International Journal of

Environmental Research

and Public Health

Article

Increasing Environmental Health Literacy through Contextual Learning in Communities at Risk

Leona F. Davis 1 , Mónica D. Ramirez-Andreotta 2,*, Jean E. T. McLain 2,3, Aminata Kilungo 4, Leif Abrell 2,5 and Sanlyn Buxner 1

1 Department of Teaching, Learning and Sociocultural Studies, University of Arizona, 1430 E. Second St, Tucson, AZ 85721, USA; [email protected] (L.F.D.); [email protected] (S.B.)

2 Department of Soil, Water and Environmental Science, University of Arizona, 1177 E. Fourth St, Tucson, AZ 85721, USA; [email protected] (J.E.T.M.); [email protected] (L.A.)

3 Water Resources Research Center, University of Arizona, 350 N. Campbell Ave, Tucson, AZ 85719, USA 4 Mel and Enid Zuckerman College of Public Health, Health Promotion Sciences Department,

University of Arizona, 1295 N. Martin Ave, Tucson, AZ 85721, USA; [email protected] 5 Department of Chemistry and Biochemistry, University of Arizona, Tucson, AZ 85721, USA * Correspondence: [email protected]; Tel.: +1-520-621-0091

Received: 16 September 2018; Accepted: 5 October 2018; Published: 9 October 2018 ���������� �������

Abstract: Environmental health literacy (EHL) has recently been defined as the continuum of environmental health knowledge and awareness, skills and self-efficacy, and community action. In this study, an interdisciplinary team of university scientists, partnering with local organizations, developed and facilitated EHL trainings with special focus on rainwater harvesting and water contamination, in four communities with known environmental health stressors in Arizona, USA. These participatory trainings incorporated participants’ prior environmental health risk knowledge and personal experiences to co-create training content. Mixed methods evaluation was conducted via pre-post participant surveys in all four trainings (n = 53). Participants who did not demonstrate baseline environmental science knowledge pre-training demonstrated significant knowledge increase post-training, and participants who demonstrated low self-efficacy (SE) pre-training demonstrated a significant increase in SE post-training. Participants overall demonstrated a significant increase in specific environmental health skills described post-training. The interdisciplinary facilitator-scientist team also reported multiple benefits, including learning local knowledge that informed further research, and building trust relationships with community members for future collaboration. We propose contextual EHL education as a valuable strategy for increasing EHL in environmental health risk communities, and for building academia-community partnerships for environmental health research and action.

Keywords: environmental health literacy; environmental health risk; informal education; contextual learning; rainwater harvesting; program evaluation; hazardous waste sites; environmental justice communities

1. Introduction

Communities that are disadvantaged by dominant social, economic, and political systems also often suffer disproportionately from environmental health risk [1–3]. Additionally, the increase of extreme and unpredictable weather events due to climate change is further widening this environmental health risk disparity [4–6]. Environmental justice (EJ) integrates the many potential layers of subjugation these communities face; for example, residents may suffer from discriminatory land use planning, limited access to health care, limited employment opportunities, and substandard sanitation infrastructure [2,7]. Though we acknowledge the four partnering communities in this study

Int. J. Environ. Res. Public Health 2018, 15, 2203; doi:10.3390/ijerph15102203 www.mdpi.com/journal/ijerph

Int. J. Environ. Res. Public Health 2018, 15, 2203 2 of 23

are EJ communities (Table 1), we use the term “environmental health risk communities” here to locate the specific study emphasis on environmental health, and to highlight environmental health risk as the focus of our partnership with these communities.

The natural interdependence of science literacy, health literacy, and environmental literacy have led to the evolution of current environmental health literacy (EHL) frameworks [8–13]. The Society for Public Health Education states that EHL “integrates concepts from both environmental literacy and health literacy to develop the wide range of skills and competencies that people need in order to seek out, comprehend, evaluate, and use environmental health information to make informed choices, reduce health risks, improve quality of life and protect the environment” [14]. As the mitigation or prevention of community-level environmental health risk often requires coordinated action, scholars have identified the need for EHL to not only include knowledge and efficacy for personal environmental health actions, but also for community level action [8,12,15]. Case studies have demonstrated increasing EHL as an effective strategy to equip environmental health risk community members to develop and implement their own contextually appropriate strategies for addressing environmental health risk [9,12,16–18]. Thus, increasing EHL in environmental health risk communities is a strategy for realizing environmental justice.

Aligning with Gray’s (2018) proposed refinement to the definition of EHL [8], illustrated in Figure 1, we define EHL here as: (1) environmental science knowledge and awareness related to the community’s specific environmental risks; (2) skills and self-efficacy for learning science, doing science, and environmental action; and, (3) community action for systemic change.

Figure 1. Three dimensions of environmental health literacy (EHL), as proposed by Gray [8].

Gray further suggests EHL as a “continuum” of learning that is often “tied to active engagement of participants throughout the research process, especially in communities directly impacted by environmental contamination” (p. 466). The contextual EHL trainings described and evaluated here represented both an educational event on their own, and a starting point for participants to continue learning alongside researchers through a citizen scientist program, if they chose to do so.

In a previous study, 74% of survey respondents reported that they get “some or a lot” of their science- and technology-related learning from “life experiences” [19]. The Contextual Model of Learning, as proposed by Falk and Dierking [20], asserts that, “learning can be conceptualized as a contextually driven effort to make meaning in order to survive and prosper within the world; an effort that is best viewed as a continuous, never-ending dialogue between the individual and his or her physical and sociocultural environment” [21]. Research has shown people learn more when

Int. J. Environ. Res. Public Health 2018, 15, 2203 3 of 23

content is grounded in local culture and context [22–24], and when participants’ knowledge, beliefs, and experiences are shared and honored in a participatory learning experience [16,17,25]. Previous EHL research has emphasized the contextual nature of defining and evaluating EHL, and the importance of matching learning content to specific local environmental health risk factors [17,26,27].

2. Materials and Methods

2.1. Contextual Training Design

The contextual EHL trainings discussed here were designed and promoted to residents of the four partnering communities as “Step 1” of participation in an environmental citizen science project titled Project Harvest. Project Harvest is a co-created citizen science project [28] aimed at assessing environmental contamination in harvested rainwater with environmental health risk communities, presently ongoing. The trainings discussed here served multiple purposes: (1) engage and educate environmental health risk community members around environmental health issues, and increase EHL in these communities, (2) provide an easy entry point for community members to volunteer as citizen scientists in Project Harvest, continuing to learn and contribute to research through local environmental monitoring, and (3) pilot the sampling methods, instructional materials, and survey methods, as formative evaluation for future Project Harvest participant materials.

In the initial project design phase of Project Harvest, principal investigators drafted the content for public community trainings to include climate change, energy conservation, air and water quality, microbiology of food and water, basic inorganic and organic chemistry, practical household actions for environmental health and environmental sustainability, and hands-on experiences that are related to collecting soil, water, and plant samples for environmental monitoring. These topics were chosen based on specific contamination issues in the partner communities, expressed concerns from community members through prior partnerships, and their importance to comprehension of relevant environmental health information.

2.2. Partner Community Selection

Four communities, located in Arizona, USA, were selected to host these trainings based on their proximity to a federal Superfund site and other toxic release site(s), expressed interest from community members or local organizations, and previously built relationships between researchers and community members. The term “community” is used here to define a group with shared local source(s) of environmental health risk, though each defined community varies in size and municipal designation. For two of these communities, Hayden/Winkelman and Globe/Miami, two adjacent municipalities were considered as one community as they share common sources of environmental health risk and associated environmental health concerns, as well as similar demographics. Although the Superfund site located in Tucson, which is a significantly larger city, poses greater health risk to residents in closer proximity, the Tucson training was not restricted from any local residents for ethical reasons, and thus the City of Tucson is defined as a community here. Table 1 describes these four communities and rationale for their partnership for this study, and Figure 2 shows geographic locations.

Int. J. Environ. Res. Public Health 2018, 15, 2203 4 of 23

Table 1. Partnering Communities Demographics, Environmental Health Risk, and Partnerships.

City/Town and County Population 1

Median Household Income 2

Predominant Races/Ethnicities

Represented 1

% Spanish- Speaking

Households 2

Sources of Environmental Health Risk Recognized by Community 5 Prior/Current Partnerships

Tucson, Pima 520,116 $37,973

White, not Hispanic/Latino:

47.2% Hispanic/Latino:

41.6

28.8

Tucson International Airport Area Superfund Site [29], where aircraft and electronics manufacturing, fire drill training, and an unlined landfill have contributed trichloroethylene (TCE) and other contaminants observed in soil, groundwater, and municipal water [30,31]

Prior collaboration with non-profit organization Sonoran Environmental Research Institute (SERI), a community participatory research institute with extensive experience working with low-income Tucsonans around environmental health issues

Hayden 3, Gila

662 $36,094 Hispanic/Latino

(of any race): 84.4%

61.8 ASARCO Hayden Plant Alternative Superfund Site, which includes the ASARCO smelter, concentrator, former Kennecott smelter and all associated tailings facilities [32] In 2016, ASARCO was involved in a $150 million settlement with the US Department of Justice and US Environmental Protection Agency for violations of the Clean Air Act [33]

National Institute of Environmental Health Sciences Superfund Research Program partnership School superintendent enthusiastic about gardening and rainwater harvesting served as local “champion” [34] to involve teachers and students in environmental health learning

Winkelman 3, Gila

353 $45,000 Hispanic/Latino

(of any race): 84.2%

55.6

Globe 4, Gila 7532 $42,557

White, not Hispanic/Latino:

55.3% Hispanic/Latino

(of any race): 36.8%

14.9 Active copper smelter, rod mill, and open pit mine in Miami [35] The Mountain View Mobile Home Estates in Globe, AZ, sits on the site of a former chrysotile asbestos mill. This site was on the Superfund Program’s National Priorities List (NPL) due to asbestos contamination of soil and groundwater until clean-up activities were completed in 1988 [36]

Gila County Cooperative Extension agent became a local “champion” for environmental health learning [34], and successfully spread the enthusiasm to the Globe-Miami community

Miami 4, Gila 1837 $36,298

Hispanic/Latino (of any race): 56%

White, not Hispanic/Latino:

40.6%

23.6

Dewey- Humboldt,

Yavapai 3894 $50,173

White, not Hispanic/Latino:

85.5% 5.2

Iron King Mine—Humboldt Smelter Superfund Site, which includes approximately four million cubic meters of mine tailings from legacy mine and smelter [37] 2012–2013 analyses of drinking water in local homes demonstrated arsenic above the US EPA drinking water standard (10 µg/L) [38]

Community members participated in past UA research projects: 1. Citizen science project Gardenroots related to soil contamination and backyard food gardens [11,39–41] 2. Biomonitoring project related to metal exposure in homes [42]

1 2010 US Census; 2 2012–2016 American Community Survey (ACS) 5 year estimate; 3 Although these neighboring municipalities joined as one community for this study, they are separated in Census data collection (Town of Hayden and Town of Winkelman); 4 Although these neighboring municipalities joined as one community for this study, they are separated in Census data collection (City of Globe and Town of Miami). 5 More information about sources of environmental health risk in these communities is available from the US EPA EJSCREEN tool, www.ejscreen.epa.gov/mapper/, and from the US EPA Toxic Release Inventory (TRI) Program database, https://www.epa.gov/toxics-release-inventory-tri-program.

Int. J. Environ. Res. Public Health 2018, 15, 2203 5 of 23

Figure 2. Partnering community locations in Arizona, USA.

2.3. Participant Recruitment

Recruitment for training participants was conducted primarily via: (1) information tables at community festivals, community group meetings, and/or town Council meetings; (2) follow-up mailings, telephone calls and emails to community members and local organizations; (3) press releases for local media outlets; (4) announcements in newsletters; and (5) the Project Harvest website. In every community, principal investigators participated in multiple events involving in-person engagement with community members. All promotional materials were produced in Spanish and English, to welcome and encourage Spanish-only speakers in addition to English speakers. See Supplemental Materials for examples of promotional flyers distributed. Table 2 details specific engagement methods that were conducted in each community.

Table 2. Training Participant Engagement Methods by Community.

Engagement Activity Tucson Hayden/Winkelman Globe/Miami Dewey-Humboldt

Press Releases X X X Local Newspaper X X X Town Newsletter X

Cooperative Extension X X X Master Gardeners X X X X

Federal Superfund Site Meetings X X X Community Advisory Boards X X X

School Superintendent & Teachers X X X SERI participants X

City of Tucson Water program participants X

2.4. Training Facilitation

Principal investigators of Project Harvest, representing research fields of Soil and Water Science, Microbiology, Organic Chemistry, Inorganic Chemistry, and Public Health, served as training facilitators, along with other specialists on climate change and rainwater harvesting. The term “facilitator-scientist” is used henceforth to refer to individuals in this role.

Community trainings were hosted in locally familiar public sites and scheduled either over 1–2 consecutive full days or over 3–5 consecutive mornings. See Supplemental Materials for the full agenda of each training. Tucson and Globe/Miami trainings were facilitated bilingually in English and

Int. J. Environ. Res. Public Health 2018, 15, 2203 6 of 23

Spanish, to accommodate Spanish-only speakers present. All of the participants in Dewey-Humboldt and Hayden/Winkelman trainings were English speakers. Four of the facilitator-scientists are Spanish-English bilingual and conducted their presentations and discussions in two languages. All other content was provided bilingually through:

1. Simultaneous translation during the lectures, with bilingual facilitator-scientists acting as translators.

2. Dual projection of slide presentations in English and Spanish (all slides were identical in content and presented and projected in English and Spanish).

3. Spanish language kit and manual during sampling hands-on education, and bilingual educators on site to provide one-on-one assistance.

4. English and Spanish take-away copies of all slides and supplemental education materials provided in a binder [43] (Supplemental Materials).

2.5. Research Methods

A 17-item survey consisting of multiple-choice and short answer questions and a 33-item Likert scale survey (see Supplemental Materials) were administered to all willing participants at the beginning and end of the training, to assess changes in participant EHL. Additionally, a pre-training survey to collect demographic information, and post-training survey to gain feedback on participants’ experience of the training, suggestions for future trainings, and current rainwater harvesting practice, were also administered. Participant feedback (both via survey and verbal) informed adjustments in training design and facilitation as the series of trainings progressed.

Of 67 total participants in the four trainings, 53 participants attended the entire training, completed both pre- and post-surveys, and consented under the University of Arizona IRB rule as an approved project. Of these, 14 participants were in Tucson, 15 were in Hayden/Winkelman, 14 were in Globe/Miami, and 10 were in Dewey-Humboldt. De-identified survey data included closed-ended question (Likert scale and multiple choice) responses that were recorded and analyzed quantitatively by one researcher and research assistant. Open-ended responses were reviewed and analyzed using validated qualitative methods by two research assistants and a supervising researcher. A coding scheme was created for each question that captured main concepts reflected in participant responses. The group met periodically throughout the coding process to revisit codebooks, sometimes choosing to create or merge categories based on emerging themes in the data. All survey data was organized in Microsoft Excel, and statistical analyses were performed while using Microsoft Excel and Statistical Package for Social Sciences (SPSS) software [44].

Table 3 outlines the specific data collected as it relates to each research question identified. The numerated list following Table 3 details specific analysis methods for each category of assessment.

Int. J. Environ. Res. Public Health 2018, 15, 2203 7 of 23

Table 3. Research goals and associated data analysis methods.

Research goal: Gain understanding about training participants’ (1) initial motivation to learn about environmental health, and (2) attitude towards the environment.

Assessment Category Data Survey Responses Coded for:

1. Motivation to learn Four short-answer questions (pre-survey only) Themes (qualitative)

2. Attitude towards the environment

Two multiple choice questions (pre- and post-survey)

Level of pro-environmental attitude (quantitative)

Research goal: Measure change in training participant EHL, as comprised of (3) environmental science knowledge, (4) skills and (5) motivation for environmental health action, (6) self-efficacy, and (7) community action for systemic change.

Assessment Category Data Survey Responses Coded for:

3. Environmental science knowledge

Four multiple choice questions, one matching question, one rank order question, seven short answer questions (pre- and post-survey)

Level of understanding (quantitative), and for themes in specific knowledge concepts (qualitative)

4. Skills for environmental health One multiple choice question, three short answer questions (pre- and post-survey)

Level of knowledge (quantitative), and for themes in specific knowledge concepts (qualitative)

5. Motivation for environmental action

Eleven Likert-scale items (pre- and post-survey) Level of motivation (quantitative)

6. Self-efficacy (SE)

Six Likert-scale items measure SE for learning science, four items measure SE for doing science, twelve items measure SE for environmental action (pre- and post-survey)

Level of self-efficacy (quantitative)

7. Community change

Two short answer questions (pre- and post-survey) one short answer question (post-survey only) facilitator-scientist survey responses

Themes of political advocacy, teaching others, meeting/talking/networking with others, or other collective strategies. 1

Research goal: Gain understanding of (8) training participants’ experiences in the training, and (9) facilitator-scientists’ experiences in the training.

Assessment Category Data Survey Responses Coded for:

8. Participant experience Three short answer questions (post-survey only) Themes (qualitative)

9. Facilitator-scientist experience Open-ended survey conducted via email with a subset of the facilitator-scientists

Themes (qualitative)

1 Volunteering as a citizen scientist for Project Harvest post-training considered as a collective strategy.

2.6. Data Analysis Methods

1. Motivation to learn was measured through four questions on the pre-program participant survey asking participants why they chose to attend, what they hoped to gain from the program, about their current rainwater harvesting practices, and what prior training on similar topics they may have received. Participant responses were coded for common themes, and occurrence of themes were aggregated to understand common motivations for choosing to attend the training.

2. Attitude towards the environment was measured by two closed-ended survey questions about preference towards environmental protection and environmental investment. Participant responses were aggregated and evaluated as a group, and pre-post change was evaluated per participant.

3. Environmental science knowledge was measured through seven open-ended and six closed-ended survey questions. Responses to four multiple choice questions were scored for level of knowledge. Responses to seven short answer questions were analyzed for both specific knowledge concepts, using qualitative coding for themes; and, for level of knowledge, by assigning a 0 (no knowledge), 1 (partial knowledge), 2 (baseline knowledge), or 3 (advanced knowledge) to each

Int. J. Environ. Res. Public Health 2018, 15, 2203 8 of 23

response. Coding rules for level of knowledge were specific to the question being analyzed. Table 4 describes the coding rules and example responses to one open-ended survey question. A dependent samples t-test was performed per question to assess significant change in participants’ mean level of environmental science knowledge for each topic area. Because participants who scored below baseline knowledge pre-training had more opportunity for learning gains, an additional dependent samples t-test was performed to look at knowledge change in those participants only. For this test, questions that could only be scored right or wrong were omitted.

Table 4. Level of knowledge coding rule and example responses to open-ended survey question.

Code 0—No Knowledge 1—Partial Knowledge 2—Baseline Knowledge

3—Advanced Knowledge

Coding Rule

Response is blank or reflects no

knowledge of key concepts.

Response suggests some correct knowledge of

topic but does not identify key concept.

Response describes key concept and is otherwise correct.

Response describes key concept with higher

complexity or details.

Example question: How is the use of energy derived from coal (electricity) and climate change related?

Example Response

“Climate Change is the Glaciers melting, Didn’t understand!”

“Dirty air and chemicals from burning coal”

“Increase in CO2 green house gases”

“Burning of fossil fuels is the main contributor of

rapidly increasing atmospheric CO2”

4. Skills for environmental health comprised a significant portion of training content. Three survey questions asked participants about specific actions they could take to (1) “curb the effects of climate change”, (2) “make a positive impact on water reliability in the future”, and (3) “protect the environment, conserve water, conserve energy, and protect the health of your family and neighbors”. Responses to these questions were coded for categories of environmental health action. The number of participants describing each skill category was compared pre- and post-training. Additionally, specific strategies described per participant pre- and post-training were averaged per participant and compared using a dependent samples t-test.

5. Motivation for environmental action was assessed by eleven Likert scale items. These items were modified from literature provided by the Cornell Lab of Ornithology [45], who use a similar measurement tool with their citizen science program participants. Five items were designed to measure external motivation for environmental action, and six items were designed to measure internal motivation. Following the recommended analysis methodology from Cornell, the sum of external motivation scores were subtracted from the sum of internal motivation item scores, to calculate an overall motivation score per participant. Positive scores indicate predominantly internal motivations, while negative scores indicate predominantly external motivations. A dependent samples t-test was used to analyze pre-post change in motivation for environmental action overall. A Wilcoxon Signed-Rank test was used to test for significance between the mean pre- and mean post-survey scores by community.