Contact Information:
Cynthia J. Cress, Ph.D., CCC-SLP, 202G Barkley Memorial Center, University of Nebraska-Lincoln, Lincoln, NE 68583-0732 Phone: (402) 472-4431, Fax: 472-7697 Email: ccress1@unl.edu

Abstract: Physical limitations can reduce children's opportunities for exploration or object-based play that comprise typical play/interaction scales. This research reports results of multiple studies involving children with physical impairments (PI). Studies address object play, social play, alternative play scales, parent directives, and child mastery. Videotapes will demonstrate characteristics of children's play interaction.

I. Background.
A. Predictions of play & parent interaction for children w/PI:

· Physical limitations can reduce opportunities for exploration and object-based play, that can restrict children's sense of mastery in play as being independent of parent directions or assistance.

· Children with PI may use strategies that are less complex on standardized play scales that may not accurately reflect their underlying cognitive skills, particularly with multiple impairments.

· Standardized tests have not been normed for this population and criterion-based tests tend to emphasize object play that is less accessible to children with physical impairments.

· If parents recognize that children with PI have difficulty with object play, they may respond by changing interaction patterns during play to help children be successful with play tasks.

· Play assessment for children with PI needs to consider if social play offers greater opportunity for children's independent initiation and exploration of strategies, and less need for directiveness.

· If children with PI show more initiation and exploration in social play, this may be a means for developing generalized mastery skills that promote cognitive and communicative development.

B. Reasons to expect different play and interaction strategies in children w/ PI (Cress, in press)

· Much of early learning involves physically acting on the world, which is hard for kids w/PI.

· Many early communicative routines familiar to parents involve hand or vocal skills, both of which may be unavailable to children

· Children may not have a clear sense of their own body's movements (kinesthetically) which makes it difficult to relate their behaviors to play or interaction effects.

· Parents may not recognize children's unusual or subtle movements as intentional or communicative, and not respond as often.

· Motor delays in children's behaviors and response may limit early awareness of the relationship between their own behaviors and other events

· Many of children's early communicative attempts may be unsuccessful (e.g. using breathing as an attention signal), and children often grow passive without results

· Learning communication from a primarily receptive role limits children's access to trial-and-error exploration with communication

II. Research Questions:
Four studies were integrated into this presentation as follows:

Study 1:
How well do existing play schemes reflect the complexity of play in children with PI?

Does cognitive estimation, amount and type of play, differ between parent and experimenter play?

Study 2:
How can we modify currently used standardized tests for object play to account for the kinds of socially negotiated object play demonstrated by children with PI?

Are there individual differences in the patterns of play for children with PI that differ from expectations for typically developing children?

Study 3:
Are parents of children with PI more directive than typically developing children in object play vs. social play that parents and children spontaneously initiate?

When parents are more directive, do their directives follow the child's interest and attention?

Study 4:
Do children with PI demonstrate different qualities of mastery and persistence with play tasks in object vs. social play with their parents?

How can play interaction supports cognition and communication in children with PI?

III. Methods
A. Subjects:
Nonspeaking children with PI (ages 1-3 years)

42 children were followed longitudinally for 18 months each (six visits, one every three months), 8 additional children received 1-3 research visits.

Data for all of the reported studies were derived from the first research visit. Randomized subsets of children from the full database were selected for each of the studies below.

Ages: Began at 12-24 months (corrected), Ended at 30-42 months

Children had primary diagnoses of physical impairments and met 2 or 4 criteria for being at risk for vocal expression (MacDonald, 1980): Cerebral Palsy=19, Inherited Condition or Syndrome=7, Oral/Motor Dysfunction =9, Postnatal Illness/Injury=11, Developmental Disability, unknown =4

Children may have had secondary diagnoses such as cognitive or sensory processing impairments, although all children showed sufficient vision and hearing to respond socially.

24% of families were ethnically diverse. Mean HOH occupational status of 46.4 was comparable to norms of 42.7 (Ganzeboom & Treiman, 1996)

B. Play interactions:
Parents and children were videotaped during a complete 2-hour session that involved multiple assessment and interactive activities.

Parent play coding tapes were created from all of the incidental and spontaneously occurring interactions that parents or children initiated together without experimenter interaction.

If such interactions were not initiated, the experimenters prompted parents to initiate familiar routines, or briefly left the parent and child alone together to sample natural interaction.

IV. Study 1: Application of Traditional Play Schemes to Children with PI (Cress & LaMontagne, 1998) - Exploratory pilot study

Methods:
Play behaviors were coded for ten infants with PI between 16 and 21 months of age, selected from the Cress longitudinal dataset, with a mean age of 18.5 months

Play skills from the Carpenter play scale (Carpenter, 1987) were coded from play samples with parents and an experimenter (the first author).

Unstructured play episodes with parent were contrasted with structured play episodes with the experimenter during either the CSBS (Wetherby & Prizant, 1993) or structured Preference screener (if children did not demonstrate sufficient intentional communication to justify CSBS).

Carpenter playscale:
1. Mouthing objects - child places an object in his/her mouth
2. Banging objects - child bangs an object or hand against another object
3. Visual regarding - child looks at an object for at least three seconds
4. Manipulating objects - child manipulates a moveable, yet attached part of an object
5. Approximating objects - child brings two or more objects (including self) into contact with each other in ways around the function of objects
6. Semi-appropriate toy use - child's action involved is appropriate to the object but is truncated, fleeting or not executed accurately
7. Nesting objects - child places, tosses, crams, or stuffs object(s) into a larger container
8. Grouping - child creates a group having at least three objects
9. Appropriate toy use - child orients objects correctly and actions are realistic and complete
 

Group Results
· Children displayed all play episodes with the exception of Nesting Objects and Grouping.

· The most frequent play type with both parent and experimenter was Visual Regarding and the most frequent complex play type was Appropriate Toy Use. Although Appropriate Toy Use was the highest level of play exhibited, minimal time was demonstrated in this skill.

· Group results for parent and experimenter play did not differ in amount and type of play.

· However, nonsignificant trends indicated that children tended to show relatively more of all play categories with parents than experimenter except for "approximating objects" (greater with experimenter), and visual regard (equivalent with both partners).

· Differences in play duration with experimenter ranged from 16% (Banging toys) to 88% (Manipulating objects) of the time spent in the same category of play with the parent.

Individual Results
· Only one of the ten children did not show visual regard as the most frequent play type.

· Several of the children who demonstrated primarily visual regard or manipulating objects with parents showed primarily semi-appropriate or appropriate use with the experimenter.

· This was possible when the experimenter presented toys with relatively simple operations that the children could operate with minimal physical effort (e.g. push button music toys).

· Because motorically simpler toys tend to be also cognitively simpler, (i.e. not inviting combinatorial or symbolic play), it may not be appropriate to automatically conclude more cognitively advanced play from these episodes.

Conclusions:
Limitations to using standardized play measures for children with PI include:

· Children with PI's strongest communication skills were not accounted for as play behaviors.

· If children with PI can approximate objects and people through communication, they may be performing at a higher level of activity than a child who is physically manipulating the toy.

· For example, one child with PI directed the experimenter to blow more bubbles using eyepointing, but this was scored no higher than visual regard on the standardized scale.

· The most complex play episodes were only briefly demonstrated, with either partner.

· Play may have been adapted or modified by partners to maximize child success, and parents and clinicians may differ in the ways they adapt play for children. If so, appropriate toy use may not show expected cognitive understanding of relationship of play to other activities.

· Parents & children w/PI tend to rely more on social play than object play in typical interaction.

V. Study 2: Development of Alternative Play Coding for Social & Object Play (Cress & Anderson, 2000) -- Pilot of alternative social/object play assessment strategy

Methods:
For children with PI, using object play to measure cognition may not be feasible

Some of the qualities of social play that reflect cognition similar to object play scales include: attention, cause and effect, object permanence, and gestural imitation (Guerette, et al, 1999).

In the Carpenter scale (1987), the important cognitive elements of object play can be represented as concepts that can be expressed in social play. For instance, associating two objects together in object play could include association between child and parent in space.

Each of the elements in the Carpenter play scale were adapted to conceptual qualities that can be measured in social play as well as object play, scored on a five-point Likert scale.

The association between adapted play scale terms and Carpenter terms is provided in the following table.

Adapted play scales were coded in free parent play for 11 children w/PI mean age 10.8 mo.

Group Observations: Best/Worst/Most Variable Play Categories

· Best skill areas for children with PI tended to have low hand/arm physical requirements: - #1 Child Initiative, #2 Anticipating Results, #3 Number of Items involved (e.g. child, adult, experimenter, toy, event), #6 Relationship of Child Action to (apparent) Purpose, e.g. more tickle

· While these skills may show multi-step actions and ability to cognitively plan actions, there were co-active play episodes where children's initiated acts could have been cued by adults.

· Child skills that were considered likely to be associated with child cognitive play intent include: consistency, directness of association with action, initiating part of act, promptness.

· Worst skills were closest to standard play scales, in which children had to independently initiate multiple different actions; many children instead may use multiple meanings of the same actions: e.g. #4- Number/type actions, #9- Play schemes, #10-- New application of play schemes

· Categories 11-12 (multiple schemes) were not scored because relevant activities did not occur, and multiple different acts were difficult for children to control without extensive partner help.

· Variability was high in categories like cognitive distance, since many augmented play skills like eyepointing require implicit cognitive distance, but is interpreted variably by the adult

· Group/spatial relation skills were variable by the number of tools or acts the child could control

· Some scales were interrelated because of the co-active nature of social play, and difficult to score separate from adult acts, e.g: play schemes, object/purpose relationship, anticipating delay.

Group Observations: Disability Characteristics Associated with Play Skills

· The coding categories of object/purpose relationships and anticipating delay tended to give children relatively higher skill estimates, and tended to be frequent with higher mental age.

· Even children with very limited physical skills but higher mental ages scored well on items like initiating independent acts, which is possible with proper positioning in familiar action routines, while the same children scored no higher than "visual regard" on the Carpenter scale.

· Children with fewer physical skills tended to be more parent-dependent for introducing and holding play objects, and interpreting child actions toward parents and toys. Some parents may have structured play to avoid child failure, and used actions or objects their child could control.

· Even children with relatively high mental age scores showed little generalization of schemes to new actions. While this is expected w/PI, it often has been associated with cognitive disabilities.

· Children with PI showed poor symbolic play, due to physical limitations in controlling multiple play schemes such as scooping, stirring, and feeding dolls used on symbolic play scales.

Clinical Implications
· If children have limited access strategies or support in communication they are unlikely to be able to demonstrate their play skills in standard assessment measures.

· We need to rely on social play as well as object play in scales like this to observe critical play skills like intent, multi-tasking, anticipation, purpose, and cognitive distance from activity.

· Social play, particularly with familiar partners and activities, is naturally co-active, and it may be difficult to determine the extent to which children initiate acts, but this is unavoidable with PI.

· If children initiate acts in some play contexts with partner support, these are key activities in which the child may be able to learn the independent impact of their behaviors in managing play.

· Pretend play for children with PI requires more complex cognitive skills than expected. For instance, directing someone else in how to feed a doll requires an abstract distance from the actual toy as well as several communicative steps between the child and the play results.

· Symbolic play may not be equally indicative of cognitive skills for children with PI for reasons beyond simple physical or sensory limitations. Symbolic play is also influenced by setting, space, length, listener behaviors, task complexity, and quantity and social nature of the toys (Rettig, 1998).

VI. Study 3: Parent Directiveness with their Children with P I
(Cress, Moskal, & Benal, 2000)

· Parent responsivity is one of the more reliable predictors of early communicative transitions.

· Directive communicative style is associated with poorer communication patterns in their children (Tannock, Girolametto & Siegel, 1992). Yet, some types of directiveness, that follows the child's action and attention, can support development (Akhtar, Dunham, & Dunham, 1991).

· Multiple researchers have reported that parents of children with disabilities tend to be more directive than expected. For instance, Hanzlik (1990) found caregivers of young children with PI tend to be more directive during interactions than caregivers of children without disabilities.

· However, these interactions involved object play that is typically difficult for these children, and increased directiveness may be a functional strategy to increase child success with tasks.

Methods:
· 25 children with PI randomly selected, mean age 17.5 mo., developmental age 9.2 mo.

· Data for two comparison groups were taken from the means and standard deviations published in Hanzlik (1990): 20 typically developing children, and 20 children with cerebral palsy (CP) of varying physical severity. Both groups had a mean developmental age of 8.8 mos. (range 3-18 mos).

· This study used Hanzlik's coding scheme for parent behaviors, including verbal and physical directiveness, holding, face-to-face positioning, and physical contact. In addition, parent initiations were coded as verbal or nonverbal, and responses as contingent or non-contingent without distinguishing "positive". In addition, categories of verbal and physical directiveness, and parent imitation of child were added to account for relevant parent behaviors observed.

· These categories of parent initiation, response, directiveness, and positioning were coded from parent free play samples of children with PI and compared statistically to the Hanzlik data.

Directiveness Results:
· Parents of children with PI were no more verbally (t(45)= -.22, p>.05) or physically directive (t(45)= .80, p>.05) in free play than parents of typically developing children, and significantly less directive than the parents of children with CP in the Hanzlik study.

· This discrepancy may relate to the physically challenging object play emphasized in the Hanzlik and other directiveness studies, and the predominantly social play seen in the present study.

· The few directive behaviors that did occur were not significantly associated with contingency, but parent non-contingency was not significantly related to any child developmental measures.

· Parent directiveness may be an appropriate compensation for children with PI in object play.

· In social play, parents tended to initiate routines that the child could successfully initiate and control without need for parent directiveness (e.g. tickling that children initiated with a body lean)

Clinical Implications:
· Parents of children with PI may structure interactions so children's gestural and vocal behaviors can play a natural role in initiating and responding to parent's communication.

· Therefore, intervention might be able to expand and elaborate existing parent-child play patterns rather than having to reduce presumed inappropriate play interactions to promote child initiations.

· Diagnosis of PI alone is insufficient to explain the directiveness differences. In this study, 11 of 27 children had cerebral palsy, with comparable relative mobility skills to Hanzlik for all diagnoses.

· When parents elicit interactions in which their child's behavior is readable, they can respond with appropriate contingency regardless of the severity of child expressive or receptive language skills.

· Some structure in routines, whether social or object play, may facilitate child cognitive and language development by reducing the cognitive load of attending to and interpreting adult actions.

· Intervention strategies involving "test questions" to which the adult already knows the answer may reflect naturally-occurring patterns of routine interaction, if the parent's question continues the child's topic. Yoder and Davies (1992) found that young children with developmental delays tended to produce more frequent language and more diverse vocabulary in routines than in non-routines.

VII: Study 4: Mastery Motivation in Social and Object Play in Children with PI (Linke & Cress, 2000)

· Mastery motivation measures how well the child's interest and goal-directed focus leads to active exploration and problem solving about a task. Qualities like persistence and the range of strategies tried in resolving problems and play tasks are typical ways of measuring mastery.

· Children show best mastery motivation on tasks that are neither too easy nor too hard, since they can solve easy problems without trying new strategies, and will give up if they are too hard.

· Jennings, et al. (1985) found PI affected children's mastery motivation, particularly when children had to structure their own activities. Play was less complex, more solitary, and less social, and parents rated their children as less motivated and more likely to seek out easy or familiar activities.

· Older children with PI at 3 years showed equivalent curiosity but less persistence than preschool peers, but by 4 years began to consistently choose easier and less novel tasks (Jennings et al, 1988). Children who were relatively less motivated at 3 years tended to show low motivation at 4 years, regardless of severity, suggesting mastery patterns even in children with PI are individually stable.

· However, Hauser--Cram (1996) found that children PI ages 1-2 did not differ in persistence, goal orientation, positive affect or competence in play from developmental peers without disabilities.

· Traits that accompanied giftedness in school-aged non-speaking children were characteristics of mastery such as goal orientation, persistence, determination, and patience (Willard-Holt, 1998).

· We need better evidence of how mastery skills are demonstrated in children with PI, to facilitate these skills in social and object play, and predict impact of these personality traits on development.

Methods:
· Randomly selected 25 children with PI, mean corrected age 17.7 mo. (range 12-26 mo.)

· Mean age equivalents were: Overall development 9.0 (range 2-19), receptive communication 13.5 (range 4-29), cognitive development 10.1 (range 2-19.5), motor development 6.3 (range 2-23).

· Free play with parents was scored on a 4-5-point scale by Seifer's Mastery scale (1996), adapted for shorter intervals and to account for reliability and variability in physical skills: Anticipatory Affect, Attention to Task, Extent & Variety of Exploration, Social Interchange with Adult, Social Reference to Adult, Positive Affect, Negative Affect, Degree of Involvement, and Persistence.

· Each segment was also coded for primarily social play, object play, or mixed play activities. Children's mastery motivation responses were compared between their social and object play.

· Because of the adaptations in the coding scheme, it was not feasible to compare children's mastery in this study to typically developing children in similar studies.

Results:
· Children were more persistent and attended to task better during object play than social play. Social Interchanges, Social Referencing, and Positive Affect were more frequent in social play.

· Because persistence in object play tends to be used as an estimate of all mastery skills, this may explain others' findings of impaired mastery skills in children with PI with poor object play skills.

· Persistence or attention alone can be inappropriately high in some children with repetitive behaviors (Hupp & Abbeduto, 1991) and social mastery can help disambiguate these behaviors.

· Ways in which children with PI show mastery skills may differ from typically developing children. For instance, Attention to Task in Seifer was changed to Attention to Task/Adult because children with PI spent considerable time in play without specific objects or tasks to which they could attend.

· Motor Age was correlated with Extent and Variety of Exploration, and Persistence. Less impaired children have more exploring behaviors in their repertoire and more varied skills at task completion.

· Cognitive Age was correlated with Extent and Variety of Exploration, Degree of Involvement, and Persistence. Only Involvement was independent of mastery skills influenced by motor impairment.

· Receptive Language was correlated with Extent and Variety of Exploration, Persistence, Positive Affect, & Social Interchange. Only the latter two skills were independent of motor or cognitive skills

Clinical Implications:
· Since mastery motivation is a characteristic distinct from other aspects of communication, reliable assessments of mastery might help to predict successful acquisition of augmented communication beyond the prediction from children's motor or cognitive skills.

· When measuring mastery skills, both social and task-related mastery variables are necessary to interpret how functional or appropriate a child's mastery behaviors are in interactions.

· One of the few factors that distinguished estimated cognitive impairment uniquely from motor impairment was degree of task involvement, which can be assessed in social as well as object play.

· It may not be representative to interpret mastery for children with PI from only object play, or persistence/attention skills alone, particularly with multiple physical and/or sensory impairments.

· To develop mastery motivation for children with PI, partners need to create environments in which children can independently initiate involvement and encourage creative task problem-solving.

· Creating meaningful interaction and play tasks that can be independently managed by children with severe PI, however, is an ongoing challenge that confronts AAC and assistive technology fields.

· In order to successfully select and manage multiple AAC modalities, children must demonstrate creativity in attempting alternative strategies to solve a problem. Persistence in conveying an exact message is necessary to overcome frequent communication breakdowns that are common in AAC.

· Helping parents gauge task difficulty and prompting child independent exploration and persistence may be helpful in fostering improved mastery motivation (Hauser-Cram & Shonkoff, 1995).

· For instance, by recognizing and responding differently to child indications of interest versus frustration with a challenging task, partners can determine when assistance may be necessary or may interfere with the child's sense of their own efficacy in achieving goals or resolving breakdowns.

Summary of Conclusions:
· Traditional play schemes tended to underestimate children's cognitive skills because of the reliance on object-based play, particularly for children with more severe PI.

·  Piagetian cognitive skills represented in object-based play scales can be represented in a conceptual scale that rates children's social (or object) play on 5-point Likert scales for quality of play.

·  In social play, parents of children with PI are no more directive than typically developing parents and less directive than the parents of children with cerebral palsy reported in the Hanzlik study.

·  Parent directiveness in the present study was no more likely than the Hanzlik study to be associated with adult contingency, possibly because directiveness was relatively infrequent.

·  Children demonstrate some of the qualities of mastery (e.g. referencing, frequency of initiation) more effectively during social play than object play. Mastery showed considerable individual variation in children with PI, but Degree of Involvement was not significantly affected by motor age.

Overall Clinical Implications:
· Children with physical impairments may learn critical cognitive and communicative skills in social play interactions, skills that typically developing children learn most often during object play.

· For instance, even simple skills like means/end control may be easier to develop in social play, since the adult is a more complex and sensitive play object that easily responds to play initiations.

· Parents tend to be skilled at initiating appropriate play activities for their children with PI, and can elicit relatively complex play behaviors in social as well as object play environments.

· We need not assume that parents need to reduce their directiveness, but instead work together to interpret children's signals, and both provide play experiences that allow children maximal control over their environment and also demonstrate how to resolve breakdowns in contexts as they occur.

REFERENCES:
Akhtar, N., Dunham, F., & Dunham, P. (1991). Directive interactions and early vocabulary development: The role of joint attentional focus. Journal of Child Language, 18, 41-49.

Carpenter, R.L. (1987). Play scale. In L.B. Olswang, C. Stoel-Gammon, T.E. Coggins, & R.L. Carpenter Assessing prelinguistic and early linguistic behaviors in developmentally young children. Seattle: University of Washington Press.

Cress, C.J. (in press). Expanding children's early augmented behaviors to support symbolic development. In J. Reichle, D. Beukelman, & J. Light (eds), Volume 2: Implementing an augmentative communication system: Exemplary strategies for beginning communicators. Baltimore: Paul H. Brookes.

Cress, C.J. & Anderson, V.A. (2000). An adapted play analysis scheme for describing complexity of play in children with physical impairments. Unpublished manuscript, University of Nebraska-Lincoln.

Cress, C.J. & LaMontagne, J. (1998). Play patterns of infants with physical impairment across interactants and play characteristics. Manuscript in submission, University of Nebraska-Lincoln.

Cress, C.J., Moskal, L., & Benal, A. (2000). Parent directiveness in social and object play with their children with physical impairments. Unpublished manuscript, University of Nebraska-Lincoln.

Ganzeboom, H. & Treiman, D. (1996). Internationally comparable measures of occupational status for the 1998 International Standard Classification of Occupations. Social Science Research, 25, 201-239.

Guerette, P., Furumasu, J., Moy, F., & Tefft, D. (1999). Development of a cognitive assessment battery for young children with physical impairments. Infant-Toddler Intervention, 9. (2), 169-184.

Hanzlik, J. (1990). Nonverbal interaction patterns of mothers and their infants with cerebral palsy. Education and Training in Mental Retardation, 25, 333-343.

Hauser-Cram, P. & Shonkoff, JP (1995). Mastery motivation: Implications for intervention. IN R.H. MacTurk & G.A. Morgan (eds), Advances in applied developmental psychology: Vol. 12. Mastery motivation: Origins, conceptualizations, and applications. Norwood, NJ: Ablex.

Hupp, S.C., & Abbeduto, L. (1991). Persistence as an indicator of mastery motivation in young children with cognitive delays. Journal of Early Intervention, 15, 219-225.

Jennings,K.D., Connors, R.E., Stegman, C.E., Sankaranarayan, P., & Mendelsohn, S. (1985). Mastery motivation in young preschoolers: Effect of a physical handicap and implications for educational programming. Journal of the Division for Early Childhood, 9, 162-169.

Jennings, KD, Connors, RE, & Stegman, CE (1988) Does a physical handicap alter the development of mastery motivation during the preschool years? Journal of American Academy of Child and Adolescent Psychiatry, 27, 312-317.

Linke, M.L. & Cress, C.J. (2000).Mastery motivation in object and social play for children with physical impairments. Unpublished manuscript, University of Nebraska-Lincoln.

McDonald, E.T. (1980). Early identification and treatment of children at risk for speech development. In R.L. Schiefelbusch (Ed.), Nonspeech language and communication: Analysis and intervention, pp. 49-80.

Rettig, M. (1998). Environmental influences on the play of young children with disabilities. Education and Training in Mental Retardation and Developmental Disabilities, 33 (2), 189-194.

Seifer, R. (1996). Mastery motivation tasks scoring manual. Presentation at the International Conference on Infant Studies. Providence, RI: Brown University.

Tannock, R., Girolametto, L. & Siegel, L.S. (1992). Language intervention with children who have developmental delays: Effects of an interactive approach. American Journal on Mental Retardation, 97 (2), 145-160.

ACKNOWLEDGMENTS:
Supported in part by research grant #1 K08 DC00102-01A1 from the National Institute on Deafness and Other Communication Disorders (NIDCD), National Institutes of Health. The authors also appreciate the critical contributions of research assistants Kelli Bartels, Alicia Koch, Megan Magnusson, Sara Havelka, Cindy Dietrich, and Jody Elliott, as well as all of the families and children that participated in these research activities.