Page 67 - Social networks of people with mild intellectual disabilities: characteristics and interventions

Network Characteristics, Satisfaction, Wishes and Quality of Life
such as loneliness questionnaires. Likewise, with respect to the wishes of the participants, a lack of knowledge or experience might be an explanation of the fact that participants did not mention the wish to expand their networks, even though their networks were not very large. People who do not have experience in making new friends may not be aware of the fact they could and may not even wish to have more friends.
We did not use multiple regression analyses in the present study to
analyze the extent to which the structural and functional characteristics of
the social networks were related to the satisfaction of people with ID with
their social networks or their perceived QOL. Our sample size (n = 33) was too 3 small in light of the number of predictor variables and some of the network characteristics were highly interrelated, which could introduce bias because
of multicollinearity. Pearson correlations were calculated for the total network
and then, where correlations were significant on p = .05 level, for specific
groups within the network (i.e. family, acquaintances and professionals). The
remaining number of tests still raised some concerns about the occurrence
of type I errors (i.e. revealing false positive results). We did not choose using
the Bonferonni correction, because in our relatively small sample size it would exacerbate the existing problem of low power (Nakagawa 2004) and it would
increase the likelihood of type II errors (i.e. the occurrence of false negative
results; Perneger, 1998). Instead the observed effect sizes (r) and the exact significance levels (p) were reported and thereby the focus was put on the
strength of the relationships between variables (Nakagawa, 2004; Perneger,
1998). Conclusions with respect to the correlations with p < .05 should be taken
with some caution. However, they do indicate many promising directions for
future research highlighting which characteristics of the social network appear
to be of major importance and therefore worthy of greater investigation. In
future research, the use of larger sample sizes in order to be able to conduct
multiple regression analyses and thereby unravel the observed interrelations
is recommended.
Only the first wish expressed by the client in our study was coded. For future research, it is nevertheless recommended that more wishes be included and a more in-depth, qualitative approach to the analysis of the information provided be adopted. Moreover, future research should focus on the impact of relationship status (e.g. dating, cohabiting, married) and relationship happiness on SWB of people with ID, as in the general population a stable
65