

228 ● S.D. Calkins et al.

TABLE 3. Descriptive Statistics for Infant Behavior Across Five Mother–Infant Tasks for Easily Frustrated (n ϭ 85) and Less Frustrated (n ϭ 77) Infants

Easily Frustrated

Less Frustrated

Total

Infant Behavior

M

SD

M

SD

Min

Max

Infant Engagement

Peek-a-boo

3.74

.75

3.83

.69

1.00

5.00

Puppet Play

4.00

.70

3.98

.70

1.67

5.00

Soothing

2.65

.93

2.70

1.04

1.00

5.00

Free Play

3.26

.71

3.09

.59

1.25

5.00

Caregiving

2.61

.88

2.98

.84

1.00

4.33

Infant Affect Peek-a-boo Puppet Play Soothing Free Play Caregiving

3.31

.96

3.81

3.11

.69

3.45

2.70

.79

3.22

2.80

.46

2.90

2.44

.72

3.01

.71

1.00

5.00

.60

1.00

5.00

.56

1.00

4.50

.44

1.00

4.17

.55

1.00

4.25

.72), intrusiveness (␣ ϭ .57), and physical stimulation (␣ ϭ .49). Each summary score was then standardized, and had a mean equal to .00 and a standard deviation equal to 1.00. The internal consistencies for intrusiveness and physical stimulation were relatively low across the five tasks, probably because the different tasks afford somewhat different opportunities for intrusiveness or physical stimulation. We examined alpha reliabilities for different combinations of the tasks. However, aggregating across the five tasks provided the best internal consistency estimates.
Infant behavior. For each of the five tasks, infant behavior was scored along two dimensions using 5-point scales. First, infant affect (␬ ϭ .86) referred to the infant’s level of positive affect (4 – 5) versus negative affect (1 – 2) and was coded with reference to the infant’s facial expressions. The infant’s engagement (␬ ϭ .83) referred to the extent to which the infant was responsive, eager, and willing to engage with the mother (4 – 5) versus the degree to which the infant was unresponsive and disengaged (1 – 2). Descriptive statistics are provided in Table 3.
Cross-episode correlations for both measures were moderate to high, and summary variables were created by standardizing and summing the individual measures. The summary variables had alpha reliabilities of .69 for infant affect and .53 for infant engagement. Each summary score was then standardized, and had a mean equal to .00 and a standard deviation equal to 1.00. As with maternal intrusiveness and physical stimulation, the relatively low internal consistency of infant engagement is probably due to different task constraints. For example, the peek-a-boo and puppet play tasks, during which the mother is actively attempting to engage the infant (and the infant is restrained in an infant seat), elicited higher levels of engagement than the other tasks.
RESULTS
Preliminary analyses were conducted to determine whether sex of infant should be included as a variable in subsequent analyses. Several repeated measures MANOVAs with sex of child as the grouping variable and (a) the three maternal-behavior measures, (b) the two infant-behavior measures, and (c) the SCL-90-R Total Symptoms score, the PSI Difficult Child Stress subscale,

Preparing to load PDF file. please wait...

0 of 0
100%
Mothers’ Interactions With Temperamentally Frustrated Infants