**6. Measures**

Our assessment of community decision-making utilized measures described in depth in Kammer-Kerwick et al. [1]. We have included a summary of the measures here for continuity.

*The Connection between Entrepreneurial Intentions and Community Member Priorities… DOI: http://dx.doi.org/10.5772/intechopen.112287*

#### **6.1 Schwartz value theory**

According to Schwartz' theory of basic human values [15], all cultures are shaped by ten distinct personal values. These values comprise self-direction, stimulation, hedonism, achievement, power, security, conformity, tradition, benevolence, and universalism, forming a circular motivational continuum that reflects the compatibility and conflict among values. Schwartz's theory proposes that this continuum is organized along two bipolar dimensions: one dimension compares "openness to change" and "conservation," while the other compares "self-enhancement" and "selftranscendence." To assess these dimensions, we have included resultant selftranscendence (self-transcendence—self-enhancement) and resultant conservation (conservation—openness to change) in our survey. This approach is commonly used in various marketing and business strategy contexts [16–19]. Additionally, we included community size, population density, and distance to the nearest major metropolitan center in Texas as additional variables to explain variance in interest in the community development concepts. Furthermore, to account for our multi-site data collection design, we included community as a random effect in our analysis. We also assessed the role of social position by including the age, race, gender, education, and income of community members in our sample.

#### **6.2 Big five personality traits**

The Big Five Inventory (BFI) is a well-established measure of five personality dimensions that have been shown to be relatively stable and distinct over the life course and applicable across cultures [20–22]. We included agreeableness and openness because the prior research indicates that these personality dimensions should predict the likelihood of a community member investing in an intervention. Agreeableness relates to a desire to engage in prosocial actions. Openness was viewed to be a personality trait that might predict respondent likelihood to invest in community development interventions. We omitted conscientiousness, extraversion, and neuroticism from the BFI to avoid participant fatigue.

More specifically, we hypothesized that agreeableness would be the personality factor that would best predict whether community members *would cooperate toward an intervention.* Openness to experience includes three higher-order measures of intellectual curiosity, active experiencing of senses and emotions, and openmindedness toward different cultural ideas and values [23]. Central aspects of openness include a willingness to entertain novel ideas and unconventional values, intellectual curiosity, and independent judgment. Openness has been studied in relation to creativity and innovation in social entrepreneurship [24], adjustment to change, identification and maintenance of specific communities [25]; Füller et al. (2008), and engagement with community development interventions [26].

#### **6.3 Asset satisfaction and importance**

Respondents rated how satisfied they were with eighteen community assets, on a 7-point scale where 1 is "not at all satisfied" and 7 is "extremely satisfied". Participants also rated the importance of these same eighteen assets in the context of considering relocating to another community (where 1 is "not at all important" and 7 is "extremely important". We performed an exploratory factor analysis (EFA) and reliability analysis to create four mean asset satisfaction subscales from the original eighteen survey


#### **Table 3.**

*Community asset themes.*

items, previously listed, see **Table 3**. This analysis is discussed in Kammer-Kerwick et al. [1]. We utilized these same four themes to calculate asset importance subscales. **Table 3** shows the eighteen community assets included in the analysis the following items, grouped by theme.

#### **6.4 Entrepreneurial intentions and antecedents**

We incorporated the measures from the entrepreneurial intention questionnaire (EIQ) [27]—based on Ajzen's [28] Theory of Planned Behavior (TPB)—to assess the level of entrepreneurial intention and its antecedents. The EIQ is a cognitive model based on the assumptions that becoming an entrepreneur is a voluntary and conscious decision and that intention is the single best predictor of behavior. The EIQ measures three motivational antecedent factors: a person's positive or negative attitudes toward being an entrepreneur (personal attitude), a person's perception of whether significant personal relationships would approve or disapprove of being an entrepreneur (subjective norms), and a person's perception of how easy or difficult it would be for them to do entrepreneurial behaviors (perceived behavioral control). These three antecedents measure the effort needed for a person to make the decision to practice

#### *The Connection between Entrepreneurial Intentions and Community Member Priorities… DOI: http://dx.doi.org/10.5772/intechopen.112287*

entrepreneurial behavior. Demographic and situational factors will influence entrepreneurial intention, like educational experiences and time constraints. We include several demographic variables, like level of education, and variables representing the social context of the community in our models predicting community intervention concepts.

In addition to entrepreneurial intention, Linan and Chen [27] include measures for personal attitudes, subjective norms, and perceived behavioral control in the EIQ, with four all measured on 7-point rating scales. Personal attitudes are measured through agreement ratings for 5 items, including "if I had the opportunity and resources, I'd like to start a firm" and "being an entrepreneur implies more advantages than disadvantages to me." Subjective norms were measured through approval rating from close family, friends, and colleagues for "if you decided to create a firm, would people in your close environment approve of that decision." Perceived behavioral control was measured through agreement ratings for 6 items, including "to start a firm and keep it working would be easy for me" and "I know how to develop an entrepreneurial project." Entrepreneurial intention was measured through agreement ratings for 6 items, including "my professional goal is to become an entrepreneur" and "I am determined to create a firm in the future." The EIQ has been utilized in diverse countries with results that demonstrate satisfactory measurement properties and strong support for the model. The EIQ has elucidated insights into how cultural values modify the way individuals in different communities perceive entrepreneurship [27].

#### **6.5 Additional explanatory variables**

To account for variability in interest across the different community development concepts, we incorporated several additional variables in our analysis. These variables encompassed community size, population density, and proximity to the nearest major metropolitan center in Texas. Moreover, we utilized community as a random effect to accommodate our multi-site data collection design. To evaluate the influence of social position, we considered age, race, gender, education, and income of community members in our sample. Lastly, as novel variables, because this study was conducted during the first year of the COVID-19 pandemic, we included variables measuring participants' perceptions of how COVID-19 impacted their community's health and economy. Specifically, we asked, "How has your community's health been affected by COVID-19?" and "How has your community's financial condition been affected by COVID-19?" Both questions used a rating scale where 1 is "not very affected" and 5 is "extremely affected".

#### **6.6 Community development concepts**

Our study focused on understanding how community members prioritize and make trade-offs among a set of community development project concepts, which were selected based on need expressed in literature as well as from comments obtained during interviews and community meetings conducted during the planning phase for this survey. These concepts were all popular options that have been considered and implemented in various community settings. The specific societal systems covered by these concepts included: Downtown Renovation for Mixed Use Facilities (Built Environment), Community Health Centers (Health System), Gigabit Fiber Broadband Downtown (Communications System), Adding more Computers and Meeting Spaces in the Public Library (Civic System), Early College Credit and

Vocational Programs for High School Students (Education System), Co-Working and Startup Working Space for Entrepreneurs (Business System).

To assess community members' preferences, we employed an exercise that asked participants to allocate 100 points across the different projects based on how well they thought the projects would fit the needs of their community. We included a seventh category for any additional concepts that participants deemed salient. The wording of the options was localized to the context of the survey, with the phrase "in my neighborhood" included in surveys conducted in major metro areas. Further details about the concepts tested can be found in Kammer-Kerwick et al. [1].

#### **7. Data analysis strategy**

We conducted descriptive statistics as well as exploratory factor and reliability analyses using SPSS 27.0. To address our research questions, we employed Generalized Linear Mixed Hurdle Models (GLMMs), using the logit link function and binomial distribution to fit each interest model. Interest in each project concept was classified as either yes (allocation of any points) or no (allocation of zero points). The allocation models were also GLMMs, with an identity link function and a Gaussian distribution, and used centered log-ratio transformation to counter skewing resulting from the point allocation exercise. The allocation model for each concept was fitted using only those participants who expressed interest in that concept. To account for multiple sites involved in the study, random intercept mixed hurdle models were used to model interest in each project concept. All predictive models were run using the glmmTMB package 1.0.2.1 in R [29].

We tested each model in several stages, starting with community characteristics, followed by social position, COVID impact, values, personality, asset satisfaction and importance, and finally entrepreneurial intention.

The improvement of the models with the addition of variables after each step was assessed with the reduction in AIC. In summary, the specification for the models employed in the present study is shown in **Table 4**. It is important to note that for the sake of parsimony and model convergence issues, we are presenting models with simplistic representations of gender and race/ethnic identities. This choice has empirical support, e.g., consistent with the communities, our rural sample includes few participants with nonbinary gender identities and fewer brown and black community members than the larger metro areas. Nonetheless, we recognize that this is a limitation in our study, which we discuss further below.

Interest (0/1) and allocation (centered log ratio) in each of:


Stage 1 Predictors: Location—RQ1


<sup>•</sup> Population (numeric, log(population))


#### **Table 4.**

*Variables included in project concept interest models.*

### **8. Results**

In **Tables 5** and **6**, we provide an overview of the results of the hierarchal hurdle models for both interest and allocation related to the six community development project concepts that address the "if at all" portions of our five research questions. These tables also include fit statistics for each model as information is added to the hierarchy. The Akaike information criteria (AIC) for each model, the change in the X2 statistic as information is added, and the significance of the change in model fit are presented.

**Table 5** specifically displays the changes in model fit as the stages of information are added to the models. The inclusion of the random intercept improved fit for all models, confirming the need to account for random differences between the communities where we collected the data (p < 0.01). The sequential addition of location variables significantly improved fit for all models except for interest in downtown renovation and early college credit/vocational training (p = 0.453 and p = 0.134, respectively) and allocation to a community health center and allocation to a public library hub (p = 0.296 and p = 0.918, respectively). Fit was improved significantly after the addition of human capital and community perceptions of the impact of COVID for all models (p < 0.01). The addition of social capital (via personal values and personality variables) significantly improved the fit of all models except interest in community health center



*The Connection between Entrepreneurial Intentions and Community Member Priorities… DOI: http://dx.doi.org/10.5772/intechopen.112287*

#### **Table 5.**

*Hierarchical modeling summary for interest and allocation models.*

(p = 0.614), broadband (p = 0.321) and allocation to a public library hub (p = 0.197). The fit of all models was improved by the addition of information about community capitals (via perceptions about satisfaction and importance of community assets), all significant at p ≤ 0.001. The fit of all models was further improved by the addition of information about business capitals (via entrepreneurial intention and its antecedents) except for interest in and allocation to community health center (p = 0.271 and p = 0.065, respectively) and interest in public library hub (p = 0.126).

These results confirm that interest in and degree of allocation to various development concepts to improve community well-being are broadly connected, with exceptions, to the characteristics of community members, the values they hold, aspects of their personality, and how they perceive the adequacy of the assets available to the community. All our research questions are at least partial answered in the affirmative.


*Caption. System-Focused Community Development Concept Key:*

*S1, Downtown Renovation for Mixed Use Facilities (Built Environment).*

*S2, Community Health Centers (Health System).*

*S3, Gigabit Fiber Broadband Downtown (Communications System).*

*S4, Adding more Computers and Meeting Spaces in the Public Library (Civic System).*

*S5 ,Early College Credit and Vocational Programs for High School Students (Education System).*

*S6 ,Co-Working and Startup Working Space for Entrepreneurs (Business System).*

*I, Predicts Interest.*

*A, Predicts Allocation among Those with Interest.*

#### **Table 6.**

*Summary of results for each research question by system.*

As a focus of the current study, most of development concepts are connected to entrepreneurial intentions, with important exceptions. Entrepreneurship did not predict interest in or allocation to community health centers nor did it predict interest in public library hubs. However, among those who are interested in public library hubs, entrepreneurial intentions are connected to a willingness to allocate resources to those hubs. These findings are summarized in **Table 5**. **Table 6** simplifies this presentation down to a mapping of capitals to systems.

After establishing the broad support for the connections posited in our research questions that connect capitals to systems, we will present the detailed results of our models that characterize the nature of these connections. The results will reveal which variables in each category increase or decrease interest and degree of allocation for each community development project concept, as perceived by communities.

**Tables 7** and **8** provide the adjusted odds ratio (AOR), standard error (SE), and significance (Sig.) for all terms in each model, and we will focus on model effects (AOR) that are significant at 0.05 or less. We organize our discussion around our research questions, starting with the results from the hierarchical modeling process across all six concepts, followed by a presentation of the results for each model taken in turn. These models correspond to research questions 1 (community characteristics), 2 (social position), 3 (personal values and personality types), and 4 (asset satisfaction and importance). We will focus on the final, or full, model in our discussion, and restrict it to only those results with a significance of p < 0.01. To provide a comprehensive qualitative review of these models, we have included **Table 6**.

*Downtown renovation of mixed-use buildings*: Now consider the results for the community development project concept of downtown renovation for mixed-use facilities, as shown in **Table 7**. Increasing age predicts a lower likelihood of interest in a downtown renovation for mixed-use facilities (AOR = 0.922, p < 0.001). A more agreeable personality predicts a higher likelihood of interest in a downtown renovation for mixed-use facilities (AOR = 1.110, p = 0.017). Greater resultant selftranscendence predicts a lower likelihood of interest in a downtown renovation for


#### *The Connection between Entrepreneurial Intentions and Community Member Priorities… DOI: http://dx.doi.org/10.5772/intechopen.112287*




### *The Connection between Entrepreneurial Intentions and Community Member Priorities… DOI: http://dx.doi.org/10.5772/intechopen.112287*

