Community and social participation among adults with mobility impairments: A mixed methods study

Abstract

Background:

Community and social participation is a complex phenomenon that is influenced by personal and environmental factors and is linked to a good quality of life and well-being. Individuals with mobility impairments are at risk of experiencing limitations in participating in community activities due to a wide range of factors.

Objective:

To understand community participation as defined by adults with mobility impairments and to examine relationships among factors that influence community participation.

Methods:

A mixed-methods study design was used. In-depth interviews of 13 adults with mobility impairments were conducted and themes related to community participation were identified. Data from the Americans’ Changing Lives Survey were used to construct variables that mimic the themes from the qualitative phase and structural equation modeling was used to examine the relationships among those variables including community participation.

Results:

Individuals with mobility impairments identified health and function, neighborhood factors and self-efficacy as possible factors influencing participation in community activities. Findings from the SEM suggest a strong causal pathway between health and function and community and social participation. Neighborhood factors and health and function had a significant impact on self-efficacy, and a possible indirect effect through self-efficacy on community and social participation.

Conclusions:

Our study provides new empirical evidence that health and function have a significant impact on community and social participation. Our quantitative findings did not support the direct influence of neighborhood factors in community and social participation, yet these factors may have an indirect role by influencing the self-efficacy of individuals with mobility impairments.

Community and social participation, conceptualized as participation in the typical civic, recreational, religious, and social activities in the community, is an important outcome of disability rehabilitation programs. Disability rehabilitation programs typically provide a variety of services to improve health, function, and self-efficacy. While improvements in these areas can lead to increased levels of participation, broader environmental factors can influence community and social participation as well. As posited in the Institute of Medicine disablement model,^1,2 an individual’s participation in life situations is influenced by the complex interaction of his/her underlying health condition, personal characteristics, and the physical/social environment.

An abundance of research examines the unique role of each of these factors in facilitating participation among persons with disabilities.^3–6 To our knowledge, however, no studies simultaneously examine the relative roles of environmental factors, health and function, and self-efficacy as barriers and facilitators to community and social participation for persons with disabilities. Understanding the mechanisms by which participation occurs and its relationship to environmental factors, health and function, and self-efficacy for adults with mobility impairments is central to developing effective intervention programs and policies that promote overall well-being. Using a mixed methods design, this study examines these interrelationships. In-depth semi-structured interviews were first conducted to explore the various dimensions of participation. Analysis of nationally representative household level data, using structural equation modeling (SEM), provides further detail about the mechanisms influencing community and social participation for adults with mobility impairments.

Literature review

In 2013, more than 20 million American adults living in the community had an ambulatory limitation.³ Ambulatory or mobility limitations may occur because of accidents or injuries, age-related cognitive or physical changes, or early onset disability. Such limitations can compromise the ability to participate in activities within several domains: community, productivity, and social. Community activities include participation in leisure, volunteer, religious and civic activities. Participation in education, employment or homemaking activities are considered part of the productivity domain. Unstructured and structured social activities are considered part of the social domain.⁴ Limited participation in any of these areas can diminish life roles, having a negative impact on individual well-being and quality of life.²

Our interest in this study is in understanding the barriers and facilitators to participation in the community and social domains. Environmental factors, health and functional status, and personal characteristics have been identified as barriers and facilitators to participation, influencing frequency and satisfaction with community and social participation.^5,6

Poorly designed communities and distressed neighborhoods have been consistently documented as an environmental barrier to participation for individuals with mobility impairments.^7–9 The classical ecological model of person-environment fit suggests that for optimal functioning to happen, a balance between environmental demands and personal competencies should be present.^10,11 Perceived safety and quality of neighborhood environments has been shown to be a robust predictor of functional health and community participation.^7,8,12,13 Neighborhood security and social capital is also associated with higher levels of social interaction and trips outside the home.^9,12 Current research on social and environmental aspects of neighborhoods are narrowly focused, however, on the univariate, unidirectional impact on health, mortality, aging, and chronic conditions.^14,15

Personal and psychosocial factors characteristics can serve as barriers or facilitators of participation as well. Functional status and cognitive functioning are frequently cited as key determinants of community based activities.^5,6,16 Gains in functional status, especially related to locomotion and self-care, are significantly associated with greater satisfaction in and participation in the community.^6,17,18 Among psychosocial factors, self-efficacy has been shown to be a strong predictor of participation in a wide range of activities such as chronic pain management, physical activities and activities of daily living.^19–22 Self-efficacy is the extent to which an individual believes in his or her capabilities and skills to perform a certain behavior.^23,24 Self-efficacy theory suggests that behavior is determined by a transactional relationship between cognitive thoughts, environmental stimuli, and feedback from performing the actual behavior.²³ Social cognition theory suggests that self-efficacy beliefs regarding health and function and the ability to navigate the neighborhood environment will most likely influence participation in community and social activities. In addition to being a mediator for health and function and neighborhood factors, self-efficacy can be further enhanced by actual participation behaviors through the process of enactive attainment.²³ Enactive attainment refers to the experience of mastery achieved by actual performance of an activity and reflects an authentic appraisal of one’s ability. Enactive attainment is the most influential source of self-efficacy when compared to encouragement by others or informational sources.²³ For individuals with mobility impairments to enact on their self-efficacy beliefs and attain mastery over community participation, they need opportunities and supportive environments to connect with others and engage in meaningful activities.²⁵ While current literature narrowly focuses on the unidirectional relationship between self-efficacy and community participation,^2,26 the literature discussed above suggests a more complex reciprocal relationship in which self-efficacy influences participation but participation experience also influences self-efficacy.

We therefore develop and test an empirical model of the intricate relationship among environmental factors, health and function, and self-efficacy with community and social participation among adults with mobility impairments. Our conceptual model is shown in Fig. 1.

Fig. 1.

Figure depicting the measurement model for community participation. Variables included in the model are neighborhood factors, health and Function, Self-efficacy, and Community and social participation and corresponding Americans’ Changing Lives survey items for each variable.

*See Table 3 for the list if item related to the Neighborhood conditions (N), Health and Functions (HF), Community and social participation (CSP), and Selfefficacy (SE).

ϵ₁- ϵ₂₈ = Error term for each item

Our analysis will specifically address the following research questions:

1. What activities do individuals with mobility impairments consider as community and social participation?

2. What factors do individuals with mobility impairments consider as barriers and facilitators to community and social participation?

3. What is the relationship among the variables identified in the qualitative phase?

Methods

We used a sequential, exploratory, mixed-methods design.²⁷ The first phase of the study involved the collection and analysis of qualitative data. The second phase built off of the results from the first phase, analyzing secondary, quantitative data. The primary purpose of the first qualitative phase was to use a phenomenological approach to examine what community and social participation meant for adults with mobility limitations and to understand the factors that shape their participation.²⁶ Thirteen in-depth, semi-structured interviews were conducted with adults with physical disabilities. Approval from the University of New Hampshire Institutional Review Board was obtained before commencing the study.

Qualitative data

In depth, interviews were conducted by graduate students in occupational therapy as part of their research coursework. Participants were recruited by word of mouth and through local area agencies using a purposive sampling method. An initial screening questionnaire was used to determine eligibility for the study and to exclude individuals with primary or co-existing cognitive impairments. Eligibility was determined based on being older than 25 and having a disability or chronic health condition that limits day-to-day activities. All participants were community dwelling adults aged 36–77.

Interviews were conducted in-person or over the phone and took approximately 45 min each to complete. A semi-structured, open ended questionnaire was used to gather data on participation in six major domains: instrumental activities of daily living, recreation/leisure activities, paid employment, socializing, volunteering, and activities of daily living. All interviews were audio recorded.

Qualitative data analyses

All interviews from the first qualitative phase were transcribed and coded by two students. Coding was verified by an independent reviewer. A phenomenological approach was used to explicate the data from the transcripts of the in-depth interviews. The goal of phenomenology is to understand a central phenomenon and to be able to explain the phenomenon as experienced by the participants. For this study, the central phenomenon of interest was community and social participation. Each student read the transcripts multiple times with the goal of understanding the overarching phenomenon of participation, while simultaneously looking for smaller units of meaning within the transcripts. Data were coded independently by each reviewer using an iterative process until no more meaningful codes could be assigned. This step was followed by a collaborative review of the codes, addressing discrepancies, and revising codes where necessary. When the two coders did not agree on a particular code, an independent third reviewer evaluated the transcripts and assigned codes.

Quantitative secondary data

The second quantitative phase was informed by the results from the qualitative work. The Americans’ Changing Lives (ACL) Survey was chosen as the data source it includes variables relevant to the concepts covered in the qualitative work. The ACL is a longitudinal survey covering a broad range of social, psychological, and behavioral factors related to health. The ACL is a publicly available, nationally representative, multistage sample of non-institutionalized adults aged 25 and older, with an oversampling of African Americans and people over the age of 60. Additional information on the purpose and demographic characteristics of the ACL is available elsewhere.²⁸ Currently five waves of the data are available, covering 1986–2012. For this study, we used a subset of the data from Wave 5 of the ACL that most closely represented the age distribution of our qualitative interviews, including individuals age 50 and over. Data were restricted to individuals who had difficulty walking or climbing stairs or who used a mobility device such as a cane, crutch or wheelchair. The sample size for the quantitative analyses was 462.

Quantitative data analyses

For the quantitative phase, data analysis was conducted using SPSS Version 21²⁹ and Stata.³⁰ Descriptive statistics were used for exploratory analysis to identify the socio-demographic, health and function, and psychological characteristics of the ACL sample and the perceived neighborhood safety, quality and accessibility.

Next, we used SEM models to examine the relationships among latent constructs depicted in Fig. 1. To simplify the model fitting procedure, we fitted the SEM model in two steps. First, the measurement model was fitted using Exploratory Factor Analysis (EFA) to derive the factor scores for each latent construct from observed indicator (or item) variables, and then the structural model was fitted through regression based analysis for all latent factor scores.

Through EFA, a health and function variable was created from the Cognitive Impairment Index, Functional Health Index, self-rated health, and CESD Depression Scale included in the ACL. Rather than a sum of raw scores, regression based factor scale scores were used to create the health and function variable. Higher factor scores indicated higher levels of health and functioning. A neighborhood variable was constructed from variables which gathered information about neighborhood condition and neighborhood cohesion. Neighborhood condition reflected physical amenities and safety. Neighborhood cohesion measured social capital, an indicator of trust and interpersonal dependence within the neighborhood.³¹ Factor scale scores were also created for the resulting ‘neighborhood environment’ variable, with higher scores indicating increased levels of cohesion and better neighborhood amenities. Similarly, the variable of self-efficacy was constructed using survey questions related to self-esteem, mastery, and competence. The self-efficacy variable included items from the Rosenberg’s self-esteem scale³² and Pearlin and Schooler’s mastery scale.³³ The measure of self-efficacy was comparable to the one used by Sherer et al³⁴ and comparable to the concept of self-efficacy as presented by Bandura.²⁴ Factor scores were again used for the resulting continuous variable of ‘self-efficacy’ where higher scores indicated higher levels of self-efficacy. The last set of EFA was conducted to identify and extract the community and social participation variable. Eleven items were included in the factor analysis and, after eliminating two items, the composite measure of community and social participation included items related to (1) formal and informal social activities; (2) participation in political, religious, and community groups, and (3) volunteering. This resulting measure of community and social participation was consistent with Chang and Coster’s⁴ conceptualization of community and social participation and also with the themes identified in the previous qualitative phase. We excluded work as an indicator of community participation because of limitations in sample size and also because conceptually, the skillsets and opportunities available to working individuals are distinct from those who are not.

After factor scores were derived for all four latent constructs, initial exploratory analysis was conducted to determine correlations among the latent variables of interest. We hypothesized that self-efficacy moderates the relationship between health and function and social and community participation. Lastly, the SEM structural model was tested to build the relationships among the latent variables shown in Fig. 1. We fitted the measurement model and structure model separately as the complexity of fitting them together caused non-convergence of model estimation.

Results

Qualitative analysis

The qualitative analysis addresses our first two research questions: (1) what activities do individuals with mobility impairments consider as community and social participation? and (2) what are the barriers and facilitators to community and social participation? Thirteen individuals with mobility limitations participated in the qualitative phase. The average age of participants was 52.4 (SD = 13.34) and 10 (77%) were female. Eight participants lived in suburban neighborhoods and four lived in a rural neighborhood. Participants talked about the frequency and the intensity at which they participated in various community and social activities and the importance of having choice and control over the type of activities they engaged in. They identified community participation as including work, leisure, social activities, volunteering and occasionally instrumental activities of daily living such as shopping.

Participants voiced several challenges navigating and participating in their local community. Their experiences were captured in four major themes related to community participation; (1) health and function; (2) neighborhood and social environment; (3) sense of mastery/self-efficacy; and (4) participation in employment, leisure, social and volunteering activities. The four themes and sample quotes from participants representing each theme is presented in Table 1. Results are summarized below.

Table 1.

Qualitative data analysis: themes related to community and social participation

Theme	Participant quotes
Health and function	• Walking around in the city is hard. I’ve found things I like to do that doesn’t take up so much energy, like reading, going out to restaurants with friends.
	• I mean I you know I get really moody and tired easily and I would only subject certain people to that. So I always hang out with particular people based on the way I’m feeling so it does hold me back from interacting with some people at sort of all times of my you know … whether or not I’m feeling good.
Environmental factors	• I mean I wish things were closer so that it didn’t exhaust me so much to drive 25–30 min to do stuff quickly and drive another 25 min home.
	• I still drive so I do not have difficulty getting to places outside of my home. If I was not able to drive I would have to move because there is no transportation services in my town.
	• There was a woman down the street, we used to walk together. We used to call each other up and say ‘‘Can you walk?’’ … ‘‘Yeah, okay,’’ and she’d come over, 2 houses away and we’d walk.
	• I think to myself, I will just go in the restaurant and get a sandwich but I might not do that because I know it’s more difficult when I’m alone. But, I’d say overall, people here go out of their way to help me.
Self-efficacy	• I think it’s just my outlook … I know I can do it. (long pause) It’s more fun walking with someone. But I got to get myself out there and do it myself.
	• The therapist says I do more than she does. She says, ‘‘I never heard of anybody doing all that in one day! Why do you do so much?’’ I said, ‘‘Because I don’t have time the rest of the week. I have to do other stuff!’’ (laughs)
	• I have no help and do not want help. I live alone and am perfectly capable right now to do everything on my own. It just hurts. I do have a handicap plate so that I can park close to buildings which makes life a little easier.
Community and social participation	• I wasn’t able to get to meetings. Now I’m not able to … like we do community gardens and for the whole past year I haven’t been able to help with any of that gardening.
	• I volunteer and work at two churches, the laundry mat, (and walk my dog, Brady).
	• I started the Potter County MS Support group, uh (pause) church I do a lot, the church up town on route 14. And then I do a lot of things like going to the hospital and nursing homes in town, so not a lot of organized volunteers.
	• Well it is important, it keep, you know, to keep your mind going, and uh … it’s kind of depressing being by yourself and having to be in the house and not being able to participate with other people

Health and function

In general, most participants attributed their inability to participate in various activities to poor functional health and disability. Many participants blamed their inability to participate in various community activities on their disability or health status. Pain and fatigue were often stated as barriers that limited them from participating in activities that they performed previously. Four participants mentioned their inability to stand for long hours and fatigue as the reasons for limited participation at work and reduced hours at work. Participants felt that given their disability and the recent changes in the economy, they did not have much choice in what jobs they had to perform.

Neighborhood and social environment

Participants seemed to rely a great deal on social support within their community to accomplish their day-to-day tasks or participate in activities. This was especially true if they lived alone or if they had limited access to transportation. Two participants talked about how they could depend on and trust their neighbors if something were to go wrong. They felt extremely confident in their neighbors’ willingness and ability to help in case of emergencies or for any health related needs.

Self-efficacy

Self-efficacy and a sense of mastery were reflected in the rhetoric of several participants. Many participants talked about how they had to ‘‘just do it’’ regardless of the challenges and a lack of assistance in venturing out in the community. One participant talked about her desire to do everything by herself and how despite severe pain and in the absence of any help she accomplishes her activities independently. Another participant talked about how she had to get herself out in the community and that her outlook on life made it possible to venture out in the community. Others acknowledged the role of friends or therapists in supporting them in going out in the community, but noted that their own self-drive and initiative actually made it possible for them to be active in the community.

Participation in employment, leisure, social and volunteering activities

Participants discussed the frequency and intensity of involvement in community activities. Socializing with friends and family was a valued activity for many participants. One participant talked about how deprived she felt when she could not go out with friends. Several participants expressed a strong desire for and active role in volunteering activities through a church, local hospital or nursing home. Many valued such volunteering activities for the social connections they offered and reported feeling good participating in such activities. For those who were not working outside the home, volunteering activities gave them ‘‘something to do’’ and ‘‘kept them fresh.’’

Quantitative analysis

Our quantitative analysis addressed our third research question: What is the relationship among neighborhood environmental factors, health and function, and self-efficacy with community and social participation for individuals with mobility impairments? Demographic characteristics of the ACL sample are shown in Table 2. The mean age of the sample was 69 years. Most (71%) of the sample was female. More than a third of the sample was married (39%). Less than a quarter of the sample was employed (23%). Most of the sample lived in households with incomes below $60,000 per year.

Table 2.

Demographic characteristics Americans’ changing lives survey sub-sample

	N = 462
Age mean (SD)	69.47 (12.56)
Gender (female)	327 (70.8%)
Married	178 (38.5%)
Employed	105 (22.7%)
Household income
Less than $19,999	176 (38.10%)
$20,000–$39,999	127 (27.49%)
$40,000–$59,999	73 (15.80%)
$60,000–$79,999	41 (8.87%)
$80,000–$99,999	11 (2.38%)
$100,000–$149,999	20 (4.33%)
$150,000–$249,999	11 (2.38%)
$250,000 or more	3 (.65%)
Health and disability
Joint replacement surgery	83 (18%)
Broken bones or fracture	34 (7.4%)
Arthritis	305 (66%)
Stroke	50 (10.9%)
Mobility device use	219 (47.4%)
Difficulty climbing stairs	259 (56.1%)
Difficulty walking several blocks	259 (56.1%)

Table 3 shows the ACL variables corresponding to the themes extracted in the qualitative phase and the results from the factor analyses.

Table 3.

Exploratory factor analysis of American’s changing lives survey (N = 451)

Items	Factor 1	Factor 2	Factor 3	Mean	Std. dev.
Health and function
Cognitive Impairment Index (HF1)	.664			2.58	1.15
Functional Health Index (HF2)	.508			5.13	.959
Self-rated health (HF3)	.703			44.53	4.32
Center for Epidemiology Depression Scale (HF4)	.752			2.66	1.01
Neighborhood factors
This is a neighborhood where I feel safe from personal attacks (N1)	.722	.375		3.19	.92
This is a neighborhood where people feel safe walking alone this area after dark (N2)	.723	.328		2.84	.996
There is no problem with vandalism and graffiti in this neighborhood (N3)	.777	.133		2.90	1.08
There are no vacant houses or storefronts in this neighborhood (N4)	.674	.061		2.87	1.21
This neighborhood is kept very clean (N5)	.578	.425		3.27	.898
If you were in trouble, there are lots of people in this neighborhood who would help you (N6)	.190	.794		3.11	.96
Most people in this neighborhood are friendly (N7)	.137	.851		3.20	.88
Most people in this neighborhood can be trusted (N8)	.337	.742		2.92	.93
Self-efficacy
I am always optimistic about my future (SE1)	.080	.752		3.09	.84
I hardly ever expect things to go my way (SE2)	.651	–.163		2.68	.99
I rarely count on good things happening to me (SE3)	.647	–.266		2.67	1.09
Inclined to feel I am a failure (SE4)	.675	.280		3.54	.80
Take positive attitude toward myself (SE5)	.058	.764		3.37	.81
I think I am no good at all (SE6)	.725	.268		3.41	.91
No way I can solve problems I have (SE7)	.542	.138		2.61	1.04
I can do anything I set my mind to do (SE8)	.053	.635		3.08	.90
I feel I am being pushed around in life (SE9)	.569	.112		3.00	1.01
Community and social participation
Talk on the telephone or Skype with friends, neighbors or relatives (CSP1)	–.270	–.178	.626	3.26	1.21
Write or email with friends, neighbors or relatives (CSP2)	.097	.393	.540	3.37	2.105
Get together with friends, neighbors or relatives (CSP3)	.195	–.112	.605	3.39	1.412
Attend meetings or programs of groups, clubs or organizations that you belong to (CSP4)	.491	.297	.311	2.85	1.80
Volunteer work for groups and organizations	–.014	.790	–.023	2.51	1.39
Help friends, neighbors, or relatives who did not live with you (CSP5)	.080	.770	–.105	2.73	1.29
Attend religious services (CSP6)	.794	–.126	.099	3.08	1.72
Activities at a church or place of worship besides religious services (CSP7)	.784	.111	–.137	3.26	1.98

Bold values: p < .05.

Results from the SEM including direct, indirect, and total path coefficients are shown in Table 4. The structural model had adequate fit statistics (χ² = 137.08, p > .05), Comparative Fit Index (CFI) = 1.00, Tucker–Lewis Index (TLI) = 1.00 and RMSEA = .00. CFI and TLI values closer to 1.0 indicate a good fit along with RMSEA less than .07, which indicates parsimony of parameters used in the model. The strength of the path coefficients and the direction of relationships are presented in Fig. 2. In examining the effects of the hypothesized independent variables on community and social participation, we found a significant direct effect of health and function on social and community participation (β = .34, p < .01). In addition, health and function had a small positive but insignificant indirect effect on community and social participation. The total effect, which is the sum of the direct and indirect effect of health and function, was strong and significant, indicating that higher levels of health and function were associated with higher levels of community participation (β = .39, p < .01). We did not observe a significant effect of neighborhood factors on community participation; the direct, indirect and total path coefficients between neighborhood factors and community participation were small, negative and insignificant. Similarly, self-efficacy had a small but insignificant direct effect on social and community participation.

Table 4.

Path coefficients in final SEM model (coefficients and std. errors)

	Direct effect	Indirect effect	Total effect
Self-efficacy as dependent variable
Neighborhood factors	.1139 (.0431)**	No path	.1139 (.0431341)**
Health and function	.6144 (.0604)**	No path	.6144 (.0604)**
Community participation as dependent variable
Self-efficacy	.0806 (.0639)	No path	.0806 (.0639)
Neighborhood factors	–.0143 (.0581)	.0092 (.0081)	–.0052 (.0578)
Health and function	.3371 (.0898)**	.0495 (.0396)	.3866 (.0809)**

RMSEA = .000, p < .05; Comparative Fit Index = 1.000; Tucker–Lewis Index = 1.000.

^**p < .01.

Source: Authors’ analysis of the Americans’ Changing Lives Wave 5 data

Fig. 2.

Figure depicting the structural model for community participation. Variables included in the model are neighborhood factors, health and Function, Self-efficacy, and Community and social participation.

Health and function (β = .61, p < .01) and neighborhood environment (β = .11, p < .01) were, however, both found to affect self-efficacy, with higher levels of health and function and better neighborhood condition and cohesion associated with higher levels of self-efficacy. The results suggest that overall, health and function have both direct and indirect effects on community and social participation. Better health and function are associated with higher levels of social and community participation.

Discussion

Our first research question was to understand what activities individuals with mobility impairments considered as community and social participation. Findings from the qualitative phase of the study indicate that leisure activities, socialization activities, volunteering, and to some extent, work, are considered important indicators of community participation for persons with mobility limitations. These findings were later supported by the factor analysis in the quantitative phase of the study.

Our study next identified several factors that individuals with mobility impairments considered as barriers and facilitators to community and social participation. Proximity of preferred destinations was a desirable feature in the neighborhood. Participants valued the presence of social supports within the community, especially to engage in physical activities and healthy behaviors. Participants also shared differences in terms of self-efficacy, where some had a strong conviction to overcome barriers related to health and function or pain, while others expressed that their limitations were insurmountable. Self-efficacy theory suggests one of the strongest mechanisms by which efficacy is achieved is through positive experiences from enactive attainment. Experiencing success through various community and social activities allows adults with mobility impairments to gain confidence in their ability. On the contrary, experiences of failure lower feelings of mastery and efficacy.

While our qualitative findings uncovered the possible mediating role of self-efficacy in determining social and community participation, our quantitative SEM analysis did not support this idea. Health and function were most important in influencing community and social participation, a finding which is consistent with several other studies.^6,17,35 According to our findings, health and function was an important factor that determined community and social participation and may influence participation both through direct and indirect mechanisms where self-efficacy acts as a moderator of health and function and participation. In our study, health and function was characterized by functional health, cognitive impairment, self-rated health, and depression. Consistent with previous studies, our findings suggest that individuals who experience depressive symptoms or view themselves in poor health are less likely to engage in community and social activities outside their homes.

The path coefficient between self-efficacy and community participation was small and insignificant. A plausible explanation for this finding is that the self-efficacy measure used in the ACL survey was a measure of general self-efficacy, which may not be sensitive to changes in community and social participation. With regards to neighborhood factors, we found a small, negative but insignificant path between neighborhood factors and community participation. This finding suggests a possible inverse relationship between neighborhood quality and community participation. This means that with any increase in neighborhood quality, there is a decrease in community participation. It is possible that individuals with mobility impairments compensate for poor quality and unsafe neighborhoods by creating close-knit communities where members volunteer, help each other, and are more engaged with each other, thereby increasing their community participation. In other words, when the physical and safety conditions of the neighborhood deteriorate, community members may be more likely to be cohesive. Similar studies have demonstrated that, in the long-term, environmental factors are not significantly related to community participation, meaning that, with time, individuals adapt to their neighborhoods and find ways to participate in valued activities.^36,37

While considering self-efficacy as the dependent variable, there was a significant path from both health and function and neighborhood factors. Consistent with previous studies, the health and function variable was a significant predictor of self-efficacy. Neighborhood factors had a small, positive impact on self-efficacy. Self-efficacy dictates the way individuals appraise their neighborhoods, make an effort to navigate the social and physical environment and anticipate success or failure in community participation.³⁸ However, in the absence of good health and function, self-efficacy alone may not be sufficient to achieve community participation. Poor health and functioning may leave individuals with very little energy for activities outside the home, even though they may wish for it. Future studies examining the relationship between neighborhood factors and community participation should further explore the role of self-efficacy in community participation.

Rehabilitation programs that seek to improve community and social participation need to consider the role of health and function and self-efficacy perceptions based on the individuals’ appraisal of their neighborhoods. Further research which examines both structural impediments as well as differences in social capital can more closely examine the influence of neighborhood characteristics on self-efficacy and participation for this population.

Limitations

Our results are limited by the use of a relatively small empirical sample to conduct the SEM. In addition, our results are conditioned on the availability of information within existing survey data. While we attempted to find a survey that included information that closely aligned to the concepts raised during our qualitative analysis, the use of survey data that captured more constructs may result in different results. Individuals with mobility impairments can be a diverse group, depending on the type of mobility devise used. Analysis by the type of device used was not possible to due to small sample sizes. The participants in the qualitative and quantitative phases of the study were not of the same age range and therefore the findings from one phase may not be fully applicable to the second.

Conclusion

Despite these limitations, our study provides new empirical evidence that health and function have a significant impact on community and social participation. Although our quantitative findings did not support the direct influence of neighborhood factors on community and social participation, environmental factors may have an indirect role by influencing the self-efficacy of individuals with mobility impairments. Improving the perceptions of self-efficacy and health and function of persons with mobility impairments is important for ensuring the full social and community participation of this population.