Does Values-Based Interpretation Make a Difference? Testing Impacts on Visitors’ Environmental Learning and Reported Adoption of Environmentally Responsible Behaviors

Abstract

This experimental field study tests the impact of values-based interpretive materials on zoo/aquarium visitors’ engagement, learning and self-reported adoption of environmentally responsible behaviors after their visit. Participants were randomly assigned to a control group (who experienced a normal visit) or a treatment group (who received additional values-based interpretive materials). A total of 719 participants in North America and 455 in Australia completed two surveys – one immediately after their visit and one four weeks later. Three hypotheses were supported by the data: Providing values-based interpretive materials influences visitors’ engagement during their visit; improves visitor learning; and increases visitors’ adoption of environmentally responsible behaviors. A fourth hypothesis was partially supported: Providing values-based interpretive materials is most beneficial for those who prioritize values other than self-transcendence. It is concluded that values-based interpretation can broaden the reach of zoos beyond traditional audiences and widen their impact on visitor uptake of environmental behavior.

Keywords:

• Environmental behavior
• environmental learning
• interpretation
• values
• wildlife tourism
• zoo visitors

Introduction

Modern zoos and aquariums place a high priority on the protection and conservation of animal species and their habitats. This includes providing interpretive experiences and activities that educate visitors about environmental issues affecting animal welfare and how visitors can help through adopting environmentally sustainable practices in their everyday lives (World Association of Zoos & Aquariums, 2015). Accordingly, mission statements of zoos and aquariums typically highlight the role they play in connecting visitors with the natural environment, enhancing visitors’ understanding of conservation issues and promoting the need to adopt sustainable environmental actions (Grajal, 2013; Luebke & Grajal, 2011; Patrick & Caplow, 2018). Underlying these mission statements is an implicit assumption that connection and inspiration prompt environmental action. While there is little doubt that zoos and aquariums are well-placed to communicate and demonstrate the importance of people adopting conservation actions (Ballantyne et al., 2021; Smith et al., 2008), there is doubt regarding whether visitation really changes the everyday environmental behavior of their visitors.

Not surprisingly, zoos and aquariums are under increasing pressure to prove their programs enhance their visitors’ environmental learning and adoption of environmentally sustainable behaviors (Malamud et al., 2010; Milstein, 2009). Scholars exploring visitor environmental learning in wildlife tourism settings have examined the impact of factors ranging from pre-visit motives and perceptions to on-site experiences and post-visit support (Ballantyne & Packer, 2011). Many of these studies focus on the role of interpretation in prompting visitors to appreciate and understand the connections between environmental issues and humans’ everyday actions. This paper reports on a study designed to examine whether providing on-site and take-home interpretation materials designed to match visitors’ values enhances the likelihood of them adopting environmentally sustainable actions after their visit.

Environmental interpretation in zoos and aquariums

The term interpretation refers to the stories, anecdotes, information and insights zoos and aquariums provide to their visitors about animals, habitats and conservation issues. Delivered via presentations, signs, shows and activities, interpretation aims to explain why particular topics, objects and events are important. As Ham (2013) explains, interpretation answers the “so what?” question. Interpretation differs from formal education in that it is tailored to suit the needs of a non-captive audience; in other words, the audience is free to choose whether or not to listen, read or interact with educational material. For these reasons, interpretation needs to be visitor-centred—it needs to consider the interests, knowledge, experiences, motivations and beliefs of target audiences and present information in a way that engages and inspires (Ballantyne et al., 2007; Hughes & Ballantyne, 2013).

There are four widely-accepted principles of interpretation—it is engaging, it is relevant to the target audience, it is organized, and it is based on an underlying theme or take-home message (Ham, 1992; Tilden, 1977). Early research assessing the role of interpretation in environmental settings demonstrated that interpretation enhances visitor experiences and contributes to visitor satisfaction (Ham & Weiler, 2007; Moscardo, 1998). It also increases visitors’ environmental knowledge (Lackey & Ham, 2003; Powell & Ham, 2008; Smith et al., 2008) and has been shown to influence visitors’ environmental behavior at a site (Brown et al., 2010; Kim et al., 2011; Marion & Reid, 2007). But do, as assumed, interpretive wildlife experiences persuade visitors to incorporate conservation actions into their everyday lives?

Studies in a variety of captive and non-captive contexts demonstrate that provision of interpretation certainly increases visitors’ intentions to adopt conservation actions. Such intentions range from picking up litter, recycling and reducing use of nonrenewable resources to donating time, money and effort to environmental causes (Hughes, 2013; Mayes et al., 2004; Smith et al., 2008; Walker & Moscardo, 2014). While it’s tempting to view this as an indicator of success, research exploring long-term behavior change indicates that for many visitors, these intentions are short-lived and do not usually translate into the adoption of actual environmental behaviors. This drop-off or gap between intentions and behavior has been reported in numerous nature-based and wildlife tourism contexts (Adelman et al., 2001; Ballantyne & Packer, 2011; Dierking et al., 2004; Hughes, 2013; Luebke & Grajal, 2011; Smith et al., 2008; Wheaton et al., 2016).

The news is not all negative, however, with recent longitudinal research suggesting there are certain elements that can be incorporated into on-site interpretation to close the gap between intentions and behavior. For example, researchers have highlighted the importance of prompting visitors to reflect during their visit (Ballantyne et al., 2011; Packer & Ballantyne, 2013; Walker & Moscardo, 2014). This can be achieved by encouraging visitors to contemplate the threats animals face, consider the impacts of their own day-to-day actions, and discuss their insights with others. Another important precursor of behavior change is emotional engagement. It seems that facilitating an emotional connection between visitors and wildlife helps to prompt a reflective response as well as subsequent action (Ballantyne et al., 2011; Ballantyne & Packer, 2011; Meyers et al., 2004).

The impact of post-visit support on long-term environmental behavior has also been examined. Ballantyne (2007) argued that the intention-behavior gap could be reduced by providing visitors with learning materials after their visit. Such materials, which he called post-visit action resources, could be designed to build on and extend on-site learning and to motivate and support visitors in the translation of their behavioral intentions into actions. Several researchers have tested this proposition and have reported promising results. Hughes (2011) found that post-visit action resources accompanied by email reminders and site updates encouraged families to adopt a range of environmentally-friendly behaviors. Similarly, Hofman and Hughes’ (2018) experimental study revealed post-visit support in the form of Facebook posts and activity sheets increased the post-viewing impact of wildlife documentaries. Bueddefeld and Van Winkle (2018) found that a treatment group who received post-visit action resources in the form of handouts and follow-up emails demonstrated significantly greater depth and mastery of learning (in relation to climate change and sustainability) than a control group. Ballantyne et al. (2018) designed and tested online action resources with zoo visitors and concluded this approach has potential to promote behavior change if suggestions for action are specific and manageable. Scholars also recommend that suggestions are framed around local issues and environments and that visitors be given a choice of relevant and achievable behaviors (Ballantyne & Packer, 2005, 2011; Grajal et al., 2018; Smith et al., 2012).

In sum, research to date suggests that the factors influencing zoo/aquarium visitors to take environmental action are many and varied, and that inspiring long-term behavioral change is challenging (Learmonth et al., 2021). At the core, however, is the notion that effective interpretation places the visitor rather than animals and conservation at the heart of the wildlife tourism experience, i.e., it designs experiences in light of visitors’ perspectives and priorities. Interpretation in zoos and aquariums tends to follow a “one size fits all” approach. But if visitors’ perspectives and preferences are at the heart of visitor-centred interpretation, perhaps an individualized approach that takes into account visitors’ value preferences might be more engaging and impactful (Ballantyne et al., 2021).

Values-based interpretation

Appealing to visitors’ predominant values is one way to individualize interpretation. Using Schwartz’s (1992) personal values model, Ballantyne et al. (2018, 2021) conducted a series of studies to explore the relationships between visitors’ values, their zoo/aquarium experiences, and their learning outcomes. Personal values are broad life goals that serve as guiding principles in major life decisions and in everyday behaviors (Schwartz, 1992). Schwartz (1992) identified 10 basic values that form a circular structure, which can be summarized with two bipolar dimensions: self-transcendence vs self-enhancement on one dimension and openness to change vs conservation on the other. Those who prioritize self-transcendence values are motivated by concern for the welfare of others and the natural environment, while those who prioritize self-enhancement values are motivated by self-interest, personal success and dominance. Those who prioritize openness to change values are motivated by independence, novelty and excitement, while those who prioritize conservation values are motivated by social expectations, predictability and maintenance of the status quo.

Visitors prioritizing different values orientations have been found to have distinctive perceptions and preferences regarding zoo/aquarium experiences and interpretive messages (Ballantyne et al., 2021). However, because interpretive materials are most likely to be designed by staff who prioritize self-transcendence values (Packer et al., 2022) they may be failing to engage those who prioritize other values. Indeed, Ballantyne et al., 2018) found that visitors who prioritized self-transcendence values were more likely to report reflective engagement, learning and behavior change outcomes than those who prioritized other values.

These studies resulted in the development of a values-based interpretive matrix (Ballantyne et al., 2021) that identifies the interpretive elements most likely to appeal to, and influence the environmental behavior of, visitors with each of the four higher order values orientations (self-transcendence, self-enhancement, openness to change and conservation, as described above). The matrix supports a more nuanced approach to interpretation that allows every values group to be accommodated. The present paper uses the matrix to create values-based interpretive materials for zoo/aquarium visitors and tests their impact on visitors’ engagement, environmental learning and self-reported adoption of long-term environmentally sustainable behavior.

Accordingly, the aim of this study is to test the following hypotheses:

• H1 Providing values-based interpretive materials influences visitors’ engagement during their visit;

• H2 Providing values-based interpretive materials improves visitors’ environmental learning;

• H3 Providing values-based interpretive materials increases visitors’ reported adoption of environmentally responsible behaviors as a result of their visit;

• H4 Providing values-based interpretive materials is most beneficial for those who prioritize values other than self-transcendence.

Method

Twelve institutions participated in this research (nine zoos in Australia and USA; three aquariums in USA and Canada). Because different recruiting and values classification procedures were used in Australia, the results for Australia and North America are reported separately, thus allowing the replicability of results to be examined. All data were collected between June 2018 and March 2019.

Each of the 12 sites selected a target exhibit that already included conservation-related interpretation in the form of information panels, video presentations, interactive displays, and/or keeper talks. A set of four supplementary interpretive materials was prepared for each target exhibit, each designed to appeal to one of the four higher-order values—self-transcendence (ST); self-enhancement (SE); openness to change (OC) and conservation values (CO). The design of the materials was informed by the values-based interpretation matrix, which in turn was derived from values-based focus groups (Ballantyne et al., 2021). The materials were presented as a laminated A3 size sheet, with an accompanying take-home postcard listing seven suggestions for actions participants could take to “help the [target species] and other wildlife”. These suggestions were the same across all values groups and target species, but the title banner and concluding remark were designed to be value-expressive. For example, the CO group was asked to “help the zoo make a difference” and encouraged to “work together to help the environment” whereas the OC group was asked “Are you prepared to take up the challenge?” and encouraged to “discover new ways to make a difference”. Examples of the materials designed for an Australian native animal exhibit are provided in the Supplementary Online Materials (Figures S1-S4).

Participants and procedure: North America

In North America, visitors were approached as they entered the target exhibit and invited to participate in the study. Those who agreed (n = 1501 across the 8 sites) were asked to complete a brief “values choice” task by choosing one of four “persons” that they considered to be most like them. Each “person” contained two or three descriptor items drawn from the Schwartz PVQ40 values instrument (see Schwartz, 2021). Each of the four options was presented in a positive light, highlighting the strengths of each particular orientation. Once they had completed this task, participants were alternately assigned to either a treatment group (who received the values-based interpretive materials) or control group (who experienced a normal zoo visit), such that the numbers of treatment vs control participants were approximately equal within each values group. The numbers in each values group varied naturally according to participants’ responses to the choice task. The initial number of participants in each group is reported in Table 1. The largest group was the self-transcendence (ST) group (40% of the sample) and the smallest was the self-enhancement (SE) group (12% of the sample).

Table 1. Number of participants in treatment (T) and control (C) groups: North America.

	Self-transcendence		Self-enhancement		Openness to change		Conservation
	T	C	T	C	T	C	T	C	Total
Recruited	310	296	97	84	213	196	155	150	1501
Completed Survey 1	296	259	97	76	213	183	152	131	1407 (94%)
Completed Survey 2	152	150	46	32	107	97	74	61	719 (48%)

Participants were instructed to explore the target exhibit as they would in a normal visit. Treatment group participants were instructed to use the sheet they had been given to consider further ideas about animal conservation. Both groups were asked to return to the data collectors when they had finished at the exhibit to complete a short questionnaire (Survey 1). On completion of the survey, participants were given a small thank-you gift and the relevant take-home postcard (value-expressive action suggestions for the treatment group; a thank you postcard listing participating zoos for the control group). Overall, 94% of those recruited completed Survey 1 (Table 1).

Invitations to complete an online follow-up questionnaire (Survey 2) were emailed approximately four weeks after the visit. Two reminder emails were sent, one week apart, to those who had not yet responded. All participants who completed both questionnaires were entered into a draw to win one of five $1000 prizes. A total of 719 participants (48% of those initially recruited) completed Survey 2 (Table 1). Again, the largest group completing both surveys was the self-transcendence (ST) group (42% of the sample) and the smallest was the self-enhancement (SE) group (11% of the sample).

Participants and Procedure: Australia

In Australia, participants were recruited through a commercial panel company and were offered free tickets to visit the zoo, as well as the standard remuneration for completing the surveys. As values data were already available for a number of panel members (based on Lee et al.’s 2019, Best-Worst Refined Values scale), we specifically targeted those in the smaller values groups (i.e., SE and CO). Participant numbers were supplemented by a general recruitment survey of panel members, using the same values choice task used in North America. A total of 1028 participants who agreed to visit the zoo and complete the associated two surveys were recruited (see Table 2). As these participants were targeted according to their value priorities, the proportion of the sample in each values group was roughly equal (ranging from 23% in the OC group to 27% in the SE group). Participants were randomly assigned to the treatment or control group.

Table 2. Number of participants in treatment (T) and control (C) groups: Australia.

	Self-transcendence		Self-enhancement		Openness to change		Conservation
	T	C	T	C	T	C	T	C	Total
Recruited	125	124	139	138	120	119	132	131	1028
Completed Survey 1	71	71	65	47	74	64	46	75	513 (50%)
Completed Survey 2	63	61	59	43	69	51	43	66	455 (44%)

On arrival at the zoo, participants were given instructions according to the group to which they had been allocated. The treatment group was given supplementary interpretive materials that matched their values priorities. Zoo entry staff were provided with color-coded participant lists to ensure that each participant was given the correct materials and instructions. All participants were asked to explore the target exhibit as they would in a normal visit. Treatment group participants were instructed to use the sheet they had been given to consider further ideas about animal conservation. Participants were given a password to access Survey 1 online and were asked to complete this as soon as possible after visiting the exhibit. As in North America, those in the treatment group were given the relevant value-expressive postcards and those in the control group were given a generic thank you version.

A total of 513 participants (50% of those recruited) visited the zoo and completed Survey 1 (Table 2). Those who completed Survey 1 were invited to complete Survey 2 approximately four weeks after the visit. Two reminder emails were sent, one week apart, to those who had not yet responded. A total of 455 participants (44% of those initially recruited) completed Survey 2 (Table 2).

Measuring visitors’ engagement

Measurement of visitors’ engagement during their visit (Survey 1, on-site) was based on Ballantyne et al.’s (2011) experiential and reflective engagement scales. Ballantyne et al.’s (2011) research demonstrated the importance of reflective engagement, which involves both cognitive and affective processing of the experience, as a precursor of learning and subsequent adoption of environmentally responsible behaviors. Ballantyne et al. (2018) further developed the reflective engagement items, and again found the construct useful in predicting learning and behavioral outcomes in the context of visitors’ personal values.

In the present study, the measurement of visitors’ engagement was further refined by distinguishing reflective engagement from emotional connection. Sixteen items (6 experiential engagement; 6 previously used reflective engagement items; and 4 new items designed to measure emotional connection, empathy and perceived interaction with the animals in the exhibit) were each rated on a 7-point scale. Using the North American data (n = 1402), the 16 items were entered into a Principal Axis Factor Analysis with Varimax rotation. The Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO = .916) and Bartlett’s Test of Sphericity (χ², 120 df = 14658.673, p < .001) confirmed that factor analysis was appropriate on these items.

Three factors with Eigenvalues greater than 1 were extracted which together accounted for 68.6% of the common variance. These factors were interpreted as Experiential Engagement, Reflective Engagement and Emotional Connection; however, in this solution, two items cross-loaded on both Reflective Engagement and Emotional Connection. A four-factor solution (with the 4th factor having an Eigenvalue of .908) provided a clearer structure. The four factors (accounting for 74.3% of the variance) were interpreted as Experiential Engagement (the original 6 items), Reflective Engagement (4 items), Emotional Connection (3 items) and Animal Interactions (3 items). Repeating the factor analysis using the Australian data (n = 513) confirmed that the four-factor structure was more stable in terms of item loadings. Four composite variables were created: Experiential Engagement, Reflective Engagement, Emotional Connection, and Animal Interaction. Based on previous research, only Reflective Engagement was expected to be related to visitor learning and subsequent adoption of environmentally responsible behaviors. Descriptive statistics, factor loadings and Cronbach alphas are reported in the Supplementary Online Materials, Table S1.

Measuring visitors’ learning

Visitor learning was measured using three variables—two on-site (Survey 1), and one at follow-up (Survey 2):

• Survey 1 rating scale: a composite of four items from Ballantyne et al. (2011) each rated on a 1-7 scale (Descriptive statistics and Cronbach alphas are reported in the Supplementary Online Materials, Table S2);

• Responses to the Survey 1 question “What, if anything, did you learn during your visit about the animals in the exhibit, or animals in the wild?” were coded as including or not including conservation issues;

• Responses to the Survey 2 question “Did you learn any new information or gain any new insights or understanding during your time at [name of exhibit]? Please write below as many things as you can remember” were coded as including or not including conservation issues.

Open-ended responses (2nd and 3rd variables above) were coded independently by two research assistants, who were blind to both the condition (treatment or control) and values group assignment of the respondents. Responses that were not consistently coded by the two assistants were examined and adjudicated by one of the authors.

Measuring reported adoption of environmentally responsible behaviors

Demonstrating long-term behavior change as a result of a zoo or aquarium visit is a major challenge and invariably relies on self-report (Esson & Moss, 2014). In the present study, a range of approaches was used in an attempt to detect the impact of the intervention on visitors’ progression toward the adoption of more environmentally responsible behaviors. These included both structured and unstructured responses and addressed changes in awareness, knowledge and attitudes (recognized as antecedents to environmental action), as well as the adoption of new behaviors and strengthening of existing behaviors.

Survey 1 items:

• Six items measured change in antecedents to environmental action, based on a progression of steps including conscious awareness of need, knowledge of viable actions, feeling a responsibility to become involved, attitudes toward action and plans to take action (Schwartz, 2010). These items were measured using a retrospective pre-post format (i.e., participants were asked to rate their position before and after the visit, but both ratings were completed at a single point in time, after the visit). The difference between the two ratings was taken as a measure of perceived change due to the visit. Descriptive statistics are reported in the Supplementary Online Materials, Table S3.

• One item measured behavioral intentions: “As a result of visiting this exhibit today, how likely is it that you will do something new, differently, or more often to help wildlife conservation?”. Five response options were offered ranging from “I am not likely to do anything new or different” through to “I am determined to do something”. A sixth option “I am already doing the maximum possible” was also provided. Only 5% of respondents selected this option, and these were excluded from the analysis of this item. Descriptive statistics are reported in the Supplementary Online Materials, Table S4.

Survey 2 items:

• Seven items measured the perceived impact of the visit on environmental knowledge, attitudes and action, each rated using a slider scale from 0 to 100. These items paralleled the six antecedents to action in Survey 1, but were asked in a perceived change format, e.g., “I am now more aware of the problems facing animals and the environment”. The seventh item was “I have already taken some actions to help animals and the environment”. Item scores ranged from 0-100, with means ranging from 57-67 in North America and 60-69 in Australia. Descriptive statistics are reported in the Supplementary Online Materials, Table S5.

• Participants were asked “Since your visit, have you done anything new to help animals or the environment? This might include doing something new, doing something you already do more frequently, or stopping doing something harmful.” Those who responded “yes” were asked to describe what they had done. Responses were coded as verified behavior change (the respondent’s “yes” response was justified as an action that was likely to benefit either species conservation or the environment more broadly) vs no behavior change (either the participant responded “no” or their “yes” response was not verified or too vague).

• Verified behavior changes were further classified as a minor change (those that would either save money or be cost neutral, or involved continuing or strengthening existing behaviors) vs major change (new behaviors that might cost money or involve some sacrifice or effort). For analysis purposes, no change was combined with minor change in order to create a binary (yes/no) measure of level of behavioral change.

Again, open-ended responses (2nd and 3rd variables above) were coded by two research assistants and any discrepancies were adjudicated by one of the authors.

Data analysis

Differences between the treatment group and control group were analyzed using t-tests for the continuous variables (four dimensions of engagement, the rating scale measure of visitor learning, antecedents to environmental action, behavioral intentions, and the follow-up measures of perceived impact), and chi-square tests for the binary measures (coding of on-site and follow-up open-ended responses regarding learning and behavior change). These analyses were conducted separately for the North American and Australian data in order to examine the extent to which the findings were generalizable across contexts. A significance level of p ≤ .05 (one-sided) was set for hypothesis testing.

A selection of the above analyses was repeated separately for each values group in order to explore whether the provision of values-based interpretive materials is more (or less) important for particular values groups. Five key dependent variables were included in these analyses: reflective engagement, on-site learning (rating scale measure); behavioral intentions; verified behavior change (binary measure); and the impact item “I have already taken some actions to help animals and the environment”. These analyses were conducted on the combined dataset (North America and Australia) in order to maximize the sample size for each values group. For these analyses, the effect sizes in relation to the difference between treatment and control groups are examined across the four values groups. (Unlike significance levels, effect size is independent of sample size.)

Results

H1 providing Values-Based interpretive materials influences visitors’ engagement during their visit

Of the four factors measuring participants’ on-site engagement with the target exhibit (Experiential Engagement, Reflective Engagement, Emotional Connection and Animal Interactions) only Reflective Engagement was expected to be related to visitor learning and subsequent adoption of environmentally responsible behaviors. Regression analyses confirmed that Reflective Engagement was the strongest and most reliable predictor of all on-site and follow-up measures of visitor learning and behavior change. The other measures of engagement added little to the prediction of learning and behavior change outcomes but were all themselves significant predictors of Reflective Engagement.

In North America, there were significant differences between the treatment and control groups on three of the four engagement measures, with the largest effect being on Reflective Engagement. In Australia, there were significant differences on all four engagement measures, with the largest effect being on Reflective Engagement. In all cases, the treatment group scored higher than the control group (Table 3). Thus, H1 is supported, especially in relation to the key variable Reflective Engagement.

Table 3. Comparing treatment (T) and control (C) groups on visitors’ engagement.

	Means (SDs)	Test statistics^a	Effect size^b
North America
Reflective Engagement	T 5.39 (1.40) C 4.99 (1.61)	t(1395) = 4.909, p < .001	0.27
Experiential Engagement	T 5.91 (1.13) C 5.72 (1.28)	t(1398) = 2.813, p = .005	0.15
Emotional Connection	T 5.19 (1.47) C 4.99 (1.52)	t(1393) = 2.412, p = .016	0.13
Animal Interaction	T 4.01 (1.81) C 3.86 (1.83)	t(1397) = 1.581, p = .114	NS
Australia
Reflective Engagement	T 5.15 (1.42) C 4.67 (1.54)	t(511) = 3.622, p < .001	0.32
Experiential Engagement	T 5.45 (1.36) C 5.21 (1.40)	t(511) = 1.958, p = .051	0.17
Emotional Connection	T 5.04 (1.40) C 4.82 (1.54)	t(511) = 1.699, p = .090	0.15
Animal Interaction	T 3.62 (1.85) C 3.34 (1.78)	t(511) = 1.730, p = .084	0.15

¹ Two-sided probabilities are reported, with p<.10 (p<.05 one-sided) considered a significant difference provided the difference is in the expected direction (T > C).

² Effect size was measured using Cohen’s d: 0.2 is considered a small effect, 0.5 moderate and 0.8 large.

H2 providing Values-Based interpretive materials improves visitor learning

In both North America and Australia, there were significant differences between the treatment and control groups on the rating scale measure of on-site learning, and the follow-up binary measure. In all cases, the treatment group scored higher than the control group (Table 4). Thus, H2 is confirmed.

Table 4. Comparing treatment (T) and control (C) groups on conservation learning.

	Means (SDs)/% conservation	Test statistics^a	Effect size^b
North America
On-site rating scale measure	T 5.59 (1.27) C 5.41 (1.35)	t(1396) = 2.540, p = .011	0.14
On-site binary measure	T 32.8% (of 758) C 30.7% (of 649)	χ^2(1,N=1407) = 0.771, p = .390	NS
Follow-up binary measure	T 48.8% (of 379) C 42.6% (of 340)	χ^2(1,N=719) = 2.744, p = .098	0.06
Australia
On-site-rating scale measure	T 5.35 (1.26) C 5.01 (1.41)	t(511) = 2.853, p = .005	0.25
On-site binary measure	T 57.0% (of 256) C 50.2% (of 257)	χ^2(1,N=513) = 2.410, p = .133	NS
Follow-up binary measure	T 57.3% (of 234) C 43.0% (of 221)	χ^2(1,N=455) = 9.269, p = .003	0.14

^a Two-sided probabilities are reported, with p<.10 (p<.05 one-sided) considered a significant difference provided the difference is in the expected direction (T > C).

^b Effect size for t-test was measured using Cohen’s d: 0.2 is considered a small effect, 0.5 moderate and 0.8 large; effect size for chi-square test was measured using Phi φ: 0.1 is considered a small effect, 0.3 medium and 0.5 large.

H3 providing Values-Based interpretive materials increases visitors’ reported adoption of environmentally responsible behaviors as a result of their visit

Significant differences were found between the treatment and control groups on five of the seven on-site variables in North America (see Table 5): four of the antecedents to action, and the behavioral intention item. Of these, the greatest effect size was found in relation to behavioral intention. In Australia, significant differences were found on all seven items, and again, the greatest effect size was found on behavioral intention (Table 5).

Table 5. Comparing treatment (T) and control (C) groups on onsite measures relating to environmentally responsible behavior.

	Means (SDs)/% reported	Test statistics^a	Effect size^b
North America
On-site changes in antecedents
• More aware of the problems	T .40 (1.00); C .32 (0.97)	t(1385) = 1.442, p = .150	NS
• Know more about actions	T .69 (1.12); C .54 (1.16)	t(1388) = 2.525, p = .012	0.14
• Convinced things will get worse	T .41 (1.01); C .34 (0.85)	t(1383) = 1.354, p = .176	NS
• Convinced of own responsibility	T .50 (0.95); C .37 (0.90)	t(1392) = 2.624, p = .009	0.14
• Positive attitude to taking action	T .41 (0.79); C .26 (0.71)	t(1392) = 3.742, p < .001	0.20
• Plan to take more action	T .77 (1.05); C .63 (1.02)	t(1389) = 2.437, p = .015	0.13
On-site behavioral intention	T 3.89 (1.11); C 3.52 (1.24)	t(1319) = 5.599, p < .001	0.31
Australia
On-site changes in antecedents
• More aware of the problems	T .73 (1.18); C .40 (1.09)	t(511) = 3.246, p = .001	0.29
• Know more about actions	T 1.12 (1.37); C .78 1.20)	t(511) = 2.985, p = .003	0.26
• Convinced things will get worse	T .77 (1.11); C .45 (0.94)	t(511) = 3.582, p < .001	0.32
• Convinced of own responsibility	T .82 (1.00); C .54 (0.89)	t(511) = 3.349, p = .001	0.30
• Positive attitude to taking action	T .57 (0.87); C .37 (0.76)	t(511) = 2.774, p = .006	0.25
• Plan to take more action	T 1.13 (1.23); C .79 (1.02)	t(511) = 3.426, p = .001	0.30
On-site behavioral intention	T 3.71 (1.19); C 3.07 (1.29)	t(476) = 5.619, p < .001	0.51

^a Two-sided probabilities are reported, with p<.10 (p<.05 one-sided) considered a significant difference provided the difference is in the expected direction (T > C).

^b Effect size for t-test was measured using Cohen’s d: 0.2 is considered a small effect, 0.5 moderate and 0.8 large.

Significant differences were found between the treatment and control groups on three of the nine follow-up variables in North America (see Table 6): slider scale items relating to the perceived impact of the visit on knowing what actions to take, planning to take more action, and already having taken action. Of these, the greatest effect size was found in relation to the item “I have already taken some actions to help animals and the environment”. In Australia, significant differences were found on eight of the nine variables. Here, the largest effect size related to the item “I now know more about the actions I can take to help”. H3 is thus partially supported, with effects being found on some of the follow-up measures in North America and most of the measures in Australia. Of note is that in both samples, there was a significant difference between the treatment and control groups in relation to the item “I have already taken some actions to help animals and the environment”.

Table 6. Comparing treatment (T) and control (C) groups on follow-up measures relating to environmentally responsible behavior.

	Means (SDs)/% reported	Test statistics^a	Effect size^b
North America
Follow-up binary behavior change	T 67% (/377); C 62% (/338)	χ^2(1,N=715) = 1.919, p = .182	NS
Follow-up level of behavior change (% major reported)	T 24% (/377); C 20% (/338)	χ^2(1,N=715) = 1.459, p = .241	NS
Follow-up impact
• More aware of the problems	T 63 (26.4); C 60 (27.4)	t(704) = 1.553, p = .121	NS
• Know more about actions	T 60 (27.8); C 54 (28.5)	t(704) = 2.715, p = .007	0.21
• Convinced things will get worse	T 65 (29.9); C 63 (30.6)	t(704) = 1.281, p = .201	NS
• Convinced of own responsibility	T 65 (29.0); C 63 (30.5)	t(704) = 0.931, p = .352	NS
• Positive attitude to taking action	T 68 (28.2); C 66 (30.1)	t(704) = 0.830, p = .407	NS
• Plan to take more action	T 68 (28.2); C 64 (30.6)	t(704) = 2.143, p = .032	0.16
• Already taken some actions	T 70 (28.2); C 62 (31.6)	t(704) = 3.514, p < .001	0.27
Australia
Follow-up binary behavior change	T 57% (/234); C 39% (220)	χ^2(1,N=454) = 15.760, p < .001	0.18
Follow-up level of behavior change (% major reported)	T 20% (/234); C 19% (/220)	χ^2(1,N=454) = 0.023, p = .906	NS
Follow-up impact
• More aware of the problems	T 68 (23.3); C 62 (26.0)	t(453) = 2.855, p = .005	0.27
• Know more about actions	T 65 (23.4); C 55 (26.2)	t(453) = 4.409, p < .001	0.42
• Convinced things will get worse	T 70 (25.8); C 65 (25.4)	t(453) = 2.036, p = .042	0.19
• Convinced of own responsibility	T 69 (25.9); C 61 (25.9)	t(453) = 3.166, p = .002	0.30
• Positive attitude to taking action	T 71 (24.5); C 64 (24.9)	t(453) = 2.810, p = .005	0.26
• Plan to take more action	T 68 (26.1); C 61 (26.6)	t(453) = 2.827, p = .005	0.27
• Already taken some actions	T 67 (27.9); C 56 (30.1)	t(453) = 4.024, p < .001	0.38

^a Two-sided probabilities are reported, with p<.10 (p<.05 one-sided) considered a significant difference provided the difference is in the expected direction (T > C).

^b Effect size for t-test was measured using Cohen’s d: 0.2 is considered a small effect, 0.5 moderate and 0.8 large; effect size for chi-square test was measured using Phi φ: 0.1 is considered a small effect, 0.3 medium and 0.5 large.

H4 providing values-based interpretive materials is most beneficial for those who prioritize values other than self-transcendence

Previous research has found that zoo and aquarium visitors who prioritize self-transcendence values are more likely to report reflective engagement, learning and behavior change outcomes than those who prioritize other values (Ballantyne et al., 2018). One possible explanation for this is that existing interpretation is largely designed by staff who prioritize self-transcendence values (Packer et al., 2022). It was thus hypothesized that the intervention in the present study would be more effective for groups other than the self-transcendence group.

In order to test this, five of the key dependent variables reported above were re-examined separately for each values group using combined North American and Australian data. The effect size for the difference between the Treatment and Control groups is used as an indicator of the extent to which the values-based interpretive materials made a difference within each values group (Table 7). The hypothesis that the least impact would be found for the self-transcendence group was supported on two of the five dependent variables: reflective engagement and follow-up verified behavior change (binary measure). Although the hypothesis was not supported on the other three measures (on-site learning, on-site behavioral intention, and actions already taken at follow-up), some interesting differences between values groups were observed on these variables.

Table 7. Effect of values-based interpretive materials for each values group.

	Self-enhancement	Conservation	Openness to change	Self-trancendence
Reflective engagement (rating scale)	.33	.35	.31	.22
On-site learning (rating scale)	.24	.09 (NS)	.18	.21
On-site behavioral intention (rating scale)	.55	.30	.34	.39
Follow-up verified behavior change (binary)	.20	.09 (NS)	.16	.03 (NS)
Have already taken some actions (slider scale)	.23 (NS)	.22	.41	.33
Average effect size	0.31	0.21	0.28	0.24

Note. Effect size for t-test (rating and slider scales) was measured using Cohen’s d: 0.2 is considered a small effect, 0.5 moderate and 0.8 large; effect size for chi-square test (binary variable) was measured using Phi φ: 0.1 is considered a small effect, 0.3 medium and 0.5 large.

On three of the five dependent variables, the greatest impact of the materials was observed in the self-enhancement group. This value priority has a motivational conflict with self-transcendence, which is typically prioritized by zoo and aquarium staff who are responsible for the design of interpretive materials and messages (Packer et al., 2022.) Of particular note for this group is the effect on behavioral intentions and the binary measure of self-reported behavior change. It is possible that the lower effect size on the third behavior change variable “I have already taken some actions to help animals and the environment”, may have been influenced by the phrase “to help animals and the environment” as this group is more likely to be motivated by self-interest, e.g., cost-saving actions.

Also of note is that on the variable “I have already taken some actions to help animals and the environment”, the greatest impact was observed in the openness to change group. As this group is by definition the most open to change, it is perhaps the group most likely to respond positively to efforts designed specifically to encourage them to make changes to their everyday behaviors and lifestyles.

Summary of findings

Compared with a control group who experienced a normal zoo visit, participants who received values-based interpretive materials:

• reported a higher level of both reflective engagement and learning during their visit;

• reported a higher level of behavioral intention in relation to helping wildlife conservation immediately after their visit;

• were more likely to mention conservation aspects when asked what they remembered four weeks later; and

• were more likely to report the visit had influenced them to take action to help animals or the environment after their visit.

All of the above findings were confirmed in two separate datasets (North America and Australia). The values-based interpretive materials were found to be effective in facilitating reflective engagement, learning and behavior change for all four values groups with the self-enhancement group benefiting most.

Discussion

Facilitating and demonstrating the positive effects of wildlife interpretation on short and long-term conservation learning is a complex and challenging task. Over the years, studies have shown that modifications to exhibit design, visitor activities and conservation messaging have enhanced the visitor experience (Ballantyne et al., 2007) and visitors’ awareness of conservation issues (Ballantyne et al., 2011), but have produced limited demonstrable effect on visitors’ long-term adoption of environmental actions (Ballantyne & Packer, 2016). A “one size fits all” approach to the design and delivery of interpretation may be partially to blame. It has been argued, for example, that environmental interpretation in wildlife tourism attractions such as zoos and aquariums is often designed by staff members who prioritize values that emphasize the protection of animals and the natural environment (Packer et al., 2022). Although such interpretation is highly engaging for like-minded visitors holding similar personal values, it does not resonate as well with those who prioritize different values.

This current study demonstrates that values-based interpretation designed to appeal to visitors’ idiosyncratic values priorities not only enhances their environmental learning, but also increases the likelihood of them adopting conservation behaviors after their visit. This effect was evident across all values groups, but was particularly notable for those prioritizing self-enhancement values (who may be neglected by existing interpretation) or openness to change values (who are potentially the most open to new ideas and new actions). Specifically, visitors who were exposed to messages that aligned to their value priorities were significantly more likely than those in the control groups to report conservation learning, intentions to engage in conservation actions, and engagement in conservation behavior after their visit.

The finding that values-based interpretation produced a significant increase in visitors’ conservation learning and intentions to adopt environmentally sustainable behavior is encouraging. More importantly, these positive intentions were followed by actual (self-reported) behavior change. The value-based interpretation both inspired visitors to express positive behavioral intentions and prompted actual behavior change. This was apparent in both the American and Australian samples - those in the treatment groups were significantly more likely to report having acted to help animals and the environment following their visit. Evidently, visitors are highly responsive to, and persuaded by, messaging framed according to their value priorities. The findings confirm that it is possible to use different interpretive materials to persuade individuals with different values priorities to adopt the same targeted responsible environmental behavioral outcomes.

Interestingly, value-based messaging had the greatest impacts on the self-enhancement group. As noted above, previous research shows that staff responsible for designing visitor experiences—zoo educators, interpreters and managers—are predominantly from the motivationally opposing self-transcendence group (Packer et al., 2022). Thus, not surprisingly, zoo and aquarium interpretation is likely to be framed around self-transcendence values. Focus group responses reported by Ballantyne et al. (2021) suggest that visitors who prioritize self-enhancement values may feel uncomfortable with, or dismissive of messages framed from a self-transcendence perspective, and thus feel out of alignment with, or even excluded by the institution they are visiting. The current study could well be the first time respondents who prioritize self-enhancement had encountered zoo messaging that appealed to their value priorities and this may have contributed to the impact of the materials. The results are encouraging as this group is not generally environmentally active and rarely comes through the zoo turnstiles. Appealing to these individuals may encourage further and more frequent visitation. While some may argue that it is not worth catering to such a small and infrequent visitor cohort, given the nature of their values it is likely that such individuals are in positions of power and influence in the community.

In support of previous studies (e.g., Ballantyne et al., 2011), the current research also shows that providing on-site prompts and opportunities for visitors to reflect upon their own behavior and how they could make a difference is key to successful environmental learning and long-term behavior change regardless of visitors’ values priorities. It seems that this reflection can be further augmented with experiences that foster experiential engagement (enjoyment, excitement, being able to view active animals), empathy/emotional connection (activities and messages that enable visitors to emotionally connect and empathize with the animals on display) and animal interactions (feeling of being able to connect with the animals on some level).

Our findings support the contention that the current generic approach (one size fits all) used in designing environmental interpretation in wildlife tourism contexts is limiting its effectiveness and long-term impact on visitor environmental learning outcomes and importantly on visitors’ adoption of responsible environmental behavior in their everyday lives. It would, however, be prohibitively expensive and complicated for zoos and aquariums to convert all their current generic interpretation into values-based offerings. Creating four values-based versions of everything, particularly if interpretation is predominantly delivered via on-site signage or presentations, would simply not be feasible. But what could be possible is to harness the power of technology to deliver the required interpretive messages through apps (Hughes & Moscardo, 2019). To illustrate, the screening questions this study used to allocate visitors into values groups proved highly effective. A similar approach could be used at the zoo entrance or when purchasing tickets for a wildlife tourism attraction. This screening could guide visitors to electronic materials and messages that match their value priorities. Thus, a group of four could visit the site and, depending on their values, be exposed to four quite different sets of messages and experiences. Our results suggest that not only would this assist with experiential engagement, learning and reflection, it could also encourage and support each visitor to take action following their visit.

Custom-designed on-site experiences could be augmented by post-visit online action resources designed to reinforce visitors’ learning and commitment (Ballantyne & Packer, 2011). These could also be designed to appeal to each values group. Resources could include links to environmentally-conscious organizations and products; details of local environmental groups, activities and events; and access to social media networks. Appeals for donations, calls for volunteer assistance and follow-up marketing could also be framed according to visitors’ value priorities. Findings from the current study suggest that such a values-based approach could help wildlife tourism to broaden its reach beyond traditional audiences to those with other values. In so doing, it may also widen its impact on visitors’ uptake of environmental behavior, though further longitudinal research to test this claim would be needed.

Conclusion

This experimental study has demonstrated that providing values-based interpretive materials influences visitors’ engagement during their visit, improves visitor environmental learning, supports their intentions to engage in environmentally responsible behavior, and (on some but not all measures) increases their self-reported adoption of environmentally responsible behaviors after their visit. These findings, confirmed in two separate datasets (North America and Australia), make an important contribution to theory and practice in relation to environmental interpretation in wildlife tourism contexts. In particular, the findings suggest that a more nuanced approach to interpretation, that takes into account the various personal values orientations prioritized by visitors, is warranted. People are more likely to pay attention to and remember environmental communications that are aligned with their values priorities (de Dreu & Boles, 1998). Thus, environmental interpretation strategies are likely to be more effective in persuading visitors to adopt sustainable environmental behaviors if interpretive materials are designed to appeal to visitors with a range of different value perspectives.

This research was ambitious in that it (a) attempted to use an experimental design in a field setting; and (b) attempted to assess the adoption of environmentally responsible behaviors as a result of a zoo or aquarium visit. Random assignment to a treatment or control group allows conclusions to be made about causation—in this case, we can conclude that the values-based interpretive materials were instrumental in achieving an increase in reflective engagement, visitor learning, and subsequent self-reported behavior change. What is not yet known, however, is which aspects of the materials were most effective, and in particular, how important it was that the materials were matched with the participants’ values priorities. It is possible that just receiving additional materials, regardless of their content, may have predisposed participants to be more mindful during their visit, or more positive in their responses. (A commonly recognized threat to the validity of research findings is when people’s responses are affected by their awareness that they are part of a study. Although both the treatment and control groups were aware of their involvement in this study, treatment group participants may have been influenced to behave or respond differently because they were given extra materials.) Further research is needed to explore these possibilities in greater depth. Further research is also needed to explore whether the provision of values-based interpretive materials is more important for some values groups than others.

Measuring environmental behavior change as a result of a zoo or aquarium visit is challenging, especially when the behaviors being targeted are lifestyle choices that are subject to multiple influences. In this study we had to rely on self-reported change, which we attempted to verify using open-ended responses. Despite this process, it is possible that participants over-reported their adoption of new behaviors as a result of their visit. It is also possible that those who had been given additional materials (the treatment group) may have been even more tempted to over-report their responses. A further limitation is the inability to control the content of the existing interpretive materials in the 12 target exhibits. If these happened to be aligned with specific values orientations, it may have reduced the ability of the treatment materials to impact the whole experience for those particular groups, thus possibly explaining some of the non-significant results and low effect sizes. Future research in this area should consider evaluating the existing interpretive materials in relation to the values orientations already represented. These issues aside, this research points to new areas and possibilities for zoo research and interpretive practice and provides support for the claims that the use of values-based environmental interpretation in wildlife tourism has considerable potential to create an environmentally aware and active citizenry.

Ethics approval

This project was approved by the University of Queensland Human Research Ethics Committee, approval number 2016000946. All participants provided informed consent.

Disclosure statement of competing interests

The authors report there are no competing interests to declare.

Additional information

Funding

This research was funded by the Australian Government through the Australian Research Council’s Linkage Projects funding scheme (project number LP150100266), in partnership with 12 zoos and aquariums in Australia, USA and Canada. The authors wish to thank the staff of the participating zoos and aquariums for their assistance with the implementation of this project.

Notes on contributors

Roy Ballantyne

Roy Ballantyne’s research interests are in free-choice environmental learning, environmental education and interpretation, wildlife and ecotourism.

Karen Hughes

Karen Hughes has research interests in the design of interpretive signs and experiences, visitor management, sustainable tourism, tourist behavior and wildlife tourism.

Jan Packer

Jan Packer’s research focuses on understanding and improving visitor experiences at natural and cultural tourism attractions.

Julie Lee

Julie Lee’s research interests include the theory, measurement and application of personal and cultural values to consumer and tourism behaviors.