The Relationship of Parent-Child Interaction to the ...

TECSE 18:1 5-17 (1998)

5

The Relationship of Parent-Child Interaction to the Effectiveness of Early Intervention Services for At-Risk Children and Children with Disabilities Gerald Mahoney, Children's Hospital Medical Center of Akron, Glenna Boyce, Utah State University Rebecca R. Fewell, University of Miami Donna Spiker, Stanford University, and C. Abigail Wheeden, Children's Hospital Medical Center of Akron

T

he results from four early intervention evaluation studies are described in relationship to their impact on parent-child interaction as measured by the Maternal Behavior Rating Scale. Intervention studies included the Infant Health and Development Program, the Longitudinal Studies of the Effects and Costs of Alternative Types of Early Intervention, the Play and Learning Strategies Program, and the Family-Centered Outcome study. Results from these analyses indicated that intervention effects on child development were unlikely to occur unless mothers modified their style of interacting with their children. Although these studies indicated that mothers of children who participated in intervention modified several different parameters of interactional style, only their level of responsiveness was positively associated with their children's developmental outcomes. These results are discussed in relationship to child-focused and relationship-focused models of early intervention.

No man can reveal to you aught but that which already lies half asleep in the dawning of your knowledge. —Kahlil Gibran During the past 30 years, numerous researchers have reported statistically significant correlations between the manner in which parents interact with their children and the rate of developmental growth children attain during the early years of life. Consistent findings have been cited for a wide range of dyads, including children from low socioeconomic (SES) families (Bradley, 1989), those born prematurely (Beckwith &c Cohen, 1989), and those with moderate to severe developmental disabilities (Brooks-Gunn & Lewis, 1984; Mahoney, Finger, & Powell, 1985). In the 1980s several early intervention demonstration projects incorporated information regarding parent-child interaction—that is, relationshipfocused intervention (Bromwich, 1981; MacDonald, 1989; Mahoney & Powell, 1988) directly into the intervention design. However, only recently has there been a more widespread recognition that research regarding

parent-child relationships may have critical implications for the design of early intervention services (Sandall, 1993; Thorp & McCollum, 1994). Up to this point, most intervention efforts have been conceptualized primarily from a child-focused, directive teaching framework (e.g., McBride & Peterson, 1997; McWilliam, Tocci, & Harbin, 1995; Meisels, Dichtemiller, &c Liaw, 1993). The acceleration of children's development, including cognition, language, and social and motor functioning, has been assumed to be linked to increasing the frequency of children's participation in specially designed learning activities and routines. Early intervention curricula have typically consisted of interrelated instructional activities designed to help children acquire the skills and concepts that are either logically (e.g., task analysis; Bagnato, Neisworth, & Munson, 1989), intuitively (e.g., preschool activities; Russell, 1994), or empirically (e.g., prerequisite developmental skills; Bricker & Cripe, 1992) related to desired developmental outcomes. Often, parents have been encouraged to implement and support intervention activities at home (i.e., as

Address: Gerald Mahoney, Family Child Learning Center, 143 Northwest Ave., Bldg. A Tallmadge, OH 44278.

@pro-ed

Topics in Early Childhood Special Education 18:1

6

supplemental interventionists). However, these home intervention activities were presumed to augment what parents typically did with their children rather than to modify the manner in which they interacted with their children. BARRIERS TO RELATIONSHIP-FOCUSED INTERVENTION A recent study of early intervention practice provided a description of the people professionals interacted with and the content of home-based early intervention sessions (McBride &c Petersen, 1997). Observations of 160 sessions indicated that interventionists primarily worked directly with children, although often in conjunction with another parent or adult: "The content of these interactions focused almost exclusively on the child's development or caretaking needs, and home interventionists spent over half of the time directly teaching the child" (p. 226). These researchers concluded that their data describe a model of child-focused intervention in which the interventionist functions as the key agent of change, and there is little, if any, emphasis on enhancing or supporting parent-child relationships. We maintain that these data are illustrative of a widespread reluctance of early intervention professionals to incorporate relationship-focused concepts directly into early intervention practice. Although this reluctance may be related partly to the fact that early intervention professionals have had limited training to implement relationship-focused procedures (McBride & Petersen, 1997; Mahoney & Wheeden, 1997), we believe that professionals have a number of legitimate concerns that have fueled skepticism about this approach. These include concerns about (a) the generalizability of research findings to populations of children with disabilities; (b) the incompatibility of these findings with the basic premises of traditional child-focused, directive teaching models; and (c) the potential that relationship-focused procedures could devalue cultural, religious, and individual values and beliefs about parenting or childrearing. These concerns are described briefly in the following paragraphs. First, there has been debate about whether research findings from studies of typically developing children generalize to dyads of children with disabilities. For example, findings have shown that high levels of parental directiveness tend to be negatively associated both with children's active participation in social and nonsocial activities and with their level of cognitive and language functioning (Field, 1980). However, some have argued that these findings do not pertain to children with disabilities and that these parents need to use qualitatively different parenting styles to address the unique

needs of their children. These researchers have proposed that the high levels of directiveness observed among parents interacting with low-functioning children may be an appropriate strategy for engaging these children in the types of experiences they need to achieve higher levels of adaptive and developmental functioning (Marfo, 1991; Maurer & Sherrod, 1987; Tannock, 1988). Second, there is a fundamental incompatibility between findings from the parent-child interaction literature and the basic premises of the predominant early intervention curricula (Mahoney, Robinson, & Powell, 1992). Parent-child interaction literature suggests that effective interactions are characterized by high responsiveness and moderate to low directiveness. In contrast, as described previously, most early intervention curricula have tended to emphasize directive, child-focused instruction with domain-specific developmental and functional goals serving as primary teaching targets. Numerous studies have reported that these types of instructional techniques (e.g., modeling, shaping, prompting, selective reinforcement) can be effective at helping young children, including children with severe disabilities, acquire a variety of developmental skills. Accepting concepts from the parent-child literature, with its emphasis on adults interacting responsively with children, would require professionals to repudiate the traditional, directive teaching model. Third, there is a tendency to reject intervention paradigms that may question the skills, competencies, cultural values, or beliefs of parents. A central theme of "family-centered service" is that parents are experts regarding their children, and their cultural folkways and mores must be valued (Baird & Petersen, 1997). Many professionals have shied away from parent-child intervention procedures that encourage parents to modify their interactions with their children in a manner that supports research findings for fear that these patterns conflict with family styles. These professionals view the parent-child literature as posing an ethical dilemma to a field of service that is fundamentally committed to respecting diversity and supporting family choice. Each of these factors pose legitimate barriers to infusing parent-child research findings into early intervention practice. Nonetheless, findings from the parent-child literature are compelling. Although the statistical relationships that have been reported between parent-child interaction and various child outcomes are modest, seldom accounting for more than 20% of variance, these relationships are equivalent to and often even greater than the effect sizes of early intervention services on child development. Furthermore, if a transactional model of development (Sameroff & Friese, 1990) is a valid characterization of the social and environmental factors that contribute to children's development, then it is likely that early intervention efforts might only be

7

Early Intervention Effectiveness

effective at enhancing children's development and functional outcomes if they encourage parents to adopt interactive qualities that research suggests are associated with children's development (Mahoney & Wheeden, 1997). Based on this rationale, it is reasonable to hypothesize that the limited effectiveness of some traditional child-focused, directive intervention models at accelerating developmental growth (e.g., Farran, 1990) is related to their failure to promote the parental interactive qualities that research suggests support the development and well-being of children. Given the relatively small child development effects reported for various early intervention efforts to date (Baumeister & Bacharach, 1996; Farran, 1990), we can no longer ignore the potential implications of the parent-child literature simply because professionals have concerns about the meaning and generalizability of this literature. REEXAMINING EARLY INTERVENTION OUTCOMES The purpose of this investigation was to reexamine the developmental outcomes attained in four independent intervention research studies in relationship to parent-child interaction: the Infant Health and Development Program (IHDP; 1990), the Longitudinal Studies of the Effects and Costs of Alternative Types of Early Intervention (Casto & White, 1993), the Play and Learning Strategies (PALS) program (Fewell & Wheeden, 1998), and the Family-Centered Outcomes study (Mahoney & Bella, in press). These studies were derived from disparate theoretical frameworks and used a variety of intervention content and procedures. They were selected because they shared a common interest in evaluating parent-child interaction. However, only one of these studies involved intervention procedures that were explicitly designed to change patterns of interaction. Thus this analysis enables us to examine the contribution of parent-child interaction to early intervention effectiveness under conditions in which this is both an intended and incidental outcome. Results from this analysis thus help us to examine the merits of the hypothesis that intervention outcomes are fundamentally dependent on parents' adopting the interactive qualities that research suggests are associated with children's development. In each of these studies, mothers' style of interacting with their children was assessed with the Maternal Behavior Rating Scale (MBRS; Mahoney, 1992, see Table 1). The MBRS is a global rating scale consisting of 12 items that have been reported in the child development literature as being significant influences on the development of young children (Mahoney, Powell, & Finger, 1986). Previous research has indicated that raters can achieve acceptable levels of interrater reliability

using this scale (Mahoney et al., 1986); that items on the scale have moderate to high levels of correlation with similarly defined constructs from other parent rating scales (Boyce, et al., 1995); that the scale is sensitive to parental characteristics that are statistically related to children's developmental functioning (Mahoney et al., 1985); and that the scale is also sensitive to changes in parents' styles of interacting with their children that were promoted through early intervention services (Mahoney & Powell, 1988; Mahoney, Wiggers, & Lash, 1996). To analyze mothers' styles of interaction, we created scale scores that were based on a factor analysis of MBRS ratings of 150 dyads (Boyce et al., 1996). This analysis identified three factors: maternal affect/animation (acceptance, enjoyment, expressiveness, inventiveness, and warmth); achievement orientation (achievement, praise); and responsiveness (effectiveness, responsiveness, sensitivity, directiveness, pace). Because the directiveness and pace items were associated negatively with the responsiveness factor, and because directiveness has been a particular focus of the parent-child literature, we divided the responsiveness factor into two subscales: responsiveness (effectiveness, responsiveness, sensitivity) and directiveness (directiveness, pace).

THE INFANT HEALTH AND DEVELOPMENT PROGRAM The IHDP was an eight-site, comprehensive early intervention program for low-birthweight premature infants (Gross, Spiker, &c Haynes, 1997). The intervention consisted of home visits (weekly in the first year, biweekly thereafter), intensive center-based services for children (25-30 hours per week) during Years 2 and 3, and bimonthly parent groups. The home visit component focused on learning activities tailored to the individual needs of the child (Sparling & Lewis, 1985) and assistance for parents in managing self-identified problems (Wasik, 1984). The day-care program consisted of a curriculum of learning activities aimed at enhancing children's intellectual and social competence. The parent groups provided information on health and safety, childrearing, and other parenting concerns. None of these intervention components were designed with the specific intent of modifying mothers' style of interacting with their children, although mothers were given specific instruction about learning activities. The major finding reported for the IHDP was that children in the intervention group showed significantly higher cognitive scores than children in the control or follow-up groups, both at 24 and 36 months (Gross et al., 1997). However, follow-up studies conducted when the children were 5 and 8 years failed to show sustained

8


TABLE 1. The Maternal Behavior Rating Scale Expressiveness

The tendency of the caregiver to express and react emotionally toward the child. It assesses the voice quality to express a range of emotions toward the child. Intensity, animation, and frequency are considered.

Enjoyment

The parent's enjoyment of interacting with the child. Enjoyment is experienced and expressed in response to the child himself—his spontaneous expressions or reactions, or his behavior when interacting with his parent.

Warmth

The demonstration of warmth is a positive attitude revealed to the child through pats, lap-holding, caresses, kisses, hugs, tone of voice, and verbal endearments. Both the overt behavior of the parent and the quality of fondness conveyed are included in this rating. It examines positive affective expression: the frequency and quality of expression of positive feelings by the parent and the parent's show of affection.

Sensitivity to child's interest

The extent to which the parent seems aware of and understands the child's activity or play interests. This item is assessed by the parent's engaging in the child's choice of activity, parent's verbal comments in reference to child's interest, and parent's visual monitoring of child's behavior or activity.

Responsivity

The appropriateness and consistency of the parent's responses to the child's behaviors, such as facial expression, vocalizations, gestures, signs of discomfort, body language, demands, and intentions.

Achievement orientation

The parent's encouragement of sensorimotor and cognitive achievement. This item assesses the amount of stimulation by the parent, which is overtly oriented toward promoting the child's developmental progress whether through play, instruction, training, or sensory stimulation.

Inventiveness

The range of stimulation the parents provide their child; the number of different approaches and types of interactions and the ability to find different things to interest the child, different ways of using toys, combining the toys and inventing games with or without toys. Inventiveness refers to a variety of behaviors that are grouped together and directed toward the child.

Praise (verbal)

The amount of verbal praise given to the child. Examples of verbal praise are "good boy," "that's a girl," and "good job." Praise in the form of smiles, claps, or other expressions of approval are not included unless accompanied by a verbal praise. Praise may be given for compliance, achievement, or the child being himself.

Effectiveness

The parent's ability to engage the child in the play interaction. It determines the extent to which the parent is able to gain the child's attention, cooperation, and participation in a reciprocal exchange characterized by balanced turntaking in play or conversation. The extent to which the parent approves of the child and the child's behavior. Acceptance is measured by the intensity of positive affect expressed toward the child and the frequency of approval expressed, either verbally or nonverbally.

Acceptance Pace

The parent's rate of behavior. The parent's pace is assessed apart from the child's; it is not rated by assessing the extent to which it matches the child's pace.

Directiveness

The frequency and intensity with which the parent requests, commands, hints, or attempts in other manners to direct the child's immediate behavior.

intervention effects (Brooks-Gunn, et al., 1994; McCarton et al., 1997). Analyses were undertaken to determine the mediating influence of mother-child interaction on the developmental outcomes children attained at 24 and 36 months (Wheeden, 1996). Subjects included 298 mother-child dyads from three of the eight IHDP sites. The sample included intervention (n = 116) and follow-up (n = 182) groups. There were no significant group differences on initial status variables (infant birthweight, gender, maternal education, and ethnicity; see Table 2). Videotapes, made when the children were 30 months corrected age for prematurity, permitted an opportunity to assess mother-child interaction during free play. Results indicated that mothers in the intervention group had higher ratings on three of the four maternal interactive style factors (affect, achievement orientation, and responsiveness) than did mothers in the follow-up group (see Table 2). Post hoc analyses of MBRS items indicated that intervention mothers had higher levels of expressiveness, enjoyment, warmth, sensitivity, responsiveness, achievement orientation, inventiveness, effectiveness, and acceptance. There were no group differences on ratings of praise, pace, and directiveness.

Stepwise multiple regressions were computed to examine the relationship of maternal style factors and intervention to children's developmental outcomes, as measured by the Stanford-Binet (Terman & Merrill, 1973) at 36 months and the Bayley (Bayley, 1969) at 24 months. Results indicated that responsiveness accounted for 2 5 % of the variance in children's Stanford-Binet scores, intervention accounted for 4% of the variance, and directiveness was negatively related to Stanford-Binet scores, accounting for 1.6% of the variance. Similarly, responsiveness accounted for 13% of the variance in children's Bayley scores, intervention accounted for 4 % of the variance, and directiveness was negatively related to Bayley scores, accounting for 1.5% of the variance. These results are noteworthy for several reasons. First, even though the IHDP curriculum did not specifically intend to modify mothers' style of interacting with their children, change in maternal style was a highly significant outcome of this intervention. Second, maternal style accounted for six times more of the variance in children's developmental outcomes than did membership in the intervention group itself. Third, the relationship of maternal style to children's development appeared to be

9


Intervention Research Institute at Utah State University. Sixteen research studies that included young children who had disabilities or were at risk for subsequent disabilities due to severe neonatal problems involved different comparisons that focused on three themes: (a) the intensity of intervention (i.e., amount of service per week), (b) the age at which intervention began, and (c) the ways in which parents were involved. Each of these themes was designed to enhance the intensity of children's exposure to developmentally stimulating activities. (See Innocenti, 1996, and White & Boyce, 1993, for details of the individual site studies.) In this investigation, we reexamined data from the six sites that included observations of parent-child interaction. To limit variability in the sample, only children between the ages of 2 and 4 years (Al = 31 months) were included, resulting in a sample of 238 dyads. All children had delays in one or more domains of development. At each site, children were randomly assigned to the ongoing intervention program (typical intervention) or to an intervention program in which services had been enhanced (expanded intervention). The intervention services that participants in the

more consistent with previous empirical findings regarding parent-child interaction (Beckwith & Cohen, 1989) than with the apparent emphasis of the IHDP curriculum. The IHDP curriculum encouraged parents to stimulate children's development by prompting their participation in specially designed activities. This emphasis was reflected by the intervention mothers being more achievement oriented in their interactions with their children. However, achievement orientation was not statistically related to children's development. Only maternal responsiveness, which was an incidental focus of the curriculum, was positively associated with children's developmental outcomes. LONGITUDINAL STUDIES OF EARLY INTERVENTION The Longitudinal Studies of the Effects and Costs of Alternative Types of Early Intervention (Casto & White, 1993) was a controlled, multisite investigation of early intervention effectiveness directed by the Early

TABLE 2. Infant Health and Development Program Group Comparisons of Child, Mother/Family, and Mother Interactive Style Characteristics Follow-up a Variable Child characteristics Gender (% female) Infant birth weight (grams) Mother/family characteristics Mother's age (in years) Mother's education (% high school completion and higher) Mother's ethnicity (% White) Marital status (% married) Maternal Behavior Rating Scale Achievement orientation Achievement Praise Affect/animation Acceptance Enjoyment Expressiveness Inventiveness Warmth Directive Directiveness Pace Responsive Effectiveness Responsiveness Sensitivity Child developmental status Bayley Mental Development Index (24 months) c Stanford-Binet IQ (36 months)

M

SD

M

SD

F

P

51 1798

452

49 1804

410

1.22

.254

1.30

.119

17.12 32.24 2.22 15.64 6.16 7.73 18.92 8.32 11.59 0.93 1.82 0.09 13.11 7.61 10.77 14.41

.000 .000 .137 .000 .014 .006 .000 .004 .001 .320 .179 .320 .000 .006 .001 .000

28.1 30.3

.000 .000

24.5 55

6.1

53 38

23.8 62

5.2

49 43

1.9 2.2 1.7 2.5 2.7 2.8 2.7 2.2 2.2 3.2 3.4 3.0 3.0 2.0 2.8 3.1

.7 .8 .9 .6 .8 .8 .7 .6 .8 .7 .9 .8 .8 .9 .9 1.0

94.1 86.7

18.0 17.5

Note. Stanford-Binet = Stanford-Binet Intelligence Scale (Terman & Merrill, 1973). N = 182. b N = 116. cBayley Scales of Infant Development (Bayley, 1969).

a

Intervention15

2.3 2.7 1.9 2.8 3.0 3.1 3.1 2.4 2.6 3.1 3.2 3.0 3.3 3.3 3.2 3.5

.7 .9 .9 .6 .7 .7 .8 .6 .9 .6 .8 .6 .8 .7 1.0 .9

106.0 98.4

20.1 18.2

10


typical intervention group received were primarily based on a child-focused, directive teaching model and were representative of curricula and service delivery patterns (e.g., center-based and home visitor programs) commonly used during the period the study was conducted (1985-1990). Parent involvement was also typical; the focus was on demonstrations of intervention strategies to accomplish specific curriculum objectives and progress in meeting those objectives. In the expanded intervention group, children and parents received the same intervention services as provided in the typical intervention group, except that services were expanded to allow for investigation of a specific intervention variable (e.g., more intervention service time per week, earlier age for beginning services, parents attending parent classes). Therefore, across all methods for expanding services, children and/or parents had more interaction with interventionists than participants in the typical intervention groups. For this study, we analyzed videotapes made at the first or second reassessment of children and their moth-

ers playing together with a standard set of toys. Child development was assessed concurrently and 1 year later. An intervention (typical vs. expanded) by site MANOVA examined whether any of the expanded interventions had unique effects on the interactional behavior of mothers and the developmental status of children. Because no significant intervention by site interaction effects was observed, data from all sites were combined into two groups (typical and expanded). Typical and expanded intervention groups were similar on most child and maternal demographic characteristics (e.g., child age, mother education, income, etiology); however, they were significantly different on children's gender and mothers' age (see Table 3). Analyses of covariance, using mothers' ages as a covariate, were used to determine the impact of intervention services on mothers' styles of interacting with their children and on children's development. Results from these analyses are displayed in Table 3. Mothers in the expanded intervention group were significantly less

TABLE 3. Longitudinal Studies: Group Comparisons of Child, Mother/Family, and Mother Interactive Style Characteristics Typical intervention 0 Variable Child characteristics Age (in months) Gender (% female) Etiology (% sample) Intraventricular hemorrhage Visual impairment Developmental delay Down syndrome Other Mother/family characteristics Mother's age (in years) Mother's education (years attended) Mother's ethnicity (% White) Family structure (% with two parents) Maternal Behavior Rating Scale Achievement orientation Achievement Praise Affect/animation Acceptance Enjoyment Expressiveness Inventiveness Warmth Directive Directiveness Pace Responsive Effectiveness Responsiveness Sensitivity Child developmental status BDI age equivalent15 BDI age equivalent (1 year later) a

M

SD

31.9 52.1

5.7

Expanded intervention 0 M

SD

F/t score

31.4 39.5

5.7

.67 1.96

.51 .05

.03 .64 .54 .88

P

55.5 10.9 8.4 5.0 20.2

50.4 11.8 10.1 5.0 22.7 28.9 13.1 74.8 74.8

5.5 2.0

30.5 13.2 78.2 75.6

6.1 2.1

-2.21 -.47 -.61 -.15

2.8 3.0 2.6 3.0 2.9 3.0 3.1 2.8 3.0 3.3 3.4 3.1 3.1 3.1 2.9 3.3

1.1 1.2 1.3 .7 .8 .7 .8 .8 .8 .7 .9 .7 .7 .9 .8 .7

2.4 2.6 2.2 3.0 3.0 3.0 3.2 2.7 2.9 3.1 3.2 2.9 3.2 3.2 3.0 3.3

.9 1.1 1.2 .6 .7 .8 .8 .7 .7 .7 .9 .8 .8 .9 .9 .8

8.92 7.70 5.04 .05 .79 .00 .02 1.00 1.01 4.67 4.24 2.85 .56 .49 .51 .39

.003 .006 .026 .824 .375 .974 .900 .318 .317 .032 .041 .093 .457 .487 .477 .535

21.3 29.9

7.9 10.9

21.4 29.8

8.1 10.6

.00 .01

.951 .927

N = 119. b BDI = Battelle Developmental Inventory (Newborg, Stock, &c Wnek, 1984).


achievement oriented and directive than mothers in the typical group. However, there were neither significant group differences in mothers' responsive or affective interactive style nor in children's level of development. Bivariate correlations were computed to examine the relationship between maternal behavioral style and children's current and subsequent development 1 year later. Results indicated that responsiveness was associated positively with the concurrent and follow-up measures of development, r = .27, p = .000, and r = .24, p = .001, respectively, whereas maternal directiveness was associated negatively with these outcomes (concurrent r = -.15, p = .02; follow-up r = -.13 p = .06). Because neither affect/animation nor achievement orientation correlated significantly with child development, these variables were excluded from subsequent analyses. Regression analyses were used to further examine the relationship of maternal interactive style, child age, and intervention group to children's development (measured concurrently and 1 year later). Maternal responsiveness and directiveness were entered into the regression, along with child chronological age and intervention (typical vs. expanded). Results indicated that the regression models accounted for a moderate, yet significant, portion of the variance in children's development (combined model, current year R2 = .23; 1 year later R2 = .12). In both analyses (current year and 1 year later), maternal responsiveness and child age were significant predictors. The betas for responsiveness were current year = .40, significance of T = .00; 1 year later = .25, significance of T = .00. Neither directiveness nor intervention were significant predictors of child development in the analyses for the current year or 1 year later. These data replicate the findings from the original Longitudinal Studies insofar as they indicate that the expanded intervention services had no measurable impact on children's development (Innocenti, 1996; White &c Boyce, 1993). In addition, they extend the findings of the original study in two ways. First, they indicate that mothers who received the expanded interventions showed lower levels of achievement orientation and directiveness. Second, they indicate that only certain aspects of maternal behavioral style were related to children's development. Neither achievement orientation nor directiveness appeared to be related to development. However, responsiveness, a feature of interactive style not affected by the expanded intervention procedures, related consistently to children's development. PLAY AND LEARNING STRATEGIES PROGRAM The Play and Learning Strategies (PALS) program was a 3-month (24 sessions, 30 minutes each) pilot inter-

11

vention program designed to help adolescent mothers acquire a more responsive, less directive style of interaction with their children. A total of 100 volunteer mothers were assigned to either the intervention or contrast groups. The intervention focused on early developmental skills, including play, language, cognition, and motherchild interaction strategies (e.g., turn taking, interactive match). Mothers were also taught strategies to handle problem behavior, give positive reinforcement, and meet child needs. Topics were addressed through the use of contrasting examples on videotapes, hands-on practice with their children, explicit feedback on their interactive styles, and discussions. Each class was limited to 8 mothers; children attend 25% of the time. Mothers were provided a simply written curriculum covering the content emphasized. Mothers in both groups continued to participate in required parenting classes. These classes met daily for 50 minutes, focused on routine childcare (e.g., feeding, bathing, diapering), and rarely included sessions with children. No other intervention was provided to the contrast group. All participants were videotaped interacting with their children in a free-play situation. Participants were excluded from the analysis if initial child assessments had questionable reliability due to the young age of the child (e.g., less than 4 months), or if the mother attended fewer than nine sessions. This resulted in 21 intervention dyads that were matched with dyads in the contrast group. Demographic characteristics and results are presented in Table 4. There were no initial differences between the two groups. Consistent with the PALS curriculum, intervention mothers had less directive, more responsive styles of interacting with their children than did mothers in the contrast group. Although the difference on the responsiveness factor score did not reach significance (p - .11), the difference as measured by the responsiveness item was significant (p = .018). A stepwise multiple regression was used to examine the contributions of children's development at pretest and mother's style of interaction at posttest to the developmental status of both child groups at the end of the program. Only maternal responsiveness entered as a significant predictor, accounting for 10% of the variance in children's Developmental Activities Screening Inventory-II (DASI-II) scores. These results suggest that PALS was effective both at modifying the manner in which mothers interacted with their children and at enhancing children's developmental functioning. Moreover, the results of the regression analysis indicated that mothers' responsiveness to their children, a major intended focus of the program, was more predictive of children's development at the end of the program than was children's development at the beginning of the program. Although responsiveness accounted for only a modest amount of the variability in

12


TABLE 4. Play and Learning Strategies Program: Group Comparisons of Child, Mother/Family and Mother Interactive Style Characteristics Contrast0 Demographic characteristics Child characteristics Age (in months) Gender (% female) Mother/family characteristics15 Mother's age (in years) Mother's education (grade level) Mother's PPVT IQb Number of children Intervention sessions attended Maternal Behavior Rating Scale Achievement orientation Achievement Praise Affect/animation Acceptance Enjoyment Expressiveness Inventiveness Warmth Directive Directiveness Pace Responsive Effectiveness Responsiveness Sensitivity Child developmental status DASIIQ (Pretest)0 DASI IQ (Posttest)c

Intervention0

M

SD

M

SD

11.5 47

7.4

11.4 57

7.7

17.1 10.7 71.6 1.3 0

.9 1.2 17.7 .5 0

17.1 10.2 71.7 1.1 14.3

1.3 1.1 16.8 .3 3.9

1.8 2.2 1.3 2.9 2.9 3.1 3.1 2.5 2.9 3.2 3.2 3.1 3.3 3.3 3.1 3.5

.9 1.2 1.0 .6 .5 .7 .7 .6 1.0 .4 .4 .4 .4 .5 .5 .5

1.8 2.2 1.4 2.8 3.0 2.9 2.8 2.8 2.6 2.8 2.8 2.8 3.6 3.4 3.6 3.7

106.7 95.7

16.9 9.1

97.5 104.6

0.37

.548

2.50

.122

.6 .9 .6 .5 .6 .6 .7 .7 .7 .5 .6 .5 .6 .6 .8 .9

0.00 0.02 0.04 0.43 0.32 1.45 2.36 1.40 1.24 7.21 5.13 5.98 2.64 0.82 6.14 0.76

1.00 .885 .848 .517 .573 .235 .133 .244 .273 .010 .029 .019 .110 .776 .018 .387

23.5 13.2

2.14 6.48

.151 .015

a

N = 21. bPPVT = Peabody Picture Vocabulary Test (Dunn & Dunn, 1981). CDASI = Development Activities Screening Inventory-II (Fewell & Langley, 1984).

children's development, the effect size of the intervention on developmental functioning (eta squared = .14) was nearly identical to the effect size of the intervention on maternal responsiveness as measured by the individual rating item (eta squared = .13). These preliminary results suggest that changes in mothers' responsiveness contributed to the enhanced rates of child development promoted through PALS. FAMILY-CENTERED OUTCOMES STUDY The Family-Centered Outcomes study was conducted to examine the impact of family-centered early intervention services on parents and their children. The movement toward family-centered intervention has been based partly on two concepts. The first is that since parents are the primary influences on their children's development, the effectiveness of intervention ultimately depends on parents becoming more active partners in the intervention process. The second is that families of children with disabilities are confronted with a number of challenges that make it difficult for them either to engage in the

early intervention process or carry out routine childcare responsibilities. Thus, family-centered intervention practices require not only changes in the attitudes of professionals toward parents, but also a fundamental shift in the focus of early intervention from working directly and exclusively with the child to collaborating with families by providing an array of supports responsive to their needs and priorities. This shift was expected to enhance intervention by enabling parents to become more effective at promoting the development and well-being of their children (Dunst, Trivette, & Deale, 1988). The Family-Centered Outcomes study was a 12month longitudinal, field-based investigation designed to test some of the assumptions underlying this familycentered intervention model. Specifically, this study attempted to determine whether services that reflected two of the key components of family-centered practice— comprehensive scope of family support services and services that corresponded to the desire or needs of families—would enable parents to interact more effectively with their children and thus be more effective at promoting children's developmental growth.

13


Participants were 47 mother-child dyads who were currently participating in early intervention programs. Because this sample came from 36 programs, the services that children and families received were assumed to reflect a wide range of services typically associated with early intervention. Most participants likely received developmental activities individually tailored to the unique characteristics of their children and designed to accelerate all domains of child functioning. The family-centered orientation of services was assessed by asking mothers to complete the Family Focused Intervention Scale (FFIS; Mahoney, O'Sullivan, & Dennebaum, 1990) at the midpoint of the study. This scale assesses the frequency or intensity of services that families receive in the domains of child information, family educational activities, systems engagement, personal family assistance, and resource assistance. Cluster

analyses were used to sort the sample into two highly differentiated groupings. The first grouping was based on the comprehensiveness of services that mothers reported receiving. Across the five subscales of the FFIS, families classified into the low family service group received an average of 43% fewer services than did families in the high family service group. The second grouping classified families into "responsive to needs" clusters based on the discrepancies between services parents reported as being important and the services they actually received. This resulted in a highly responsive service group that reported almost no discrepancies between their ratings of the importance of services and types of services they received, and a low responsive group that had discrepancies in which mothers' ratings of the importance of services were, on an average, twice as great as their ratings of services they received.

TABLE 5. Family-Centered Outcomes Study: Group Comparisons of Child, Mother/Family, and Mother Interactive Style Characteristics

Time 1a Variable Child characteristics Age (in months) Gender (% female) Etiology Cerebral palsy (% of sample) Down syndrome (% of sample) Premature (% of sample) Other (% of sample) Mother/family characteristics Mother's age (in years) Mother's education (years attended) Mother's ethnicity (% White) Marital status (% married) Maternal Behavior Rating Scale Achievement orientation Achievement Praise Affect/animation Acceptance Enjoyment Expressiveness Inventiveness Warmth Directive Directiveness Pace Responsive Effectiveness Responsiveness Sensitivity Child developmental status Bayley Mental Development Ageb REEL Expressive Language Agec REEL Receptive Language Agec REEL Combined Language Agec Vineland Social Maturity Age5 a

Time 2°

M

SD

17.3 63.8

.3

M

SD

F

P

.8 1.0 1.0 .7 .7 .8 .9 .8 .8 .7 .8

0.23 1.49 0.06 2.49 0.72 5.05 5.53 0.16 1.54 2.12 1.00 2.58 0.95 3.64 1.03 0.07

.630 .229 .811 .120 .399 .029 .023 .695 .221 .150 .322 .115 .340 .062 .315 .796

12.5 38.3 11.1 38.3 32.3 14.5 93.6 85.1

5.9 2.4

2.8 3.2 2.5 3.0 2.9 3.1 3.3 2.6 3.0 3.5 3.7 3.4 3.0 2.8 2.9 3.2

.8 .9 1.1 .7 .7 .9 1.1 .7 1.0 .8 1.0 .8 .8 .9 .9 .9

2.9 3.4 2.4 2.8 2.8 2.8 3.0 2.6 2.8 3.4 3.5 3.2 3.1 3.0 3.1 3.3

.8 .7 1.0 .9

10.8 10.8 11.8 11.3 11.5

4.7 4.9 4.6 4.6 3.1

18.5 18.6 21.5 20.0 17.7

7.3 8.6 8.5 8.2 5.8

139.8 78.2 112.4 115.0 103.4

.000 .000 .000 .000 .000

N = 47. bBayley Scales of Infant Development (Bayley, 1969). CREEL = Receptive Expressive Emergent Language Scale (Bzoch & League, 1991). Vineland Adaptive Behavior Scales (Sparrow, Balla, & Cicchetti, 1984).

d

14


Assessments of parent-child interaction and child and family functioning were collected at the beginning and end of a 12-month period (see Table 5). Pre-/postcomparisons indicated that over the course of intervention there were no changes in the interactive style of mothers. Changes in children's developmental ages were significant, yet their rates of development during the study appeared to be equivalent to their rates of development prior to this study. That is, children's proportional change indices (Wolery, 1983) for the three developmental measures used in this study—Bayley Mental Development Index (Bayley, 1969), Receptive Expressive Emergent Language Scale (Bzoch & League, 1991), and Vineland Social Maturity Test (Sparrow, Balla, & Cicchetti, 1984)—ranged from an average of 69% to 115% during intervention. Furthermore, analyses of these data as a function of the family-centered orientation of services indicated that neither of the parameters of familycentered service assessed in this study had a significant impact on encouraging mothers to modify their style of interacting with their children or on enhancing children's rate of developmental growth. The Family-Centered Outcomes study thus presents another example of an intervention evaluation in which there was no observable acceleration of children's development over a 1-year period. Although it is assumed that parents will become more engaged in the intervention process and/or more effective at interacting with their children when they are provided a wide array of intervention supports, our assessments of parent-child interaction produced no evidence to support this assumption. Over the course of this study, there were no systematic changes in parents' styles of interacting with their children that were related to the type of family-centered services they received. DISCUSSION At the outset of this article, we discussed two contrasting models for conducting early intervention. The child-focused, directive teaching model emphasizes the acquisition of specific skills or developmental milestones offered on the IFSP and derived from tests and/or curricula. The relationship-focused model assumes that intervention should encourage and support the patterns of parenting and caregiving that developmental research indicates are associated with the optimal development of children. Because the child-focused model emphasizes the unique instructional qualities of specific activities and experiences, it encourages adults to guide or direct children to engage in developmentally stimulating activities. In contrast, the relationship-focused model encourages adults to primarily support and sustain children's involvement in activities they have selected.

The four studies we presented provide a rich database that sheds light on the relative merits of these two models. If the child-focused model is an appropriate framework for conducting early intervention, then the effectiveness of early intervention should be related to the intensity or frequency of services children receive. Insofar as parental involvement is associated with intervention effectiveness, parents' patterns of involvement should be consistent with the characteristics of adult interaction associated with the child-focused, directive teaching model (i.e., achievement orientation, directiveness). On the other hand, if the relationship-focused model is an appropriate framework for early intervention, early intervention effectiveness should be dependent on parental involvement, particularly as characterized by their responsiveness with their children, more than on the intensity or frequency of services. Two of the projects described in this paper were primarily based on a child-focused model: the Longitudinal Studies and the IHDP. A primary characteristic of the design in both of these studies was intensity of service (i.e., amount of services provided). In the Longitudinal Studies, each of the expanded interventions reflected an increased frequency of service. In the IHDP, intensity was reflected by the treatment group receiving a comprehensive and systematic array of home- and center-based services that reflected the upper limits of services that low-birthweight children might normally receive (Baumeister & Bacharach, 1996). Results from the Longitudinal Studies provide no evidence that increased intensity (i.e., number of services received) had a significant impact on intervention effectiveness. Although the IHDP appeared to have an impact on enhancing children's development at least for the short term, it is debatable whether these effects were a by-product of the intensity or frequency of services children received. Three studies have been published arguing that the intensity of IHDP services contributed to its effectiveness (Blair, Ramey, &C Hardin, 1995; Ramey et al., 1992; Sparling et al., 1991). However, none of these studies analyzed indices of intensity in relationship to parent-child interaction styles. We believe that our findings indicating that maternal responsiveness—an unplanned outcome of the IHDP—was the predominant predictor of children's development and raises the possibility that some of the more intensive child-focused services of this intervention that did not involve parents (e.g., the 25- to 30-hour-per-week childcare program) may not have been the major factors contributing to the developmental effects of this program. This conclusion is compatible with another reanalysis of the IHDP data conducted by Baumeister and Bacharach (1996), who argue that a familial variable (i.e., maternal IQ) was the major contributor to the IHDP outcomes. The other two evaluations reported in this article, the Family-Centered Outcomes study and the PALS pro-

15


gram, examined intervention efforts that focused explicitly on parents. Both studies were based on intervention concepts derived from the transactional model of development. However, the Family-Centered Outcomes study emphasized factors related to a wide array of supports responsive to family priorities and needs, but not systematically focused on the manner in which parents interacted with their children. In this evaluation we found no evidence either that parents modified their style of interacting with their children or that children accelerated their rate of development. In contrast, the PALS program worked with mothers by addressing interactional processes directly related to their children's development. This study showed significant intervention effects not only on mothers' styles of interaction, but also on their children's development. As in the IHDP, these changes in child development were directly related to mothers' levels of responsiveness. The contrasting results from these parent-directed studies suggest that the critical determinant of intervention effectiveness may not simply be that parents are involved, or that the intervention focuses on family needs, but rather that parents are encouraged and supported to engage in highly responsive interactions with their children. Parent-professional collaboration, responding to the concerns and priorities of families, and other family support activities that are hallmarks of family-centered intervention, are undoubtedly crucial for engaging parents in the intervention process. Yet our findings suggest that these types of activities may be ineffective at enhancing children's development, unless they also encourage parents to engage in more responsive interactions with their children. In general, the pattern of findings reported in these four intervention studies provide relatively consistent support for assumptions underlying the relationshipfocused intervention model—that intervention effectiveness is fundamentally dependent on parent-child interaction. Our findings suggest that the impact of interventions on children's development was directly related to their effectiveness at supporting and encouraging parents to engage in responsive interactions with their children, regardless of whether this was an intended or unintended outcome. This was indicated by two findings. First, intervention effects on child development did not appear to occur unless mothers modified their style of interacting with their children. Second, the one parameter of maternal style that was consistently associated with significant child outcomes was responsiveness—the very same interactive feature indicated by child development research to be a significant influence on child development. In contrast, our findings provide little, if any, support for the child-focused, directive teaching model. That is, we found little evidence either that intensity or frequency of early intervention services was a major predic-

tor of effectiveness and/or that the features of interactive style associated with the child-focused, directive teaching (e.g., achievement orientation and directiveness) contributed to intervention effectiveness. Although these results point to the centrality of parent-child interaction to the early intervention process, they do not lessen the seriousness of some of the concerns raised about the relationship-focused model discussed in the introduction. However, we believe that many of these concerns arise from our limited understanding of the interactional phenomena being discovered through parent-child research. There is a need for future research to continue to explore the relationship between parent-child phenomena and theories of learning and development. We also need to determine how parent-child phenomena affect the development of lowincidence populations, such as children with autism, as well as better understand how interactional principles apply to minority and nontraditional families. Finally, in a previous paper we raised concerns about translating findings from the parent-child interaction literature into clinical assessment procedures (Mahoney, Spiker, & Boyce, 1996). Despite the results reported in the present article, we maintain serious reservations about using constructs derived from the parent-child literature as bases for making evaluative judgments about the childrearing skills of parents. The true value of the information derived from the parent-child literature does not depend on our developing another classification taxonomy. Rather, the value of this literature is more likely to emerge from serious inquiries focused on discovering both the potential and limitations of parent-child processes as a basis for developing and evaluating intervention strategies and procedures. • REFERENCES Bagnato, S., Neisworth, J., & Munson, S. (1989). Linking developmental assessment and early intervention: Curriculum-based prescriptions. Rockville, MD: Aspen. Baird, S., & Petersen, J. (1997). Seeking a comfortable fit between familycentered philosophy and early intervention practice: Time for a paradigm shift. Topics in Early Childhood Special Education, 17(2), 139-164. Baumeister, A. A. & Bacharach, V. (1996). A critical analysis of the Infant Health and Development Program. Intelligence, 23, 79-104. Bayley, N. (1969). Bayley scales of infant development. New York: Psychological Corp. Beckwith, L., & Cohen, S. E. (1989). Maternal responsiveness with preterm infants and later competency. In M. H. Bornstein (Ed.), Maternal responsiveness: Characteristics and consequences. New Directions for Child Development, 43, 75-87. Blair, C , Ramey, C. T., & Hardin, J. M. (1995). Early intervention for low birthweight, premature infants and intellectual development. American Journal on Mental Retardation, 99(5), 542-554. Boyce, G. C , Casto, G., Mahoney, G., Marfo, K., Wilfong-Grush, E., & Spiker, D. (1995, December). A comparison of four observational instruments for coding interactions between parents and their children with disabilities. A poster presented at the Zero to Three 10th National Training Institute, Adanta, GA.

16

Boyce, G. C., Marfo, K., Mahoney, G., Spiker, D., Price, C., 6c Taylor, M. J. (1996, March). Parent-child interaction in dyads with children at risk for developmental delays: A factor analytic study. Poster presented at the Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Gatlinburg, TN. Bradley, R. (1989). HOME measurement of maternal responsiveness. In M. H. Bornstein (Ed.), Maternal responsiveness: Characteristics and consequences. New Directions for Child Development, 43, 63-74. Bricker, D., 6c Cripe, J. J. W. (1992). An activity-based approach to early intervention. Baltimore: Brookes. Bromwich, R. M. (1981). Working with parents and infants: An interactional approach. Baltimore: University Park Press. Brooks-Gunn, J., 6c Lewis, M. (1984). Maternal responsivity in interactions with handicapped infants. Child Development, 55, 782-793. Brooks-Gunn, J., McCarton, C. M., Casey, P. H., McCormick, M. C , Bauer, C. R., Bernbaum, J. C , Tyson, J., Swanson, M., Bennett, F. C , Scott, D. T., Tonascia, I., 6c Meinert, C. L. (1994). Early intervention in low birthweight, premature infants. Journal of the American Medical Association, 272, 1257-1262. Bzoch, E. R., 6c League, R. (1991). Receptive-expressive emergent language scale (2nd ed.). Austin, TX: PRO-ED. Casto, G., & White, K. R. (1993). Longitudinal studies of alternative types of early intervention: Rationale and design. Early Education and Development, 4, 224-237. Dunn, L. M., & Dunn, L. M. (1981). Peabody picture vocabulary test-Revised. Circle Pines, MN: American Guidance Systems. Dunst, C. J., Trivette, C. M., 6c Deale, A. G. (1988). Enabling and empowering families: Principles and guidelines for practice. Cambridge, MA: Brookline Books. Farran, D. (1990). Effects of intervention with disabled and disadvantaged children: A decade review. In S. J. Meisels 6c J. P. Shonkoff (Eds.), Handbook of early childhood intervention (pp. 501-539). New York: Cambridge University Press. Fewell, R. R., 6c Langley, M. B. (1984). Developmental activities screening inventory-H. Austin, TX: PRO-ED. Fewell, R., & Wheeden, C. A. (1998). A pilot study of intervention with adolescent mothers and their children: A preliminary examination of child outcomes. Topics in Early Childhood Special Education, 18(1), IS-25 Field, T. M. (1980). Interactions of preterm and term infants with their lower- and middle-class teenagers and adult mothers. In T. M. Field, S. Goldberg, D. Stem, & A. M. Sostek (Eds.), High-risk infants and children: Adult and peer interactions (pp. 113-132). New York: Academic Press. Gross, R. T , Spiker, D., 6c Haynes, C. (Eds.). (1997). Helping low birthweight, premature babies: The Infant Health and Development Program. Stanford, CA: Stanford University Press. Infant Health and Development Program. (1990). Enhancing the outcomes of low birthweight, premature infants: A multisite randomized trial. Journal of the American Medical Association, 263, 3035-3042. Innocenti, M. S. (1996). 1991-1996 Final report of the effects of early intervention for children with disabilities (Contract No. HS90010001). Logan: Utah State University, Early Intervention Research Institute. McBride, S. L. 6c Peterson, C. (1997). Home-based intervention with families of children with disabilities: Who is doing what? Topics in Early Childhood Special Education, 17(2), 209-233. McCarton, C. M., Brooks-Gunn, J., Wallace, I., Bauer, C. R., Bennett, F. C , Bernbaum, J. C , Broyles, S., Casey, P. H., McCormick, M. C , Scott, D. T., Tyson, J., Tonascia I., & Meinert, C. L. (1997). Results at age 8 years for early intervention in low birthweight, premature infants. Journal of the American Medical Association, 277, 126-132.


McWilliam, R. A., Tocci, L., 6c Harbin, G. (1995). Services are childoriented and families like it that way—But why? (Findings: Early Childhood Research Institute on Service Utilization). Chapel Hill: University of North Carolina. MacDonald, J. D. (1989). Becoming partners with children: From play to conversation. San Antonio, TX: Special Press. Mahoney, G. (1992). Maternal behavior rating scale (revised). Tallmadge, OH: Family Child Learning Center. Mahoney, G., 6c Bella, J. (in press). Family-centered outcomes: A field based study. Topics in Early Childhood Special Education. Mahoney, G. J., Finger, I., 6c Powell, A. (1985). Relationship of maternal behavioral style to the development of organically impaired mentally retarded infants. American Journal of Mental Deficiency, 90, 296-302. Mahoney, G., O'Sullivan, P., &C Dennebaum, J. (1990). Maternal perceptions of early intervention services: A scale for assessing familyfocused intervention. Topics in Early Childhood Special Education 10(1), 1-15. Mahoney, G., 6c Powell, A. (1988). Modifying parent-child interaction: Enhancing the development of handicapped children. Journal of Special Education, 22, 82-96. Mahoney, G., Powell, A., 6c Finger, I. (1986). The maternal behavior rating scale. Topics in Early Childhood Special Education, 6, 44—56. Mahoney, G., Robinson, C , 6c Powell, A. (1992). Focusing on parentchild interaction: The bridge to developmentally appropriate practices. Topics in Early Childhood Special Education, 12(1), 105-120. Mahoney, G., Spiker, D., 6c Boyce, G. (1996). Clinical assessments of parent-child interaction: Are professionals ready to implement this practice? Topics in Early Childhood Special Education, 16(1), 26-50. Mahoney, G., 6c Wheeden, C. A. (1997). Parent-child interaction: The foundation for family-centered early intervention practice: A response to Baird and Peterson. Topics in Early Childhood Special Education, 17(2), 165-184. Mahoney, G., Wiggers, B., 6c Lash, S. (1996). Using a relationship focused intervention program to enhance father involvement. Infant-Toddler Intervention, 6(4), 295-308. Marfo, K. (1991). The maternal directiveness theme in mother-child interaction research: Implications for early intervention. In K. Marfo (Ed.), Early intervention in transition: Current perspectives on programs for handicapped infants (pp. 177-203). New York: Praeger. Maurer, H., 6c Sherrod, K. B. (1987). Context of directives given to young children with Down syndrome and nonretarded children: Development over two years. American Journal of Mental Deficiency, 91, 579-590. Meisels, S. J., Dichtemiller, M., 6c Liaw, F. (1993). A multidimensional analysis of early childhood intervention programs. In C. H. Zeanah (Ed.), Handbook of infant mental health (pp. 361-385). New York: Guilford Press. Newborg, J., Stock, J. R., 6c Wnek, L. (1984). Battelle developmental inventory. Allen, TX: DLM Teaching Resources. Ramey, C , Bryant, D. M., Wasik, B. H., Sparling, J. J., Fendt, K. H., 6c Lavange, L. M. (1992). The Infant Health and Development Program for low birthweight premature infants: Program elements, family participation and child intelligence. Pediatrics, 3, 454-465. Russell, K. M. (1994). Kids are kids: An activity based curriculum guide. Sevierville, TN: International Network Publishers. Sameroff, A. J., 6c Friese, B. (1990). Transactional regulation and early intervention. In S. Meisels 6c J. Shonkoff (Eds.), Handbook of early childhood intervention (pp. 119-159). New York: Cambridge University Press. Sandall, S. R. (1993). Curricula for early intervention. In W. Brown, S. K. Thurman, 6c L. F. Pearl (Eds.), Family-centered early intervention with infants and toddlers: Innovative cross-disciplinary approaches (pp. 129-172). Baltimore: Brookes.


Sparling, J., & Lewis, I. (1985). Partners for learning. Winston-Salem, NC: Caplan Press. Sparling, J. S., Lewis, I., Ramey, C. T., Wasik, B. H., Bryant, D. M., & Lavange, L. M. (1991). Partners: A curriculum to help premature, low birth weight infants get off to a good start. Topics in Early Childhood Special Education, 11(1), 36-55. Sparrow, S., Balla, D., & Cicchetti, D. (1984). Vineland Adaptive Behavior Scales. Circle Pines, MN: American Guidance Service, Inc. Tannock, R. (1988). Mothers' directiveness in their interactions with their children with and without Down syndrome. American Journal on Mental Retardation, 93, 154-165 Terman, L. M., & Merrill, M. A. (1973). Stanford-Binet intelligence scale: Manual for the third revision form L-M. Boston: Houghton Mifflin. Thorp, E. K., & McCollum, J. A. (1994). Defining the infancy specialization in early childhood special education. In L. J. Johnson, R. J. Gallagher, M. J. LaMontague, J. B. Jordon, J. J. Gallagher, P. L. Hutinger, & M. B. Karnes (Eds.), Meeting early intervention challenges: Issues from birth to three (pp. 167-183). Baltimore: Brookes. Wasik, B. (1984). Coping with parenting through effective problem-solving: A handbook for professionals. Chapel Hill, NC: Frank Porter Graham Child Development Center. Wheeden, C. A. (1996). The effects of a comprehensive early intervention program on the quality of mother-child interaction with low birthweight, premature infants. Unpublished doctoral dissertation, University of Miami, Miami, FL. White, K. R., & Boyce, G. C. (Eds.). (1993). Comparative evaluations of early intervention alternatives [Special issue]. Early Education and Development, 4(4). Wolery, M. (1983). Proportional change index: An alternative for comparing child change data. Exceptional Children, 50, 167-170.

17