Investigation of the effects of a model of physical therapy on mother-child interactions and the motor behaviors of children with motor delay.

Parent-child interaction is an important

component of a family-focused approach to early

intervention.[1,2] The parent-child relationship

contributes to the child's development.[3-9] In

addition, parents and practitioners reported

enhancement of parent-child relationship as an expected family

outcome of early intervention.[10] In a meta-analysis of the

efficacy of early intervention for children with

disabilities, Shonkoff and Hauser-Cram[11] concluded that

programs that targeted their efforts at both the child and

the family appeared to be most effective. The positive

influence of relationship-focused early intervention on

mother-child interaction and child development has

long been recognized.[12-25] A relationship-focused model

of early intervention strives to enhance positive

parent-child interaction and emphasizes that these interactions

provide the foundation for a satisfying relationship as

well as for the development of the child and the

parents.[12,14] Intervention programs are designed to guide

parents in understanding and responding to their

child's behaviors, interests, and needs. Although positive

parent-child interaction is considered part of

state-of-the-art pediatric physical therapy practice for young

children with motor disabilities,[1,2] the effectiveness of

physical therapy in promoting parent-child interaction

has not been investigated.[26,27]

There is evidence that a child's motor skill level

influences mother-child interactions, particularly for

children with motor disabilities.[28-34] As a child's motor

developmental age increases, physical contact decreases,

maternal warmth and friendliness increase, and the

child's responsiveness and clarity of cues increase during

mother-child interactions.[28-30] Children with motor

disabilities have been reported as being less active

participants during interactions with their mothers.[28,29,31-33]

Mothers of children with disabilities have consistently

been reported to be more directive; however, mothers'

activity level, responsiveness, and affect have been

variable.[28,29,31-34]

Maternal directiveness refers to the mother's use of verbal

and nonverbal controls to lead her child's behaviors

during interaction. It is a complex phenomenon and

may have both positive and negative implications.[35-38]

Maternal directiveness may be related to parent training

models that emphasize the need for parents to teach

their children developmental skills. Maternal

directiveness, if provided with sensitivity to the child's needs, may

help to foster the child's participation, interest, and

development.[38] If maternal directiveness, however, is

coupled with insensitivity and a lack of warmth, then

maternal control may inhibit a child's responsiveness

and play. The implications of maternal directiveness on

child outcomes have not been thoroughly investigated.

Parent-child interaction is a sensorimotor activity as well

as a social-emotional experience.[2] Motor control,

sensory integration, and social skills are all components of

parent-child interactions.[39] Infancy has been described

as the sensorimotor period of development.[40] Young

children use motor actions to engage the attention of

their parents and to respond to requests and social

interactions. Limitations on the ability to control the

head and trunk, maintain positions, move in and out of

positions, locomote, manipulate objects, and respond to

sensory input can alter the dynamics of social

interactions and may influence the degree to which a child can

successfully interact with the environment.

To promote parent-child interactions, physical therapists

need to provide intervention to the child to promote

motor function as well as provide intervention to the

parents to assist them in optimizing the child's abilities.

Families have indicated that meeting their information

needs regarding their child's disability, development,

and care is a primary expectation of early intervention.[10]

Physical therapists provide parents with information on

handling, positioning, and adaptive equipment that may

assist the parents in the physical needs that arise when

they interact with their child.

The Individuals With Disabilities Education

amendments of 1991 (Public Law 99-457 and its

reauthorization, Public Law 102-119) support the provision of early

intervention services in the natural environment. With

infants and toddlers, the context of learning is their

home.[41] The home environment fosters spontaneous

use of skills by providing natural cues and

reinforcement.[42] Parents have rated home visits as the most

helpful service component in early intervention,[43] and

mothers have reported a desire to participate in their

children's play.[44] Play is related to many areas of child

development and is now accepted as an appropriate

medium for teaching children new skills.[42,45,46]

We propose that a functional and meaningful model of

service delivery is one that focuses on enhancing a

child's motor abilities and performance within the

context of mother-child interactions during play. Our model

is based on the following concepts: (1) Improvement of

sensorimotor function may foster a child's ability to

participate more actively during parent-child

interactions, (2) providing parents with information on their

child's sensorimotor abilities enables them to adapt their

expectations and interactions to fit the child's

developmental abilities and functional needs, (3) home

environments foster spontaneous use of motor abilities by

providing natural cues and reinforcement,[42] and (4)

play is an effective medium for both parent-child

interaction and therapeutic intervention.

The purpose of our study was to test our model in a

preliminary fashion by examining the effects of

home-based physical therapy provided in the context of motor

play on mother-child interactions and the motor

behaviors of children with motor delay. In addition, we wanted

to explore maternal satisfaction with the home-based

physical therapy intervention. An experimental and

control group design was used in which children receiving

early intervention were randomly assigned to either

receive home-based physical therapy (experimental

group) or not receive home-based physical therapy

(control group). We hypothesized that the change in

occurrence of behaviors from the preintervention

assessment to the postintervention assessment would be

greater for the experimental group than for the control

group for the following behaviors: (1) children's motor

behaviors during play with their mothers, (2) mothers'

behaviors to promote the occurrence or quality of their

children's motor behaviors during play, and (3)

mothers' and children's pleasant reciprocal interactions. In

addition, we hypothesized that the mothers of children

in the experimental group would report satisfaction with

the intervention program.

Method

Subjects

The subjects for this study were 38 mothers and their

infants and toddlers, aged 6 to 34 months (X = 18.8,

SD = 7.2), who were enrolled in early intervention

programs. None of the children in the study were receiving

home-based physical therapy when the study was begun.

Twenty-six children were receiving center-based physical

therapy, and 12 children (5 children in the experimental

group and 7 children in the control group) were not

receiving any physical therapy services. All mothers

signed an informed consent form. Each child had a

documented motor delay, no major visual or auditory

handicap, and a cognitive level of at least 4 months,

including being able to demonstrate a social response to

people. Each child had a Psychomotor Developmental

Index (PDI) of less than 73 (greater than 1.5 standard

deviations below the mean PDI for their age) on the

Motor Scale of the Bayley Scales of Infant

Development.[47] Thirty-one children had a PDI less than 50,

which is greater than 3 standard deviations below the

mean for children without motor delays. The children

were unable to ambulate independently when the study

was begun. Nine children (5 children in the

experimental group and 4 children in the control group) had a

diagnosis of Down syndrome, 12 children (6 children in

each group) had cerebral palsy, and 2 children (both in

the experimental group) had myelomeningocele. The

remaining 15 children demonstrated delayed motor

development and had one or more of the following

conditions: developmental delay, infantile spasms,

hydrocephalus, myotonic dystrophy, deletion of

chromosome 9, agenesis of the corpus callosum, prematurity,

cytomegalovirus, and Pelizaeus-Merzbacher disease.

The children in the study were recruited from eight early

intervention programs. In an attempt to ensure that the

groups were comparable on motor development and

prior intervention, children were assigned to

experimental and control groups in the following manner. All

children recruited from a particular center were listed by

rank order of motor developmental age using the Bayley

Motor Scale.[47] The first child on the list was randomly

assigned to the experimental group or the control

group, and the remaining children were then alternately

assigned to each group. In this manner, half of the

children recruited from a particular center were in the

experimental group and half of the children were in the control

group.

The demographic data for the experimental and control

groups are presented in Table 1. Statistical analyses

revealed that the two groups did not differ on these

characteristics. Families of children in the experimental

group had a median socioeconomic status score of 51, as

compared with a median socioeconomic status score of

42 for families of children in the control group

(P = .068). The median scores for the families in both

groups, however, were in the same social stratum (AB

Hollingshead, unpublished research).

[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]

Protocol

Prior to group assignment, all subjects received a

preintervention assessment in their home. The mother and

child were videotaped during 12 to 15 minutes of free

play. Mothers were given the following standardized

instructions: "I would like for you to play with your child

as you usually do at home. You can use whatever toys you

or your child likes, but you do not have to use toys if you

prefer other games or activities. Feel free to use any

adaptive equipment for positioning, and you can play in

any positions you and your child usually use for play. I

would like to request that other family members not be

in the room during the videotaping because it may

distract you and your child from playing. Because I am

interested in how mothers and children interact during

playtime, I will not participate in the interaction. Before

or after the videotaping, I will answer any questions you

may have."

The decision not to use a standardized set of toys was

based on the primary investigator's (LAC's) pilot work.

When a set of toys were provided, mothers often felt

compelled to have their children play with the variety of

toys provided and thus frequently directed their

children from one activity to another. The purpose of the

assessment was to examine how a mother and child

routinely play together in their home environment. The

choice of play materials was thus left to the mother and

child. In addition, Mash and Terdal[48] reported that

although variations in play materials may prompt

different kinds of play interaction, which in essence was part

of the intervention process, the social content and

structure of the mother-child interaction was a more

salient feature than the play materials.

After the observation session, the mother was given the

Maternal Observation Interview[49] to complete. The

Maternal Observation Interview is an 11-item interview

schedule designed to assess a mother's reactions

regarding the validity of the observation session. The mother

was asked to ascertain how the observed interaction

compares with her usual behavioral style and to report

any behavioral occurrences that were not representative

of typical patterns. In one instance, the Maternal

Observation Interview indicated that the mother believed the

observation session was not representative of her usual

interactions with her child. A second mother-child

observation session was conducted, the mother indicated

that the session was representative of their routine play

time, and the responses from the second session were

used for data analysis.

At the end of the initial session, the Bayley Motor Scale[47]

was administered to the child. Interrater reliability

between the primary investigator and another pediatric

physical therapist was established on five children prior

to the start of the study and on two children during

the data collection phase of the study. The intraclass

correlation coefficient (ICC[2,1]) for the total raw score

was .99.

During the postintervention assessment, mother and

child pairs in the experimental and control groups were

again videotaped in their home during a 12- to

15-minute free-play session. A modified version of the

Client Satisfaction Questionnaire[50] was used to assess the

degree of maternal satisfaction with the experimental

physical therapy intervention program. Larsen et al[50]

reported that the original scale had a high degree of

internal consistency, as evidenced by an alpha

coefficient of .93. The questionnaire was modified to reflect

the components of the program. The questionnaire was

used to determine the consumers' perception of how

they benefited from the developed model for

intervention. With a family-focused framework, consumer

feedback and collaboration were essential components of the

study to confirm whether the model was meeting the

needs of the mother and the child. The parent

satisfaction questionnaire was left with the mothers of children

in the experimental group, who were asked to complete

the questionnaire anonymously and to return it in a

preaddressed and stamped envelope to the primary

investigator.

Physical Therapy Interventions

During the study, all subjects continued to receive early

intervention services as designated by their respective

Individualized Family Service Plans. Through a

questionnaire, the therapists from the centers indicated that

enhancing mother-child interactions through play was

not the explicit focus of the center-based therapy

sessions. The presence and participation of the mothers

and the format and availability of physical therapy

services in the center-based programs were variable,

depending on the staffing and philosophy of the early

intervention programs.

In our study, the experimental group received weekly

home visits for 5 weeks from a pediatric physical

therapist (LAC). We believed that five weekly sessions was an

appropriate intensity to provide the mothers and their

children with guidance in their motor play that

addressed the children's current needs. The short-term

intervention for the experimental group had a specific

focus that was unique compared with the center-based

services that all the subjects were receiving. The

emphasis was on collaboration with the mothers to identify

enjoyable interactive motor play activities in the home

and to provide recommendations for incorporating

therapeutic strategies into interactive play. The desired effect

of the intervention was enhancement of the interactive

play process between mothers and their children.

Appendix 1 delineates the components of the

experimental intervention model. Each intervention session

lasted approximately 1 hour. During the first

intervention visit, the physical therapist and the mother

discussed the child's strengths and needs, the concerns and

priorities of the mother, and the intervention plan. This

discussion included an exchange of information on both

the child's motor status and mother-child interactions.

The session was informal, allowing the therapist, the

mother, and the child an opportunity to get to know one

another and to establish rapport.

Physical therapist intervention was individualized. The

mother was present for the sessions and was invited and

encouraged to participate in the activities. The mother's

level of involvement in the session was left up to the

mother. Conventional physical therapist interventions

were incorporated into play activities. Therapeutic

techniques included active range of motion, weight bearing,

strengthening and endurance activities, use of prompts,

and guiding movement. Other activities to enhance

motor function included practice of functional motor

skills and environmental adaptations. Instruction in

positioning and handling techniques was provided by the

pediatric physical therapist (LAC).

The play activities were selected in an effort to enhance

gross motor and fine motor function. The intervention

was designed to structure the home setting to provide

movement opportunities. Activities were not restricted

to play with toys or objects but also included simple

games emphasizing play with motion and language.

Typical play activities in a standing position included

moving to the music of popular children's singing games

such as "Monkeys Jumping on the Bed," "Ring Around

the Rosy," and "Musical Chairs"; a sit-to-stand transition

during "The Kids on the Bus"; and choo-choo train

pretend play. Recommendations were made to modify

equipment, articles, or toys found in the home.

Throughout each session, the therapist demonstrated

and provided the mother with opportunities to

participate in positive interactive behaviors such as allowing the

child time to respond, providing the child with choices,

praising and encouraging the child's efforts, selecting

developmentally appropriate activities, and following the

child's direction. The therapist attempted to generate a

relaxed and enjoyable atmosphere to enhance the

interpersonal aspects of the interaction. The mother and

child were provided with positive reinforcement and

ongoing feedback regarding the child's motor function

and the mother's and child's interactive behaviors.

These strategies reflect the principles and guidelines

proposed by several leaders in early intervention.[6,14,21,51,52]

Procedural Reliability

Procedural reliability was examined during the study to

determine whether the assessment and intervention

procedures were performed in the intended manner.

The second author observed one preintervention

assessment and four intervention sessions, one of which was an

initial intervention visit. A checklist completed during

each observation indicated that the therapist completed

100% of the outlined steps developed from the

intervention model (Appendix 1) during a preintervention

assessment, an initial intervention visit, and three

subsequent treatment visits.

In addition to completing a checklist of the outlined

procedures, information on the frequency of key

elements of the intervention was also determined.

Procedural reliability results indicated that treatment

procedures were consistent with the described method.

Mothers participated in the play activities during 60% of

the observed session. Children were involved in motor

activities during 65% of the observed session. The

therapist provided information during 15% of the observed

session, promoted the child's motor skills without

handling during 41.7% of the observed session, and

promoted the child's motor skills with handling during

18.3% of the observed session. The context of the

sessions was interactive play during 61.7% of the

observed session.

Measurement Tool

The following dependent variables were defined prior to

data collection: (1) the child's voluntary physical

behavior, (2) the child's changes in position, (3) the child's

locomotion (eg, creeping, cruising), (4) maternal

promotion of the child's motor skills (eg, providing verbal

cues, setting up equipment), (5) the mother's use of

therapeutic positioning, (6) maternal holding, (7) the

child's summary measure, which was a combination of

the child's positive, interactive, and voluntary physical

behaviors, and (8) the maternal summary measure,

which was a combination of maternal positive and

interactive behaviors and promotion of the child's motor

skills. The child's voluntary physical behavior, changes in

position, and locomotion were used as an index of the

child's motor behaviors during play with the mother.

Maternal promoting of the child's motor skills, use of

therapeutic positioning, and holding were used as an

index of the mother's promotion of the occurrence or

quality of her child's motor behaviors during play.

Several behaviors were combined to form the summary

measures, which were used as an index of pleasant

reciprocal interactions.

Mother-child interactions were measured from the

videotapes using the response-class matrix (EJ Mash, L Terdal,

K Anderson; unpublished coding manual),[53] an interval

recording observational method. The matrix was used to

record the proportion of intervals in which verbal and

nonverbal interactive behaviors occurred before and

after intervention. The response-class matrix was

modified to reflect the nature of the study. The behaviors

measured were chosen based on the original matrix,

studies of mother-child interactions with children with

motor delay,[25,28,31,32,54] and a philosophy of pediatric

physical therapy.[1] The child behaviors that were

measured were negative, solitary, interactive, positive,

physically directed, and voluntary physical behaviors. The

maternal behaviors that were measured were negative,

solitary, interactive, positive, and directive behaviors and

promoting the child's motor skills. Definitions of the

child's and mother's behaviors that were measured are

provided in Appendix 2.

An independent observer who was unaware of group

assignment recorded the behaviors from the videotapes.

Each videotape was divided into 10-second intervals,

marked by a time signal on the videotape and an

audiotape. For each interval, the observer viewed and

noted three behaviors: the mother's first (antecedent)

behavior, the child's response, and the mother's next

(consequent) behavior. This standard procedure

enabled the researchers to examine behaviors in

context. The observer noted the first mother-child-mother

interaction sequence that occurred during each

10-second observation segment. If an interval could not be

noted due to obstruction of view, the interval was

considered invalid for data collection. The first valid 60

intervals (10 minutes) were used for data analysis.

General rules for scoring the behaviors were as follows. If

either the mother or the child performed several

behaviors in succession, the behavior closest in time to the

corresponding behavior of the other member of the

dyad was noted. If a nonverbal behavior and a verbal

behavior occurred simultaneously, the verbal behavior

was noted. The original coding scheme of the

response-class matrix reflected general classes of behaviors as

opposed to discrete behaviors (EJ Mash, L Terdal,

K Anderson; unpublished coding manual). We were

interested in two specific behaviors: maternal promoting

of a child's motor skills and the child's voluntary physical

behavior. If a behavior met the definition for one of

these two specific behaviors, it was noted as such and was

not noted as an interactive category behavior.

The noted behaviors were recorded with two matrices.

The first matrix was used to record the mother's first

(antecedent) behavior and the child's response. The

second matrix was used to record the mother's behavior

consequent to the child's response. In total, 120

maternal behaviors and 60 child behaviors were recorded.

In addition to recording the behaviors listed in the

matrix, a tally of the following four behaviors was kept:

the child's changes in position, the child's locomotion,

the mother's use of therapeutic positioning, and

maternal holding. A distinction was made between changes in

position that were performed independently by the

child, with participation by the child but requiring the

assistance of the mother, or by the mother with the child

totally dependent. This distinction was made to reflect

opportunities for active movement by the child. A

distinction was also made between therapeutic and

nontherapeutic holding to reflect incidences in which

physical contact may be beneficial to the child's motor needs

and enhance the interaction. A tally was recorded if a

behavior occurred any time during each of the 60

10-second intervals. Definitions of these behaviors are

provided in Appendix 2. A sample recording sheet is

provided in Appendix 3.

Following the guidelines of the coding manual, the

primary investigator and the independent observer

participated in 26 hours of collaborative training.

Videotapes of interactions during free play of mothers and

their children with motor delays who were not subjects

in the study were viewed. An additional 14 hours of

observation and interval recording was used to

determine reliability. Prior to the start of the study, interrater

reliability in scoring the response-class matrix was

established between the independent observer and the

primary investigator. Reliability was analyzed by occurrence

agreement for each maternal and child behavior that

was to be used for analysis in this study. Occurrence

agreement of 80.1% was achieved for the maternal

category "mother promotes the child's motor behavior."

Occurrence agreement for the other categories ranged

from 93.1% to 98.8%. Interrater reliability was

rechecked during the study using four observations.

Occurrence agreement for these observations ranged

from 93.3% to 100%.

Data Analysis

The measurement of each dependent variable was

expressed as a proportion of the total number of

recorded behaviors. The total number of maternal

behaviors was 120, and the total number of child

behaviors was 60. The measurement of the four dependent

behaviors that were tallied (child's changes in position,

child's locomotion, mother's use of therapeutic

positioning, and maternal holding) was reported as a

proportion, with the total number of intervals equal to 60. For

example, if maternal promoting of the child's motor

skills was counted on the response-class matrix for 20 of

the 120 recorded behaviors, then the occurrence of this

behavior would be expressed as the proportion .17.

For each dependent variable, a two-factor (group X

time) analysis of variance for repeated measures on the

second factor was used to analyze group differences in

child and mother interactive behaviors. The group X

time interaction was examined to determine whether

there was a differential change between the two groups.

An alpha level of .05 was used for all analyses. This

decision increased the risk of a Type I error because

multiple tests were performed. Because this research was

exploratory in nature, that is, to determine which aspects

of mother-child interaction may be influenced by

physical therapy, we did not believe that a correction factor

was necessary. All statistical analyses were carried out

using the SPSS/PC+ version 2.0 (1988) software

program.

Results

The change in the occurrence of the child's voluntary

physical behaviors, changes in position, and locomotion

between the preintervention and postintervention

assessments did not differ between the experimental

and control groups (Tab. 2). Although the F value for

the interaction term (F=3.76; df = 1,36; P=06) did not

reach the alpha level (P=.05), children in the

experimental group changed positions an average of 8.7 more

intervals at postassessment than children in the control

group did.

[TABULAR DATA 2 NOT REPRODUCIBLE IN ASCII]

Mothers of children in the experimental group and

mothers of children in the control group did not differ

in their change in the proportion of behaviors during

which they promoted children's motor skills between

the preintervention and postintervention assessments

(Tab. 3). A child's voluntary physical behavior followed

maternal promoting of motor behavior greater than

70% of the time. Mothers of children in the

experimental group and mothers of children in the control group

did not differ in their change in the proportion of

intervals in which they used adaptive positioning

between the preinter-vention and postintervention

assessments. An interaction, however, was found for

maternal holding (F=4.69; df= 1,36; P [is less than] .05). Mothers of

children in the experimental group demonstrated an

increase in the proportion of intervals in which they held

their children in an appropriate manner from the

preintervention assessment to the postintervention

assessment (from 6.1 to 12.2), whereas mothers of

children in the control group demonstrated a decrease

(from 16.8 to 11.9).

[TABULAR DATA 3 NOT REPRODUCIBLE IN ASCII]

The change in the interactive summary measure scores

between assessments did not differ for the children in

the experimental and control groups (Tab. 4). Children

in both groups demonstrated pleasant, reciprocal

interactions in an average of at least 84% of the play intervals.

Negative, solitary, and physically directed behaviors

collectively were observed in an average of less than 17% of

the play intervals.

The change in the interactive summary measure scores

between assessments differed for the mothers of the

children in the experimental and control groups

(F=9.95; df=1,36; P [is less than] .01) (Tab. 4).

[TABULAR DATA 4 NOT REPRODUCIBLE IN ASCII]

The interactive summary measure score decreased from

77.2% to 66.2% for the mothers of children in the

experimental group and increased from 72.5% to 77.2%

for the mothers of children in the control group. To

clarify this finding, five additional two-factor (group X

time) repeated-measures analyses of variance were

performed to determine the specific behaviors where the

change in frequency differed between the two groups

(Tab. 5). From the preintervention assessment to the

postintervention assessment, the mothers of children in

the experimental group demonstrated an increase in

directive behaviors from 21.4% to 33.1% and the

mothers of children in the control group demonstrated a

decrease from 26.1% to 21.6% (F=11.3, df=1,36;

P [is less than] .01). Mothers of children in the experimental group

demonstrated a corresponding decrease in general

interactive behaviors from 36.3% to 28.7% between

assessments compared with mothers of children in the

control group, who demonstrated an increase from

30.7% to 41.2% (F=8.49; df=1,36; P [is less than] .02). The changes

in the occurrence of negative, solitary, and positive

behaviors between assessments did not differ for the

mothers of children in the experimental and control

groups.

[TABULAR DATA 5 NOT REPRODUCIBLE IN ASCII]

Fifteen of the 19 mothers of children in the

experimental group returned the Client Satisfaction

Questionnaires, a return rate of 79%. The mothers indicated that

they were very satisfied with the services they received

and that they believed the services addressed their needs

(Tab. 6). They commented that the communication

between the therapist and themselves was effective and

provided them with an understanding of play. The

mothers expressed their beliefs about how physical

therapy positively influenced both their interactions and

their children's interactions during mother-child play.

Overall, the mothers expressed that both they and their

children enjoyed the program.

[TABULAR DATA 6 NOT REPRODUCIBLE IN ASCII]

Discussion

The purpose of our study was to examine, in a

preliminary fashion, a model for provision of home-based

physical therapy within the context of motor play.

Although our eventual aim is to determine whether this

treatment model enhances development of motor

function in young children with motor delays, the first step

was to examine the immediate effects on mother-child

interactions and children's motor behaviors. Our model

is based on the assumption that an intervention

provided within a family-focused framework and in the

context of play would support and enrich the child's

natural environment and ultimately enhance

development of motor function. The intervention was positively

received by mothers and children in this study. Although

our hypotheses were not fully supported, the findings

indicate that mothers can promote physical activity

through play while being sensitive to positive social

interactions.

Statistical analysis did not support our hypothesis that

children in the experimental group would achieve a

greater change in the occurrence of changes in position

during play than children in the control group would

achieve. Further investigation may be warranted to

determine whether physical therapists can be influential

in teaching mothers how to integrate gross motor

activities and movement transitions into play. In our study,

the children in the experimental group changed

positions an average of 8.7 more intervals than did the

children in the control group. Mothers of children in

the experimental group were more apt to include gross

motor play in a variety of positions. On the Client

Satisfaction Questionnaire, mothers commented that

they developed an awareness of the use of play to

promote their children's motor abilities. This added

practice and repetition may be clinically important in

the generalization of motor skills.

Despite the emphasis on motor behaviors during playful

interactions, no change was noted in the proportion of

voluntary physical behaviors between the

preintervention and postintervention assessments. Two factors

may have contributed to this result. First, the proportion of

voluntary physical behaviors was high when the study

began. For the most part, children were active

participants. This finding is in contrast to many reports that

children with disabilities are passive.[21,32,33] Second, this

category was general and included any motor response.

In retrospect, the definition for voluntary physical

behaviors may have been more appropriate for detecting

change in children with severe motor impairment who

demonstrate a paucity of movement. Although the

children in our study had delayed motor development, they

all demonstrated some method of self-initiated

movement during play. A detailed rating scale would have

been needed to measure qualitative changes and the

success of the children's movements during play.

The mothers of the children in the experimental group

did not demonstrate a greater change in the proportion

of behaviors that promoted their children's motor

abilities or in the proportion of use of adaptive positioning

compared with the mothers of the children in the

control group. In comparison, Hanzlik[19] found an

increase in the use of adaptive positioning by mothers of

children with cerebral palsy after 1 hour of maternal

instruction on interactive techniques and positioning. In

our study, the families had participated in an average of

I year of early intervention, and a ceiling effect may have

occurred.

Our results support the role of the physical therapist in

providing instruction to mothers on therapeutic

handling to promote their children's play and movement.

The mothers of the children in the experimental group

demonstrated an increase in the proportion of therapeutic

holding between assessments, whereas the mothers of

the children in the control group demonstrated a

decrease. This finding appeared to be a result of the

strategies provided to the mothers to give their children

guidance in supported positions. The mothers

successfully provided support for their children that enabled

them to play in appropriate positions to promote gross

motor activity. The mothers of the children in the

control group showed a higher proportion of holding

during the preintervention assessment than during the

postintervention assessment. This finding may have been

attributable to a few mothers holding their infants for a

large percentage of the play session. This possibility is

substantiated by the large variability in the control

group's holding scores.

The results of our study suggest that our treatment

model, which integrated therapeutic activities into play

and social interactions, was well received by the children

and their mothers. The finding that the children's

interactive summary measure did not differ between the

two groups during the postintervention assessment can

be interpreted as a positive finding. The children in both

groups rarely demonstrated negative or solitary

behaviors during either assessment. Unlike the results of

earlier studies in which mothers displayed negative

behaviors and children cried during mother-child

therapy sessions,[29,55] the children and mothers in our study

were positive and interactive during play activities in

which the mothers promoted and directed their

children's motor behaviors.

We believe the finding that the mothers of children in

the experimental group became more directive after the

5-week period than the mothers of children in the

control group may represent a desired effect of physical

therapy, although that was not our view when the study

began. The connotation of maternal directiveness that

we adopted from research on mother-child interactions

may not be directly applicable for children with motor

delays. A focus of early intervention is providing

information and guidance to parents on how to enhance

their child's development. As parents become more

knowledgeable about how play fosters development,

they are likely to be more directive when interacting with

their children during play. In our study, the increase in

the directiveness of the mothers of children in the

experimental group did not appear to have a negative

effect on the children, who were still active and pleasant

participants. Rather, the increased directiveness appears

to reflect (1) the increased ability of the mothers to use

play activities as therapeutic strategies, (2) the children's

need for directiveness to participate in play activities that

were developmentally challenging, and (3) the mothers'

desire for their children to perform optimally. Mothers

of children in the experimental group learned to set up

play situations to encourage creeping, pulling the stand,

and cruising and used verbal commands to direct the

children to the activity instead of allowing the children

to explore freely. Our new perspective is supported by

Marfo[35] and Tannock,[37] who stated that maternal

directiveness, warmth, sensitivity, and responsiveness are not

incompatible and that maternal directiveness may serve

to encourage child participation.

Controversy still exists over the significance of maternal

directiveness. We believe that therapists should be aware

that their interventions may foster maternal

directiveness and place a greater emphasis on motor

performance and less of an emphasis on interactive play. We

contend that therapists need to consider maternal

directiveness in the context of a mother's affective behaviors

and how this interactive style enhances or limits a child's

responsiveness. Further research is needed to determine

the effect of physical therapy on fostering maternal

directiveness and to explore the interrelationships

between maternal directiveness and responsiveness and

their influence on the child.

The response-class matrix did not provide information

on two important characteristics of maternal interactive

behavior: mothers' developmental awareness and

sensitivity. Furthermore, mothers and children in both

groups appeared to have demonstrated effective

interaction abilities during the preintervention assessment.

This finding was not anticipated, and the measuring tool

may not have been sensitive enough to document the

mothers' incorporation of therapeutic strategies into

pleasant play interactions. Most importantly, limitations

in internal validity need to be emphasized. Even though

a control group was used in the design, both groups were

receiving center-based early intervention and may have

been exposed to some aspects of our intervention

model. While not examined, the center-based model of

early intervention may have positively influenced the

children and mothers in the study. In addition, it is

acknowledged that the amount of therapy was a

confounding variable that was not controlled. The

experimental group received five intervention visits which were

not balanced by additional services for the control

group.

Conclusion

A model for provision of home-based physical therapy

within the context of interactive play was well received by

the children and mothers in this study. The results of the

influence of this model on children's motor behaviors

during play with their mothers were inconclusive.

Children in the experimental group changed positions an

average of 8.7 more inter-vals than did children in the

control group. Children in the experimental group,

however, did not demonstrate a greater change in

overall voluntary physical behaviors, as was hypothesized.

The proportion of voluntary behaviors for both groups

was high during the preintervention assessment, which is

in contrast to reports that children with disabilities are

passive.[29,33,34] Additionally, the definition of voluntary

physical behaviors was too general for the subjects in this

study, all of whom demonstrated some method of

self-initiated mobility.

The results of the influence of this model on mothers'

behaviors to promote their children's motor behaviors

were inconsistent. Mothers of children in the

experimental group demonstrated an increase in therapeutic

holding but did not demonstrate a greater change in

overall behaviors to promote their children's motor

behaviors. Mothers of children in both groups were

already promoting their children's motor behaviors at

the time the study was begun.

The hypothesis that children in the experimental group

and their mothers would demonstrate a greater change

in pleasant reciprocal interactions was not supported.

Children in both groups rarely demonstrated negative or

solitary behaviors at either the preintervention

assessment or the postintervention assessment. In comparison

with the mothers of children in the control group, the

mothers of children in the experimental group

demonstrated an increase in directiveness, but they were not

less positive or more negative when interacting with their

children. The directiveness appeared to reflect the

mothers' abilities to incorporate therapeutic strategies

into play and to direct their children to participate in

play activities that were developmentally challenging.

Additional research is necessary to examine whether our

treatment model is effective in promoting the

generalization and practice of motor skills during play without

interfering with positive reciprocal interactions. A next

step is to determine whether short-term, home-based

intervention results in carry-over of parent-child motor

play interactions as part of daily routines. This evidence

is needed prior to investigation of whether a

family-focused, ecological model of physical therapy is effective

in promoting the development of motor function in

children with motor delays.

References

[1] Effgen SK, Bjornson K, Chiarello LA, et al. Competencies for

physical therapists in early intervention. Pediatric Physical Therapy.

1991;3(2): 77-80.

[2] Chiarello LA, Effgen SK, Levinson M. Parent-professional

partnership in evaluation and development of individualized family service

plans. Pediatric Physical Therapy. 1992;4(2):64-69.

[3] Ainsworth MD, Bell SM. Mother-infant interaction and the

development of competence. In: Connolly K, Brunner J, eds. The Growth of

Competence. New York, NY: Academic Press Inc; 1974:97-118.

[4] Beckwith L, Cohen SE, Kopp CB, et al. Care-giver infant interaction

and early cognitive development in preterm infant. Child Dev. 1976;47:

579-587.

[5] Bee HL, Barnard KE, Eyres SJ, et al. Prediction of IQ and language

skill from perinatal status, child performance, family characteristics.

and mother-infant interactions. Child Dev. 1982;53:1134-1156.

[6] Clarke-Stewart KA. Interactions between mothers and their young

children: characteristics and consequences. Monogr Soc Res Child Dev.

1973;38(6-7):1-109. Serial no. 153.

[7] Hartup W. Social relationships and their developmental significance.

Am Psychol. 1989;44:120-126.

[8] Mahoney G, Finger I, Powell A. Relationship of maternal behavioral

style to the development of organically impaired mentally retarded

infants. Am J Ment Defic. 1985;90:296-302.

[9] Yarrow LJ, Rubenstein JL, Pedersen F.A Jankowski JJ. Dimension of

early stimulation and their differential effects on infant development.

Merrill-Palmer Quarterly. 1972;18:205-218.

[10] Summers JA, Dell'Oliver C, Turnbull AP, et al. Examining the

individualized family service plan process: What are family and

practitioner preferences? Topics in Early Childhood Special Education. 1990;

10(1):78-99.

[11] Shonkoff JP, Hauser-Cram P. Early intervention for disabled infants

and their families: a quantitative analysis. Pediatrics. 1987;80:650-658.

[12] Affleck G, McGrade BJ, McQueeney M, Allen D. Promise of

relationship-focused early intervention in developmental disabilities.

Journal of Special Education. 1982;16:413-430.

[13] Barrera ME, Rosenbaum PL, Cunningham CE. Early home

intervention with low-birth-weight infants and their mothers. Child Dev.

1986;57:20-33.

[14] Bromwich RM. Working With Parents and Infants: An Interactional

Approach. Baltimore, Md: University Park Press; 1981.

[15] Bromwich RM, Parmelee AH. An intervention program for preterm

infants. In: Field TM, Sostek A, Goldberg S, Shuman HH, eds. Infants

Born at Risk: Behavior and Development. New York, NY: SP Medical and

Scientific Books; 1979:389-411.

[16] Calhoun M. Rose T, Hanft B, Sturkey C. Social reciprocity

interventions: implications for developmental therapists. Physical &

Occupational Therapy in Pediatrics. 1991; 11 (3):45-56.

[17] Dawson PM, Robinson JL, Butterfield PM, et al. Supporting new

parents through home visits: effects on mother-infant

interaction. Topics in Early Childhood Special Education. 1990;10(4):29-43.

[18] Field F. Interaction coaching for high-risk infants and their parents.

Prevention in Human Services. 1982;1:5-24.

[19] Hanzlik JR. The effect of intervention on the free-play experience

for mothers and their infants with developmental delay and cerebral

palsy. Physical & Occupational Therapy in Pediatrics. 1989;9(2):33-51.

[20] Kelly JF. Effects of intervention on caregiver-infant interaction when

the infant is handicapped. Journal of the Division for Early Childhood.

1982;5:53-63.

[21] Mahoney G, Powell A. Modifying parent-child interaction:

enhancing the development of handicapped children. Journal of Special

Education. 1988;22:82-96.

[22] McCollum JA. Stayton VD. Infant/parent interaction: studies and

intervention guidelines based on the SIAI model. Journal of the Division

for early Childhood. 1985;9:125-135.

[23] Rosenberg SA, Robinson CC. Enhancement of mothers'

interactional skills in an infant education program. Education and Training of

the Mentally Retarded. 1985;20:163-169.

[24] Seitz S, Hoekenga R. Modeling as a training tool for retarded

children and their parents. Ment Retard. 1974;12:28-31.

[25] Tyler N, Kogan K. Reduction of stress between mothers and their

handicapped children. Am J Occup Ther. 1977;31:151-155.

[26] Harris SR. Efficacy of physical therapy in promoting family

functioning and functional independence for children with cerebral palsy.

Pediatric Physical Therapy. 1990;2(3):160-164.

[27] Stuberg W, Harbourne R. Theoretical practice in pediatric physical

therapy: past, present, and future considerations. Pediatric Physical

Therapy. 1994;6(3):119-125.

[28] Hanzlik JR. Nonverbal interaction patterns of mothers and their

infants with cerebral palsy. Education and Training of the Mentally

Retarded. 1990:25(4):333-343.

[29] Kogan KL, Tyler N, Turner P. The process of interpersonal

adaptation between mothers and their cerebral-palsied children. Dev Med

Child Neurol. 1974;16:518-527.

[30] Palisano RJ, Chiarello LA, Haley SM. Factors related to

mother-infant interaction in infants with motor delay. Pediatric Physical

Therapy. 1993;5(2):55-60.

[31] Barrera ME, Vella DM. Disabled and nondisabled infants'

interactions with their mothers. Am J Occup Ther. 1987;41:168-172.

[32] Hanzlik JR, Stevenson MB. Interaction of mothers with their infants

who are mentally retarded, retarded with cerebral palsy, or

nonretarded. Am J Ment Defic. 1986;90:513-520.

[33] Kogan KL, Tyler N. Mother-child interaction in young physically

handicapped children. Am J Ment Defic. 1973;77:492-497.

[34] Lieberman D, Padan-Belkin E, Harel S. Maternal directiveness and

infant compliance at one year of age: a comparison between mothers

and their developmentally delayed infants and mothers and their

nondelayed infants. J Child Psychol Psychiatry. 1995;36:1091-1096.

[35] Marfo K. Correlates of maternal directiveness with children who are

developmentally delayed. Am J Orthopsychiatry. 1992;62:219-233.

[36] Mahoney G, Fors S, Wood S. Maternal directive behavior revisited.

Am J Ment Retard. 1990;94:398-406.

[37] Tannock R. Control and reciprocity in mothers' interaction with

Down syndrome and normal children. In: Marfo K, ed. Parent-Child

Interaction and Developmental Disabilities: Theory, Research, and

Intervention. New York, NY: Praeger; 1988:163-180.

[38] Rosenberg SA, Robinson CC. Interactions of parents with their

young handicapped children. In: Odom S, Karnes M, eds. Early

Intervention for Infants and Children With Handicaps: An Empirical Base.

Baltimore, Md: Paul H Brookes Publishing Co; 1988.

[39] Ayres A. Sensory Integration and the Child. Los Angeles, Calif: Western

Psychological Services; 1981.

[40] Piaget J. Piaget's theory. In: Kessen W, ed. Handbook of Child

Psychology. New York, NY: John Wiley & Sons Inc; 1983:103-128.

[41] Rosenberg SA, Robinson CC, Beckman PJ. Measures of

parent-infant interaction: an overview. Topics in Early Childhood Special

Education. 1986;6(2):32-43.

[42] Grabowski K. Best Practices for Therapy in Preschool Settings.

Morganton, NC: North Carolina Division for Early Childhood of the Council

for Exceptional Children; 1991.

[43] Upshur CC. Mothers' and fathers' rating of the benefits of early

intervention services. Journal of Early Intervention. 1991;15:345-357.

[44] Hinojosa J. How mothers of preschool children with cerebral palsy

perceive occupational and physical therapists and their influence on

family life. Occupational Therapy Journal of Research. 1990; 10: 144-162.

[45] Garvey C. Play. Cambridge, Mass: Harvard University Press; 1990.

[46] Gowen JW, Goldman BD, Johnson-Martin N, Hussey B. Object play

and exploration of handicapped and nonhandicapped infants. Journal

of Applied Developmental Psychology. 1989;10:53-72.

[47] Bayley N. The Bayley Scales of Infant Development. New York, NY: The

Psychological Corporation; 1969.

[48] Mash EJ, Terdal L. Play assessment of noncompliant children with

the response-class matrix. In: Schaefer CE, Gitlin K, Sandgrund A, eds.

Play Diagnosis and Assessment. New York, NY: John Wiley & Sons Inc;

1991:283-315.

[49] Eheart B. A Comparative Observational Study of Mother-Child Interactions

With Nonretarded and Mentally Retarded Children. Madison, Wis:

University of Wisconsin-Madison; 1976. Doctoral dissertation.

[50] Larsen DL, Attkisson CC, Hargreaves WA, Nguyen TD. Assessment

of client/patient satisfaction: development of a general scale. Education

and Program Planning. 1979;2:197-201.

[51] Johnson-Martin N, Jens KG, Attermeier SM, Hacker BJ. Carolina

Curriculum for Infants and Toddlers With Special Needs. Baltimore, Md:

Paul H Brookes Publishing Co; 1991.

[52] Shere B, Kastenbaum R. Mother-child interaction in cerebral palsy:

environmental and psychological obstacles to cognitive development.

Genetic Psychological Monographs. 1966;73:255-335.

[53] Mash EJ, Terdal L, Anderson K. The response-class matrix: a

procedure for recording parent-child interaction. J Consult Clin Psychol.

1973;40:163-164.

[54] Brooks-Gunn J, Lewis M. Maternal responsivity in interactions with

handicapped infants. Child Dev. 1984;55:782-793.

[55] Tyler N, Kogan KL, Turner P. Interpersonal components of therapy

with young cerebral palsied. Am J Occup Ther. 1974;28:395-400.

Appendix 1.

Physical Therapy Experimental Intervention Model

Service Delivery Components

* Home-based

* Individual treatment

* Integration of child-centered instruction for parents and

relationship-focused approaches

Family-Focused Early Intervention Principles

* Respect mother's position as a consumer and the child's mother

* (Listen to mother's concerns and priorities

* Provide information from assessment

* Establish rapport with mother and child

* Encourage mother's participation based on her choice

* Collaborate on intervention activities

* Build on the resources of the family

Principles of Play

* Activities must be of interest and fun for mother and child

* Play with objects and movement

* Repetition of experiences

* Exploration

* Balance between interactive and independent play

Mother-Child Interaction Principles

* Enjoyment

* Encourage active participation of child

* Reciprocity

* Sensitivity to child's interests, cues, and needs

* Balance between holding and opportunities for exploration

* Model and encourage positive interactive patterns as opposed to

negative or directive patterns

Physical Therapy Components

* Principles of motor learning

* Instruction on motor needs for activities

* Functional training for activities of daily living

* Environmental adaptation to enhance motor function

* Range of motion to achieve flexibility for movement

* Strengthening and endurance activities

* Use of prompts

* Guiding child's mobility

* Weight bearing, approximation, and handling from key points of

control to enhance posture and stability in a variety of positions

* Discuss application of intervention strategies into daily routines

Appendix 2.

Operational Definitions for Child and Maternal Interactive Behaviors

Coded in the Response-Class Matrix(a)

Child:

Negative behavior: Any verbal or nonverbal behavior that was negative

in orientation, indicating displeasure.

Solitary behavior: Any verbal or nonverbal behavior that was not aimed

at or in response to the mother.

Interactive behavior: Any verbal or nonverbal behavior that was neutrol

in orientation, that was aimed at or in response to the mother, that was

an attempt to initiate or maintain some type of mutual contact, or that

communicated attention.

Positive behavior: Any verbal or nonverbal behavior that indicated

pleasure or demonstrated affection.

Physically directed behavior: Any nonverbal behavior that was

physically directed or controlled by the mother, defined as "the child did

not actively participate in the movement."

Voluntary physical behavior: Any nonverbal behavior that was initiated

by the child during interaction with the mother. This behavior included

moving his or her arms or legs during play, reaching for or

manipulating a toy, moving around the room, or changing position during

interaction with the mother.

Change in position(b) : Any time that the child changed position. Subtle

changes in posture within a position were not scored. indicated I for

the child performed the transition independently, A for the child

assisted with the transition, or D for the child was dependent in the

transition.

Locomotion(b): Anytime that the child was mobile (sequential rolling,

pivoting, crawling, creeping, scooting, cruising, assisted ambulation).

Mother:

Negative behavior: Any verbal or nonverbal behavior that was negative

in orientation, indicating displeasure in the interaction or

nonacceptance and disapproval of the child's behavior.

Solitary behavior: Any verbal or nonverbal behavior that was not aimed

at or in response to the child.

Interactive behavior: Any verbal or nonverbal behavior that was neutral

in orientation, that was aimed at or in response to the child, that was

an attempt to initiate or maintain some type of mutual contact, or that

communicated attention.

Positive behavior: Any verbal or nonverbal behavior that demonstrated

enjoyment of interaction, reinforced or encouraged the child's

behavior, indicated approval or acceptance of the child's behavior,

or demonstrated affection.

Directiveness: Any verbal or nonverbal behavior in which the mother

was intending to control the child's behavior, such as commands or

statements that included imperatives, questions, or gestures that implied

a command, or physically controlling the child's motor behavior.

Promoting child's motor skills: Any verbal or nonverbal behavior in

which the mother encouraged or promoted a child's specific motor

response while being sensitive to the child's abilities, interests, cues,

and needs, such as providing verbal cues, demonstrating an activity,

setting up the environment, use of a tactile prompt, or physically

guiding.

Use of adaptive seating or positioning(b): Mother positioned child in

appropriate postural alignment with use of support to promote

function during play.

Holding(b): Mother physically held her child with her hands, had the child

on her lap, or carried the child in her arms. Indicated A for an

appropriate posture in reference to child's alignment and the ease

with which the posture encouraged interaction or N for a

nonappropriate posture in reference to the child's alignment or for a posture

that restricted interaction.

(a) Condensed. Complete operational definitions available from the authors.

(b) Indicates a behavior that was not part of the matrix but was tallied as a

separate category.

Appendix 3

Matrix of Maternal and Child Behaviors

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]