1
Mainstream Students’ Attitudes To Playing Unified Sports With Students
Who Have An Intellectual Disability
Michael Townsend and John Hassall
University of Auckland, New Zealand
Paper presented at the European Conference on Educational Research, Crete, September 2004. For further information write to Michael Townsend, School of Education, University of Auckland, Auckland, New Zealand,
Mainstream Students’ Attitudes To Playing Unified Sports With Students
Who Have An Intellectual Disability
This study was an examination of regular students’ attitudes toward the inclusion of students with an intellectual disability in an integrated sports programme within their school in New Zealand. Primary school students (n = 63, aged 10 years) and secondary school students (n = 107, aged 16-17 years) at four schools in Auckland completed an attitude scale assessing their acceptance of unified sports at their school, a test of their knowledge about Special Olympics, and were given an opportunity to write any other comments about unified sports. Two focus groups, one each at the primary level and secondary levels, were held to further explore students’ attitudes to unified sports in schools. Students were found to have positive attitudes toward involvement alongside students with an intellectual disability in unified sports. These attitudes were moderated by age and gender, but not knowledge about Special Olympics.
The progress toward normalization for people with an intellectual disability has been remarkable. In little more than 25 years segregation and institutionalization have given way to more productive engagement with the community in education, work, housing and leisure (Keith & Schalock, 2000; Pretty, Rapley & Bramston, 2002; Emerson & Hatton, 1996). However, despite these advances social acceptance of those with intellectual disability remains problematic (Freeman & Alkin, 2000; Piercy, Wilton & Townsend, 2002; Siperstein, Leffert & Widaman, 1996).
In education, increased social acceptance has been viewed as an inevitable outcome of contact with regular children under policies of inclusion in schools. Although early research indicated that simple contact did not increase the social acceptance of children with an intellectual disability (Zigler & Hodapp, 1986), recent work has emphasised the importance of contact through cooperative activities (Dovidio, Gaertner & Kawakami, 2003). Classroom-based cooperative activities have been found to increase the social acceptance of children with intellectual disability (e.g., Jacques, Wilton & Townsend, 1998). Outside of the classroom, some schools have recently introduced cooperative contact in the form of integrated organised sport.
The individual and social benefits of sport have been available to children (and adults) with intellectual disability through the Special Olympics since 1968 (Songster, 1984), but this movement has increasingly been criticised for being segregated (Orelove, Wehman & Wood, 1982; Poretta, Gillespie & Jansma, 1996). In reaction to this criticism, Unified Sports is a relatively new Special Olympics programme that combines similar numbers of athletes with an intellectual disability with athletes without a disability (called ‘partners’) on sports teams for training and competition. This is meaningful inclusion within a valued community activity (Janes & Goad, 1995). Although such teams may be initiated from any community group, schools are uniquely placed to act as change agents if they have already accepted the principles of normalisation and inclusion, and organised sport is an existing part of school life.
Since its beginning in the United States in 1989, unified sports has been adopted by a growing number of countries where the Special Olympics movement is present. New Zealand, however, is not one of those countries. This seems anomalous in a country that places strong cultural value on participation in sports and outdoor activities, has a national educational policy of inclusion, and has a highly developed programme of organised sport in schools. Given that any initiatives toward unified sports are likely to be based in schools, there is a need to explore the opportunities, and the likely barriers, to its introduction. Notwithstanding the considerable social, political, economic and educational issues to be resolved, one critical element in the successful introduction of unified sport is its acceptance by the end-users, the students. The current study was undertaken to assess the attitudes of regular students in New Zealand to the inclusion of children with intellectual disability in regularly scheduled organised sports activities.
Method
Participants
The participants were 170 primary and secondary school students attending public schools in Auckland, New Zealand. The primary school sample contained 35 females and 28 males in three Year 6 classes (aged approximately 10 years) at two primary schools. The secondary school sample contained 58 females and 49 males in six Year 11 and Year 12 classes (aged 16-17 years) at two high schools. At both the primary and secondary level, schools were selected from both higher and lower socioeconomic areas of the city to provide socioeconomic and ethnic diversity in the sample.
Instruments
Attitude Toward Unified Sports: A questionnaire was designed to assess regular students’ attitudes toward playing organised team sports with students with intellectual disability (unified sports). Initially, a group of people representing teachers of both special needs and regular children, and the Special Olympics programme, met to discuss issues that might be of concern to regular students. These issues included concerns about elements of the games themselves (such as differences in skill level, commitment to winning, and ‘not going hard out’ on players with intellectual disability), socialising with the students with intellectual disability (such as using the same changing facilities, training together, and being at after-match functions), and public identification (such as appearing in school team photos with students with intellectual disability, and travelling together in a van marked with a unified sports logo).
Fourteen possible items measuring the issues just noted were discussed and written as statements to be responded to using a 6-point Likert scale of agreement (Strongly Agree, Agree, Unsure/Probably Agree, Unsure/Probably Disagree, Disagree, and Strongly Disagree), and scored so that a higher score represented a more positive attitude. Approximately two-thirds of the items were positively worded (e.g., “I would like to play on a sports team with students who have an intellectual disability”), while five were worded negatively (e.g., “I would not like to share my playing time on the field/court with teammates who have an intellectual disability”). The negatively worded items were included to reduce the likelihood of response set (Nunnally, 1978), although their inclusion was known to carry some risks. Negatively worded items may be difficult for younger students, and may lead to measurement problems that are often ignored in the interpretation of results (Gable & Wolf, 1993). However, a small sample of primary and secondary school students found the questionnaire easy to read, understand and fill out.
Following administration of the questionnaire to the 170 students in this sample the 14 items were subjected to a factor analysis using a maximum likelihood extraction method and an oblimin with Kaiser normalisation rotation method. In this analysis the sample size exceeded the suggested ratio (Nunnally, 1978) of at least 10 cases for each item to be factor analysed. Prior to conducting the factor analysis, inspection of the correlation matrix revealed the presence of a large number of coefficients of .30 and above. Further information supporting the suitability of the correlation matrix for factor analysis was found in the significant test of sphericity (χ2 (91) = 814.56, p < .001) and in a Kaiser-Meyer-Olkin index of sampling adequacy of .819 (exceeding the suggested minimum value of .60 recommended by Tabachnick & Fidell, 1996).
The factor analysis extraction revealed three factors with eigenvalues exceeding 1.0, explaining 33.49, 15.01 and 7.60 percent of the variance, respectively. However, because the scree plot showed a clear beak after the second factor, and because the factor matrix indicated that one of the factors was composed of a single item, it was decided to retain only two factors for further investigation. The subsequent two-factor extraction analysis revealed that all items had substantial loadings on only one of the factors, one composed of 9 items (with loadings ranging from .55 to .77) accounting for 29.23% of the variance, and the other composed of five items (with loadings ranging from .47 to .62) accounting for 10.78% of the variance. However, the first factor was composed of all of the positively-worded items, while the second factor was composed of all of the negatively-worded items. Although the two factors had only small shared variance (r = .33, p <.01), items from the two factors did not appear to be measuring different attitudinal constructs. For example, the Factor 1 item, “I would be happy to be in a team photo with intellectually disabled students” appeared to mirror the idea in the Factor 2 item, “I would not like to have my picture taken with students who have an intellectual disability”. Thus, it appears likely that the presence of two factors was an outcome of the use of both positively and negatively worded items in the attitude measure. In a discussion of the effects of including positive and negative items in attitude instruments, Gable and Wolf (1993) caution against their combined use, concluding that negative items may not only reduce item- and scale-level reliabilities but they may also lack proper definition and support for the targeted constructs. In light of this caution, and the analysis of the factor loadings just described, a decision was made to use only the nine positively-worded items as a single-factor scale in this study. Scores on this resultant attitude measure of 9-items could vary between 9 and 54. It was found to have high internal consistency reliability in the entire sample (Cronbach alpha = .86). Similar estimates of reliability were found in both the primary (.82) and secondary (.87) samples. An abbreviated form of each item is shown in Table 1 (below).
At the end of the attitude questionnaire an open-ended opportunity was given to students to write any other comments relating to playing sports with students with intellectual disability.
Knowledge of Special Olympics: A 5-item questionnaire was designed to assess the level of knowledge that students possessed about the Special Olympics programme. Written responses were requested for open-ended questions asking about the purpose, nature, scope and participants of the Special Olympics (e.g., “What is Special Olympics for? Who is allowed to take part in Special Olympics?); one question asked students to describe the unified sports programme. Responses to the questions were scored 1 (correct), .5 (partially correct) and 0 (wrong or no response). Scale analysis revealed modest reliabilities for the five item scale, total sample α = .52, secondary = .48, and primary = .62.
Focus Group Interview: An interview was designed to further probe mainstream students’ attitudes toward participation in organised team sports with students with an intellectual disability. The session was relatively unstructured, beginning with a brief account of the concept of unified sport and then inviting students to say what they thought about it. Prompting (e.g., “Tell me more about that”) was used to encourage greater elaboration where appropriate. The discussions were designed to be relaxed settings in which similarly-aged students could share their views openly.
Procedure
The knowledge and attitude instruments were administered under no time restrictions to intact classes in a single session. These sessions took 30-40 minutes, including time for the free-response section. Subsequently, two focus groups (one primary, one secondary) were formed by inviting approximately 20 percent of students in two classes to participate; classes and students were selected randomly. The two focus group sessions (involving a total of 9 children) were held in informal settings at each school, where the discussions were audio-tape recorded for later transcription and analysis. Discussions took approximately 15 minutes.
Prior to each session students were given a brief outline of the nature and purpose of the tasks and were reassured of confidentiality. Students were requested to record only their age and gender on the forms. For the discussion groups students were encouraged to be frank and to “just jump in whenever and wherever you want to”. All data was collected by the researchers and the study was conducted with the approval of the research ethics committee of the University of Auckland.
The study provided three sources of data. Data from the knowledge and attitude questionnaires were analysed using SPSS (Version 12). Written free-response comments from the end of the attitude questionnaire, and verbal responses from the focus group discussions were analysed using procedures outlined by Strauss & Corbin (1998). This involved transcribing and scanning all comments, establishing apparent themes and sub-themes, and coding the material in terms of those themes.
Results
Knowledge and Attitude
A preliminary analysis was made of possible effects associated with socioeconomic status. Schools were classified as either higher or lower in socioeconomic status and students’ scores on the attitude and knowledge scales were examined in a School Level (primary/secondary) by Gender by School SES (higher/lower) multivariate analysis of variance. Socioeconomic status was involved in a significant multivariate main effect, F(2,163) = 6.11, p < .01, Wilks’ Lambda = .92, partial eta2 = .08, but not in any multivariate interaction effects. Examination of the univariate effects revealed a significant main effect of socioeconomic status, with knowledge about Special Olympics being greater for children attending higher socioeconomic status schools (M = 1.54, SD = 0.90) than lower socioeconomic schools (M = 1.05, SD = 0.86), F(1,164) = 12.25, p < .01, partial eta2 = .07. The univariate effect for the attitude scale was not significant (F = 0.25). Because socioeconomic status was not related to attitude, and was not involved in any interactions with school level or gender in either attitude or knowledge, further analysis of socioeconomic effects was discontinued.
A school level by gender multivariate analysis of variance was then carried out on the attitude and knowledge scores. There were significant multivariate main effects for both school level, F(2,165) = 13.47, p < .001, Wilks’ Lambda = .86, partial eta2 = .14, and gender, F(2,165) = 4.07, p < .05, Wilks’ Lambda = .95, partial eta2 = .05. However, the school level by gender interaction was not significant (F = 1.10). The univariate effects associated with attitude and knowledge were then examined separately.
The major factor of interest in this study was regular students’ attitude towards involvement in unified sports with children with an intellectual disability. The univariate results of the MANOVA just described revealed that there was a significant main effect for school level, with the mean attitude score for primary school students being higher (M = 46.97, SD = 8.12) than that for secondary school students (M = 41.50, SD = 6.63), F(1,166) = 21.65, p <.001, partial eta2 = .12. The main effect for gender was also significant, with the mean attitude score for girls being higher (M = 44.82, SD = 7.21) than that for boys (M = 41.97, SD = 7.95), F(1,166) = 8.09, p <.01, partial eta2 = .05. The interaction effect of school level and gender was not significant (F = 2.21). Overall, the mean score for the sample (M = 43.53, SD = 7.67) was considerably above the mid-point of the scale (31.50), indicating that students were markedly positive about possible closer involvement with children with intellectual disability in sport activities. However, as just noted, this was more likely for younger students and female students than for older students and male students.
It was of some interest to examine students’ knowledge about the nature and scope of Special Olympics. The univariate results of the MANOVA described above revealed no significant differences in students’ knowledge as a function of school level (F = 3.39), gender (F = 0.00), or their interaction (F = 0.04). It should be noted, however, that although the school level main effect just described was not significant, there was a non-significant trend (p < .10) suggesting greater knowledge about Special Olympics in the secondary level students (M = 1.34, SD = 0.89) than in the primary level students (M = 1.10, SD = 0.93). Overall, the mean score for the sample was 1.27 (SD = 0.92), with only six percent of students achieving a score equivalent to more than 50% of the items correct. Most students (71.2%) could provide either a partial or full description of the purpose of the Special Olympics (usually a recognition that it involves children with “a handicap”), but the proportion of students obtaining partial or full credit on the remaining items ranged between 9% (for knowledge about the unified sports programme) and 48% (knowledge of sports involved in Special Olympics). In brief, students had relatively little knowledge about Special Olympics and unified sports.
Although knowledge about Special Olympics was low it may have influenced students’ attitudes. To examine this possibility, scores on the attitude scale were re-examined in a school level by gender analysis of covariance, with knowledge scores introduced as a covariate. The pattern of results was identical to that already reported. The correlation between the attitude scores and knowledge scores for the entire sample was not statistically significant, r = .05, p = .58.
Following the analyses just described an examination was made of the scores on the individual items comprising the attitude scale. Scores for the nine items were examined in a school level by gender multiple analysis of variance. The mean scores are shown in Table 1. The multivariate main effect for school level was significant, F(9,158) = 8.37, p < .001, Wilks’ Lambda = .68, partial eta2 = .32, but neither the multivariate main effect for gender (F = 1.64) nor the school level by gender interaction effect (F = 1.00) was significant. In view of the gender effect already reported in the total scale scores, the univariate effects for both school level and gender were examined in each of the nine items.
[Insert Table 1 about here]
Significant school level effects (all p < .001) were found for six of the items (1, 3, 4, 5, 7 and 8); in all cases the primary school students were more positive than the secondary school students. Although generally positive, older students were most likely to be less positive about out-of-hours participation (item 9, eta2 = .19), interacting at school (item 1, eta2 = .13), and in appearing at special schools or units (item 5, eta2 = .12). Generally, the younger students appeared to be more positive concerning their general interactions with students with an intellectual disability (in talking with them, playing sport with them, travelling together, sharing a uniform, and so on). Age differences were not apparent for more “public” interactions, such as being in team photos, whether published or not, and going to parties.