Table of Contents | Previous | Next |
PARENT RESPONSIVENESS AND CHILDREN’S DEVELOPMENTAL OUTCOMES
Jane Atwater, Judith Carta, Jean Ann Summers, And Martha
Staker
University of Kansas
One of the primary goals of Early Head Start (EHS) is to support parents in fostering their children’s development. EHS programs attempt to engage families in a variety of experiences to help them learn how to interact with their children in ways that will stimulate language and cognitive development and that will provide emotional support. Compared to children whose parents face fewer challenges, children from families with multiple risks – such as substance abuse, poverty, and limited education – are more likely to experience negative interactions and non-responsive parenting (Booth, Barnard, Mitchell & Spieker, 1987; Kelly, Morriset, Barnard, Hammond & Booth, 1996). Because they miss out on critically important opportunities to interact with their parents, the children in these families often have slower rates of cognitive and language development in the early years and, thus, often begin school at a disadvantage (Beckwith, 1971; Downey & Coyne, 1990; Furstenberg, Brooks-Gunn & Chase-Lansdale, 1989).
Project EAGLE, an EHS program in Kansas City, Kansas, has identified responsive parent-child interaction as an optimal and essential context for promoting children’s development and for fostering families’ well being. The present analyses were designed to support this program focus by examining parent responsiveness (close involvement and verbal response) as a predictor of early development for children in multi-risk families. In addition, for EHS families, we asked whether the level of engagement in home-based services, which were designed to enhance parent-child relationships, would be related to the level of parents’ responsiveness with their children and to children’s developmental progress.
Method
Participants
The analysis sample includes 74 families randomly assigned to the EHS Program Group and 79 control families. All families in the Program Group were offered home-based EHS services; and, for families with child care needs, the EHS program also provided placement in developmentally appropriate, community-based child care. Comparison families were free to access community services other than those provided by Project EAGLE. A stratified random sample was selected to represent the ethnic diversity of the community: 59 percent African American, 20 percent European American, and 20 percent Hispanic/Latino. When the focus children were born, maternal age ranged from 12 to 39 years (mean = 21.8).
Measures
Parent responsiveness. Parent responsiveness was assessed during 1-hour home-based observations scheduled to occur when children were 8, 14, 18, 24, 30 and 36 months of age, with actual ages ranging from 7 to 42 months. Data were collected during typical, unstructured home activities using the Code for Interactive Recording of Children’s Learning Environments (CIRCLE) (Atwater, Montagna, Creighton, Williams & Hou, 1993). The CIRCLE system is a computer-based direct observation instrument that provides a sequential record of parent and child behaviors, as well as the context of their interactions, and includes 90 specific behavior and context codes. Inter-observer agreement, assessed during six percent of observations, averaged 91.1 percent across all CIRCLE codes (range = 80.7-96.30). For the specific behaviors used in these analyses, mean percentage agreement was 85.4 percent for parent involvement, 87.7 percent for parent verbal responses, and 80.2 percent for child social behavior.
For the present analyses, we identified a subset of behaviors relevant to the concept of parent responsiveness and constructed five measures to reflect different aspects of responsiveness. Each measure represents the percentage of intervals the behavior occurred during observations, averaged across all observations for a given family. First, we identified two composite measures that provide an index of the parent’s general responsiveness with the child.
-
Parent talk to the child–Any parent talk directed specifically to the focus child.
-
Close involvement–Any time the parent is in close proximity and attending to the child.
Second, we selected three specific measures that describe qualitative features of Parent Talk and Close Involvement. These behaviors are subsets of the two general measures described above and were selected because of their relevance to supporting children’s language and learning.
-
Prompt/expansion of child communication–The parent (a) requests a communicative response from the child, or (b) expands or elaborates on the child’s communication.
-
Positive/exuberant response–The parent (a) praises, affirms, or expresses affection to the child, or (b) speaks in a warm, enthusiastic manner to engage the child.
-
Shared activity–The parent is closely involved, and is also participating with the child in an activity (e.g., playing with a toy, making cookies together). The parent is a co-participant rather than simply directing the child.
Children’s development. To track developmental progress, we analyzed children’s cognitive and language growth over time, using hierarchical linear modeling (HLM) (Raudenbush, Bryk, Cheong & Congdon, 2000). We assessed children’s cognitive development with the Bayley Scales of Infant Development. Like the observations, Bayley assessments were scheduled at 8, 14, 18, 24, 30 and 36 months of age, with actual age of administration ranging from 7 to 42 months. The measure of language development was the percentage of time focus children talked (using words or signs) to other children or adults during CIRCLE observations.
For half of the assessment periods (14, 24 and 36 months), Bayley and CIRCLE assessments were conducted during the separate home visits. At other age levels, the two assessments typically were completed during the same visit.
Family risk factors. In previous studies, family risk factors have been associated with a higher risk of developmental delay (e.g., Sameroff & Fiese, 1990). Thus, to control for the possible confound of family risk status in the present analyses, a Cumulative Risk Index was calculated for each family, composed of factors assessed at enrollment: low parent education, parent not employed or in school, single parent status, adolescent mother, large family, minority status, and limited English proficiency.
Parent engagement in the EHS program. Active engagement in EHS services was examined as a possible predictor of parents’ responsiveness with their children. We hypothesized that more highly engaged parents would carry through on program goals by being more responsive during parent-child interactions outside the intervention context. The engagement summary score is a composite of ratings that represent three different aspects of families’ participation in EHS services: the level and consistency of parent participation over time, active interest and involvement during home visits, and follow through on individual program goals between visits. Program staff rated parent engagement after children aged out of the EHS program at age 3.
Results and Discussion
Indicators of Family and Child Risk
Many of the families in our sample experienced multiple risk factors in addition to poverty: 57 percent of mothers had not finished high school, 59 percent were neither employed nor attending school, 14 percent were minors when their children were born, 73 percent were single parents, 13 percent had large families (more than five members), 10 percent did not speak English, and 79 percent were from minority groups that are at increased risk for limited opportunities in American society. On average, families experienced 3.1 risk factors in addition to low income (range = 0-6). Children’s standard scores for cognitive development also indicate the level of risk in this sample (see Table 1). Although we found significant variation in children’s cognitive development, their average scores were approximately one standard deviation below the mean and tended to decline over time.
| Child Age | N | Mean | Range |
|---|---|---|---|
| 8 months | 54 | 90 | 74 - 111 |
| 14 months | 116 | 89.6 | 53 - 110 |
| 18 months | 93 | 82.5 | 55 - 117 |
| 24 months | 79 | 83.1 | 52 - 118 |
| 30 months | 92 | 85.8 | 55 - 120 |
| 36 months | 95 | 85 | 52 - 105 |
What Was the Relationship Between Parent Responsiveness and Children’s Development?
To answer this question, we examined developmental trajectories for children in both the EHS Group and the Comparison Group. As a preliminary step, we first used HLM analyses to determine whether family risk status and group assignment (EHS vs. comparison) were significant predictors of the developmental measures. To control for the number of analyses conducted, results were evaluated at a .01 significance level. In contrast to expectations, risk status and group assignment were not significant predictors of children’s Bayley performance or verbal communication. Thus, those variables were not included in further tests of parent responsiveness as a predictor of child outcomes.
Next, the five measures of responsiveness were examined individually as possible predictors of children’s Bayley performance and verbal communication. HLM analyses revealed that every measure of verbal responsiveness (Parent Talk, Prompt/Expansion, and Positive/Exuberant Response) was a significant predictor of Bayley outcomes (see Table 2). Figure 1 illustrates growth trajectories in cognitive development for children whose parents talked to them more often (the highest quartile for parent talk), compared to those who experienced the lowest level of parent talk (lowest quartile). Although the general measure of Close Involvement was not a significant predictor, Shared Activity was related positively to cognitive outcomes and was the only significant predictor of growth in cognitive development from 8 to 36 months.
Results for children’s verbal communication were even more striking and consistent. Every measure of responsiveness was a significant predictor of both communication outcomes and increases in verbal communication from 8 to 36 months (see Table 3). Figure 2 illustrates developmental trajectories in communication for children who received the highest and lowest levels of parent talk. Thus, when parents were more verbally responsive and involved in their children’s activities, children not only talked more; their use of words also increased more rapidly over time.
| Predictors | df | Intercept | Slope | ||||
|---|---|---|---|---|---|---|---|
| Coefficient | T | P | Coefficient | T | P | ||
| Unconditional | 50 | 113.2 | 94.302 | 0 | 2.78 | 27.983 | 0 |
| Parent Talk to Child | 150 | 0.19 | 3.883 | .000** | -0.002 | -0.405 | - |
| Unconditional | 150 | 113.45 | 73.28 | 0 | 2.65 | 21.598 | 0 |
| Close Involvement | 150 | 0.09 | 2.515 | 0.012 | 0.003 | 0.956 | - |
| Unconditional | 150 | 115.37 | 177.45 | 0 | 2.83 | 48.992 | 0 |
| Prompt/Expansion | 150 | 0.46 | 4.117 | .000** | -0.02 | -1.803 | - |
| Unconditional | 150 | 115.92 | 193.826 | 0 | 2.7 | 55.431 | 0 |
| Positive/Exuberant Response | 150 | 0.87 | 3.703 | .000** | 0.04 | 2.352 | 0.019 |
| Unconditional | 150 | 116.68 | 222.025 | 0 | 2.72 | 66.061 | 0 |
| Shared Activity | 150 | 0.42 | 2.799 | .006* | 0.04 | 3.933 | .000** |
| Note: The unconditional model is the Level 1
HLM model without predictor variables. The intercept represents developmental
level at the midpoint of the age range (23.5 months). The slope represents
developmental change per month. Significance levels of .05 or better
are listed in the table; asterisks indicate those Level 2 predictors
that meet the .01 standard.
* p < .01 |
[D] |
| Predictors | df | Intercept | Slope | ||||
|---|---|---|---|---|---|---|---|
| Coefficient | T | P | Coefficient | T | P | ||
| Unconditional | 151 | 0.42 | 0.616 | - | 0.04 | 0.587 | - |
| Parent Talk to Child | 151 | 0.24 | 6.741 | .000** | 0.02 | 6.002 | .000** |
| Unconditional | 151 | 0.57 | 0.678 | - | 0.11 | 1.598 | - |
| Close Involvement | 151 | 0.12 | 5.121 | .000** | 0.01 | 4.561 | .000** |
| Unconditional | 151 | 2.89 | 5.779 | 0 | 0.29 | 5.578 | 0 |
| Prompt/Expansion | 151 | 0.62 | 4.719 | .000** | 0.04 | 3.053 | .003* |
| Unconditional | 151 | 3.74 | 10.667 | 0 | 0.35 | 9.919 | 0 |
| Positive/Exuberant Response | 151 | 1.33 | 3.95 | .000** | 0.11 | 3.629 | .001** |
| Unconditional | 151 | 4.78 | 14.121 | 0 | 0.41 | 13.261 | 0 |
| Shared Activity | 151 | 0.58 | 3.577 | .001** | 0.08 | 4.713 | .000** |
| Note: The unconditional model is the Level 1
HLM model without predictor variables. The intercept represents developmental
level at the midpoint of the age range (23.5 months). The slope represents
developmental change per month. Significance levels of .05 or better
are listed in the table; asterisks indicate those Level 2 predictors
that meet the .01 standard.
* p < .01 |
|
[D] |
What Was the Relationship of Program Engagement to Parent Responsiveness and Child Development?
Given the results of the previous analyses, our next question was whether responsive parent behavior would be more frequent among those families who had participated most actively and consistently in EHS services. Although group differences were modest, parents with the highest level of program engagement had higher rates of verbal responsiveness with their children (see Table 4). In other words, those parenting behaviors that were most clearly related to child outcomes occurred more frequently in families who were highly engaged in the EHS program. Moreover, engagement in the program was predictive of more positive outcomes in children’s cognitive development and verbal communication and of growth over time in verbal communication (see Table 5). Thus, the present results are consistent with previous evidence of a positive relationship between program engagement and developmental progress at 24 months (Atwater, Carta, Summers & Staker, 2001) and suggest that responsive interactions might be one of the processes that underlie that relationship.
Taken together, these analyses provide empirical support for the EHS program’s emphasis on responsive parent-child interactions as a key component of intervention for children and families who experience multiple risks.
| Responsiveness Measure | Level of Program Engagement | df = 2,70 | ||
|---|---|---|---|---|
| Low (n=21) |
Moderate (n=27) |
High (n=26) |
||
| Parent Talk to Child | 16.37 | 19.81 | 25.71 | F = 4.799** |
| Close Involvement | 43.02 | 38.01 | 46.63 | - |
| Prompt/Expansion | 2.55 | 3.28 | 5 | F = 3.990* |
| Positive/Exuberant Response | 0.8 | 0.98 | 1.91 | F = 3.491* |
| Shared Activities | 0.12 | 0.78 | 0.96 | - |
| Note: The Cumulative Risk Index was entered as
a covariate in these analyses.
* p < .05 |
| Predictors | df | Intercept | Slope | ||||
|---|---|---|---|---|---|---|---|
| Coefficient | T | P | Coefficient | T | P | ||
| Cognitive Development | |||||||
| Unconditional | 75 | 115.66 | 49.673 | 0 | 3.04 | 11.825 | 0 |
| Risk | 75 | -0.96 | -1.961 | 0.05 | -0.04 | -0.761 | - |
| Engagement | 75 | 0.33 | 2.681 | .008* | -0.01 | -0.951 | - |
| Verbal Communication | |||||||
| Unconditional | 71 | -0.67 | -0.336 | - | 0.1 | 0.512 | - |
| Risk | 71 | 0.85 | 1.66 | - | 0.05 | 0.89 | - |
| Engagement | 71 | 0.31 | 3.103 | -.002* | 0.02 | 2.686 | .008* |
| Note: The unconditional model is the Level 1
HLM model without predictor variables. The intercept represents developmental
level at the midpoint of the age range (23.5 months). The slope represents
developmental change per month. Significance levels of .05 or better
are listed in the table; asterisks indicate those Level 2 predictors
that meet the .01 standard.
* p < .01 |
References
Atwater, J., Carta, J., Summers, J. A., & Staker, M. (2001). Relationships between services and child outcomes in an urban Early Head Start program. In Building their futures: How Early Head Start programs are enhancing the lives of infants and toddlers in low-income families. Washington, DC; Administration on Children, Youth and Families.
Atwater, J., Montagna, D., Creighton, M., Williams, R., & Hou, S. (1993). CIRCLE-II: Code for Interactive Recording of Caregiving and Learning Environments - Infancy through Early Childhood. Kansas City, KS, USA: Early Childhood Research Institute on Substance Abuse, Juniper Gardens Children's Project.
Beckwith, L. (1971). Relationships between attributes of
mothers and their infants’ IQ scores.
Child Development, 42, 1083-98.
Booth, C. L., Barnard, K. E., Mitchell, S. K., & Spieker, S. J. (1987). Successful intervention with multiproblem mothers: Effects on the mother-infant relationship. Infant Mental Health Journal, 8, 288-306.
Downey, G., & Coyne, J.C. (1990). Children of depressed
parents: An investigative review.
Psychological Bulletin, 108, 50-76.
Furstenberg, F. F., Jr., Brooks-Gunn, J., & Chase-Lansdale, L. (1989). Teenaged pregnancy and childbearing. American Psychologist, 44, 313-320.
Kelly, J. F., Morriset, C. E., Barnard, K. E., Hammond, M., & Booth, C. L. (1996). The influence of early mother-child interaction of preschool cognitive/linguistic outcomes in a high-social-risk group. Infant Mental Health Journal, 17(4), 1-11.
Raudenbush, S., Bryk, A., Cheong, Y. F., & Congdon, (2000). HLM5: Hierchical linear and non-linear modeling. Lincolnwood, IL: Scientific Software International.
Sameroff, A. J., & Fiese, B. H. (1990). Transactional regulation and early intervention. In S. J. Meisels & J. P. Shonkoff (Eds.), Handbook of early childhood intervention (pp. 119-149). Cambridge: Cambridge University Press.
| Table of Contents | Previous | Next |