Social Participation, Self-Perception, and Social …...ii Social Participation, Self-Perception, and Social Support in Boys Living with Hemophilia Aubrey Serah Chiu Master of Science
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Social Participation, Self-Perception, and Social Support in Boys Living with Hemophilia
by
Aubrey Serah Chiu
A thesis submitted in conformity with the requirements for the degree of Master of Science
Institute of Health Policy, Management, and Evaluation University of Toronto
restriction (28-40), severe restriction (41-66), and extreme restriction (67-115).
Additionally, three open-ended questions were included probing subjects for their thoughts on
what participation consisted of, the effect (if any) of hemophilia on their participation, and the
effect (if any) of personality on participation. The additional questions are listed in Table 4.
Table 4. Participation Scale for Kids Questions (Open-Ended) Question 1. What do you think participation is? 2. Do you feel that hemophilia stops you from participating like others? If so, how? 3. Do you feel that personality changes how you participate with others? (e.g. your thoughts, feelings, and behaviors.)
The primary reason why this measure was chosen for this study was because for every item on
the scale, there was an option to indicate the relevancy of the activity to the participant. The
measure has good face validity and seemed relevant to the study. Also, this instrument was
designed based on the participation domains of the ICF, was cross-cultural in nature, aimed to
assess patient-perceived participation, and it was not disease-specific (van Brakel et al., 2006).
The original scale had high internal consistency (Cronbach’s α 0.9) and high inter-tester
reliability (0.8) (van Brakel et al., 2006). The authors felt that there were no existing tools that
could be used to validate this instrument under the prevailing conditions, thus they chose to
validate the results against expert opinions. They found moderate correlations between the
expert opinions and the scores of participants (rho = 0.4).
3.7.4 Self-Perception
Self-perception was measured using items from the Self-Perception Profile for Children or the
Self-Perception Profile for Adolescents (Harter, 1985a; 1988; 2012a; 2012b), often considered
the gold standard for measuring self-perception. Five specific domains of perceived competence
are tapped by the original measure: Scholastic Competence, Social Acceptance, Athletic
Competence, Physical Appearance, and Behavioral Conduct. A sixth subscale measures the
child or adolescent’s perception of their Global Self-Worth. There are two versions of the scale:
one for children (ages 7-12) and one for adolescents (ages 13-18). By having different versions,
the scales ensured that questions were relevant to the participant’s age group.
The questionnaire first asked participants to choose between two statements describing
conflicting types of kids or teenagers, e.g. “some kids have a lot of friends” or “other kids don’t
65
have very many friends.” Once the participant made his choice, he then had to indicate if the
chosen statement was ‘really true for me’ or ‘sort of true for me.’ Items on each of the subscales
were scored 4, 3, 2, or 1; higher scores indicated higher perceived competence on that item.
There was no overall score for the scale as the measure yielded six subscale means. Within each
subscale, the items were counter-balanced to avoid participants yea-saying or acquiescing when
filling out the questionnaires (Streiner & Norman, 2008a).
To reduce burden on participants, three of the six domains were selected for inclusion in this
study. The questionnaire manuals note that some subscales can be lifted from the instrument as
long as all items of the given subscales are administered to the participant (Harter, 2012a;
2012b). Other researchers have successfully administered individual subscales to their samples
Details of participants’ family situation were collected from the participant’s primary guardian
and are listed in Appendix C.
4.5 Treatment
Parents were asked about the participant’s treatment history, which was verified through medical
charts. Further details of the participant’s treatment history are listed in Appendix D. As
prophylaxis is the standard of care for severe hemophilia patients at The Hospital for Sick
Children, when recounting treatment frequency for the previous six months over two-thirds of
the participants were on prophylaxis (receiving treatment two or three times a week, or every
other day). Of the severe hemophilia patients, 29 (83%) patients were on prophylaxis. Nearly a
quarter of all participants received treatment on-demand.
75
There was a wide range of ages at which infusions began with some participants requiring factor
infusions immediately after birth and some not being infused with factor until the age of 15 (M =
1.7, SD = 2.9).
Clinic staff supported the use of self-infusion when appropriate. In this sample, 15 (30.0%)
participants reported they had learned how to self-infuse and the average age that self-infusion
began was 11.2 years old (Minimum = 8.0, Maximum = 15.5, SD = 2.2). Some participants who
reported an age at which self-infusion began only identified a parent as the infuser.
4.6 Internal Consistency
The reliability of the modified Participation Scale for Kids was tested with Cronbach’s alpha (see
Appendix E). Both the overall raw alpha (α = 0.7) and the standardized alpha (α = 0.7) were on
par with the 0.7 cut-off previously specified (Nunnally, 1978, as cited in Streiner & Norman,
2008a).
4.7 Outcomes
4.7.1 Social Participation
The mean score on the Participation Scale for Kids was 6.5 (SD = 6.2). Forty-six (92.0%)
participants had no significant restriction (scores between 0-15), 3 (6.0%) participants had mild
restriction (scores between 16-27), and one (2.0%) participant had moderate restriction (scores
between 28-40). No participants indicated that they had severe restriction (scores between 41-
66) or extreme restriction (scores between 67-115). Item responses for each item are displayed
in Appendix F.
Although the mean score was still within the mild restriction range, some respondents reported
that they had “large problems” feeling comfortable meeting new people (item 17, n = 3) and felt
that they did not have the same respect in the community as their classmates (item 8, n = 3).
Each of the following items were endorsed once by a respondent as being a “large problem,”
though not necessarily by the same respondent:
2) Do you work as hard as your classmates do?
76
6) Do you take part in casual recreational/social activities as much as your classmates do?
(e.g. sports, chats, meetings, etc.)
7) Are you as socially active as your classmates? (e.g. in religious/community affairs,
etc.)
13) In your neighbourhood, do you visit public places as often as other people do? (e.g.
schools, shops, tea/coffee shops, etc.)
20) Do you use social networking sites as often as other people do? (e.g. Facebook,
Twitter, Myspace, etc.)
Some participants felt that they had “medium problems” when it came to participating as their
peers. These included being able to take part in casual recreational/social activities (item 6, n =
3), having the same respect in the community (item 8, n = 3), working as hard (item 2, n = 3),
helping other people (item 16, n = 3), playing games as often as their peers (item 19, n = 3), and
visiting other people in the community (item 11, n = 2).
Other items endorsed by at least one participant as a “medium problem” included:
3) Do you contribute to the household financially in a similar way to your classmates?
5) Do you take part in major events as your classmates do? (e.g. weddings, funerals,
religious festivals, etc.)
10) Do you have the same opportunities as your classmates to start or maintain a
relationship?
17) Are you comfortable meeting new people?
Children’s responses on the Participation Scale for Kids elicited a mean score of 7.6 (SD = 7.2)
while adolescents had a mean score that was slightly lower at 4.9 (SD = 3.7). Based on these
scores, all adolescents indicated that they had no significant restriction when it came to
participation. However, when comparing the means of the children and the adolescents,
allowing for unequal variances, the difference between the two groups was not statistically
significant (t(45.4) = 1.8, p = 0.1).
77
Participants were asked what they thought participation was and 34 (68.0%) participants said that
it involved being included or involved in an activity. Other responses as to what participation
was included working hard (n = 4, 8.0%), volunteering (n = 3, 6.0%), being allowed to take part
in an activity (n = 3, 6.0%), learning new things (n = 2, 4.0%), cooperating/helping (n = 1,
2.0%), and “what you do” (n = 1, 2.0%). Only two participants (n = 2, 4.0%) replied that they
did not know what participation was and both of these participants were quite young (7.8 and 8.8
years of age).
When participants were asked if their hemophilia prevented them from participating like others,
29 (58.0%) said that hemophilia only prevented them a very little bit (“barely” or “not really”) or
not at all. For the remaining 21 participants (42.0%), hemophilia primarily prevented them from
participating in contact sports and physical activity, particularly hockey and football.
Nearly half (n = 24, 48.0%) of the participants felt that personality did not affect how they
participated with others, but the rest reported that it did. Most of the responses revolved around
how being angry or being in a bad mood could lead to one participating less with others. Some
participants responded that when they participated with others, they felt happier.
4.7.2 Self-Perception
Overall, the participants scored highly when it came to self-perception subscales. Given that the
highest score on any of the subscales was 4.0, the mean scores for each of the subscales are
displayed in Table 6.
Table 6. Self-Perception Subscale Scores for Children & Adolescents (N = 50) Subscale Mean Standard Deviation Minimum-Maximum Social Acceptance 3.3 0.7 1.2-4.0 Athletic Competence 3.2 0.7 1.7-4.0 Global Self-Worth 3.5 0.6 1.7-4.0
Since the average grade of the sample was six, the scores from this study were compared to the
subscale means of sixth grade boys assessed by Susan Harter (Harter, 1985a). The means of this
sample are actually higher than the means of the sixth grade sample assessed by Susan Harter as
listed in Appendix G. Independent samples t-test were conducted and the domains of Social
Acceptance (t(67.8) = 2.2, p = 0.03) and Global Self-Worth (t(77.5) = 3.7, p = 0.0004) showed a
significant difference between the Harter reference groups and our sample. There was no
significant difference for Athletic Competence (t(68.17) = 0.2, p = 0.9).
78
When looking at the self-perception subscale scores for both children and adolescents, a strong
correlation was found between social participation and Social Acceptance (r = -0.5, p = 0.0003)
as well as between social participation and Global Self-Worth (r = -0.6, p = < 0.0001). The
negative correlation suggests that when children had better self-perception in terms of Social
Acceptance and Global Self-Worth, they tended to report better social participation. The
correlation between social participation and Athletic Competence was smaller and not
statistically significant (r = -0.2, p = 0.1), therefore the null hypothesis cannot be rejected that
there was no correlation between children and adolescents when it came to self-perceived
Athletic Competence and social participation restrictions.
Figure 3. Social Participation and Social Acceptance (Children & Adolescents) (N = 50)
79
Figure 3 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Social
Acceptance as measured by the Self-Perception Profile for Children or Adolescents, where
higher scores indicated more adequate self-perception. The confidence elliptical curve shows
95% prediction.
Figure 4. Social Participation and Global Self-Worth (Children & Adolescents) (N = 50)
Figure 4 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Global
Self-Worth as measured by the Self-Perception Profile for Children or Adolescents, where higher
80
scores indicated more adequate self-perception. The confidence elliptical curve shows 95%
prediction.
Figure 5. Social Participation and Athletic Competence (Children & Adolescents) (N = 50)
Figure 5 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Athletic
Competence as measured by the Self-Perception Profile for Children or Adolescents, where
higher scores indicated more adequate self-perception. The confidence elliptical curve shows
95% prediction.
81
Although there were differences in the correlations found between social participation and self-
perception for children and adolescents, when comparing the means of the children and
adolescent groups, allowing for unequal variances, no significant differences were found for any
of the subscales. Table 7 displays the results of a test that compared the means of children and
adolescents on the self-perception subscales.
Table 7. Comparing the Means of the Self-Perception Subscale Scores for Children &
Adolescents (N = 50) Subscale df t p
Social Acceptance 40.7 -1.4 0.2 Athletic Competence 38.3 -0.2 0.8 Global Self-Worth 42.0 -0.8 0.4
4.7.2.1 Children
The subscale mean scores on the Self-Perception Profile for Children are displayed in Table 8.
Table 8. Self-Perception Subscale Scores for Children (n = 30) Subscale Mean Standard Deviation Minimum-Maximum Social Acceptance 3.2 0.7 1.7-4.0 Athletic Competence 3.2 0.7 1.7-4.0 Global Self-Worth 3.5 0.6 1.7-4.0
The subscale scores for children revealed that there was a strong relationship between social
participation and Social Acceptance (r = -0.6, p = 0.002) as well as between social participation
and Global Self-Worth (r = -0.7, p = < 0.0001). Thus, if a child reported better self-perception
on the subscales of Social Acceptance or Global Self-Worth, he tended to report better social
participation. Again, the correlation between social participation and Athletic Competence was
smaller and not statistically significant (r = -0.1, p = 0.5) and thus the null hypothesis could not
be rejected that there was no correlation between self-perceived Athletic Competence and social
participation for children.
82
Figure 6. Social Participation and Social Acceptance (Children) (n = 30)
Figure 6 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Social
Acceptance as measured by the Self-Perception Profile for Children, where higher scores
indicated more adequate self-perception. The confidence elliptical curve shows 95% prediction.
83
Figure 7. Social Participation and Global Self-Worth (Children) (n = 30)
Figure 7 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Global
Self-Worth as measured by the Self-Perception Profile for Children, where higher scores
indicated more adequate self-perception. The confidence elliptical curve shows 95% prediction.
84
Figure 8. Social Participation and Athletic Competence (Children) (n = 30)
Figure 8 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Athletic
Competence as measured by the Self-Perception Profile for Children, where higher scores
indicated more adequate self-perception. The confidence elliptical curve shows 95% prediction.
4.7.2.2 Adolescents
The adolescent age group had slightly higher mean scores for Social Acceptance and Global
Self-Worth subscales than the children, however the Athletic Competence subscale mean score
for adolescents was slightly lower than for children. The mean scores for the subscales are
displayed in Table 9.
85
Table 9. Self-Perception Subscale Scores for Adolescents (n = 20) Subscale Mean Standard Deviation Minimum-Maximum Social Acceptance 3.5 0.7 1.2-4.0 Athletic Competence 3.2 0.7 2.0-4.0 Global Self-Worth 3.6 0.6 2.2-4.0
Unlike the children, the adolescents had a strong correlation between social participation and
Athletic Competence (r = -0.6, p = 0.005), but the correlations for Social Acceptance (r = -0.3, p
= 0.2) and Global Self-Worth (r = -0.3, p = 0.2) were moderate and not statistically significant.
This result indicates that adolescents who report higher scores in Athletic Competence will also
report that they have better social participation.
86
Figure 9. Social Participation and Athletic Competence (Adolescent) (n = 20)
Figure 9 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Athletic
Competence as measured by the Self-Perception Profile for Adolescents, where higher scores
indicated more adequate self-perception. The confidence elliptical curve shows 95% prediction.
4.7.3 Social Support
Table 10 displays the subscale scores for the Social Support Scale for Children and Adolescents.
Table 10. Social Support Subscale Scores for Children & Adolescents (N = 50) Children (n = 30)
Subscale Mean Standard Deviation Minimum-Maximum Parental Support/Regard 3.6 0.6 1.8-4.0 Teacher Support/Regard 3.4 0.7 1.2-4.0 Classmate Support/Regard 3.4 0.6 1.3-4.0 Close Friend Support/Regard 3.5 0.5 1.8-4.0
Adolescents (n = 20) Subscale Mean Standard Deviation Minimum-Maximum Parental Support/Regard 3.7 0.4 2.5-4.0 Teacher Support/Regard 3.3 0.8 1.7-4.0 Classmate Support/Regard 3.5 0.6 1.3-4.0 Close Friend Support/Regard 3.5 0.7 1.0-4.0
Appendix H details the correlations for children and adolescents separately. Although
significant correlations were found between social participation and social support for children
only, when comparing the means of the children and adolescent groups, allowing for unequal
variances, no significant differences were found for any of the subscales. Table 11 displays the
results of a test that compared the means of children and adolescents on the social support
subscales. Given that there were no significant differences between the two group means, further
analyses discussed below combined both children and adolescent scores.
Table 11. Comparing the Means of the Social Support Subscale Scores for Children &
Much like self-perception, when the self-reported social support of the sample was compared to
the sample assessed by Susan Harter, it was apparent that the mean subscale scores for sixth
grade boys were higher in the sample of boys with hemophilia (see Appendix I). Indeed, when
87
independent samples t-tests were conducted, there were significant differences between the
Harter reference group and our sample in all of the domains (see Appendix J).
A strong correlation was found between social participation and Parental (r = -0.6, p = < 0.0001),
Teacher (r = -0.5, p = < 0.0001), and Classmate (r = -0.6, p = < 0.0001) Support/Regard.
Moderate correlations were found between social participation and Close Friend (r = -0.4, p =
0.002) Support/Regard. These results indicated that children who endorsed better support or
regard from their parents, classmates, teachers, and close friends also endorsed that they have
better social participation.
88
Figure 10. Social Participation and Parental Support/Regard (N = 50)
Figure 10 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Parental
Support/Regard as measured by the Social Support Scale for Children, where higher scores
indicated more perceived social support. The confidence elliptical curve shows 95% prediction.
Figure 11. Social Participation and Teacher Support/Regard (N = 50).
Figure 11 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Teacher
89
Support/Regard as measured by the Social Support Scale for Children, where higher scores
indicated more perceived social support. The confidence elliptical curve shows 95% prediction.
Figure 12. Social Participation and Classmate Support/Regard (N = 50)
Figure 12 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Classmate
Support/Regard as measured by the Social Support Scale for Children, where higher scores
indicated more perceived social support. The confidence elliptical curve shows 95% prediction.
90
Figure 13. Social Participation and Close Friend Support/Regard (N = 50)
Figure 13 illustrates the correlation between social participation as measured by the Participation
Scale for Kids, where higher scores indicated more restrictions with participation, and Close
Friend Support/Regard as measured by the Social Support Scale for Children, where higher
scores indicated more perceived social support. The confidence elliptical curve shows 95%
prediction.
When it came to disclosing their hemophilia to others, the participants in this sample tended to
disclose to most people in their life. Of the 50 participants, 47 indicated that their family
members knew (94.0%), 31 had told their close or best friends (62.0%), 35 had told their
teachers (70.0%), and 24 had told their classmates (48.0%).
91
4.7.4 Family Impact
When it came to the impact on family members, the scores on the PedsQL™ Family Impact
Module were high, which suggested that overall the participants had a low negative impact on
the family.
There was a moderate correlation between social participation and the subscales tapping family’s
Social Functioning, which indicated that when parents/guardians reported better social
functioning, their children tended to have better social participation.
All other subscales did not have statistically significant correlations. The results of the PedsQL™
Family Impact Module are in Table 12.
Table 12. PedsQL™ Family Impact Module Results and Correlation with the Participation Scale
for Kids (N = 50) Variable Mean Standard Deviation Minimum-Maximum Pearson r Total Score 3074.0 571.1 1275.0-3600.0 -0.2 Physical Functioning 91.4 14.0 37.5-100.0 -0.2 Emotional Functioning 80.8 19.7 25.0-100.0 -0.2 Social Functioning 89.8 18.9 25.0-100.0 -0.3* Cognitive Functioning 90.5 19.9 20.0-100.0 -0.2 Communication 80.8 20.2 25.0-100.0 -0.2 Worry 73.2 22.5 15.0-100.0 0.0 Daily Activities 88.8 19.1 33.3-100.0 -0.3 Family Relationships 87.9 18.1 20.0-100.0 -0.2 Note. *Correlation is significant at p < 0.05.
4.8 Model Checking
For each Pearson correlation coefficient, a number of tests were run to verify that data
assumptions were held.
4.8.1 Normality of Residuals
Residuals were considered normally distributed if the Shapiro-Wilk test was non-significant.
The normality of residuals are listed in Appendix K. Nearly all the Shapiro-Wilk tests were
significant (except for Athletic Competence for Adolescents and Parental Support/Regard), thus
the data do not meet the assumption of homoscedasticity. However, the central limit theorem
states that given a large enough sample size, the means of the data will be normally distributed,
even if the original distribution is not. Therefore, since the data is being used to make inferences
on means, the use of parametric statistics such as Pearson correlations is justified.
92
4.8.2 Influential Outliers
Observations were identified as influential outliers if the DFFIT value was greater than 2, a
common rule of thumb (Vogt & Johnson, 2011). No influential outliers were identified using
this test.
4.8.3 Straight-Line Relationship and Homoscedasticity
Residual plots constructed for the relationship between social participation and self-perception
(Social Acceptance, Athletic Competence, and Global Self-Worth subscales), social support
(Parental, Teacher, Classmate, Close Friend Support/Regard subscales), and Impact on the
Family (Physical Functioning, Emotional Functioning, Social Functioning, Cognitive
Functioning, Communication, Worry, Daily Activities, and Family Relationships subscales)
identified some relationships and patterns, suggesting that perhaps the residuals were not
normally distributed. Residual plots are displayed in Appendix L.
4.8.4 Non-Parametric Tests
Non-parametric tests were conducted for comparison purposes given that some assumptions for
the use of Pearson correlations were violated. Spearman’s rank-order correlations were
compared to Pearson correlation coefficients and some changes were observed regarding the
strength of the correlation, however the outcomes were essentially unchanged. See Appendix M
for a detailed analysis.
The fact that the residuals are non-normal suggests that there may be some deviations from the
linear model. These changes may be partially due to the fact that the data collected was ordinal
in nature rather than interval or ratio. Ordinal data is often assessed using parametric tests and
given that the results of the non-parametric tests were not substantially different from those of
the parametric tests, we have chosen to report the Pearson correlation coefficients.
4.9 Sensitivity Analysis
To test the robustness of this analysis, sensitivity analysis of a worst- and best-case scenario were
performed for missing data. Minimum values were specified as the 10th percentile, while
maximum values were specified as the 90th percentile. In keeping with the previously reported
data, Pearson correlation coefficients were performed on the data.
93
Since less than 5% of the data was missing for any one variable, there were minimal changes
witnessed in the re-analysis. Only data that reached statistical significance with single
imputation were discussed. Data are presented in Appendix N.
4.9.1 Worst-Case Scenario
When imputing the 10th percentile for missing values, correlations were essentially unchanged
for all three of the self-perception subscales for children and adolescents, children alone, and
adolescents alone.
Nearly imperceptible changes were observed between social participation and social support for
all four subscales.
A very minor change was seen between social participation and impact on the family. Social
Functioning no longer met significance (r = -0.3, p = 0.1) while Daily Functioning met
significance (r = -0.3, p = 0.05). Again, these changes were very minor and the correlations
moderate at best.
4.9.2 Best-Case Scenario
Using the 90th percentile for missing values yielded fewer changes to the analysis. Social
Functioning no longer met significance (r = -0.3, p = 0.1).
4.10 Linear Modeling
Multiple linear regressions were constructed to examine if social participation was affected by
self-perception and social support. Although the ICF model suggests that the relationships
between all domains are bidirectional, the data seemed appropriate for exploratory analysis to
determine if any of the variance in social participation could be accounted for by these variables.
4.10.1 Self-Perception
When looking at social participation and self-perception for all participants, the omnibus F-test
was significant and thus individual covariates were further evaluated for significance, as shown
in Table 13.
94
Table 13. Parameter Estimates for Linear Regression of Social Participation and Self-
Perception Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 10.4 (3, 46) < 0.0001 Social Acceptance -1.4 1.3 -1.1 0.3 Athletic Competence -0.6 1.1 -0.6 0.6 Global Self-Worth -5.3 1.5 -3.5 0.001 Note. R2 = 0.4
This model suggested that 40.2% of the variance in social participation could have been
explained by self-perception; in particular it was Global Self-Worth that exerted an effect.
4.10.1.1 Model Checking
The use of linear modeling requires that the data meet several assumptions. The data for self-
perception and social participation did not meet all the required assumptions; therefore a linear
model may not be best suited for this data.
4.10.1.1.1 Normality of Residuals
Residuals initially appeared to be normally distributed when using graphical tools (histograms
and Q-Q plots, see Appendix O) and were not skewed (skewness = 1.0). However, when model
assumptions of normality were determined using the Shapiro-Wilk test, the significant test result
(W = 0.9, p = 0.01) suggested that the residuals were in fact from a non-normally distributed
population. Since the residuals were not normally distributed, there may have been a non-linear
relationship between social participation and self-perception and linear regression may not have
been the best fit.
4.10.1.1.2 Influential Outliers
No influential outliers were identified when using tests of DFFIT.
4.10.1.1.3 Straight-Line Relationship and Homoscedasticity
To verify the straight-line relationship and homoscedasticity, residual plots were created. There
was no obvious relationship or pattern visible from the residual plot. However when variation
around the predicted line was plotted, the residuals did not follow a straight line nor did they
appear to be homoscedastic, suggesting that a linear line may not have been appropriate for the
data.
95
Tests for collinearity used a Variance Inflation Factor cut-off of > 4.00 to indicate collinearity
(Victor, 2012). The tests showed that Social Acceptance (VIF = 1.8), Athletic Competence (VIF
= 1.2), Global Self-Worth (VIF = 1.6) were not collinear with one another.
4.10.2 Social Support
The omnibus F-test for social participation and social support was significant and therefore
individual covariates were further examined as shown in Table 14.
Table 14. Parameter Estimates for Linear Regression of Social Participation and Social Support Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 11.5 (4, 45) < 0.0001 Parental Support/Regard -4.2 1.9 -2.2 0.0 Teacher Support/Regard -1.1 1.3 -0.8 0.4 Classmate Support/Regard -4.2 1.8 -2.3 0.0 Close Friend Support/Regard 0.6 1.7 0.4 0.7 Note. R2 = 0.5.
According to this data, social support explains 50.6% of the variance in social participation and
the domains of parental and classmate support/regard are the domains that have the strongest
influence.
4.10.2.1 Model Checking
Again, the data for social support and social participation did not meet the assumptions required
for the use of linear modeling. Linear models may not best describe the data presented.
4.10.2.1.1 Normality of Residuals
Graphical tools such as histograms and a Q-Q plot initially suggested that the residuals were
relatively normal (see Appendix P). However, the Shapiro-Wilk test used to determine
normality was significant and therefore the residuals were not normally distributed (W = 1.0, p =
0.05). Residuals were skewed (0.6) but within an acceptable range; kurtosis was also acceptable
(1.2).
4.10.2.1.2 Influential Outliers
No influential outliers were identified when using tests of DFFIT.
96
4.10.2.1.3 Straight-Line Relationship and Homoscedasticity
Residual plots again did not indicate any relationships however variation around the predicted
line showed a non-linear relationship and lack of homoscedasticity.
Again, tests for collinearity using a cut-off of > 4.0 on the Variance Inflation Factor were used.
The results showed that none of the domains were collinear (Parental VIF = 2.3, Teacher VIF =
2.3, Classmate VIF = 2.7, Close Friend VIF = 2.5).
4.10.3 Stratification Variables
Multiple linear regressions were conducted using age and factor level as predictor variables of
social participation and self-perception. None of the models achieved significance. See
Appendix Q.
97
Chapter 5 Discussion
5 Interpretation of Findings Hemophilia is a condition that can influence many factors in one’s life. The ICF model suggests
that there is a dynamic interaction between hemophilia and the personal and environmental
factors in an individual’s life. These factors are engaged in a bilateral relationship, where
changes in one factor can lead to changes in the other factors. Using this model, the data support
the conclusion that the personal factor self-perception contributes to his social participation.
Similarly, social support is an environmental factor that can impact social participation. Given
the bilateral nature of the model, social participation will also exert an influence on these
contextual factors as well. It is the interplay observed between all of these variables that
determines one’s functioning and disability, including one’s social participation.
The results of this study indicated that boys with hemophilia in Toronto, Ontario, had minimal
social participation restrictions, high levels of perceived competence, and perceived that other
supported them. This study was highly generalizable given the sampling strategy. There appears
to have been moderate to strong relationships between social participation and self-perception as
well as social participation and social support. There was little relationship observed between
social participation and impact on the family.
5.1 Sample
There was a high participation rate in this study with very few patients declining to participate.
Research staff was cognizant of the burden of participating in the research study and thus
structured the study visit to coincide with a regularly scheduled clinic appointment. This ensured
that extra trips to the hospital were not required. Research staff was also aware that participants
may fatigue throughout the multiple assessments and encouraged participants and their family
members to take breaks as needed. By offering to remove as many barriers to participation as
possible, research staff was able to increase recruitment.
The participants came largely from married or common-law households and identified both their
mother and father as their primary guardians. In spite of the extreme stressors and pressures of
98
illness on the family, the findings of this study support the works of others who have found a
high rate of marriage amongst parents of children with hemophilia and other illnesses (Cadman,
Rosenbaum, Boyle, & Offord, 1991; Kellerman et al., 1980; Varekamp et al., 1990; Wiedebusch,
Pollmann, Siegmund, & Muthny, 2008).
Treatment for the majority of participants in the study cohort happened in the home with
infusions typically being administered by oneself or by one’s mother. Other studies have also
found that mothers tended to be more involved in the daily care of the patient (Beeton et al.,
2007) and that fathers were less likely to infuse (Herrick, Nussbaum, Holtzman, & Wissow,
2004). Possible reasons are that mothers may be more likely to stay at home with an ill child
(Lindvall et al., 2012) and therefore more likely to attend doctors appointments. The mother
may feel more comfortable administering the treatment initially due to increased exposure and
this may continue to perpetuate with the mother administering treatment more often.
Furthermore, mothers in particular have been identified as being overprotective of the child with
hemophilia, perhaps due to feelings of guilt about the genetic transmission of the disease (Agle,
1984).
Attending school increases the opportunities for social participation. Nearly all of the
participants in this sample attended school, which was in keeping with research completed in the
United States that found approximately 98% of their sample of boys with hemophilia attended
school and 2% were home-schooled (Shapiro et al., 2001). Shapiro et al. (2001) found a
significant difference in school attendance between participants who had a higher number of
bleeding episodes (defined as 12 or more bleeding episodes in the year before enrollment) than
the participants who had fewer bleeding episodes. Although there may be differences in school
attendance within the hemophilia population, research conducted as early as 1987 in Finland
between persons with hemophilia and controls found no significant differences in the number of
school hours absent (B. Kvist et al., 1990).
School attendance today is in sharp contrast to data from 1971, which found that persons with
hemophilia were absent from school more than double the hours of controls (B. Kvist et al.,
1990). The increased school attendance may be attributed to advances in treatment. This sample
was composed of participants primarily treated with prophylactic therapy, which essentially
converts a severe to a moderate phenotype (Feldman, Pai, Rivard, & Israels, 2006). Having
99
access to treatment and a prophylactic treatment regimen can minimize bleeds, allowing the
person to participate more and engage in regular activities including school.
5.2 Social Participation
Participation was defined as “involvement in a life situation” (World Health Organization, 2001)
and it was important that the measure used took into account an assessment of the subject’s
limitations or restrictions against the norm. The measure chosen included an element of
desirability in its measurement of social participation. Although much of the research conducted
to date on participation has focused on sport, the goal of this study was to take a more inclusive
approach to participation and assess all aspects of social life. It is evident that the concept of
participation encompasses many facets in the respondents’ minds. The Participation Scale for
Kids did not have an explicit focus on sport, which allowed for participants not interested in
sport to still express their ability to participate.
An important finding of this study was that most of the sample reported that they experienced
very few social participation restrictions. This may be due to the fact that the majority of the
sample was on prophylaxis, allowing the study subject to engage in activities similar to their
peers. This is comparable to the Western European sample studied by Gringeri et al. (2004),
who found that patients on prophylaxis were less impaired than those using on-demand
treatment. Similar to Khair et al.’s (2011) United Kingdom sample, most of this sample felt that
their hemophilia did not prevent them from participating like their peers. As discussed earlier,
samples that were studied in the 1970s prior to the advent of prophylaxis would have had
markedly different social participation as patients may not have been able to attend school
regularly and may have suffered from hemophilic arthropathy at an early age.
When this sample was asked specifically if their hemophilia prevented them from participating
like their peers, the majority of positive answers related to inability to participate in certain
sports. Since sport is desirable in Western culture (Heijnen et al., 2000), it was not a surprise
that many of the boys wished to be able to participate similar to their peers. More importantly,
early adolescence is a time to develop a sense of industry or mastery of skills, in physical,
academic, and social tasks (Manning, 1988). Spending time in activities that help the child build
his competencies may be critically important in shaping the child’s development (Mahoney,
2005, as cited in Slutzky & Simpkins, 2009). Being able to competently engage in sport may be
100
especially salient for boys, as boys tend to spend more time in sports and report higher sport self-
concepts and stronger beliefs in the value of sports than girls (Slutzky & Simpkins, 2009).
Denying children the opportunity to engage in sport as their peers may thwart opportunities to
gain competence in that area and could lead to lower self-perception. Slutzky and Simpkins
(2009) found that the more time children participated in team sports, the better they felt about
their athletic competency, which was positively associated with self-esteem. Sport participation
for children with low peer acceptance may be particularly beneficial as sport has the potential to
address both peer relationships and individual development (Smith, 2003; Weiss & Duncan,
1992, as cited in Slutzky & Simpkins, 2009).
Only a small number of participants indicated that they experienced mild or moderate
participation restrictions. Significantly, not one participant endorsed experiencing severe
participation restrictions. This finding is consistent with study results from Heijnen et al. (2000),
who found that 95% of boys with hemophilia in their study in the Netherlands were able to
participate in physical education classes. It appears that although the boys may be prevented
from participating in certain sports, they were still able to socially participate on par with their
peers. In fact, many boys with hemophilia actively engage in sports such as swimming, cycling,
tennis, skating, and skiing (Heijnen et al., 2000).
Children had a slightly higher number of participation restrictions than adolescents due to a
couple of influential outliers who had mild to moderate participation limitations. These four
participants skewed the mean and they were different than the rest of the sample because the
majority of the sample indicated that they experienced no participation restrictions. These four
participants were also some of the youngest participants in the sample, and it may be that these
participants felt more limited or overprotected by their parents/guardians or teachers. Hamburg
(1983, as cited in Spirito et al., 1990) noted that the risk of parental overprotection was
particularly striking for children with chronic illnesses and that this could lead to limited
opportunities for socialization. Well-meaning friends and family members may hamper the
child’s ability to fend for himself and make it difficult for the patient to feel a sense of
empowerment and capability (Thomas & Gaslin, 2001). While taking inappropriate risks are not
encouraged, avoiding any activity with the potential for injury would lead to a sheltered life that
would not allow for adequate development of the self and could lead to feelings of inadequacy
and inferiority (Manning, 1988).
101
Adolescents reported fewer social participation restrictions than the children in this sample and
all adolescent scores fell below the mild restriction range. This data suggests that as children
age, the number of restrictions to their social participation decreases. It may be that as children
get older, they adapt their activities to be in line with things that they have fewer participation
restrictions with such as less physically strenuous activities or more academically inclined
activities. Research conducted by Köiter et al. (2009) found that the choice of sports that
patients with hemophilia engaged in appeared to be based on adaptation to their disease. Heijnen
et al. (2000) had similar findings as they concluded that most boys and men with hemophilia
were sensible in choosing a sport that accommodated their hemophilia, as many indicated they
would like to play soccer but did not because of their disease. Our sample echoed these thoughts
as those who commented that their hemophilia restricted them identified contact sports such as
American football and ice hockey as being off-limits. As stated earlier, however, most of the
participants felt that they did not experience social participation restriction due to their
hemophilia.
Another possibility as to why adolescents did not experience as many social participation
restrictions as children is that adolescents may engage in activities that their parents forbid
(Buzzard, 1996). Thus the adolescent participants may have reported fewer participation
restrictions because they were able to engage in those activities that they enjoyed, regardless of
whether or not their parents supported it. This arises in part because children begin to conform
to their peers as they progress through their elementary school years (Spirito et al., 1990). Peers
become the benchmark upon which the child measures his success and peers’ opinion of their
success or failure may be of far more importance to the child than the opinions of his parents or
teachers (Maier, 1969, as cited in Manning, 1988). Adolescents are also beginning to search for
their identity at this age and may be concerned about how they appear to others as compared to
how they feel they are (Manning, 1988).
As the Participation Scale for Kids scale was modified and administered for the first time to this
sample, there was no reference group for comparison. It is also possible that some of the
younger participants did not understand the open-ended questions at the end of the instrument
and it may have been more appropriate to explore these topics with the older participants instead.
However, the internal consistency of the scale was good and the results of the correlations
between social participation, self-perception, and social support suggest good convergent
102
construct validity. Given the very few identified participation restrictions, it can be concluded
that children with hemophilia in Toronto, Canada, seem to have relatively few social
participation limitations. Difficulties that arise with social participation may impact and be
impacted by factors such as self-perception and social support, as well as issues with body
function and structure.
It is important to remember that although the majority of participants with hemophilia in this
sample were able to participate on par with their peers, there were some participants who
reported very low self-perception scores in each of the domains. In spite of all that has been
achieved, it is important to bear in mind that chronic illness is a part of these boys’ lives and
sometimes it can be a very difficult illness to cope with (Thomas & Gaslin, 2001).
5.3 Self-Perception
Self-perception is defined as one’s perception of themselves, which may include his or her
feelings of competence and their overall self-worth (Harter, 1985a). Despite living with a
chronic illness, self-perception in this sample was actually higher than the reference groups
studied by Harter et al. (1985) in Colorado, United States. Previous research has not come to a
unanimous conclusion about self-perception in boys with hemophilia. Some have suggested that
boys with hemophilia are more likely to have lower feelings of perceived competence (Agle,
1984), while others have said that self-esteem and self-perception are mostly comparable to
healthy peers (Hegeman et al., 2011; Kellerman et al., 1980). Regardless, Stein and Jessop
(1984, as cited in Thomas & Gaslin, 2001) note that self-esteem deteriorates and health status
subsequently worsens if children are unable to perform age-appropriate roles and tasks.
Global Self-Worth in this sample was the perceived competence that had the highest rating. If a
child indicated higher scores on the Global Self-Worth subscale, this meant that the child liked
themselves as a person, were happy with the way they were leading their life, and were generally
happy with the way they were (Harter, 1985a). Social Acceptance was the next highest scoring
subscale and this suggested that the boys with hemophilia generally felt accepted by their peers
and/or felt popular. The lowest scoring subscale was Athletic Competence, which examined the
boys’ sports and outdoor games competency, for which most boys felt adequate. The reference
scores for sixth grade boys (Samples A and B) were similarly ordered with the highest subscale
103
scores in the domain of Global Self-Worth, followed by Athletic Competence, and then Social
Acceptance (Harter, 1985a).
The higher mean score in this sample may be due to the fact that this sample had a wider range
of grades completing the survey (participants ranged in grade from grades 2-12) and older
participants tended to report higher scores on the self-perception subscales than the younger
participants. There may be more variability of scores within one grade (reference scores were
based on sixth grade boys) than across ten grades (this sample). Furthermore, the reference
scores were gathered thirty years earlier (1980s), drew on different neighborhoods (lower middle
class to upper middle class), was mostly Caucasian (approximately 90% of the sample), and from
a different country (United States). All these factors may have contributed to differing mean
scores.
Alternatively, it may be that these findings support the notion that children with hemophilia are
no more susceptible to lower levels of self-perception than their peers. This is in line with recent
research by Hegeman et al. (2011) who found that in general, self-perception scores were
comparable between children with hemophilia and healthy peers. Older research that compared
multiple illnesses (oncology, cardiology, diabetes, cystic fibrosis, nephrology, and
rheumatology) and healthy controls also found no differences in self-esteem between the various
illness groups or between the healthy controls and the ill groups (Kellerman et al., 1980).
Hemophilia shares some similarities with diabetes, as both illnesses allow for patients to learn
self-injection/infusion techniques. Kellerman et al.’s (1980) study also looked at locus of
control, which the authors defined as one’s attitude towards his ability to control his own destiny.
They found that there was no difference in locus of control between the healthy group and the
diabetes or cystic fibrosis groups. The nature of chronic illness “often opposes children’s
developmental striving for independence” (Coupey and Cohen, 1984, as cited in Thomas &
Gaslin, 2001), but mastering the task of self-injection/infusion may allow for the patient to feel
that he is more in control of his future and feelings of industry can contribute to greater self-
perception.
While earlier research theorized difficulties for boys with hemophilia, including low self-esteem
(Agle, 1984), self-perception was found to be quite high for our sample. The difficulties that
boys with hemophilia faced in the 1980s may be quite different than the challenges experienced
104
today. Patients today have access to ultra-safe, effective therapy and prophylactic regimens that
allowed for safe engagement in nearly all activities. Participating in activities like one’s peers
may allow the boy with hemophilia to not only develop a sense of mastery, but also be accepted
by his peers. Both of those may contribute to feelings of competence and self-perception and
vice versa.
Alternatively, it is possible that this sample was less likely to endorse feelings of negative self-
worth in the presence of their medical team or their family members, who were often present
when study questionnaires were being completed. Participants may not have wanted to worry
their medical team or family members and thus endorsed more positive feelings of self-worth.
That being said, steps were taken to reduce social desirability, as participants were encouraged to
complete the questionnaires in a self-report manner whenever possible and individual
questionnaire results were not shared with their medical team. Furthermore, the setting in which
this research was conducted was likely similar to other research carried out in hospital clinics.
The results of this study are consistent with those reported by Trzepacz et al. (2003), who found
differences between boys with hemophilia and comparison peers in the United States on the
dimensions of Scholastic Competence, Athletic Competence, and Global Self-Worth, but the
scores were still within the normal range. Hegeman et al. (2011) in the Netherlands found that
for the most part, boys with hemophilia had perceived competence on par with a normative
sample. The only significant differences found were in the domains of Global Self-Worth for
children and Close Friendship for adolescents but these scores were still relatively high,
indicating better self-perception. The results of this sample support the findings of other studies
in developed countries and clearly demonstrate that boys with hemophilia treated with
prophylaxis did not have reduced levels of self-perception.
5.3.1 Primary Hypothesis
The primary study hypothesis was that an association existed between self-perception and social
participation. The present data shows that overall there were moderate to strong correlations
between social participation and the three self-perception subscales used in this study. Thus,
participants who reported more adequate self-perception in terms of Social Acceptance, Athletic
Competence, and Global Self-Worth also reported fewer social participation restrictions.
105
Accordingly, those participants who experienced more difficulties with social participation had
lower scores on self-perception subscales.
Social Acceptance was moderately associated with social participation for younger boys with
hemophilia, which shows that peer integration may be related to social participation. The
associations between social participation and Social Acceptance were lower and not statistically
significant for adolescents, which would seem to indicate that participating socially and having a
sense of being accepted by one’s peers may be less related for adolescents than for children.
This result was unexpected because it was initially hypothesized that adolescents who felt
competent in the domain of Social Acceptance would also experience fewer restrictions in social
participation. It may be that our sample of adolescents was too small (n = 20) to find a
significant effect size, or it may be that as adolescents aged, they were faced with more
participation restrictions compared to their peers but continued to have perceived competence in
the area of Social Acceptance.
Indeed, Heijnen et al. (2000) found that 68.0% of their sample of older boys with hemophilia in
the Netherlands reported that they sometimes did not attend physical education sessions because
of bleedings, a clear restriction on their social participation. Other work in the Netherlands by
Hegeman et al. (2011) found that adolescents with hemophilia reported high levels of Social
Acceptance, albeit with scores lower than the Dutch reference group. Our data appear to reflect
the work carried out in the Netherlands and show that even when faced with social participation
restrictions, one can have good self-perception in the domain of Social Acceptance.
There was a lack of correlation between social participation and Athletic Competence when
children and adolescents were combined in the analysis. There was a significant correlation
when only adolescents were examined, which would seem to indicate that sports and outdoor
games are more intimately tied to social participation for adolescents than children. For
adolescents, Athletic Competence was the most strongly associated subscale with social
participation, suggesting that as children get older Athletic Competence becomes more strongly
associated with social participation. This finding is not surprising as older children may become
more involved in team sports and gain competence and mastery of the skills required to
participate in sports and outdoor games, which could then significantly affect one’s social
participation.
106
As described earlier, sporting activity was the participation restriction cited most often by this
sample. In line with previous recommendations (Buzzard, 1996), participants noted that their
hemophilia tended to prevent them from participating in contact sports. Our data show that some
children with hemophilia had very high levels of athletic competence but some had very low
levels. It is possible that since Western culture desires sport participation (Heijnen et al., 2000),
participation in sport as a child is expected and accepted as part of daily life. Those who do not
thrive in sport may self-select themselves out of sport-related activities as they get older but
participate fully in other activities. Thus, one may find children with high levels of social
participation and low levels of Athletic Competence and vice versa. Alternatively, younger
children in the sample may be less engaged in organized sport than their adolescent counterparts
and therefore Athletic Competence may still be in the developing stages for some participants.
The Global Self-Worth subscale showed differences in the sample between the children and
adolescents. For children, social participation was most strongly associated with the subscale of
Global Self-Worth, indicating that when children feel generally happy with themselves as a
person, they also tend to engage in social participation more. On the other hand, the data for
adolescents was not sufficient to reject the null hypothesis that there was no association between
social participation and Global Self-Worth. As discussed above, it appears that as children get
older, other aspects (such as Athletic Competence) become more salient and more associated
with social participation than Global Self-Worth and this may partially explain the differences
observed.
It has been suggested that children with strong self-esteem are better able to handle challenges of
daily life (White, 1998, as cited in Hegeman et al., 2011; Thomas & Gaslin, 2001). Thus,
developing strong self-esteem and self-worth may be crucial to an individual’s success and social
participation, especially in the case of children with hemophilia where some may be vulnerable
to lower self-worth.
5.4 Social Support
Overall, boys in this study felt that they were well supported by their parents, teachers,
classmates, and close friends. This is in contrast to previous research that has found that
adolescents and adults with hemophilia report difficulties with perceived support (Gringeri et al.,
2004; Iannone et al., 2012). One reason for the differences observed may be that this sample was
107
predominantly composed of children, who may be exposed to a wider support network than
adults.
The importance of a social support network has far-reaching consequences. If a patient
perceives that he has a good social support network, he may feel less depressed, have better
compliance, and higher health-related quality of life (Gringeri et al., 2004; Iannone et al., 2012;
Petrini & Seuser, 2009). Developing a strong social support network for the patient may prevent
patients from concealing dangerous or risk-taking behaviors. Buzzard (1996) notes that if a
patient is forbidden from participating in certain sports, “there is a good chance that he…will
play it anyway without the parents’ knowledge.”
There is evidence that perceived emotional social support may be protective in the psychological
adjustment of children with cancer and their siblings. In the United States, it was found that
classmate social support was the most important factor in the psychological adjustment of newly
diagnosed cancer patients (Varni, Katz, Colegrove, & Dolgin, 1994). Meanwhile it was found
that Canadian siblings with stronger perceived social support reported fewer depressive and
anxiety symptoms, and their parents reported fewer behavioral problems (Barrera, Fleming, &
Khan, 2003). Children who have good social support may also have better psychological
adjustment, as they would have resolved the identity versus role confusion stage and developed a
strong sense of self. This may, in turn, increase the likelihood that they have higher levels of
social participation.
As children progress through the normal developmental stages, they begin to carve out identities
for themselves and may adopt idols and heroes to whom they look up to. Parents and teachers
may find themselves in these roles and may need to provide support to the child as he discovers
inevitably that his ideological thinking is not realistic. Parents and teachers may also need to
adjust to the changing identities of the child and they can nurture these relationships to help the
adolescent pass through the identity versus role confusion stage.
Peers can also have a substantial impact on the adolescent’s search for identity. Erikson (1963,
as cited in Manning, 1988) noted that cliques and crowds can be “remarkably clannish and cruel
in their exclusion of those who are different.” The boy with hemophilia may already feel
different than his peers and has to overcome that hurdle in addition to the normal adolescents
worries and concerns. Although there is little that can be done to abolish cliques and clans from
108
the schoolyard, having the support of just one close friend or classmate may be enough for the
child to feel he is cared for and loved.
One difference between the reference groups and this sample is that in the reference groups,
parent and teacher support was systematically higher than support from classmates and close
friends (Harter, 1985a). Our sample found the highest levels of perceived support to come from
parents, close friends, and classmates. Teacher support was lower overall, but the scores were
still quite high. Harter (1985b) suggests that teacher support tends to decline as a function of
grade level. Given that our sample had a wider grade range (participants in the study were in
grades 2-12), the larger proportion of older participants may have had an influence on teacher
support overall. Indeed, our sample indicates that younger participants felt more supported by
their teachers than the older participants, which may have lowered the overall mean score in the
study cohort.
Participants in our sample tended to disclose their illness to others, although not at the same rate
as in other studies. Less than two-thirds of the sample had disclosed their illness to their close or
best friends, in contrast to the 94.0% who had disclosed their illness to friends in a study by
Lindvall et al. (2006). Participants were not probed as to why they had not disclosed their illness
to others, but perhaps since the participants in this sample encountered relatively few
participation restrictions, they felt it unnecessary to disclose their illness to others.
5.4.1 Secondary Hypothesis
Regarding the second aim of this study, results indicate that social support and social
participation were associated. There was a moderate correlation between social participation and
support from close friends, but a strong correlation between social participation and support from
classmates. Since close friends at this age are also likely one’s classmates, it is not surprising
that the correlation is higher for classmates. In fact it is not uncommon for classmates to be used
as a peer control group when studying children (e.g. B. Kvist et al., 1990; Noll, LeRoy,
Bukowski, Rogosch, & Kulkarni, 1991; Trzepacz et al., 2003). Participants with close friends
and classmates whom they feel supported by will thus have peers with whom they can socially
participate alongside.
109
A strong correlation was found between social participation and support from parents. Strong
perceived social support from one’s parents might enable a child to participate more readily if
given permission or encouraged to participate in activities. Indeed it has been documented that
parents may struggle with wanting their child to have a normal active lifestyle but also wishing
to protect the child from risk, leading to restrictions on the child’s participation (Shaw & Riley,
2008). However, as discussed earlier, not having the support of one’s parents may not be
sufficient to prevent children from participating in those restricted activities. Having parental
support may include open communication between the patient and the parent about activities he
wishes to participate in. A compromise may be possible that allows the patient to engage in said
activity safely, such as using protective gear or modifying prophylaxis appropriately. Although
the parent may ultimately object to the patient’s participation in said activity, the patient may still
benefit from feeling that his opinion was important, another aspect of social participation.
A modest correlation was found between social participation and support from teachers. Support
from one’s teacher may be especially important in physical education classes. Knowledge of the
patient’s strengths and limitations would be an asset to encourage appropriate levels of
participation (Heijnen et al., 2000). Furthermore, there has been some evidence that patients
with hemophilia may be subject to bullying (B. Kvist et al., 1990) and having a teacher that one
can open up to about such experiences may make the child feel safer at school and allow for
appropriate intervention, if necessary.
Manning (1988) described the implications for educational practice when it came to assisting
students through the early adolescent stages. The author concluded that the significant adults in
the student’s life would greatly impact whether a young person emerged from adolescence with
feelings of industry and the development of a strong personal identity. Without support from
adults whom they can trust, children may feel inferior and confused about demonstrating roles,
and this may hamper their pursuit of social participation.
5.5 Impact on the Family
According to the present data, hemophilia had a relatively limited impact on the family with few
parents or guardians reporting any difficulties. Parents reported the most difficulties in the
domain of Worry, but the mean score for Worry was still high, indicating good family
functioning. This finding is not surprising as Wiedebusch et al. (2008) found parents of children
110
with hemophilia reported fewer psychosocial strains when compared to parents of children with
diabetes and rheumatologic diseases.
An earlier study in the Netherlands found that activities or relations were almost never disturbed
because of extra care required for patients with hemophilia; hemophilia was perceived as a minor
burden on daily practical problems (Varekamp et al., 1990). More recently, an international
multi-center cohort study in Malmö, Stockholm, London, and Toronto found that caregivers of
boys with hemophilia and inhibitors reported the disease was a medium burden, while parents of
children without inhibitors reported a low burden (Lindvall et al., 2012). There was also a
significant difference in scores between those with inhibitors and those without on the Impact on
Family Scale, with caregivers of children with inhibitors reporting more negative impact on the
family. Another earlier study in the United States found that approximately one-fifth of parents
had difficulty adjusting to hemophilia, but it is important to note that approximately one-fifth of
that sample also had HIV co-infection, which could have drastically changed the patient’s health
trajectory (Drotar et al., 1997). However it appears that in this sample, where children with co-
morbidities were excluded and none of the participants had inhibitors at the time of the study,
overall family functioning was good.
Some authors have noted that the initial diagnosis can be very stressful for the family of a child
with hemophilia (Agle, 1984; Coppola et al., 2011), especially if there is no family history of
hemophilia. It is possible that the time of diagnosis may be the most stressful event that a family
has yet to encounter. However, given that the minimum age for participation in this study was 7
years old, most of the participants in the study had been diagnosed for several years prior to
participation in the study. It may be that parents and participants alike had habituated to the
illness and thus the illness was seen to have relatively little impact on the family in the four
weeks prior to the assessment. Furthermore, this study was a subset of a larger study and one of
the criteria for inclusion was that participants could not have had an acute bleed within two
weeks of the clinical assessment. Thus we were studying participants when they may have been
at their best and consequently the illness would have had minimal impact on the family in the
previous four weeks.
111
5.5.1 Tertiary Hypothesis
When it came to the third hypothesis that impact on the family and social participation were
associated, the data mostly did not support this hypothesis and thus the null hypothesis that there
was no association cannot be rejected. Only modest correlations were found between social
participation and Social Functioning and Daily Activities. These correlations show that
participants who reported better social participation tended to have parents who felt they had
better functioning in social and daily activities. If parents felt that they had a good social
network and did not feel greatly distressed by daily activities, it stands to reason that they would
encourage their children to expand their social activities as well.
5.6 Exploratory Analyses
Recognizing that the ICF model suggests that there is a bidirectional relationship between
personal factors and social participation, linear modeling was used to explore the possibility that
personal and environmental factors may exert influence on one’s social participation.
5.6.1 Self-Perception
The linear model strengthens the finding of moderate associations between social participation
and self-perception, as 40.2% of the variance in social participation was explained by self-
perception in this model. Specifically, it was the competence of Global Self-Worth that exerted
the most effect. Given this data, the conclusion was that if a participant exhibited more adequate
Global Self-Worth, he would then perceive fewer social participation restrictions. This finding
was supported by the work of Katz (1970, as cited in Agle, 1975), who noted that if a patient had
an image of being disabled and inadequate, he may have had an inability to assume social and
vocational responsibilities as an adult.
5.6.2 Social Support
When it came to social support, the linear model found that 50.6% of the variance in social
participation could be explained by social support. The domains of parental and classmate
support/regard had the strongest effects on this model. This finding lead to the conclusion that if
a boy with hemophilia had good support from parents and classmates, he would be more likely to
engage in social participation. Other studies have found that with adequate social support, levels
of depression are lower (Iannone et al., 2012) and quality of life increases (Bullinger, Von
112
Mackensen, Haemo-QoL Group, 2003), which may lead to reports of increased social
participation.
5.7 Study Limitations
Despite random sampling efforts to increase the representativeness of the sample, this study was
a voluntary study and therefore may have been susceptible to sampling bias because participants
who agreed to participate in the study may have varied in a systematic way from those who
refused to participate. However, there was a very low rate of refusal and the random sampling of
participants increased the generalizability of this study.
There may have been some element of social desirability in the participants’ responses, as the
individuals may not have wanted to report severe limitations to their healthcare team or in front
of their family members. Participants may have wanted to appear self-assured and equivalent to
their peers and thus may have reported higher scores than they truly felt. To try and account for
this, all participants were assured that their responses would be strictly confidential and
participants were encouraged to complete self-report questionnaires on their own.
Due to the cross-sectional nature of this design, the data from this study were not sufficient to
make causal inferences. However, the benefits of a cross-sectional design far outweighed the
limitations. Strengths included the need for fewer resources, no loss to follow-up or migration
bias, and minimized participant burden, as there was no follow-up period in which participants
had to be monitored.
As temporality cannot be determined when using a cross-sectional design, reverse causation bias
was possible and thus the linear models may not be valid because it may have been social
participation that exerted an effect on self-perception and not the other way around. It is
important to bear in mind that in this exploratory study, we were seeking to understand if there
was a possible relationship between self-perception, perceived social support, and social
participation, rather than trying to determine the causes of impaired social participation. We
cannot therefore deduce underlying mechanisms as it may be that social participation difficulties
lead to lower self-perception and perceived social support, or vice versa. That being said, when
exploring possible models to determine what variables may impact social participation, we found
113
that both self-perception and social support fit the model nicely and thus have been included in
this analysis.
This study was limited to English-speaking patients and this may impact the generalizability of
the results. Given that Toronto is a multi-cultural city, it would have been ideal if we could have
included participants who were not English-speaking. It is unknown if patients who do not speak
English in Toronto have different levels of social participation than their English-speaking peers,
as the questionnaires in this study were unable to assess these patients. However, this study
strived to include a representative sample of the entire hemophilia population and only four
patients were excluded from the research study due to language barriers.
The sample had a relatively high-income level and this may have skewed the results in a positive
direction, as households with higher income may have been able to afford more opportunities to
participate socially (e.g. enroll children in activities outside of school such as organized sports or
clubs). Again, given the random sampling efforts and the low rate of refusal, the generalizability
of this study to other similar settings is still good.
The small sample size increased the impact of influential observations, however the results were
essentially unchanged when influential observations were removed and the data were re-
analyzed.
Since self-perception was measured differently for children and adolescents, the sample was split
into two groups for some analyses. Although there were no significant differences found
between the two groups, the sample sizes of the split groups may have been too small to detect
any differences.
Single imputation was used to handle missing data in this study, as this was the technique
advised by all the authors of the outcome measures. More sophisticated techniques for handling
missing data such as multiple imputation may have accounted for the uncertainty of the
imputations (Stuart, 2010). However, since data was missing from just eight variables, with a
maximum of 2 participants (4.0%) missing values for any one variable, single imputation was
deemed appropriate for the analysis. Missing values appeared to be questions skipped
accidentally by participants when filling out questionnaires, and not systematically unanswered.
114
Another limitation this study faced was that the Participation Scale for Kids (i.e. the
questionnaire modification used in this study) was untested. However tests for internal
consistency of the modified measure met the minimum cut-offs recommended by Nunnally
(1978, as cited in Streiner & Norman, 2008a). Current participation scales are often targeted at
adults and those scales that did exist for children did not adequately measure social participation
in our opinion. The original Participation Scale upon which modifications were made had the
advantage of including an option for participants to select if the item was neither relevant or of
interest to him. The Participation Scale addresses the personal and subjective nature of quality of
life (Feldman, 2013) and this feature was the primary reason why this scale was selected for
inclusion in this study.
By changing some of the wording of the items on the original Participation Scale, the validity
and reliability of the scale may have been reduced. It is possible that younger children did not
fully understand the scale or the concept of peer comparison. However, modifications were
made in consultation with the author of the original scale and it was felt that the amendments
were relatively minor and that the Participation Scale for Kids still maintained face validity.
Future studies may wish to re-validate this modified scale to test the new items and determine the
scale’s psychometric properties.
Finally, the questionnaires used to measure social participation may not have adequately
assessed how much of an impact hemophilia can have on an individual. Although hemophilia
may be more of an inconvenience than a disability on a daily basis, there may be times when
hemophilia is severely limiting. For example, if a patient suffers a joint bleed he may experience
restricted mobility and the inability to engage like his peers. How can one fully embrace a
disease that is unpredictable, uncomfortable, or limiting? These scenarios do not appear to be
the norm for our participants, but they can occur and may not be fully captured by our
questionnaires. Patients with hemophilia today can live a relatively normal life, but
intermittently the disease can cause pain associated with bleeds and the burden of illness
becomes apparent.
5.8 Future Directions
With the knowledge gained from this study, future works on social participation can be
contemplated. Social participation research in the hemophilia population is very limited and is
115
often focused on adults or specifically on sport participation. It was recently noted that for
patients and family members, participation is increasingly becoming the most salient outcome
(Blanchette, O'Mahony, McJames, & Mahlangu, 2014). This study was the first step in
understanding social participation in children with hemophilia outside the realm of sport.
The next step will be to develop a children-specific scale to more accurately measure the
construct of social participation. Psychometric testing and standardization of the Participation
Scale for Kids would allow researchers to develop a better tool for assessing social participation.
Other studies may choose to explore the construct of social participation at a cross-cultural level.
The Participation Scale for Kids could be administered to samples of patients in different
countries to see what social participation is like in other parts of the world. Our sample had few
patients with arthropathy or co-infection, but this is not necessarily the case for hemophilia
patients around the world. Indeed, the Participation Scale for Kids is currently being
administered at some centers in São Paulo, Brazil, where prophylaxis has only recently been
used more readily.
Future works may also include administering the Participation Scale for Kids to children with
and without chronic illnesses to compare patients with hemophilia to others. Such a study would
inform efforts to understand what other contextual factors are related to social participation and
if children with other chronic illnesses have levels of social participation on par with their peers.
Longitudinal studies that assess temporality may be able to isolate factors that directly contribute
to social participation. With the knowledge gained from such studies, interventional studies may
be developed to attempt to effect change on social participation.
5.9 Conclusions
In conclusion, the data show that boys with hemophilia have minimal social participation
restrictions, adequate self-perception, good social support, and their illness has minimal impact
on the family. This study is the first, to our knowledge, that has examined social participation
and its relationship to self-perception, social support, and the impact of the disease on the family
in boys with hemophilia.
116
In keeping with the ICF model, the data illustrate that the interaction between personal and
environmental factors as well as the health condition of hemophilia are associated with one’s
social participation. The data show that self-perception and social support were moderately to
strongly correlated with social participation. This study shows that participants who have higher
levels of self-perception and social support will also report higher levels of social participation.
Although we do not know the directionality of this relationship, it would seem appropriate to
encourage patients to take part in activities to increase self-perception, strengthen their social
support network, and increase their social participation. Doing so may result in an increase in
one or more of these domains, all of which would increase the patient’s health-related quality of
life.
Psychologists have been identified as crucial members in the comprehensive care team, as they
can assist patients with the self-management of their disease (Coppola et al., 2011) and
consequently foster the development of the patient’s independence. Monitoring for depression in
patients is important because health outcomes can be compromised when patients are depressed,
leading to increased risk taking behaviors or poor treatment compliance (Iannone et al., 2012).
Incorporating routine screening into usual care for hemophilia may allow comprehensive care
teams to easily identify patients at higher risks for low self-worth. Indeed, St. Michael’s
Hospital in Toronto, Canada, has rolled out such an intake assessment in its hemophilia clinic
and depression screening has been tested in Tucson, United States (Iannone et al., 2012).
Identifying patients who struggle with depression or low self-worth will enable healthcare
providers to offer appropriate support to the patients.
117
References
2011 National Household Survey. (2013). 2011 National Household Survey (pp. 1–10). Toronto. Retrieved from http://www1.toronto.ca/city_of_toronto/social_development_finance__administration/files/pdf/nhs_backgrounder.pdf
Agle, D. (1975). Psychological Factors In Hemophilia? The Concept Of Self-Care. Annals of the New York Academy of Sciences, 240(1 Recent advance), 221–225. doi:10.1111/j.1749-6632.1975.tb53351.x
Agle, D. P. (1984). Hemophilia - Psychological Factors and Comprehensive Management. Scandinavian Journal of Haematology, 33(S40), 55–63. doi:10.1111/j.1600-0609.1984.tb02545.x
Bagheri, S., Beheshtipoor, N., Rambod, M., Karimi, M., Zare, N., & Hashemil, F. (2012). The Quality of Life of Children with Hemophilia in Shiraz, Iran. The International Journal of Community Based Nursing and Midwifery, 1(2), 110-120.
Barrera, M., Fleming, C. F., & Khan, F. S. (2003). The role of emotional social support in the psychological adjustment of siblings of children with cancer. Child: Care, Health, and Development.
Beeton, K., Neal, D., Watson, T., & Lee, C. A. (2007). Parents of children with haemophilia--a transforming experience. Haemophilia, 13(5), 570–579. doi:10.1111/j.1365-2516.2007.01494.x
Binnema, M., Schrijvers, L. H., Bos, R., Schuurmans, M. J., & Fischer, K. (2014). Coping in adult patients with severe haemophilia. Haemophilia. doi:10.1111/hae.12366
Blanchette, V. S., O'Mahony, B., McJames, L., & Mahlangu, J. N. (2014). Assessment of outcomes. Haemophilia, 20 Suppl 4, 114–120. doi:10.1111/hae.12426
Blanchette, V. S., Sparling, C., & Turner, C. (1991). Inherited bleeding disorders. Baillière's Clinical Haematology, 4(2), 291–332.
Breakey, V. R., Young, N. L., Burke, T., Wakefield, C., Hilliard, P., Stain, A.-M., et al. (2012, July 10). Understanding the Impact of Hemophilia on the Family: A Pilot Study of the PedsQL-Family Impact Module. WFH 2012. Paris. doi:10.3252/pso.eu.WFH2012.2012
Broderick, C. R., Herbert, R. D., Latimer, J., & van Doorn, N. (2013). Patterns of physical activity in children with haemophilia. Haemophilia, 19(1), 59–64. doi:10.1111/j.1365-2516.2012.02904.x
Browne, W. J., Mally, M. A., & Kane, R. P. (1960). Psychosocial Aspects Of Hemophilia: A Study Of Twenty‐Eight Hemophilic Children And Their Families. American Journal of Orthopsychiatry, 30(4), 730–740. doi:10.1111/j.1939-0025.1960.tb02089.x
Bullinger, M., Von Mackensen, S., Haemo-QoL Group. (2003). Quality of life in children and families with bleeding disorders. Journal of Pediatric Hematology/Oncology, 25 Suppl 1, S64–7.
Buzzard, B. M. (1996). Sports and hemophilia: antagonist or protagonist. Clinical Orthopaedics and Related Research, (328), 25–30.
Cadman, D., Rosenbaum, P., Boyle, M., & Offord, D. R. (1991). Children With Chronic Illness: Family and Parent Demographic Characteristics and Psychosocial Adjustment. Pediatrics, 87(6), 884–889.
Calman, K. C. (1984). Quality of life in cancer patients--an hypothesis. Journal of Medical Ethics, 10, 124–127.
118
Carcao, M. D. (2012). The diagnosis and management of congenital hemophilia. Seminars in Thrombosis and Hemostasis, 38(7), 727–734. doi:10.1055/s-0032-1326786
Coppola, A., Cerbone, A. M., Mancuso, G., Mansueto, M. F., Mazzini, C., & Zanon, E. (2011). Confronting the psychological burden of haemophilia. Haemophilia, 17(1), 21–27. doi:10.1111/j.1365-2516.2010.02280.x
Crocker, P. R. E., Sabiston, C. M., Kowalski, K. C., McDonough, M. H., & Kowalski, N. (2006). Longitudinal Assessment of the Relationship Between Physical Self-Concept and Health-Related Behavior and Emotion in Adolescent Girls. Journal of Applied Sport Psychology, 18(3), 185–200. doi:10.1080/10413200600830257
Crockett, L. J., Moilanen, K. L., Raffaelli, M., & Randall, B. A. (2006). Psychological profiles and adolescent adjustment: A person-centered approach. Development and Psychopathology, 18(01). doi:10.1017/S0954579406060111
De Kleijn, P., Heijnen, L., & Van Meeteren, N. (2002). Clinimetric instruments to assess functional health status in patients with haemophilia: a literature review. Haemophilia, 8(3), 419–427.
De Veaux, R. D., Velleman, P. F., Bock, D. E., Vukov, A. M., & Wong, A. (2012). STATS: Data and Models (Canadian). Toronto, ON: Pearson Canada.
DeKoven, M., Karkare, S., Kelley, L. A., Cooper, D. L., Pham, H., Powers, J., et al. (2014). Understanding the experience of caring for children with haemophilia: cross-sectional study of caregivers in the United States. Haemophilia, 20(4), 541–549. doi:10.1111/hae.12379
Drotar, D., Eckl, C. L., Beitzel, M., Gil, E., Kocik, S., Kuekes, K., et al. (1997). Psychosocial Services for Children and Adolescents With Hemophilia and Their Families: Result of a National Survey. Children's Health Care, 26(3), 137–150. doi:10.1207/s15326888chc2603_1
Elander, J., Robinson, G., & Mitchell, K. (2009). An assessment of the relative influence of pain coping, negative thoughts about pain, and pain acceptance on health-related quality of life among people with hemophilia. Pain, 145, 169–175.
Emiliani, F., Bertocchi, S., Potì, S., & Palareti, L. (2011). Process of normalization in families with children affected by hemophilia. Qualitative Health Research, 21(12), 1667–1678. doi:10.1177/1049732311417456
Feldman, B. M. (2013). Issues in the measurement of quality of life in hemophilia. Revista Brasileira De Hematologia E Hemoterapia, 35(5). doi:10.5581/1516-8484.20130118
Feldman, B. M., Funk, S., Lundin, B., Doria, A. S., Ljung, R., & Blanchette, V. (2008). Musculoskeletal measurement tools from the International Prophylaxis Study Group (IPSG). Haemophilia, 14(s3), 162–169. doi:10.1111/j.1365-2516.2008.01750.x
Feldman, B. M., Pai, M., Rivard, G. E., & Israels, S. (2006). Tailored prophylaxis in severe hemophilia A: interim results from the first 5 years of the Canadian Hemophilia Primary Prophylaxis Study. Journal of Thrombosis and Haemostasis, 4(6), 1228–36.
Fischer, K., van der Bom, J. G., & van den Berg, H. M. (2003). Health-related quality of life as outcome parameter in haemophilia treatment. Haemophilia, 9 (Suppl 1), 75–81– discussion 82.
Franchini, M., Frattini, F., Crestani, S., & Bonfanti, C. (2012). Haemophilia B: current pharmacotherapy and future directions. Expert Opinion on Pharmacotherapy, 13(14), 2053–2063. doi:10.1517/14656566.2012.721780
Fromme, A., Dreeskamp, K., Pollmann, H., Thorwesten, L., Mooren, F. C., & Völker, K. (2007). Participation in sports and physical activity of haemophilia patients. Haemophilia, 13(3), 323–327. doi:10.1111/j.1365-2516.2007.01456.x
119
Giannelli, F. (1989). Factor IX. Baillière's Clinical Haematology, 2(4), 821–848. Gill, T. M., & Feinstein, A. R. (1994). A critical appraisal of the quality of quality-of-life
measurements. JAMA-Journal of the American Medical Association-US Edition, 272(8), 619–626.
Globe, D., Young, N. L., Von Mackensen, S., Bullinger, M., Wasserman, J., & Group, On Behalf Of The Healthrelated Quality Of Life Expert Working Group Of The International Prophylaxis Study Group. (2009). Measuring patient-reported outcomes in haemophilia clinical research. Haemophilia, 15(4), 843–852. doi:10.1111/j.1365-2516.2008.01961.x
Gringeri, A., & Von Mackensen, S. (2008). Quality of life in haemophilia. Haemophilia, 14(s3), 19–25. doi:10.1111/j.1365-2516.2008.01709.x
Gringeri, A., Von Mackensen, S., Auerswald, G., Bullinger, M., Garrido, R. P., Kellermann, E., et al. (2004). Health status and health-related quality of life of children with haemophilia from six West European countries. Haemophilia, 10 (Suppl 1), 26–33.
The WHOQOL Group. (1995). The World Health Organization quality of life assessment (WHOQOL): position paper from the World Health Organization. Social Science & Medicine (1982), 41(10), 1403–1409.
Harter, S. (1982). The perceived competence scale for children. Child Development, 53, 87–97. Harter, S. (1985a). Manual for the Self-Perception Profile for Children: Revision of the
Perceived Competence Scale for Children. Denver: University of Denver. Harter, S. (1985b). Manual for the Social Support Scale for Children. Denver: University of
Denver. Harter, S. (1988). Manual for the Self-Perception Profile for Adolescents. Denver: University of
Denver. Harter, S. (2012a). Self-Perception Profile for Adolescents: Manual and Questionnaires (pp. 1–
50). Denver: University of Denver. Harter, S. (2012b). Self-Perception Profile for Children: Manual and Questionnaires (Grades 3-
8) (pp. 1–47). Denver: University of Denver. Hartl, H. K., Reitter, S., Eidher, U., Ramschak, H., Ay, C., & Pabinger, I. (2008). The impact of
severe haemophilia on the social status and quality of life among Austrian haemophiliacs. Haemophilia, 14(4), 703–708. doi:10.1111/j.1365-2516.2008.01684.x
Hassan, T. H., Badr, M. A., Fattah, N. R. A., & Badawy, S. M. (2011). Assessment of musculoskeletal function and mood in haemophilia A adolescents: a cross‐sectional study. Haemophilia, 17(4), 683–688. doi:10.1111/j.1365-2516.2010.02475.x
Hegeman, A. K., Van Genderen, F. R., Meijer, S., Van Den Briel, M. M., Tamminga, R. Y. J., & Van Weert, E. (2011). Perceived competence in children and adolescents with haemophilia: an explorative study. Haemophilia, 17(1), 81–89. doi:10.1111/j.1365-2516.2010.02357.x
Heijnen, L., Mauser-Bunschoten, E. P., & Roosendaal, G. (2000). Participation in sports by Dutch persons with haemophilia. Haemophilia, 6(5), 537–546. doi:10.1046/j.1365-2516.2000.00422.x
Hemphill, J. F. (2003). Interpreting the magnitudes of correlation coefficients. American Psychologist, 58(1), 78–79. doi:10.1037/0003-066X.58.1.78
Herrick, E. K., Nussbaum, R., Holtzman, N. A., & Wissow, L. (2004). Asking fathers: a study of psychosocial adaptation. Haemophilia, 10(5), 582–589. doi:10.1111/j.1365-2516.2004.00958.x
Hilberg, T., Czepa, D., Freialdenhoven, D., & Boettger, M. K. (2011). Joint pain in people with hemophilia depends on joint status. Pain, 152(9), 2029–2035. doi:10.1016/j.pain.2011.04.030
120
Iannone, M., Pennick, L., Tom, A., Cui, H., Gilbert, M., Weihs, K., & Stopeck, A. T. (2012). Prevalence of depression in adults with haemophilia. Haemophilia, 18(6), 868–874. doi:10.1111/j.1365-2516.2012.02863.x
Kellerman, J., Zeltzer, L., Ellenberg, L., Dash, J., & Rigler, D. (1980). Psychological effects of illness in adolescence. I. Anxiety, self-esteem, and perception of control. The Journal of Pediatrics, 97(1), 126–131. doi:10.1016/S0022-3476(80)80152-1
Kern, M., Blanchette, V., Stain, A. M., Einarson, T. R., & Feldman, B. M. (2004). Clinical and cost implications of target joints in Canadian boys with severe hemophilia A. The Journal of Pediatrics, 145(5), 628–634. doi:10.1016/j.jpeds.2004.06.082
Khair, K., Holland, M., & Carrington, S. (2011). Social networking for adolescents with severe haemophilia. Haemophilia, 18(3), e290-6. doi:10.1111/j.1365-2516.2011.02689.x
Kim, S.-Y., Kim, S.-W., Kim, J.-M., Shin, I.-S., Baek, H.-J., Lee, H.-S., et al. (2013). Impact of personality and depression on quality of life in patients with severe haemophilia in Korea. Haemophilia, 19(5), e270–5. doi:10.1111/hae.12221
Klein, R. H., & Nimorwicz, P. (1982). Psychosocial aspects of hemophilia in families: 1. Assessment strategies and instruments. Clinical Psychology Review, 2(2), 153–169.
Köiter, J., Van Genderen, F. R., Brons, P. P. T., & Nijhuis-van der Sanden, M. W. G. (2009). Participation and risk-taking behaviour in sports in children with haemophilia. Haemophilia, 15(3), 686–694. doi:10.1111/j.1365-2516.2009.02006.x
Kvist, B., Kvist, M., & Rajantie, J. (1990). School Absences, School Achievements and Personality Traits of the Haemophilic Child. Scandinavian Journal of Public Health, 18(2), 125–132. doi:10.1177/140349489001800206
Lindvall, K., Colstrup, L., Wollter, I.-M., Klemenz, G., Loogna, K., Grönhaug, S., & Thykjaer, H. (2006). Compliance with treatment and understanding of own disease in patients with severe and moderate haemophilia. Haemophilia, 12(1), 47–51. doi:10.1111/j.1365-2516.2006.01192.x
Lindvall, K., Von Mackensen, S., Elmståhl, S., Khair, K., Stain, A.-M., Ljung, R., & Berntorp E. (2014). Increased burden on caregivers of having a child with haemophilia complicated by inhibitors. Pediatric Blood & Cancer, 61(4), 706-11. doi: 10.1002/pbc.24856
Von Mackensen, S. (2007). Quality of life and sports activities in patients with haemophilia. Haemophilia, 13(s2), 38–43. doi:10.1111/j.1365-2516.2007.01505.x
Manning, L. (1988). Erikson's psychosocial theories help explain early adolescence. NASSP Bulletin, 72(509), 95–100.
McCrum-Gardner, E. (2008). Which is the correct statistical test to use? British Journal of Oral and Maxillofacial Surgery, 46(1), 38–41. doi:10.1016/j.bjoms.2007.09.002
McGee, R., Williams, S., Howden-Chapman, P., Martin, J., & Kawachi, I. (2006). Participation in clubs and groups from childhood to adolescence and its effects on attachment and self-esteem. Journal of Adolescence, 29(1), 1–17. doi:10.1016/j.adolescence.2005.01.003
Miners, A. H., Sabin, C. A., Tolley, K. H., Jenkinson, C., Kind, P., & Lee, C. A. (1999). Assessing health-related quality-of-life in individuals with haemophilia. Haemophilia, 5(6), 378–385. doi:10.1046/j.1365-2516.1999.00347.x
Mulder, K., Cassis, F., Seuser, D., & Narayan, P. (2004). Risks and benefits of sports and fitness activities for people with haemophilia. Haemophilia, 10(Suppl. 4), 161–163.
Naraine, V. S., Risebrough, N. A., Oh, P., Blanchette, V. S., Lee, S., Stain, A.-M., et al. (2002). Health-related quality-of-life treatments for severe haemophilia: utility measurements using the Standard Gamble technique. Haemophilia, 8(2), 112–120.
Nathwani, A. C., Tuddenham, E. G. D., Rangarajan, S., Rosales, C., McIntosh, J., Linch, D. C.,
121
et al. (2011). Adenovirus-associated virus vector–mediated gene transfer in hemophilia B. New England Journal of Medicine, 365(25), 2357–2365. doi:10.1056/NEJMoa1108046
Noll, R. B., LeRoy, S., Bukowski, W. M., Rogosch, F. A., & Kulkarni, R. (1991). Peer Relationships and Adjustment in Children with Cancer. Journal of Pediatric Psychology, 16(3), 307–326. doi:10.1093/jpepsy/16.3.307
Norman, G. R., & Streiner, D. L. (2003). PrettyDarnedQuick (PDQ): Statistics (3rd ed.). Shelton: People's Medical Publishing House.
Petrini, P., & Seuser, A. (2009). Haemophilia care in adolescents–compliance and lifestyle issues. Haemophilia, 15, 15–19.
Plug, I. (2004). Thirty years of hemophilia treatment in the Netherlands, 1972-2001. Blood, 104(12), 3494–3500. doi:10.1182/blood-2004-05-2008
Plug, I., Peters, M., Mauser-Bunschoten, E. P., de Goede-Bolder, A., Heijnen, L., Smit, C., et al. (2008). Social participation of patients with hemophilia in the Netherlands. Blood, 111(4), 1811–1815. doi:10.1182/blood-2007-07-102202
Rolland, J. S. (2012). Mastering family challenges in serious illness and disability. In F. Walsh, Normal Family Processes (4 ed., pp. 452–482). New York: Guilford Press.
Rosendaal, F. R., Smit, C., Varekamp, I., Bröckervriends, A. H. J. T., Dijck, H., Suurmeijer, T. P. B. M., et al. (1990). Modern haemophilia treatment: medical improvements and quality of life. Journal of Internal Medicine, 228(6), 633–640. doi:10.1111/j.1365-2796.1990.tb00291.x
Saviolo-Negrin, N., Cristante, F., Zanon, E., Canclini, M., Stocco, D., & Girolami, A. (1999). Psychological aspects and coping of parents with a haemophilic child: a quantitative approach. Haemophilia, 5(1), 63–68. doi:10.1046/j.1365-2516.1999.00213.x
Shapiro, A. D., Donfield, S. M., Lynn, H. S., Cool, V. A., Stehbens, J. A., Hunsberger, S. L., et al. (2001). Defining the Impact of Hemophilia: The Academic Achievement in Children With Hemophilia Study. Pediatrics, 108(6), e105–e105. doi:10.1542/peds.108.6.e105
Shaw, D., & Riley, G. A. (2008). The impact on parents of developments in the care of children with bleeding disorders. Haemophilia, 14(1), 65–67. doi:10.1111/j.1365-2516.2007.01598.x
Sherlock, E., O’Donnell, J. S., White, B., & Blake, C. (2009). Physical activity levels and participation in sport in Irish people with haemophilia. Haemophilia, 16(1), e202–e209. doi:10.1111/j.1365-2516.2009.02111.x
Slutzky, C. B., & Simpkins, S. D. (2009). The link between children's sport participation and self-esteem: Exploring the mediating role of sport self-concept. Psychology of Sport & Exercise, 10(3), 381–389. doi:10.1016/j.psychsport.2008.09.006
Spirito, A., DeLawyer, D. D., & Stark, L. J. (1990). Peer relations and social adjustment of chronically ill children and adolescents. Clinical Psychology Review.
Stevelink, S. A. M., Terwee, C. B., Banstola, N., & Van Brakel, W. H. (2012). Testing the psychometric properties of the Participation Scale in Eastern Nepal. Quality of Life Research, 22(1), 137–144. doi:10.1007/s11136-012-0116-8
Streiner, D. L., & Norman, G. R. (2008a). Introduction. In Health Measurement Scales: A Practical Guide to Their Development and Use (p. 431). Oxford University Press. doi:10.1093/acprof:oso/9780199231881.001.0001
Streiner, D. L., & Norman, G. R. (2008b). Selecting the items. In Health Measurement Scales: A practical guide to their development and use. Oxford University Press. doi:10.1093/acprof:oso/9780199231881.001.0001
Stuart, E. A. (2010, June 25). Recent Advances in Missing Data Methods: Multiple Imputation by Chained Equations. AcademyHealth Annual Research Meeting. Boston, MA.
122
Szende, A., Schramm, W., Flood, E., Larson, P., Gorina, E., Rentz, A. M., & Snyder, L. (2003). Health-related quality of life assessment in adult haemophilia patients: a systematic review and evaluation of instruments. Haemophilia, 9(6), 678–687.
Tagariello, G., Iorio, A., Santagostino, E., Morfini, M., Bisson, R., Innocenti, M., et al. (2009). Comparison of the rates of joint arthroplasty in patients with severe factor VIII and IX deficiency: an index of different clinical severity of the 2 coagulation disorders. Blood, 114(4), 779–784. doi:10.1182/blood-2009-01-195313
Tencer, T., Friedman, H. S., & Li-McLeod, J. (2007). Medical costs and resource utilization for hemophilia patients with and without HIV or HCV infection. Journal of Managed Care Pharmacy, 13(9), 790–798.
Thomas, D., & Gaslin, T. C. (2001). “Camping Up” Self-Esteem in Children with Hemophilia. Issues in Comprehensive Pediatric Nursing, 24(4), 253–263. doi:10.1080/014608601753260344
Trzepacz, A. M., Vannatta, K., Davies, W. H., Stehbens, J. A., & Noll, R. B. (2003). Social, emotional, and behavioral functioning of children with hemophilia. Journal of Developmental and Behavioral Pediatrics: JDBP, 24(4), 225–232.
Tuddenham, E. (1989). Factor VIII and haemophilia. Baillière's Clinical Haematology, 2(4), 849–877.
van Brakel, W. H., Anderson, A. M., Mutatkar, R. K., Bakirtzief, Z., Nicholls, P. G., Raju, M. S., & Das-Pattanayak, R. K. (2006). The Participation Scale: measuring a key concept in public health. Disability & Rehabilitation, 28(4), 193–203. doi:10.1080/09638280500192785
van der Net, J., Vos, R. C., Engelbert, R. H. H., van den Berg, M. H., Helders, P. J. M., & Takken, T. (2006). Physical fitness, functional ability and quality of life in children with severe haemophilia: a pilot study. Haemophilia, 12(5), 494–499. doi:10.1111/j.1365-2516.2006.01307.x
van Dijk, K., Fischer, K., van der Bom, J. G., Grobbee, D. E., & van den Berg, H. M. (2005). Variability in clinical phenotype of severe haemophilia: the role of the first joint bleed. Haemophilia, 11(5), 438–443. doi:10.1111/j.1365-2516.2005.01124.x
Van Genderen, F. R., van Meeteren, N. L. U., Fischer, K., & Helders, P. J. M. (2012). Determinants of participation in patients with severe haemophilia. Haemophilia, 18(6), 860–867. doi:10.1111/j.1365-2516.2012.02884.x
Varekamp, I., Suurmeijer, P., Rosendaal, F. R., van Duck, H., Bröcker-Vriends, A., & Briët, E. (1990). Family burden in families with a hemophilic child. Family Systems Medicine, 8(3), 291.
Varni, J. W., Katz, E. R., Colegrove, R., & Dolgin, M. (1994). Perceived social support and adjustment of children with newly diagnosed cancer. Journal of Developmental and Behavioral Pediatrics : JDBP, 15(1), 20–26.
Varni, J. W., Limbers, C. A., & Burwinkle, T. M. (2007). Parent proxy-report of their children“s health-related quality of life: an analysis of 13,878 parents” reliability and validity across age subgroups using the PedsQL™ 4.0 Generic Core Scales. Health and Quality of Life Outcomes, 5(1), 2. doi:10.1186/1477-7525-5-2
Varni, J. W., Sherman, S. A., Burwinkle, T. M., Dickinson, P. E., & Dixon, P. (2004). The PedsQL™ family impact module: preliminary reliability and validity. Health and Quality of Life Outcomes, 2(1), 55. doi:10.1186/1477-7525-2-55
Victor, J. C. (2012, April 25). Applied Linear Regression Analysis. Biostatistics II: Advanced Methods in Applied Regression Analysis. Toronto, Ontario.
Vogt, W. P., & Johnson, B. (2011). Dictionary of Statistics & Methodology: A Nontechnical
123
Guide for the Social Sciences. SAGE Publications. White, G. C., Rosendaal, F., Aledort, L. M., Lusher, J. M., Rothschild, C., Ingerslev, J., Factor
VIII and Factor IX Subcommittee. (2001). Definitions in hemophilia. Recommendation of the scientific subcommittee on factor VIII and factor IX of the scientific and standardization committee of the International Society on Thrombosis and Haemostasis. Thrombosis and haemostasis, 85(3), 560.
Wiedebusch, S., Pollmann, H., Siegmund, B., & Muthny, F. A. (2008). Quality of life, psychosocial strains and coping in parents of children with haemophilia. Haemophilia, 14(5), 1014–1022. doi:10.1111/j.1365-2516.2008.01803.x
Williams, K. A., & Chapman, M. V. (2011). Social challenges for children with hemophilia: child and parent perspectives. Social Work in Health Care, 50(3), 199–214. doi:10.1080/00981389.2010.527790
World Health Organization. (1997). WHOQOL: measuring quality of life. World Health Organization.
World Health Organization. (2001). International classification of functioning, disability and health: ICF. World Health Organization.
World Health Organization. (2014). Adolescent development. World Health Organization. Retrieved August 3, 2014, from http://www.who.int/maternal_child_adolescent/topics/adolescence/dev/en/
Young, N. L., Bradley, C. S., Wakefield, C. D., Barnard, D., Blanchette, V. S., & McCusker, P. J. (2006). How well does the Canadian haemophilia Outcomes-Kids' Life Assessment Tool (CHO-KLAT) measure the quality of life of boys with haemophilia? Pediatric Blood & Cancer, 47(3), 305–311. doi:10.1002/pbc.20618
Zou, K. H., Tuncali, K., & Silverman, S. G. (2003). Correlation and Simple Linear Regression. Radiology, 227(3), 617–628. doi:10.1148/radiol.2273011499
124
Appendices
125
Appendix A: Research Ethics Board Approval Letters
Copy of Approval by The Hospital for Sick Children
126
127
128
Copy of Approval by the University of Toronto
129
Appendix B: Data Collection Instruments
Copy of Participation Scale for Kids (modified)
130
131
132
Copy of Self-Perception Profile for Children
133
134
Copy of Self-Perception Profile for Adolescents
135
136
Copy of Social Support Scale for Children
137
138
Copy of PedsQL™ Family Impact Module
139
140
141
Appendix C: Family Situation Family Situation (N = 50) Variable n (%) Marital Status
Married/Common-law Separated Unmarried Divorced
41 (82.0)
4 (8.0) 2 (4.0) 3 (6.0)
Primary Guardian Mother & Father Mother Father
37 (74.0) 12 (24.0)
1 (2.0) Living Area
Rural Urban
Household Income Low (less than CAD$30,000) Medium (CAD$30,000-$80,000) High (greater than CAD$80,000)
7 (14.0)
43 (86.0)
6 (12.2) 17 (34.7) 26 (53.1)
Appendix D: Treatment
Table D1 Treatment Modalities Variable n (%) Where treatment occurs
Home Clinic or Hemophilia Treatment Center Other
37 (74.0) 10 (20.0)
3 (6.0) Who infuses treatment
Mother Father Mother & Father Self-Infusion Clinic staff Mother & Self-Infusion Mother, Father & Self-Infusion
15 (31.3)
4 (8.3) 4 (8.3)
8 (16.7) 12 (25.0)
4 (8.3) 1 (2.1)
Treatment frequency within the last 6 months On-demand Once per week 2 times per week 3 times per week Every other day Every day Immune tolerance induction or bypassing agents
12 (24.0)
3 (6.0) 13 (26.0)
3 (6.0) 18 (36.0)
1 (2.0) 0 (0.0)
Table D2 describes the treatment differences between moderate and severe participants.
Participants with moderate severity were more likely to receive on-demand therapy than severe
participants. If moderate participants required prophylaxis, it was far more likely to be
secondary in nature. On the other hand, participants with severe phenotype were less likely to
receive on-demand therapy as the majority of participants were on primary or secondary
prophylaxis.
142
Table D2 Frequency of treatment over entire life by participant factor level (N = 50) Variable Moderate (n (%)) Severe (n (%)) On-demand
Appendix F: Item Responses for Social Participation Scale for Kids
See Appendix B for Social Participation Scale for Kids items.
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 1
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 2
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 3
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 4
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 5
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 6
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 7
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 8
144
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 9
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 10
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 11
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 12
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 13
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 14
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 15
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 16
145
Appendix G: Reference Scores from the Manual for the Self-Perception Profile for Children
(Harter, 1985a) Reference Scores for Self-Perception for Sixth Grade Boys Sample Social Acceptance (M (SD)) Athletic Competence (M (SD)) Global Self-Worth (M (SD)) A (n = 206) 3.1 (0.6) 3.2 (0.6) 3.2 (0.6)
Appendix H: Social Support Scale for Children Scores by Age Group Pearson Correlations between Social Participation and Social Support by Age Group Variable Children Adolescents
Parental Support/Regard -0.8*** -0.2 Teacher Support/Regard -0.7*** -0.4 Classmate Support/Regard -0.7*** -0.3 Close Friend Support/Regard -0.5** -0.4 Note. *Correlation is significant at p < 0.05. **Correlation is significant at p < 0.01. ***Correlation is significant at p < 0.001.
Appendix I: Reference Scores from the Manual for the Social Support Scale for Children
(Harter, 1985b) Reference Scores for Social Support for Sixth Grade Boys Sample Parent (M (SD)) Teacher (M (SD)) Classmate (M (SD)) Close Friend (M (SD)) Elementary (n = 35) 3.0 (0.8) 2.9 (0.7) 2.9 (0.7) 2.9 (0.6)
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 17
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 18
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 19
0 10 20 30 40 50
0 1 2 3 5 Freq
uenc
y
Score
Item 20
146
Appendix J: Comparison with Reference Scores Comparison with Reference Scores for the Social Support Scale for Children Subscale t df p Parental Support/Regard 4.0 53.1 0.0002 Teacher Support/Regard 3.0 77.1 0.004 Classmate Support/Regard 4.3 67.2 <0.0001 Close Friend Support/Regard 4.9 74.9 <0.0001
Appendix K: Normality of Residuals Normality of Residuals Variable Skewness Kurtosis Shapiro-Wilk Self-Perception Profile for Children/Adolescents Social Acceptance
Children & Adolescents (n = 50) Children (n = 30) Adolescents (n = 20)
1.0 0.6 0.9
0.9 0.2
-0.6
0.9*** 0.9** 0.8**
Athletic Competence Children & Adolescents (n = 50) Children (n = 30) Adolescents (n = 20)
1.9 1.4 0.8
3.9 1.5
-0.2
0.8*** 0.8***
0.9* Global Self-Worth
Children & Adolescents (n = 50) Children (n = 30) Adolescents (n = 20)
1.1 1.4 0.7
1.6 3.3
-0.8
0.9** 0.9** 0.9*
Social Support Scale for Children Parental Support/Regard (n = 50) 0.3 1.0 1.0 Teacher Support/Regard (n = 50) 0.7 0.4 0.9** Classmate Support/Regard (n = 50) 0.5 1.0 0.9* Close Friend Support/Regard (n = 50) 1.3 1.7 0.9*** PedsQL™ Family Impact Module (n = 50) Total Score 1.9 4.1 0.8*** Physical Functioning 1.8 3.8 0.8*** Emotional Functioning 1.8 4.0 0.8*** Social Functioning 1.9 4.1 0.8*** Cognitive Functioning 1.7 3.6 0.8*** Communication 1.7 3.4 0.8*** Worry 1.6 2.6 0.8*** Daily Activities 1.9 4.2 0.8*** Family Relationships 1.7 3.0 0.8*** Note. *Normality test is significant at p < 0.05. ** Normality test is significant at p < 0.01. *** Normality test is significant at p < 0.001.
147
Appendix L: Residual Plots
Residual plots for Social Participation and Self-Perception
Figure L1. Social Acceptance
Figure L2. Athletic Competence
148
Figure L3. Global Self-Worth
Residual plots for Social Participation and Social Support
Figure L4. Parental Support/Regard
149
Figure L5. Teacher Support/Regard
Figure L6. Classmate Support/Regard
150
Figure L7. Close Friend Support/Regard
Residual plots for Social Participation and Family Impact
Figure L8. Total Score
151
Figure L9. Physical Functioning
Figure L10. Emotional Functioning
152
Figure L11. Social Functioning
Figure L12. Cognitive Functioning
153
Figure L13. Communication
Figure L14. Worry
154
Figure L15. Daily Activities
Figure L16. Family Relationships
155
Appendix M: Non-Parametric Tests
Self-Perception
When the data was re-analyzed using Spearman’s rank-order correlation coefficients, a strong
correlation was found between social participation and Social Acceptance (rho = -0.6, p = <
0.001), when looking at the scores of both children and adolescents. Moderate correlations were
found between social participation and Athletic Competence (rho = -0.4, p = < 0.01) and Global
Self-Worth (rho = -0.4, p = < 0.01).
When examining Children alone, social participation was strongly associated with Social
Acceptance (rho = -0.6, p = < 0.001) and Global Self-Worth (rho = -0.6, p = < 0.001), and not
significantly correlated with Athletic Competence (rho = -0.3, p = > 0.05). For adolescents, a
strong correlation was found between social participation and Athletic Competence (rho = -0.5, p
= < 0.05), but no statistically significant relationship was found between social participation and
Social Acceptance (rho = -0.4, p = > 0.05) or Global Self-Worth (rho = -0.2, p = > 0.05). Both
the Spearman and Pearson correlations are listed in Table M1.
These relationships suggested that overall children and adolescents with more adequate levels of
self-perception tended to have fewer social participation restrictions.
Table M1 Correlations between Social Participation and Self-Perception Variable Spearman rho Pearson r Social Acceptance
Children & Adolescents Children Adolescents
-0.6*** -0.6***
-0.4
-0.5*** -0.5**
-0.3 Athletic Competence
Children & Adolescents Children Adolescents
-0.4**
-0.3 -0.5*
-0.2 -0.1
-0.6** Global Self-Worth
Children & Adolescents Children Adolescents
-0.4**
-0.6*** -0.2
-0.6*** -0.7***
-0.3 Note. *Correlation is significant at p < 0.05. **Correlation is significant at p < 0.01. ***Correlation is significant at p < 0.001.
Social Support
Spearman rank-order correlations continued to find correlations between social participation and
social support (see Table M2). There was a strong correlation between social participation and
support/regard from classmates (rho = -0.5, p = < 0.001), but only moderate correlations between
156
social participation and support/regard from parents (rho = -0.4, p = < 0.01), close friends (rho =
-0.4, p= < 0.01), and teachers (rho = -0.3, p = < 0.05).
These correlations show that children and adolescents who reported more social support from
their parents, classmates, teachers, and close friends tended to report fewer social participation
restrictions.
Table M2 Correlations between Social Participation and Social Support Variable Spearman rho Pearson r Parent -0.4** -0.6*** Teacher -0.3* -0.5*** Classmate -0.5*** -0.6*** Close Friend -0.4** -0.4** Note. *Correlation is significant at p < 0.05. **Correlation is significant at p < 0.01. ***Correlation is significant at p < 0.001.
Family Impact
Much like the Pearson correlations, few associations were observed between social participation
and family functioning. There continued to be a moderate correlation between social
participation and social functioning (rho = -0.3, p = < 0.05), and there was also a moderate
correlation between social participation and daily activities (rho = -0.3, p < 0.05). These
correlations indicated that parents/guardians who reported less impairment in social functioning
and daily activities would in turn have children who were less likely to report social participation
restrictions (see Table M3).
Table M3 Correlations between Social Participation and Family Impact Variable Spearman rho Pearson r Total Score -0.2 -0.2 Physical Functioning -0.2 -0.2 Emotional Functioning -0.3 -0.2 Social Functioning -0.3* -0.3* Cognitive Functioning -0.2 -0.2 Communication -0.1 -0.2 Worry -0.1 0.0 Daily Activities -0.3* -0.3 Family Relationships -0.2 -0.2 Note. *Correlation is significant at p < 0.05.
157
Appendix N: Sensitivity Analysis Minimum and Maximum Imputed Values for Missing Data Variable Minimum r Maximum r Self-Perception Profile for Children/Adolescents Social Acceptance
Children & Adolescents Children Adolescents
-0.5*** -0.6***
-0.3
-0.5*** -0.5**
-0.3 Athletic Competence
Children & Adolescents Children Adolescents
-0.2 -0.2
-0.6**
-0.2 -0.1
-0.6** Global Self-Worth
Children & Adolescents Children Adolescents
-0.6*** -0.7***
-0.3
-0.6*** -0.7***
-0.3 Social Support Scale for Children Parental Support/Regard -0.6*** -0.6*** Teacher Support/Regard -0.5*** -0.5*** Classmate Support/Regard -0.6*** -0.6*** Close Friend Support/Regard -0.4** -0.4** PedsQL™ Family Impact Module Total Score -0.2 -0.2 Physical Functioning -0.2 -0.2 Emotional Functioning -0.2 -0.2 Social Functioning -0.3 -0.3 Cognitive Functioning -0.2 -0.2 Communication -0.2 -0.2 Worry 0.0 -0.0 Daily Activities -0.3* -0.3 Family Relationships -0.2 -0.2 Note. *Correlation is significant at p < 0.05. **Correlation is significant at p < 0.01. ***Correlation is significant at p < 0.001.
158
Appendix O: Model Checking for Self-Perception
Figure O1. Histogram for Residuals
Figure O2. Q-Q Plot for Residuals
159
Appendix P: Model Checking for Social Support
Figure P1. Histogram for Residuals
Figure P2. Q-Q Plot for Residuals
160
Appendix Q: Linear Modeling
Table Q1. Parameter Estimates for Linear Regression Parameter Estimates with Age and Factor Level as Predictor Variables Social Participation Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 2.8 (2, 47) 0.1 Age -0.7 0.3 -2.4 0.0 Factor Level -0.3 1.8 -0.2 0.9 Self-Perception: Social Acceptance Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 2.4 (2, 47) 0.1 Age 0.1 0.0 2.2 0.0 Factor Level 0.1 0.2 0.4 0.7 Self-Perception: Athletic Competence Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 0.0 (2, 47) 1.0 Age 0.0 0.0 0.3 0.8 Factor Level 0.0 0.2 0.0 1.0 Self-Perception: Global Self-Worth Predictor Parameter Estimate Standard Error Test Statistic p-value Omnibus F-test (F(ndf,ddf)) 0.8 (2, 47) 0.5 Age 0.0 0.0 0.9 0.4 Factor Level 0.2 0.2 0.9 0.4
None of the above models met overall significance with an omnibus F-test and therefore