The results of the present study demonstrate that the brief scale that we developed, the CCS to measure CC, showed acceptable internal consistency and concurrent validity among both the local volunteers and the general residents in Japanese urban areas. CC measured by the CCS was related to a similar concept, SOC, and also to self-efficacy for helping elderly neighbors to prevent elderly social isolation, and their concurrent validity was identified.
Our scale development process was solid, as we tested the CCS through population-based surveys of 859 local volunteers as a target sample and 3,484 general residents as a control sample, after pilot testing with 266 local volunteers. One of the strengths of our study is to clarify that the two domains of “belonging” and “socializing” were composed of exactly the same items in both types of sample. The present results suggest that the structure of the CCS is strongly confirmed.
The results also suggest that the statistical evaluation of the CCS was adequate. The distribution including skewness and kurtosis was parametric and the internal consistency was sufficient, as shown by Cronbach’s alphas of 0.75 and 0.78, both achieved at an adequate level for an 8-item scale
Since CC was highly associated with SOC which has been measured in other countries, concurrent validity of the CCS was verified. Compared with the SOC based on McMillan and Chavis
[15, 16], which includes needs fulfillment, group membership, influence and emotional connection, CC is more focused on mutual relationship between people living in the local community rather than SOC.
The CCS is aiming to measure CC, which relates to helping elderly neighbors in our study perspective. We had assumed that general residents would not actually experience helping elderly neighbors very often. However, the results have shown that local volunteers as well as general residents who are committed to their communities were more confident in both forms of helping elderly neighbors to prevent their social isolation. Previous studies have also shown the relationships between strong SOC and being a volunteer
[12, 13]. Although relationships between self-efficacy and actual practices of helping elders should be confirmed in future, our results suggest the possibility that enhanced CC among community-dwelling people may facilitate the helping of elderly neighbors and could be measured as an outcome indicator of any intervention for facilitating such helping by utilizing this scale.
The present study has several limitations that suggest caution about generalizing the results. First, response rates of the two surveys, although quite high at about 60%, do not suggest that local volunteers or residents who did not respond to survey have less commitment to their community than the participants who did respond.
Second, the present study setting is limited to Japanese suburban areas close to a megalopolis. People may gather from various local areas in Japan to those cities and have diverse cultures. CC can be especially affected by culture or values in the community and can differ by community characteristics according to whether the area is rural or urban. There may be specific Japanese characteristics in CC compared to other countries. Future studies need to investigate the validity of the CCS in residents living in rural areas in Japan or in other countries, who may have different characteristics from the present participants. Moreover, in using the CCS English version, its validity or reliability should be verified, because the face validity of the CCS was limited.
Third, our study subjects were relatively older residents, but younger people could be local volunteers, so that a cross-generational investigation in CC is needed.
Finally we found evidence regarding the face validity and concurrent validity of the CCS. Other types of concurrent validity or criterion validity including predictive validity or discriminant validity need to be investigated. In particular, the known-group validity of the CCS, for example, according to the groups of local volunteers or general residents, or years of living in the area, should be identified in further analyses.