This was a cross-sectional study of children with and without CP based on questionnaires completed by parents/foster parents. Ethical approval was received from the National Research and Ethics Committee of the service provider. All parents gave written informed consent for the study subjects at the time of first contact.
Children living with CP
The South of Ireland Cerebral Palsy Register (SICPR) is a register of all child residents of the counties of Cork and Kerry who, at the age of four years or older, were diagnosed as living with CP. During the study period (2004–2005), the families of all 120 children on the SICPR aged 8–12 years were contacted, first by telephone and then by letter. Ninety-eight parents participated (response rate = 82%) and completed structured questionnaires that measured participation, quality of life and impairment and a range of other factors, in the presence of the study researcher, usually in their home and taking a duration of 90–120 minutes.
Children not living with CP
Via the school principals, the Frequency of Participation Questionnaire (FPQ) was sent to the parents of all 650 children aged 8–12 years attending two Cork schools. A total of 448 completed questionnaires were received by the researcher (response rate = 69%). None of the children had a physical disability.
Participation in everyday activities
The ICF recognises potential overlap in the concepts of activity with participation and hence classifies it across the same domain. The Sparcle group considered the concepts to be separate and wanted to examine participation separately independent of adaptations or assistance required by a child in doing the activity.
Frequency of participation (FPQ) in 14 everyday activities was measured using the FPQ instrument, an instrument developed by Sparcle from the Life Habits questionnaire [23, 24]. The Life-H instrument was used as the conceptual framework for the frequency items. It was designed for disabled children and has been used in children with CP . Our instrument differs from the Life-H because it captures how frequently activities are done. The Sparcle group chose the items from the Life-H instrument so that they would be relevant to the general population also. The FPQ has face validity as the questions were derived from the content of the Life-H instrument. We did not undertake intra and inter observer reliability. The 14 FPQ activities are: eating out, leisure, computer use, helping with housework, riding a bicycle, tricycle or wheelchair, running errands, joining organized activities outside school (community), school activities, playing sports, playing non-sports, watching sports, doing arts and crafts, going to cinema (culture), and taking part in tourist activities. Frequency of participation in each activity was assessed using a six-point Likert scale (scored 0–5): never, less than once a month, about once a month, once every two weeks, about once a week a few times a week. Because of limited numbers at some levels of participation, each of the 14 items was collapsed to two levels for the data analysis relating to the 14 activities. The sum of the original 14 items (scored 0–5) was used to give an overall measure of participation in everyday activities (possible range 0–70). For respondents who failed to answer one, two or three items, the average of the answered items was imputed for the unanswered items. Cronbach's alpha was 0.63 for the overall measure of participation which indicates a moderate level of reliability [26, 27]. Overall participation scores followed a Normal distribution.
Quality of life
Quality of life was measured using KIDSCREEN, a 52-item generic health-related quality of life measure applicable to healthy and chronically ill children and adolescents aged 8–18 years and designed for child or parent report . The KIDSCREEN instrument was used because it assesses across both healthy and ill children. It has been well validated psychometrically with 22,110 European children from the general population . KIDSCREEN assesses ten domains of quality of life: physical well-being, psychological well-being, moods and emotions, self-perception, autonomy, parental relations, financial resources, social support and peers, school environment and social acceptance (bullying). The 52 items have a five-point Likert scale with two sets of possible responses: never, seldom, quite often, very often, always and not at all, slightly, moderately, very, extremely. For each domain, the relevant items are summed and scaled to yield a score in the range 0–100 with higher scores indicating better quality of life. Cronbach's alpha was greater than 0.7 for all ten quality of life domains except one at 0.69. These results suggest a high level of reliability for the domains [26, 27]. Domain scores did not follow a Normal distribution because of the ceiling effect whereby high proportions of the children scored close to 100.
Level of impairment of children living with CP
Parents provided information about their child's gross motor function  and fine motor function . CP type was available from the SICPR. Seizure activity, vision level, hearing level, feeding, communication and intelligence quotient (IQ) were recorded. Gross motor function and two-hand fine motor function were recorded according to the Gross Motor Function Classification System (GMFCS) and the Bimanual Fine Motor Function (BFMF) level. For the latter, levels I and II and levels III and IV were collapsed.
An overall, three-category measure of impairment was derived from the core impairment variables of gross motor function and IQ. Mild impairment was defined as GMFCS level I-III and IQ > 70. Moderate impairment was indicated by GMFCS level IV-V or IQ <= 70. Severe impairment was defined as GMFCS level IV-V and IQ <= 70.
Data were analysed using SPSS v14. For the total study sample (i.e. children with CP and non-CP children), chi-square tests were used to assess the association between level of impairment and level of participation in each of the 14 everyday activities. Because overall participation scores followed a Normal distribution, the parametric t-test and one-way analysis of variance (ANOVA) were used to assess between-group differences with respect to two groups and more than two groups, respectively. Following statistically significant one-way ANOVA tests, Tukey's post hoc tests were used to identify the differing pairs of groups. A multiple linear regression model was estimated with overall participation in everyday activities as the dependent variable and gender, age and level of impairment as the independent variables. Diagnostic tests were used to check for violations of the assumptions inherent in linear regression models.
For the sample of children with CP, Mann-Whitney and Kruskall-Wallis tests were used to examine between-group differences in relation to the ten quality of life domains between two groups and more than two groups, respectively. For each quality of life domain, a multivariate linear regression model was estimated with quality of life as the dependant variable and gender, age, level of impairment and overall participation as the independent variables. Diagnostic tests were used to check for violations of the assumptions inherent in linear regression models.