care, health and development Child: Original Article doi:10.1111/j.1365-2214.2010.01089.x Out-of-school time activity participation profiles of children with physical disabilities: a cluster analysis cch_1089 726..741 G. King,* T. Petrenchik,† D. DeWit,‡ J. McDougall,§ P. Hurley† and M. Law† *Bloorview Research Institute, Toronto †CanChild Centre for Childhood Disability Research, Hamilton ‡Centre for Addiction and Mental Health, London, and §Thames Valley Children’s Centre, London, ON, Canada Accepted for publication 30 January 2010 Abstract Keywords child, disability, leisure, out-of-school, participation Correspondence: Gillian King, Bloorview Research Institute, 150 Kilgour Road, Toronto, ON, Canada M4G 1R8 E-mail: gking27@uwo.ca Objective To determine out-of-school activity participation profiles of school-aged children with physical disabilities. Methods Activity participation profiles were determined by cluster analysing 427 children’s responses on multiple dimensions of participation (intensity, location, companionship, enjoyment, preference) in five activity types (recreational, active physical, social, skill-based, self-improvement). Socio-demographic, child, parent, family and environmental predictors of group membership were determined, along with child functioning, socio-demographic, self-concept and social support variables significantly associated with group membership. Results The cluster analysis revealed four groups, labelled Social Participators (a highly social and neighbourhood-focused group), Broad Participators (a group of high participators who enjoy participation), Low Participators (a group with low enjoyment and weak preferences) and Recreational Participators (a group of younger children who participate in recreational activities with family members). The groups showed meaningful differences across a range of socio-demographic, child, parent, family and environmental variables. Conclusions The findings support an affective and contextual view of participation, indicating the importance of motivational theory and a person–environment approach in understanding the complexity of children’s out-of-school activity participation. There is growing interest in determining the nature of children’s activity profiles. Examining the determinants and consequences of membership in different leisure or school-based activity profiles may lead to a deeper understanding of the factors influencing participation and the mechanisms by which participation influences development (Feldman & Matjasko 2005). Practically speaking, children fitting the different profiles may benefit from different types of interventions. Several recent studies have used cluster analysis to determine the activity profiles of adolescents who are typically developing, based on their out-of-school time use or frequency of partici- 726 pation in different sets of leisure activities (Raymore et al. 1999; Bartko & Eccles 2003; Roth et al. 2005). These studies have determined clusters of adolescents based on what they do, rather than other dimensions of participation such as activity setting (location), nature of their companions or enjoyment. For example, Raymore and colleagues (1999) derived clusters from time use data for 954 typically developing adolescents and young adults. Although the leisure patterns were slightly different for males and females, both sexes displayed a Positive Active pattern of participation in socially valued activities, such as volunteer work, clubs and religious activities. Both sexes also © 2010 Blackwell Publishing Ltd Activity participation profiles 727 displayed a Risky pattern defined by higher frequencies of drug and alcohol use, and a Diffused pattern characterized by low participation in activities. Roth and colleagues (2005) derived five distinct clusters based on adolescents’ non-school time use in 21 activities, including a group that watched television and a group that took part in social activities. Bartko and Eccles (2003) identified profiles for adolescents without disabilities based on their frequency of participation in organized and relaxed leisure activities, and then examined psychosocial indicators for the groups. Adolescents’ activity involvement was found to be related to their psychological and behavioural functioning. Little is known, however, about the profiles of recreation and leisure activity participation for school-aged children who have physical disabilities (i.e. chronic conditions associated with physical functional limitations, such as cerebral palsy, spina bifida and non-progressive muscular disorders). Two recent studies examined the preschool or school participation of children with physical and developmental disabilities (Almqvist & Granlund 2005; Almqvist 2006), clustering variables thought to be associated with degree of participation (e.g. children’s autonomy scores, parent support and socio-economic status) in addition to aspects of the participation experience itself. Almqvist (2006) examined patterns of home and school engagement of young children ages 1–4 years with and without developmental delay. The cluster analysis of demographic and biopsychosocial variables revealed five clusters, labelled High-Level Engagement, Interaction-Related Engagement, Low-Level Engagement, Activity-Related Engagement and Conflict-Level Engagement. Children with developmental delay were found in all the clusters, but particularly Low-Level Engagement, indicating large within-group variability in developmental outcome. Almqvist and Granlund (2005) derived clusters for students with visual, motor and multiple disabilities, ages 7–17, and then determined relationships between these clusters and students’ degree of participation in school activities considered as a whole. The groups with a high degree of participation had high scores in autonomy, greater perceived interaction with peers and teachers, and internal locus of control, thus pointing to the role of intrapersonal factors in influencing extent of school participation. Type and degree of disability did not predict cluster group membership, indicating that participation is better predicted by patterns of child functioning and environmental factors. These findings support a noncategorical approach emphasizing commonalities in the issues and experiences of children with disabilities (Pless & Pinkerton 1975). The present study contributes to this sparse literature by examining the activity profiles of a large, representative sample of children ages 6–14 with physical functional limitations, based on their self-reports of multiple dimensions of participation in recreation and leisure activities (i.e. intensity, location, companionship, enjoyment and preference), rather than frequency or time-use data, or parent/teacher ratings of teacher–child or child–peer interaction. The present study examined participation in five types of activities (recreational, active physical, skill-based, social and self-improvement) derived from a comprehensive set of 49 out-of-school time activities. Studies have typically examined frequency of participation in small sets of activities (Feldman & Matjasko 2005). For example, Raymore and colleagues (1999) examined clusters of leisure behaviour patterns based on frequency of adolescents’ participation in what the authors themselves described as a limited set of 12 activities. Dimensions of activity participation To obtain a rich understanding of different patterns in which children participate in their worlds, we considered five fundamental dimensions of participation – behavioural aspects (intensity), contextual or environmental aspects (where and with whom activities take place), and affective or motivational aspects (enjoyment and preference). To have a thorough understanding of meaningful participation, it is important to assess different dimensions of participation (Law 2002; King et al. 2004). Intensity, enjoyment and preferences are typically examined, but location (where) and companionship (with whom) have received less attention. Companionship is, however, considered to be as important as the nature of activities in which children participate (Whiting & Edwards 1988) and peer companionship has been found to lead to benefits of extracurricular participation for typically developing adolescents (Stone et al. 2005). As well, there is increasing acknowledgement of the importance of understanding the ‘lived experiences’ of children with disabilities (James et al. 1998), including their enjoyment of activities and preferences. The present study makes a unique contribution to the literature by determining activity participation profiles for children with physical disabilities based on self-reports of affective features of participation (i.e. enjoyment and preference) and contextual features (i.e. location and companionship) in addition to participation intensity. Our previous research using this dataset of 427 children with physical functional limitations has examined the predictors of formal and informal participation intensity, using structural equation modelling (King et al. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 728 G. King et al. 2006a), and the predictors of change over time in activity participation, using latent growth curve modelling (King et al. 2009), but has not determined children’s profiles based on multiple dimensions of participation. By examining relationships among multiple dimensions of experience, our approach reflects a holistic view, in which participation is considered to be a complex and multiply determined process, changing in response to emerging developmental needs and shifting opportunities, contexts and settings (Overton 1998; Wapner & Demick 1998). To understand a complex adaptive system such as children’s participation, it is important to understand recurring patterns in the system and the role played by contextual features (Plsek & Wilson 2001). Examining participation profiles reflects a person-oriented approach (Magnusson & Stattin 2006) and helps us to understand activity experiences in a holistic, interrelated and dynamic manner. Cluster analytic techniques, seldom used in paediatric rehabilitation research, are the ideal tool to interrelate aspects of the participation experience into meaningful constellations. Objectives The primary objective of this secondary analysis of a large survey dataset was to determine whether meaningful activity participation profiles could be identified (using cluster analysis) for a sample of children ages 6–14 years with physical disabilities, based on self-reports of participation intensity, location, companionship, enjoyment and preference in five types of activities. The second objective was to determine significant socio-demographic, child functioning, parent health, family functioning and environmental predictors of group membership, using multinomial regression. The third objective was to determine the independent contributions of child functioning variables to group membership by examining whether children in the groups differed in psychosocial, physical and communicative functioning, controlling for the significant predictors of group membership. The fourth objective was to describe the socio-demographic characteristics, self-concepts and perceived social support of children in the groups. Method Determinants of activity profiles To the best of our knowledge this is the first study to examine factors associated with comprehensive profiles of out-of-school activity participation for children with disabilities. In the broader literature on childhood participation, few studies have examined factors associated with children’s activity profiles. The factors that have been examined include children’s sex (Raymore et al. 1999) and problem behaviour and mental health (Bartko & Eccles 2003). A broader consideration of factors will provide greater insight into the nature of profiles derived through cluster analysis. Our selection of predictive factors was informed by: (i) a conceptually and empirically based model of predictors of children’s participation (King et al. 2003); and (ii) significant predictors of children’s participation intensity at a single point in time (King et al. 2006a) and longitudinally (King et al. 2009). Five sets of factors were of interest: (i) socio-demographic factors (e.g. child age and sex, parent ethnicity); (ii) child functioning factors (psychosocial, physical and communicative functioning); (iii) parent health factors (physical and mental health); (iv) family functioning factors (e.g. family environment); and (v) environmental factors (parents’ perceptions of the community environment). These sets of factors provide a more detailed categorization of the child, family and environmental factors outlined in Bronfenbrenner’s (1979) socioecological model. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 Participants The participants were 427 school-aged children (229 boys, 198 girls), ages 6–14 years, with physical disabilities. Ethical approval for the study was obtained from McMaster University. Eleven regional children’s rehabilitation centres and one children’s hospital in the province of Ontario, Canada, assisted with recruitment. Working with the recruitment sites, a list was compiled of all children with physical disabilities born between 1 October 1985 and 30 September 1994 inclusive. Children with the following primary diagnoses/conditions were included: amputation; cerebral palsy; cerebral vascular accident/stroke (vascular brain disorders); congenital anomalies; hydrocephalus; juvenile arthritis; non-progressive muscular disorders; neuropathy; orthopaedic conditions (e.g. scoliosis); spinal cord injury; spina bifida; and traumatic brain injury. For purposes of analysis, children were grouped into those with central nervous system disorders (79.6%) and musculoskeletal disorders (20.4%), as done by Law and colleagues (2004). A total of 3062 children met the inclusion criteria with respect to age and physical functional limitation. Of the 3062 families sent packages, 84 did not have a valid mailing address or the child was deceased, and 510 were determined to be ineligible for various reasons (based on information they provided at the recruitment stage), including the capability of their child to take part in research. Of the remaining 2468 families, 1442 Activity participation profiles 729 Table 1. Child, parent, and family characteristics at time 1 (n = 427) Characteristic Frequency Child’s sex Male 299 Female 198 Child’s age (M = 10.37, SD = 2.37, range = 6–15 years) 6–10 years 225 11–15 years 202 Child’s primary health and development problem Central nervous system disorders 340 Musculoskeletal disorders 87 Family type Two-parent 348 Single-parent 67 Missing 12 Ethnic background Caucasian 345 Non-Caucasian 79 Missing 3 Respondent age (years) 20–29 19 30–39 154 40–49 218 50–59 35 Missing 1 Respondent sex Male 48 Female 375 Missing 4 Educational attainment of respondent Completed elementary, some high school 47 Completed high school, some college/ 137 technical training Completed college/technical training, some 139 university Completed university 104 Total family income (Can $) <30 000 68 30 000–44 999 79 45 000–59 999 74 60 000–89 999 114 >90 000 87 Missing 5 Valid % 53.6 46.4 parent households (83%). Participants were predominantly of Caucasian background (81%). Just over 53% of families reported annual household incomes of less than $60 000. The median family income in the province of Ontario at the time of data collection was $61 000 (Statistics Canada 2001). 52.7 47.3 79.6 20.4 81.5 15.7 2.8 80.8 18.5 0.7 4.5 36.2 51.2 8.2 NA 11.3 88.7 NA 11.0 32.1 32.6 24.4 16.1 18.7 17.5 27.0 20.6 NA made no response (58.4%) and 557 were not interested in participating (22.6%), leaving 469 consenting families (a 19.0% consent rate). Of the 469 enrolled families, 28 withdrew before data collection and 14 were judged to be unsuitable by the interviewer, leaving 427 children in the study (a 17.3% useable response rate). We broadly sampled from rehabilitation centres in both rural and urban areas of the province. The majority of children with physical disabilities have the opportunity to be served by these centres. Table 1 presents the characteristics of the participating children, parents and families. Mothers were the primary parent respondents (89%) and the majority of children lived in two- Procedure A package of questionnaires was mailed to the family, with measures to be completed by either the parent or child prior to a home visit. The parent-completed measures included a demographic questionnaire, the Family Environment Scale (FES; Moos & Moos 1994), the Impact on Family Scale (IOF; Stein & Riessman 1980), the Strengths and Difficulties Questionnaire (SDQ; Goodman 1997), the Craig Hospital Inventory of Environmental Factors (CHIEF; Whiteneck et al. 2004) and the MOS 36-Item Short-Form Health Survey (SF-36; Ware et al. 1993). The child-completed measures included phase one of the Children’s Assessment of Participation and Enjoyment (CAPE; King et al. 2004) and the Activities Scale for Kids (ASK; Young et al. 2000). Parents also completed the Communication Domain of the Vineland Adaptive Behavior Scales (VABS; Sparrow et al. 1984) via a phone interview. Families then were contacted by one of 15 experienced study interviewers to arrange a home-based interview in which the child completed phase two of the CAPE (King et al. 2004), the Preferences for Activities of Children (PAC; King et al. 2004), the Self-Perception Profile for Children (Harter 1985a) or the Self-Perception Profile for Adolescents (Harter 1988), depending on the age of the child/youth, and the Social Support Scale for Children (Harter 1985b). Measures of dimensions of participation Children’s Assessment of Participation and Enjoyment The CAPE is a reliable and valid self-report measure of participation for children/youth ages 6–21 that includes both formal and informal domains, and five activity types: recreational, active physical, social, skill-based and self-improvement (King et al. 2004; 2006b). The CAPE assesses five dimensions of children’s participation in recreation and leisure activities (i.e. diversity, intensity, location, companionship, enjoyment). The conceptual strengths of the CAPE include its measurement of multiple dimensions of participation (Imms 2008). The CAPE has been used to examine the participation of children/youth with physical disabilities, primarily those with cerebral palsy, in © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 730 G. King et al. Table 2. CAPE activities in the activity type groupings Recreational (n = 12) Active physical (n = 9) Social (n = 9) Skill-based (n = 9) Self-improvement (n = 10) Collecting things Crafts, drawing or colouring Pretend or imaginary play Doing puzzles Going for a walk or hike Playing board or card games Playing computer or video games Playing on playground equipment Playing with pets Playing with things or toys Taking care of a pet Watching TV or a video Bicycling, in-line skating or skateboarding Doing a paid job Martial arts Team sports Water sports Snow sports Playing games Racing, track or field School club participation Going on a full day outing Going to a live event Going to a movie Going to a party Hanging out Listening to music Making food Talking on the phone Visiting with others Dancing Doing gymnastics Horseback riding Learning to dance Learning to sing Participating in community organizations Playing a musical instrument Swimming Taking art lessons Attending church, temple or a religious activity Doing a chore Doing homework Doing volunteer work Getting tutored for schoolwork Going to the public library Reading Shopping Writing a story Writing letters CAPA, Children’s Assessment of Participation and Enjoyment. Australia (e.g. Imms et al. 2008), Canada (e.g. Law et al. 2006; King et al. 2006a; Majnemer et al. 2008) and the USA (e.g. Orlin et al. 2010; Kang et al. 2009). The CAPE items are inclusive of voluntary activities outside of school. The 49 specific activities assessed by the scales are presented in Table 2. In initial testing, the internal consistency reliabilities of the scales were found to range from 0.30 (skillbased) to 0.65 (recreational), and test–retest reliabilities for the activity type intensity scores ranged from 0.72 (social) to 0.81 (active physical) (King et al. 2004). In the present study, the CAPE was administered in two phases: (i) a self-administered questionnaire (with parent/ caregiver assistance, as needed) to determine what activities the child took part in (in the previous 4-month period) and how often (on a 7-point scale from 1 = ‘1 time in the past 4 months’ to 7 = ‘1 time a day or more’), followed by (ii) a home interview, which gathered information about with whom the child took part, where, and enjoyment. Participation intensity is calculated by dividing the sum of item frequency by the number of possible activities in each activity type scale. Intensity scores provide a relative indicator of participation frequency for a set of activities, which is useful for comparing participation across activity types (Imms 2008). The location of participation is scored on a 6-point scale [1 = home, 2 = relative’s home, 3 = neighbourhood, 4 = school (but extracurricular), 5 = community, 6 = beyond community]. Median scores were calculated for each activity type for each child, with lower scores indicating participation closer to home and higher scores indicating more community-based participation. Companionship of participation is scored on a 5-point scale (1 = alone, 2 = with family, 3 = with other relatives, 4 = with friends, 5 = with others). Median scores were calculated for each child, with lower scores indicating more solitary engagement in activi- © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 ties and higher scores indicating more social engagement. Enjoyment is measured on a 5-point scale ranging from 1 (not at all) to 5 (love it). Preferences for Activities of Children The PAC (King et al. 2004) was used to assess children’s selfreported preferences for recreational, active physical, social, skill-based and self-improvement activities. Activity preference is scored on a 3-point scale (1 = would not like to do at all, 2 = would sort of like to do, 3 = would really like to do). Mean scores were calculated for each child, with lower scores indicating lower preference and higher scores indicating greater preference. The PAC has demonstrated good internal consistency and construct validity (King et al. 2004; 2006a). Socio-demographic and child functioning variables Demographic questionnaire This questionnaire was completed by the parent or caregiver. It provided measures of child sex and age (in years), parent ethnicity, number of parents in the family, highest household educational attainment and annual household income. Strengths and Difficulties Questionnaire The SDQ (Goodman 1997) was completed by the parent. It consists of 25 questions providing measures of the child’s emotional problems, conduct problems, hyperactivity, peer Activity participation profiles 731 problems and prosocial behaviour. The SDQ has satisfactory internal consistency and test–retest reliability (Goodman 2001). Activities Scale for Kids The ASK (Young et al. 2000) is a child-report instrument measuring physical functioning for children 5–15 years of age. The ASK has excellent reliability (internal consistency, test–retest, inter-rater and intra-rater reliabilities of 0.94 or greater) and good construct and criterion validity (Young et al. 2000). Vineland Adaptive Behavior Scales The VABS (Sparrow et al. 1984) is a widely used measure, with excellent reliability and content, construct and criterion validity. The VABS is completed by the parent and provides a measure of the child’s adaptive function in the areas of communication, daily living skills, socialization and motor skills. We used the overall communication raw scores. health and well-being, including a Physical Functioning Scale and Mental Health Scale, which were used in the present study (Ware & Sherbourne 1992). The scales are widely used and have evidence of construct validity (Mchorney et al. 1993). Family Environment Scale The FES is a parent-completed instrument, providing measures of Family Cohesion, Active-Recreational Orientation and Intellectual-Cultural Orientation (Moos & Moos 1994). The Family Cohesion Scale measures the degree of commitment, help and support that family members provide to one another. Active-Recreational Orientation taps the extent of the family’s participation in social and recreational activities, and Intellectual-Cultural Orientation assesses the family’s degree of interest in political, social, intellectual and cultural activities. The FES has been widely used and has adequate test–retest reliability (scores for the 10 subscales range from 0.52 to 0.89) and adequate internal consistency reliability (scores ranging from 0.61 to 0.78) (Moos & Moos 1994). Self-Perception Profiles for Children and Adolescents The Self-Perception Profile for Children (Harter 1985a), appropriate for children in grades 3–8, taps domain-specific judgments of competence in five domains, as well as global self-worth. The measure has adequate internal consistency reliabilities, ranging from 0.71 to 0.86. The Self-Perception Profile for Adolescents (Harter 1988), recommended for youth in grades 9–12, includes the scales in the child version along with three additional scales relevant to adolescents. Internal consistency reliabilities of the scales range from 0.58 to 91. We calculated Z-scores for children in the sample on global selfworth, social acceptance, scholastic competence and athletic competence. Social Support Scale for Children This scale is completed by the child and measures parent support, classmate support, teacher support and friend support. The measure has adequate reliability and validity (Harter 1985b; Appleton et al. 1994). Parent health and family variables MOS 36-Item Short-Form Health Survey The SF-36 assesses parent reported health. The instrument contains 36 items and yields an eight-scale profile of functional Impact on Family The 24-item IOF yields a total score, used in the present study in accordance with the most current scoring protocol (Stein & Jessop 2003), and four subscores measuring the impact of living with a child with a disability on various components of family life. It has adequate internal consistency (Stein & Riessman 1980; Stein & Jessop 2003). Environmental variables Craig Hospital Inventory of Environmental Factors The 25-item CHIEF (Whiteneck et al. 2004) was completed by the parent. It quantifies the degree to which aspects of the physical, social and political environment act as barriers to full participation. The CHIEF contains five subscales (Policies, Physical/Structural, Work/School, Attitudes/Support and Services/Assistance), which have good test–retest and internal consistency reliability, and evidence of content, construct and discriminant validity (Whiteneck et al. 2004). We used the frequency-magnitude product scores (overall impact scores) from the CHIEF, which range from 0 to 8; these are the product of a frequency score (a 5-point scale ranging from ‘daily’ to ‘never’) and a magnitude score (a dichotomous scale of ‘big problem’ or ‘little problem’). © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 732 G. King et al. Data analyses Overview of analyses Hierarchical cluster analysis was performed to determine activity profiles. To interpret the groups, analyses were then conducted to determine (i) the predictors of group membership; (ii) the child functioning variables that discriminated the groups, over and above the predictors; and (iii) the sociodemographic, self-concept and social support variables associated with group membership. Missing values on the dependent variables were handled using mean substitution. There were very few missing cases (<6% of the sample). Table 3. Predictor variables Measure Socio-demographic variables Demographic questionnaire Child functioning variables Strengths and Difficulties Questionnaire Hierarchical cluster analysis procedure Cluster analysis is an exploratory classification method (Gorsuch 1983), which identifies types or classes of individuals by grouping together variables that are most alike. We used Ward’s method, which maximizes differences between clusters and is one of the most robust cluster methods (Milligan 1996; Henry et al. 2005). The 427 cases were classified based on their pattern of scores across 25 participation variables (intensity, location, companionship, enjoyment and preference scores for each of five types of activities) using the SPSS cluster analysis procedure, with squared Euclidean distances for each variable and standardization on the range. This procedure yields a tree diagram, or dendrogram, showing the arrangement and progression of the clustering. Activities Scale for Kids Vineland Adaptive Behavior Scales Parent health variables MOS 36-Item Short-Form Health Survey Family functioning variables Family Environment Scale Impact on Family Environmental variables Craig Hospital Inventory of Environmental Factors Variables Child sex Child age Parent/caregiver ethnicity Number of parents in the family Highest household educational attainment Annual family income Type of health condition (central nervous system or musculoskeletal) Child emotional functioning (Emotional Problems scale) Child behavioural functioning (Conduct Problems scale) Child hyperactivity Child peer problems Child prosocial behaviour Child physical functioning Child communicative functioning (overall communication raw scores) Parent/caregiver physical functioning Parent/caregiver mental health functioning Family cohesion Family active-recreational orientation Family intellectual-cultural orientation Impact on family (total score) Policies environment Physical/structural environment Work/school environment Attitudes/support environment Services/assistance environment Interpretation of clusters First, the role of 24 socio-demographic, child, parent, family and environmental variables (see Table 3) was examined using multinomial regression. Second, analyses of covariance (ancovas) were used to examine whether the groups differed on child psychosocial, physical and communicative functioning variables. Following the approach of Bartko and Eccles (2003), we selected covariates for these analyses that were significant predictors of group membership in the regressions. Third, differences between the groups were examined using analyses of variance (on child age, self-concept and social support) and chi-square tests (for child sex and type of health condition). Results Cluster analysis findings Following recommendations in the literature (Everitt 1974), the number of clusters present in the data was determined by exam- © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 ining the dendrogram and meaningfulness of each additional cluster in providing distinctly new and relevant patterns across the participation variables. A four-cluster solution provided the best fit to the data and provided a reasonable number of subjects within each cluster (Curry & Thompson 1982). All children were validly classified into one of the four groups. Table 4 presents mean scores for the four groups on dimensions of participation for each activity type. Post hoc analyses (Hochberg’s multiple comparison test) were performed to determine variables on which the groups significantly differed from one another, with alpha set at 0.01 to control for Type II error rate. As shown in Table 4, here were four substantial groups of children: Social Participators, Broad Participators, Low Participators and Recreational Participators. The Social Participators (n = 41) were socially inclined and neighbourhood-focused, with high social participation intensity and high enjoyment of social activities, who were most likely to take part in activities Activity participation profiles 733 Table 4. Dimensions of participation predicting group membership Predictors Intensity Recreational activities Active physical activities Social activities Skill-based activities Self-improvement activities Location Recreational activities Active physical activities Social activities Skill-based activities Self-improvement activities Companionship Recreational activities Active physical activities Social activities Skill-based activities Self-improvement activities Enjoyment Recreational activities Active physical activities Social activities Skill-based activities Self-improvement activities Preferences Recreational activities Active physical activities Social activities Skill-based activities Self-improvement activities Social Participators (n = 41) Broad Participators (n = 140) Low Participators (n = 122) Recreational Participators (n = 124) 19.17*** 4.22* 2.18 11.05*** 2.14 3.52ab 1.72 3.34 1.13 2.96 4.43ac 1.98a 3.34 1.31a 3.22 3.79cd 1.62a 3.06 0.78ab 2.96 4.45bd 1.66 3.19 1.15b 3.06 4.70* 0.58 1.75 5.60** 7.28*** 1.17 3.15 2.62 3.13 2.18abc 1.25a 3.06 2.33 2.66a 1.63a 1.04a 3.01 2.35 3.06a 1.50b 1.11 2.98 2.45 2.77 1.53c 54.54*** 16.17*** 15.98*** 118.56*** 4.21* 3.45abc 3.92ab 3.59abc 3.71a 2.15a 2.16ad 3.42c 2.59a 4.22bc 1.78 1.62bde 3.09a 2.64b 3.61bd 1.67a 1.96ce 2.75bc 2.44c 2.06acd 1.81 29.62*** 16.73*** 40.49*** 21.47*** 31.46*** 3.73ab 4.19 4.29a 3.94a 3.12 4.22ac 4.41a 4.44b 4.38ab 3.54a 3.74cd 3.83ab 3.78abc 3.65bc 2.69ab 4.16bd 4.27b 4.29c 4.20c 3.56b 65.92*** 19.15*** 22.68*** 60.84*** 40.38*** 2.15ab 2.25 2.59 1.90ab 1.98a 2.58ac 2.47a 2.72a 2.37ac 2.16bc 2.23cd 2.12ab 2.43ab 1.80cd 1.80bd 2.61bd 2.43b 2.67b 2.34bd 2.31acd F value Similar superscripts indicate pairs of means that are significantly different (P < 0.01) using Hochberg’s multiple comparison test. *P < 0.01;**P < 0.001;***P < 0.0001. with friends or relatives outside the immediate family. The defining features were the nature of their companionship and their low preference for and participation in recreational activities, which are often done alone at home. The Broad Participators (n = 140) were high participators who enjoyed participation and had strong preferences. They had the highest intensity of participation in active physical and skill-based activities, and high levels of enjoyment and preference for all activities. The Low Participators (n = 122) had particularly low participation in recreational, active physical and skill-based activities, and reported the lowest levels of enjoyment and the weakest preferences. The Recreational Participators (n = 124) had the highest level of participation in recreational activities, relatively high enjoyment and preference for all activity types, and were most likely to participate with family members. Interpretation of clusters To assist with interpretation, we first examined predictors of group membership. A series of socio-demographic, child functioning, parent health, family functioning and environmental variables (24 variables in all) were entered into a multinomial logistic regression with membership in one of the clusters as the dependent variable. The eight significant predictors of group membership are presented in Table 5. The multinomial regressions give the odds of a child falling into a group, in contrast to the last group (reference group). An odds ratio less than 1.0 indicates that group members are less likely to have high scores than the reference group, whereas a ratio greater than 1.0 indicates a greater likelihood of group members having higher scores. As shown in Table 5, in comparison with the Recreational Participators (the reference group), the other three groups were © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 734 G. King et al. Table 5. Significant predictors of relative group membership Odds ratios Predictors Socio-demographic variables Child age (in years) Parent ethnicity (non-Caucasian) Child functioning variables Peer problems Prosocial behaviour Parent health variables Parent physical health Family functioning variables Family intellectual-cultural orientation Environmental variables Work/school environmental barriers Attitudes/support environmental barriers Social Participators (n = 41) Broad Participators (n = 140) Low Participators (n = 122) Recreational Participators (n = 124) (reference group) 1.72*** ns 1.16* 0.36** 1.29*** ns 1.00 1.00 0.73** ns ns ns ns 0.80** 1.00 1.00 0.89* 0.90** 0.92* 1.00 0.72* ns ns 1.00 1.40* ns ns ns ns 0.74* 1.00 1.00 *P < 0.05;**P < 0.01;***P < 0.001. significantly older and had parents with poorer physical health. In addition, the Social Participators had a lower level of peer problems and their parents perceived greater school environment barriers to their child’s participation. The Broad Participators were more likely to be Caucasian. The Low Participators displayed lower levels of prosocial behaviour and their parents were less likely to perceive the presence of attitudinal barriers to participation. Table 6 presents the child functioning variables significantly differentiating the groups. Following the approach of Bartko and Eccles (2003), the ancovas controlled for six of the eight significant predictors of group membership determined in the regression analyses (i.e. child age, parent ethnicity, parent physical functioning, family intellectual-cultural orientation, parent perceptions of work/school environment and parent perceptions of attitudes/support environment). The two significant child psychosocial functioning predictors were not controlled for in these analyses because they were part of the series of dependent variables. The results indicate a differential pattern of relations between activity groups and child functioning, even after controlling for the six variables. There were significant differences between the cluster groups on social relational variables (peer problems and prosocial behaviour) and emotional difficulties, but not on behaviour problems, hyperactivity, or physical or communicative functioning. The Low Participators had significantly greater levels of emotional difficulties, greater peer problems and lower levels of prosocial behaviour. Social Participators displayed the lowest level of peer problems. As shown in Table 7, the groups were significantly different on child age, social acceptance, athletic competence and © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 classmate support. There were no significant differences on child sex or type of health condition. Discussion Children’s self-report data on five dimensions of participation (intensity, location, companionship, enjoyment and preference) in five types of activities (recreational, active physical, social, skill-based and self-improvement) were cluster analysed to determine the activity participation profiles of a representative sample of children with neurological and musculoskeletal conditions ages of 6–14 yeas. This analysis revealed four distinct and substantially sized groups, labelled as Social Participators (a highly social and neighbourhood-focused group), Broad Participators (a group of high participators who enjoy participation), Low Participators (a group with low enjoyment and weak preferences), and Recreational Participators (a group of younger children who participate in recreational activities with family members). The findings enhance our understanding of different ways in which physical disability can be experienced by showing, in a province-wide sample, that some children are highly social with relatively high self-concepts, others broad-based in their participation, and others more at risk for poorer developmental outcomes across childhood and adolescence. There were important and meaningful differences in the participation patterns of children with physical disabilities that were not associated with their type of health condition or level of physical or communicative functioning, but were associated with their emotional functioning, peer difficulties and prosocial behaviour. The Activity participation profiles 735 Table 6. Child functioning variables and group membership (analyses of covariance†) Variables Emotional difficulties Social Participators Broad Participators Low Participators Recreational Participators Behaviour problems Social Participators Broad Participators Low Participators Recreational Participators Hyperactivity Social Participators Broad Participators Low Participators Recreational Participators Peer problems Social Participators Broad Participators Low Participators Recreational Participators Prosocial behaviour Social Participators Broad Participators Low Participators Recreational Participators Physical functioning Social Participators Broad Participators Low Participators Recreational Participators Communicative functioning Social Participators Broad Participators Low Participators Recreational Participators Adjusted means† F value for cluster membership 2.86 3.07a 3.83a 3.11 3.66* 1.60 1.70 1.92 1.41 1.79 (ns) 3.81 4.35 4.56 4.11 1.12 (ns) 2.11ab 3.19a 3.48b 3.12 4.49** 8.17 8.05a 7.33ab 8.24b 5.03** 67.89 70.41 73.09 68.20 0.95 (ns) 109.88 111.07 111.60 110.55 0.19 (ns) findings parallel the large within-group variability in developmental outcomes reported by Almqvist (2006), who examined the home and school engagement patterns of young children with and without developmental delay. They also echo the findings of Bartko and Eccles (2003), who found typical adolescents’ activity involvement to be related to their psychological and behavioural functioning. We expected the profiles to reflect different intensities of participation in different types of activities, as found by researchers who have cluster analysed frequency data (e.g. Raymore et al. 1999; Bartko & Eccles 2003). Although this was the case, especially for recreational, active physical and skillbased activities, we also found the participation profiles to be meaningfully distinguished by affective or motivational dimensions of participation – enjoyment and preference – along with one of the contextual dimensions, namely companionship. Of all the measured dimensions, companionship was by far the most varied for the clusters. By deriving profiles based on multiple dimensions, rather than only intensity, the present study provides a richer understanding of children’s lived experiences. In addition, different constellations of predisposing and descriptive factors were associated with the profiles. There are distinct profiles of children that may resonate with the observations of experienced paediatric therapists. Interpretation of activity participation profiles Social Participators Similar superscripts indicate pairs of means that are significantly different (P < 0.05) with Bonferroni adjustments for multiple comparisons. †These analyses controlled for child age, parent ethnicity, parent physical functioning, family intellectual-cultural orientation, parent perceptions of work/school environment, and parent perceptions of attitudes/support environment. *P < 0.01; **P < 0.005. This was a group of socially inclined and neighbourhoodfocused older children (mean age 11.8 years) with the lowest level of peer problems, the highest self-perceived social acceptance and athletic competence, and the highest perceived classmate support. This group reported a high intensity of participation in social activities (e.g. going to a party, talking on the phone) and high enjoyment of social activities compared Table 7. Significant socio-demographic characteristics, self-perceptions and social support scores for cluster groups Variable F value Social Participators Broad Participators Low Participators Recreational Participators Mean age Social acceptance (Z-scores) Athletic competence (Z-scores) Classmate support F(3,423) = 12.8; P < 0.0001 F(3,423) = 2.7; P < 0.05 F(3,423) = 5.0; P < 0.001 F(3,312) = 3.0; P < 0.02 11.8a (n = 41) 0.27b (n = 41) 0.35c (n = 41) 3.42e (n = 35) 10.3 (n = 140) -0.01 (n = 140) 0.11d (n = 140) 3.18 (n = 105) 11.1 (n = 122) -0.18b (n = 122) -0.25cd (n = 122) 3.03e (n = 97) 9.8a (n = 124) 0.10 (n = 124) 0.01 (n = 124) 3.17 (n = 79) Means with similar superscripts are significantly different from one another. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 736 G. King et al. with other types of activities. They reported the lowest preference for recreational activities (e.g. doing puzzles, watching TV or a video), which are commonly done at home and often alone. This group was the most likely to participate in activities with friends. They participated with their peer network or others (instructors, coaches, etc.) in neighbourhood and school settings. They had higher odds of having parents who perceived work/school environmental barriers for their children, and their families were least likely to be academically inclined. The overall picture is of a group of older children who are relatively well adjusted psychosocially, with good self-concepts and classmate support, and who participate in more structured extracurricular school activities and activities in the neighbourhood, therefore experiencing more attitudinal or discriminatory barriers. The profile may have been influenced by this group’s somewhat higher mean age (11.8 years) because, as children transition into the teen years, activity participation tends to become more socially focused and complex (Eccles 1999). We speculate that some members of this group may be socially unskilled and unaware of how others perceive them, because this group reported the highest self-perceived athletic competence and yet the worst physical functioning (although not significantly so). Alternatively, the positive social relationships and higher classmate support experienced by this group may enhance their perceived athletic efficacy (Bandura 1997). intensity in active physical and skill-based activities. Group members reported the lowest levels of activity enjoyment and the weakest preferences, and had the highest levels of psychosocial difficulties. They had the greatest emotional difficulties and peer problems, as well as significantly lower prosocial behaviour. They were more likely to be male (62%), although there were no significant sex differences among the groups. They also had the lowest self-perceived social acceptance and athletic competence, and the lowest perceived classmate support. We speculate that this group may feel alienated. They may be the group to feel most worried about, because of their high levels of social and emotional problems, and lower prosocial behaviour. Children with disabilities who have higher levels of functioning may be more likely to compare themselves to children without disabilities, thus experiencing greater stigma and negative psychosocial consequences (Harter et al. 1986; Crocker & Major 1989). These experiences may contribute to negative self-appraisals and social withdrawal (Bandura 1997), leading them to choose to participate in home-based recreational activities. Their parents reported the least difficulty in barriers to participation because of attitudes at home and in the community, perhaps because these children are more home-based in their activities. Recreational Participators Broad Participators This was a large group of high participators (mean age 10.3 years) who enjoyed participation, were relatively well adjusted (compared with the other groups) and had relatively high selfperceptions of athletic competence. This group was more likely to be Caucasian. Members of the group reported higher levels of intensity of participation across all types of activities than did the other groups, particularly with respect to active physical and skill-based activities. What really distinguished this group, however, was their high enjoyment and preference for all activities. The high activity intensity of this group may be partly because of their Caucasian background, because Caucasian families appear to value high levels of participation in comparison with other families (Posner & Vandell 1999). Beliefs about what constitutes valuable childhood experiences may influence the opportunities and encouragement provided by families. Low Participators This group (mean age 11.1 years) displayed the lowest intensities of participation across the board, with particularly low © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 This younger group (mean age 9.8 years) participated the most in recreational activities, reported high levels of enjoyment and strong preferences, and had a high level of prosocial behaviour, as rated by parents. Their parents had better physical health and were the youngest of the four groups. Children in this group participated more intensely in recreational activities than did the other groups (with the exception of the Broad Participators). They participated in active physical, social and skill-based activities with family or other relatives more so than did the other groups (especially the Social Participators). Of the four groups, they had the least differentiated preferences and enjoyment of the activity types, as shown in the small ranges of their enjoyment and preference ratings (ranges of 0.7 and 0.4 respectively). As they were a young group, their high involvement in recreational activities and relatively undifferentiated patterns of activity preferences and enjoyment seem developmentally appropriate. However, this group can also be considered to be socially isolated. It is the group with the narrowest out-ofschool time social network and the only group to report no activity participation with friends either at home or outside of Activity participation profiles 737 the home. These children are primarily participants in informal recreational activities, which are home-based, either alone or with family. Affective and contextual dimensions of participation The participation dimensions that distinguished the cluster groups were intensity of involvement in recreational activities, enjoyment, preferences and companionship. All four groups were generally alike in the location of their out-of-school activity participation, other than the highly social group being more likely to engage in self-improvement activities in the neighbourhood rather than at home or at the homes of relatives. This suggests that children with disabilities utilize and perhaps require more structured settings and programmes for leisure participation, whereas children without disabilities can possibly take part in these activities more informally (i.e. closer to home or in their neighbourhoods). Overall, the findings support an affective and contextual view of participation, indicating the importance of motivational theory and a person–environment approach in understanding the complexity of children’s out-ofschool activity participation (Bartko & Eccles 2003). Predictors of group membership The significant predictors of group membership were a mix of socio-demographic factors (child age, parent ethnicity), child psychosocial factors (peer problems and prosocial behaviour difficulties), parent and family factors (parent physical health, family intellectual-cultural orientation) and environmental factors (parent perceptions of barriers in the work/school environment and attitudes/support environment). Different combinations of these factors distinguished the groups, supporting the importance of a holistic consideration of determinants of participation patterns (Bartko & Eccles 2003). Over and above these factors, the groups were significantly different in children’s levels of psychosocial functioning. These findings correspond to previous research linking better emotional, behavioural and social functioning to greater child participation (Byrne et al. 1988; Rae-Grant et al. 1989). As well, Bartko and Eccles (2003) found that adolescents’ activity involvement was related to their psychological and behavioural functioning. These relationships may be bidirectional rather than causal. For example, psychosocial difficulties may make children less motivated to participate in activities, and lower levels of participation may in turn result in psychosocial difficulties. It is noteworthy that physical functioning (severity of impairment) was not associated with differences in group membership. Almqvist and Granlund (2005) also did not find type or degree of disability to predict cluster group membership with respect to school activities. Meta-analyses and literature reviews consistently indicate that chronic physical health status per se has little effect on adaptation (e.g. Lavigne & Faier-Routman 1992) and that psychosocial factors are relatively more important with respect to children’s outcomes (Wallander et al. 1989). Study strengths, limitations and directions for future research Study strengths include use of children’s self-report data on a comprehensive set of leisure activities and clustering of multiple participation dimensions. The present study contributes to the literature by indicating that meaningful activity profiles can be created based on the participation intensity, location, companionship, enjoyment and preferences of children with physical disabilities. Past studies have not taken such a comprehensive approach. This study has taken a preliminary look at the nature of different groups of children with disabilities. Unintended stereotypes may arise from clustering children into groups; however, children with disabilities are often treated as homogeneous and as having a unitary identity, which ignores important dimensions of identity such as sex, age and personal preference (Priestley 1998). Clustering multiple dimensions of participation and considering a broad set of predictive factors takes us away from a one-dimensional perspective. With respect to study limitations, there is first of all a need to replicate the groups with different samples, to determine generalizability. The Social Participators and Low Participators are similar to social and low participating groups of typically developing adolescents, found by both Raymore and colleagues (1999) and Roth and colleagues (2005). The Broad Participators and Recreational Participators may be unique to children with disabilities, the ages of children in our sample, or influenced by our focus on activity types. A related limitation is the 19% participation rate in the study. However, this response rate is typical of survey studies and it should be noted that the focus of the study was on relationships between variables rather than on issues of prevalence. A second limitation is the cross-sectional nature of the data. We cannot be certain of the direction of associations between the predictor variables and the groups, and do not know whether the groups have a developmental aspect and will continue to evolve. For example, some Low Participators may © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 738 G. King et al. become more social over time. It would be interesting to see which groups display appreciable changes in the trajectories of their participation patterns, and also to examine the meaning of activity participation on a deeper level than ‘enjoyment’ and ‘preference’. We need greater understanding of the experiences of disabled children, who can be viewed as social actors negotiating complex identities within disabling environments (Priestley 1998). We need to determine the meaning and valuing of activity from the child’s perspective and understand how participation acts as an expression of the child’s hopes and sense of meaning in life. Research is also needed to understand the role of children’s temperament in influencing participation enjoyment and preferences, and to clarify the role of parent values and support in determining and/or reinforcing children’s activity profiles. To what extent are the profiles a reflection of parents’ values and views of what is important to do in life? The present study found that children in the broad participation cluster were more likely to be of Caucasian background. Aside from work on differences in children’s time use across the world (Larson & Verma 1999) and a study reporting that Hispanic youth are less likely to participate in school-based extracurricular activities (Feldman & Matjasko 2007), little is known about the role of ethnicity and cultural beliefs and values in shaping children’s participation profiles. Last, research is needed to examine processes by which profiles of activity participation develop and evolve over time, and to develop person-process context models to better understand the dynamic nature of activity participation (Bartko & Eccles 2003). The present findings direct attention to the importance of exploring processes that mediate links between sociodemographic, child, parent, family, and environmental variables and participation profiles. Enjoyment and preference were important contributors to the nature of the groups, and perceptions of athletic and social competence distinguished the groups, suggesting that motivational processes driven by perceptions of competence and self-concept may play a fundamental role in determining the pattern of what a child does, where, and with whom. A more in-depth consideration of how identity and self-perceptions determine activity profiles is warranted, including study of how the development of self-awareness and discovery of uniqueness affect choice of activities and lead to the development of differentiated interests or activity preferences. Implications for clinical practice and theory development There are implications for services for each group. The Social Participators may prefer interventions that allow them greater © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 access to participation opportunities with both disabled and non-disabled peers. Service organizations can collaborate with community organizations to develop these opportunities (King et al. 2002). The group of Broad Participators participates highly, appears well adjusted and supported, and may require less intervention. Children who fit the Low Participator profile may benefit most from exploration of their satisfaction with community participation, and from services addressing selfdiscovery, awareness of community activities, and transition planning. The younger group of Recreational Participators seems to be participating in a developmentally appropriate manner, but may benefit from services designed to promote peer social interaction and friendships, such as social skills training. Perhaps the most obvious implication is that not all children with disabilities are alike. Studies have repeatedly shown that children experience disability differently with respect to their participation in social roles and activities. Some, depending on inner resources and environmental supports and opportunities, are resilient and do well. Others, because of greater risk, fewer protective factors, or some combination of the two, experience social exclusion and psychosocial difficulties. Work needs to be done to more fully understand the psychosocial and contextual factors and processes underlying different profiles of participation. It is important for clinicians to understand children’s and parents’ motivations regarding activity participation, to make a meaningful difference in children’s lives. With respect to theory, the findings inform biopsychosocial and biopsychospiritual models of participation and inclusion. They indicate how children’s out-of-school activity profiles reflect differences in affective and motivational variables (enjoyment and preferences) and contextual features (companionship), and are associated with different socio-demographic, child, parent, family and environmental variables. The findings highlight the context-dependent nature of activity participation, and the importance of recognizing roles played by multiple, activity-specific determinants (King et al. 2009). Understanding how characteristics of children, families and environments interact to influence engagement in activities throughout childhood is important for the development of appropriate and effective policies and programmes. Rather than examining the influence of discrete variables conceptualized within an ecological or systems framework, a fundamental challenge facing the field is to develop new ways of conceptualizing and capturing the aggregated essences of the antecedents, correlates and processes that characterize participation in interrelated sets of activities. The activity profiles presented here contribute to this effort. The language of complexity Activity participation profiles 739 science (Gleick 1987; Kaufman 1992) may provide us with insights needed to move our understanding of activity participation to a more holistic and contextual level, by allowing us to envision self-organizing relationships between dimensions of participation and types of activities, which lead to meaningful profiles of activity participation. Key messages • This study found that children with disabilities exhibited four distinct and meaningful activity profiles: Social Participators, Broad Participators, Low Participators and Recreational Participators. • By deriving profiles based on multiple self-reported dimensions, rather than only participation intensity, the present study provides a richer understanding of children’s lived experiences. • The groups were significantly different in their levels of psychosocial functioning but not in their physical functioning. • Children who fit the Low Participator profile may benefit most from intervention, given their high levels of social and emotional problems, and lower prosocial behaviour. • Research is needed to examine processes by which profiles of activity participation develop and evolve over time, and to better understand the dynamic nature of activity participation. Acknowledgements This study was supported by a research grant from the National Institutes of Health (Grant HD38108-02) and by CanChild Centre for Childhood Disability Research, funded in part by an operating grant from the Ontario Ministry of Health and Longterm Care. We acknowledge the contributions of Susanne King, Peter Rosenbaum, Marilyn Kertoy, Nancy Young and Steven Hanna to the project on which this article is based. Preparation of this article was supported by a senior research fellowship awarded by the Ontario Mental Health Foundation to Gillian King. Mary Law holds the John and Margaret Lilli Chair in Childhood Disability Research. Appreciation is extended to Patricia Baldwin for her clinical perspective on the findings, and to the many interviewers, organizations and families who contributed their time and commitment to this research project. References Almqvist, L. (2006) Patterns of engagement in young children with and without developmental delay. Journal of Policy and Practice in Intellectual Disabilities, 3, 65–75. Almqvist, L. & Granlund, M. (2005) Participation in school environment of children and youth with disabilities: a person-oriented approach. Scandinavian Journal of Psychology, 46, 305–314. Appleton, P. L., Minchom, P. E., Ellis, N. C., Elliott, C. E., Boll, V. & Jones, P. (1994) The self-concept of young people with spina bifida: a population-based study. Developmental Medicine and Child Neurology, 36, 198–215. Bandura, A. (1997) Developmental analysis of self-efficacy. In: The Exercise of Control (ed. A. Bandura), pp. 162–211. W. H. Freedman and Company, New York, NY, USA. Bartko, W. T. & Eccles, J. S. (2003) Adolescent participation in structured and unstructured activities: a person-oriented analysis. Journal of Youth and Adolescence, 32, 233–241. Bronfenbrenner, U. (1979) The Ecology of Human Development. Harvard University Press, Cambridge, MA, USA. Byrne, E. A., Cunningham, C. C. & Sloper, P. (1988) Families and Their Children with Down’s Syndrome: One Feature in Common. Routledge, London, UK. Crocker, J. & Major, B. (1989) Social stigma and self-esteem: the self-protective properties of stigma. Psychological Review, 96, 608–630. Curry, J. F. & Thompson, R. J. (1982) Patterns of behavioral disturbance in developmentally disabled children: a replicated cluster analysis. Journal of Pediatric Psychology, 7, 61–73. Eccles, J. S. (1999) The development of children ages 6 to 14. The Future of Children, 9, 30–44. Everitt, B. S. (1974) Cluster Analysis. Heinemann Educational Books, London, UK. Feldman, A. F. & Matjasko, J. L. (2005) The role of school-based extracurricular activities in adolescent development: a comprehensive review and future directions. Review of Educational Research, 75, 159–210. Feldman, A. F. & Matjasko, J. L. (2007) Profiles and portfolios of adolescent school-based extracurricular activity participation. Journal of Adolescence, 30, 313–332. Gleick, J. (1987) Chaos: The Making of A New Science. Heinemann, London, UK. Goodman, R. (1997) The strengths and difficulties questionnaire: a research note. Journal of Child Psychology and Psychiatry, 38, 581–586. Goodman, R. (2001) Psychometric properties of the strengths and difficulties questionnaire. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345. Gorsuch, R. L. (1983) Factor Analysis, 2nd edn. Erlbaum, Hillsdale, NJ, USA. Harter, S. (1985a) Self-Perception Profile for Children. Department of Psychology, University of Denver, Denver, CO, USA. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 740 G. King et al. Harter, S. (1985b) Social Support Scale for Children. Department of Psychology, University of Denver, Denver CO, USA. Harter, S. (1988) The Self-Perception Profile for Adolescents. Department of Psychology, University of Denver, Denver, CO, USA. Harter, S. (1986) Processes underlying the construction, maintenance, and enhancement of the self-concept in children. In: Psychological Perspectives on the Self, Vol. 3 (eds J. Suls, A. Greenwald), pp. 136–182. Erlbaum, Hillsdale, NJ, USA. Henry, D. B., Tolan, P. H. & Gorman-Smith, D. (2005) Cluster analysis in family psychology research. Journal of Family Psychology, 19, 121–132. Imms, C. (2008) Review of the children’s assessment of participation and enjoyment and the preferences for activity of children. Physical and Occupational Therapy in Pediatrics, 28, 389–404. Imms, C., Reilly, S., Carlin, J. & Dodd, K. (2008) Diversity of participation in children with cerebral palsy. Developmental Medicine and Child Neurology, 50, 363–369. James, A., Jenks, C. & Prout, A. (1998) Theorizing Childhood. Polity Press, Cambridge, MA, USA. Kang, L., Polansky, M., Palisano, R., Orlin, M., Chiarello, L. & Maggs, J. (2009) Social engagement and community participation among children and youth with cerebral palsy. Pediatric Physical Therapy, 21, 92. Kaufman, S. A. (1992) Origins of Order: Self Organization and Selection in Evolution. Oxford University Press, Oxford, UK. King, G., Tucker, M. A., Baldwin, P., Lowry, K., Laporta, J. & Martens, L. (2002) A life needs model of pediatric service delivery: services to support community participation and quality of life for children and youth with disabilities. Physical and Occupational Therapy in Pediatrics, 22, 53–77. King, G., Law, M., King, S., Rosenbaum, P. & Kertoy, M. (2003) A model of the factors affecting the participation of children with disabilities. Developmental Medicine and Child Neurology, 45, 52–52. King, G., Law, M., King, S., Hurley, P., Rosenbaum, P., Hanna, S., Kertoy, M. & Young, N. (2004) Children’s Assessment of Participation and Enjoyment (CAPE) and Preferences for Activities of Children (PAC). Harcourt Assessment, San Antonio, TX, USA. King, G., Law, M., Hanna, S., King, S., Hurley, P., Rosenbaum, P., Kertoy, M. & Petrenchik, T. (2006a) Predictors of the leisure and recreation participation of children with physical disabilities: a structural equation modeling analysis. Children’s Health Care, 35, 209–234. King, G. A., Law, M., King, S., Hurley, P., Hanna, S., Kertoy, M. & Rosenbaum, P. (2006b) Measuring children’s participation in recreation and leisure activities: construct validation of the CAPE and PAC. Child: care, health and development, 33, 28–39. King, G., McDougall, J., Dewit, D., Petrenchik, T., Hurley, P. & Law, M. (2009) Predictors of change over time in the activity participation of children and youth with physical disabilities. Children’s Health Care, 38, 321–351. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741 Larson, R. W. & Verma, S. (1999) How children and adolescents spend time across the world: work, play and developmental opportunities. Psychological Bulletin, 125, 701–736. Lavigne, J. V. & Faier-Routman, J. (1992) Psychological adjustment to pediatric physical disorders: a meta-analytic review. Journal of Pediatric Psychology, 17, 133–157. Law, M. (2002) Participation in the occupations of everyday life. American Journal of Occupational Therapy, 56, 640–649. Law, M., Finkelman, S., Hurley, P., Rosenbaum, P., King, S., King, G. & Hanna, S. (2004) Participation of children with physical disabilities: relationships with diagnosis, physical function, and demographic variables. Scandinavian Journal of Occupational Therapy, 11, 156–162. Law, M., King, G., King, S., Kertoy, M., Hurley, P., Rosenbaum, P., Young, N. & Hanna, S. (2006) Patterns of participation in recreational and leisure activities among children with complex physical disabilities. Developmental Medicine and Child Neurology, 48, 337–342. McHorney, C. A., Ware, J. E. & Raczek, A. E. (1993) The MOS 36-Item Short-Form Health Survey (SF-36): Ii. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Medical Care, 31, 247–263. Magnusson, D. & Stattin, H. (2006) The person in the environment: towards a general model for scientific inquiry. In: Theoretical Models of Human Development. Volume 1 of Handbook of Child Psychology, 6th edn (ed. R. M. Lerner), pp. 100–464. Wiley, Hoboken, NJ, USA. Majnemer, A., Shevell, M., Law, M., Birnbaum, R., Chilingaryan, G., Rosenbaum, P. & Poulin, C. (2008) Participation and enjoyment of leisure activities in school-aged children with cerebral palsy. Developmental Medicine and Child Neurology, 50, 751– 758. Milligan, G. W. (1996) Clustering validation: results and implications for applied analyses. In: Clustering and Classification (eds P. Arabie, L. J. Hubert & G. De Soete), pp. 341–375. World Scientific, River Edge, NJ, USA. Moos, R. H. & Moos, B. S. (1994) Family Environment Scale Manual, 3rd edn. Consulting Psychologists Press, Inc., Palo Alto, CA, USA. Orlin, M. N., Palisano, R. J., Chiarello, L. A., Kang, L., Polansky, M., Almarsi, N. & Maggs, J. (2010) Participation in home, extracurricular, and community activities among children and youth with cerebral palsy. Developmental Medicine and Child Neurology, 52, 160–166. Overton, W. F. (1998) Developmental psychology: philosophy, concepts, and methodology. In: Handbook of Child Psychology. Theoretical Models of Human Development, Vol. 1, 5th edn (ed. R. M. Lerner), pp. 107–188. John Wiley & Sons, New York, NY, USA. Pless, I. B. & Pinkerton, P. (1975) Chronic Childhood Disorder: Promoting Patterns of Adjustment. Yearbook Medical Publishers, Chicago, IL, USA. Plsek, P. E. & Wilson, T. (2001) Complexity science: complexity, leadership, and management in healthcare organizations. BMJ (Clinical Research Ed.), 323, 746–749. Activity participation profiles 741 Posner, J. K. & Vandell, D. L. (1999) After-school activities and the development of low-income urban children: a longitudinal study. Developmental Psychology, 35, 868–879. Priestley, M. (1998) Childhood disability and disabled childhoods: agendas for research. Childhood, 5, 207–223. Rae-Grant, N., Thomas, B. H., Offord, D. R. & Boyle, M. H. (1989) Risk, protective factors, and the prevalence of behavioral and emotional disorders in children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 28, 262–268. Raymore, L. A., Barber, B. L., Eccles, J. S. & Godbey, G. C. (1999) Leisure behavior pattern stability during the transition from adolescence to young adulthood. Journal of Youth and Adolescence, 28, 79–103. Roth, J. L., Linver, M. & Brooks-Gunn, J. (2005) Who Does What: A Person-Oriented Analysis of Children’s Non-School Time-Use. Society for Research in Child Development, Atlanta, GA, USA. Sparrow, S. S., Balla, D. A. & Cicchetti, D. V. (1984) Vineland Adaptive Behavior Scales. American Guidance Service, Inc, Circle Pines, MN, USA. Statistics Canada (2001) Median Family Income in Constant Dollars (2000) for All Census Families for Canada, Provinces and Territories. Report for Statistics Canada (Ottawa ON). Stein, R. E. & Jessop, D. J. (2003) The impact on family scale revisited: further psychometric data. Journal of Developmental and Behavioral Pediatrics, 24, 9–16. Stein, R. E. K. & Riessman, C. K. (1980) The development of an impact on family scale: preliminary findings. Medical Care, 18, 465–472. Stone, M. R., Barber, B. L. & Eccles, J. S. (2005) Positive Friendship Network Characteristics Associated with Varying Types of Extracurricular Activity Participation. Society for Research on Child Development Conference, Atlanta, GA, USA. Wallander, J. L., Varni, J. W., Babani, L., Banis, H. T., Dehaan, C. B. & Wilcox, K. T. (1989) Disability parameters, chronic strain, and adaptation of physically handicapped children and their mothers. Journal of Pediatric Psychology, 14, 23–42. Wapner, S. & Demick, J. (1998) Developmental analysis: a holistic, developmental, systems-oriented perspective. In: Handbook of Child Psychology. Theoretical Models of Human Development, Vol. 1, 5th edn (ed. R. M. Lerner), pp. 761–806. John Wiley & Sons, New York, NY, USA. Ware, J. E. & Sherbourne, C. D. (1992) The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual framework and item selection. Medical Care, 30, 473–483. Ware, J. E., Snow, K. K., Kosinski, M. & Gandek, B. (1993) SF-36 Health Survey: Manual and Interpretation Guide. The Health Institute, New England Medical Center, Boston, MA, USA. Whiteneck, G. G., Harrison-Felix, C. L., Mellick, D. C., Brooks, C. A., Charlifue, S. B. & Gerhart, K. A. (2004) Quantifying environmental factors: a measure of physical, attitudinal, service, productivity, and policy barriers. Archives of Physical Medicine and Rehabilitation, 85, 1324–1335. Whiting, B. B. & Edwards, C. P. (1988) Children of Different Worlds: The Formation of Social Behavior. Harvard University Press, Cambridge, MA, USA. Young, N. L., Williams, J. I., Yoshida, K. K. & Wright, J. G. (2000) Measurement properties of the activities scale for kids. Journal of Clinical Epidemiology, 53, 125–137. © 2010 Blackwell Publishing Ltd, Child: care, health and development, 36, 5, 726–741