Factors that contribute to community members’ support of local nature centers

Abstract

Nature centers can serve as valuable community institutions if they are seen as providing important services to the community. Through survey research in communities surrounding 16 nature centers in the United States, we examine the attitudes, behaviors, and beliefs that drive hypothetical support for nature centers from local residents. Respondents who recognized centers as institutions that provide opportunities for leisure and connection with nature; contribute to civic engagement; and bolster local economies, community pride, and aesthetics reported the greatest likelihood of donating, volunteering, or responding to threats at local centers. Additional predictors of support included positive evaluations of staff members, perceptions of positive attitudes toward the center held by other community members, familiarity with center activities, pro-environmental attitudes, and past donation or volunteering. The findings highlight the potential returns for centers that expand their activities and operations beyond more traditional roles of providing nature-based outdoor recreation and environmental education experiences.

Keywords:

• Nature centers
• community support
• philanthropy
• volunteering
• organizational threats

Acknowledgements

The authors are grateful to our National Audubon Society and Association of Nature Center Administrators colleagues who assisted with and facilitated this work. We also appreciate the nature center directors, staff, and community members who gave generously of their time to participate in the interviews and surveys.

Notes

1. Discrepancy exists with regard to the number of nature centers that exist in the United States. This number was estimated by a Duke University team of researchers who compiled a list of US nature centers using print materials and online directories (e.g. those published by the Association of Nature Center Administrator or ‘ANCA’), Wikipedia lists, and state and national park websites and directories that described parks as having nature centers. We believe this 1800 number to be the most comprehensive. However, the National Environmental Education Foundation, as reported in McNeil (2011), suggests the number of nature centers in the United States is closer to 2000. By contrast, ANCA estimates the number of centers to be approximately 1200; they base this estimate on membership numbers and active participants in network activities (personal communication, Jen Levy, September 2014).

2. These included both urban and rural centers across the United States (two in the Midwest, two in the East, and two in the West). Local community demographics ranged from low to high socio-economic status and from predominantly non-Hispanic whites to predominantly minorities.

3. We desired a range of urbanity levels because past research suggests programming at nature centers is conducted differently in urban versus rural settings (Kostka 1976).

4. Our criterion for ‘natural area’ was broad (e.g. parks, greenways, community gardens, woodlands, and prairies) and excluded only one center: a school bus retrofitted into a scientific laboratory that travelled from one natural area to the next for environmental education programming. We assumed this center was fundamentally different in its relationship with local communities because of its lack of a permanent, stationary interpretive center with adjacent lands on which it engaged in land conservation and recreational or educational programming.

5. We received 1381 completed surveys from those who received a $2 incentive but only 350.

6. We aimed for approximately 400 respondents per center and initially assumed a response rate of 10 to 13%, based on previously reported response rates from general population web-based survey studies (Link and Mokdad 2005, 2006) and pre-paid incentive studies (Dillman, Smyth, and Christian 2014; Göritz 2006). We attempted to follow a modified version of the Tailored Design Method (Dillman, Smyth, and Christian 2014) and contact each person five times (pre-notice, invitation, and three reminders); however, we were limited by our Institutional Review Board to contacting each person a maximum of three times. We adjusted our calculated response rate for the 51,072 unique email and letter bounce-backs we received. We have no way of counting how many emails were filtered by spam folders. As such, we do not know how many respondents actually received our email invitations.

7. The community resilience factor was built from social-ecological systems and psychology of development and mental health literatures (see Berkes and Ross 2013) which discuss how strong economies, senses of pride, and favorable environmental features (e.g. landscape aesthetics) – among other elements – provide a ‘community resilience’ which Magis (2010, 401) defined as ‘the existence, development and engagement of community resources by community members to thrive in an environment characterized by change, uncertainty, unpredictability and surprise.’

8. The importance and performance levels of each value were each moderately strongly correlated (r > .40), and importance-performance analyses suggest these concepts are difficult to de-couple in survey research (Oh 2001).

9. The sample size for these regression analyses was smaller than the total sample due to some respondents indicating ‘I don’t know’ or not responding to one or more survey items. Most prominently, 25% of our sample (n = 354) did not provide performance scores for survey items related to at least one service provision factor (environmental connection, leisure provision, civic engagement, and/or community resilience). An additional 195 respondents did not respond to one or more other predictor or outcome variables. We excluded these cases for regression analyses. The resulting sample size used was 850.

10. Census populations contained 50% males, 29% people with children in their home, and 71% non-Hispanic Whites. Other races/ethnicities present were Hispanic/Latino (19%), black or African American (9%), Asian (5%), and mixed (3%). The average age in census tracts was 38 (census data included people under the age of 18, but we only invited people 18 years old or older to our survey, so a direct comparison between census tract age and sample age was not possible). Of these populations, 14% had less than a high school diploma while 25% had earned their high school diploma, 26% had attended some college, 21% had completed a bachelor’s degree, and 16% had completed a graduate degree.

11. We report adjusted R² values, because they are more conservative than R² values as they account for the number of parameters in the regression equation.