Public attitudes regarding trophy hunting

In the United States, hunting remains a primary wildlife conservation and management strategy even as the number of active hunters has declined [1, 2]. With this demographic shift, wildlife managers, conservation organizations, and hunting associations increasingly work to mitigate conflict and maintain support for hunting among stakeholders not directly familiar with the practice, a task often accomplished through public outreach [1, 3, 4]. Despite the diminishing hunter population, surveys of U.S. residents have consistently indicated majority approval (>70%) for the practice since the 1990s [2, 3, 5]. However, this support varies according to hunters’ motivations: while hunting for subsistence or protection is broadly accepted, trophy hunting receives minimal support in the United States [3, 5, 6].

Trophy hunting describes the practice in which individual animals with particularly desirable physical characteristics or features (e.g., body size, horn or antler measurements) are hunted for preservation and display. Otherwise known as “sport” or “safari” hunting, it is usually practiced by hunters paying for the right to hunt, the services of a guide, and requisite equipment or staff [7]. Males of large, charismatic species are the most typical targets, from polar bears in northern Canada [8] to ungulates and large cats in sub-Saharan Africa [7].

Responsibly managed trophy hunting can serve as a tool to benefit local communities and surrounding ecosystems by improving socioeconomic and conservation outcomes [7–13]. Hunting license sales and land use fees can fund wildlife conservation initiatives, including population monitoring, habitat enhancement, and anti-poaching efforts [9]. Further, as trophy hunters are often willing to hunt in marginal or remote areas, their activities can underwrite wildlife conservation in locations that do not receive other types of tourist spending [10]. Trophy hunting can also reframe species that are otherwise considered nuisances or hazards as potentially valuable, thereby incentivizing the conservation of these animals and their associated ecosystems [7, 11]. Meanwhile, the socioeconomic benefits of trophy hunting include job opportunities, marketable skill training for local residents, and funding for community programs (e.g., infrastructure, education, healthcare) [8, 11, 12]. Additionally, game meat sold or donated to communities by trophy hunting operators can be a culturally significant source of protein [13].

Despite the global ubiquity of trophy hunting and its potential benefits when appropriately implemented, American attitudes toward the practice have been broadly negative for decades. In the 1980s, over three-quarters of surveyed U.S. residents disapproved of hunting for a trophy [14]. Recent studies indicate that American support for trophy hunting remains minimal; typically, only about a third of the public approves of the practice [3, 5]. Concerns about the morality or legitimacy of trophy hunting may contribute to this lack of acceptance [6]. In addition, hunting communications professionals have noted that efforts to discuss trophy hunting with members of the public are often stymied by misinterpretations of terminology and misunderstandings regarding the practice [1]. Negative misconceptions of trophy hunting are ubiquitous, including assertions that the practice is a threat to endangered species, a wasteful use of wildlife, or a form of animal cruelty [1, 4, 15, 16].

The pervasive antipathy for trophy hunting is evident in the popularity and influence enjoyed by anti-trophy hunting campaigns; these efforts shape public discourse even without enacting substantial policy changes [17, 18]. Citing the potential drawbacks of trophy hunting and highlighting instances of poor sportsmanship or violations of hunting regulations, activists call for national or global prohibitions on the practice [19]. Meanwhile, conservation practitioners’ ability to effectively respond to such efforts is hindered by the real and perceived difficulties associated with public outreach about this contentious form of hunting. In particular, a prevailing recommendation to separate attempts to change public attitudes toward trophy hunting through dedicated communications strategies from broader discussions of hunting in general may only serve to reinforce the stigma against the practice, thereby exacerbating the issues surrounding its place in the popular discourse [1]. Given the prevalence of anti-trophy hunting sentiments in the United States, hunters, wildlife managers, and other interested parties seek to improve their outreach efforts to change attitudes regarding the practice and maintain well-managed trophy hunting as a tool for socioeconomic development and wildlife conservation and management.

Improving trophy hunting communications

Well-designed messaging programs could build support for trophy hunting in the face of widespread animosity toward the practice. One measure of success for such efforts is their capacity to address existing misbeliefs, a major challenge when communicating with the public about trophy hunting [1]. Information provided via messaging can dispel misconceptions and change attitudes about controversial or emotional subjects if audiences are willing to listen [20, 21]. Conventional anti-misinformation strategies include inoculation (warnings about misinformation alongside preemptive counterarguments), debunking of erroneous information, presentation of plausible alternative viewpoints, and encouraging careful evaluation of message and source credibility [21]. These tactics may be applied to trophy hunting communications to improve public attitudes by refuting the numerous misconceptions regarding the practice.

The success of a messaging campaign also depends on its ability to relate information to different audiences. To effectively persuade audiences, message designers must account for how the various identities of potential recipients may affect their response to a message’s contents and presentation [22, 23]. Social identity theory states that people formulate self-images according to group memberships. According to this theory, individual group members’ attitudes and behaviors are determined based on perceived ideals (norms) for the larger group [24]. As an example, in the United States, political affiliation strongly influences reasoning processes for many issues (e.g., gun control, vaccination); such ideologically motivated cognitions demonstrate adherence to party priorities and signal group affinities [25]. For environmental concerns, social identity can be a major determiner of individuals’ positions on given issues (e.g., the “conservative white male effect” seen in some climate change research, [26, 27]), thereby necessitating that related communications account for identity-driven cognitions to maximize efficacy [25, 28, 29]. In debates over wildlife management and hunting, polarization and high tension make appeals to group membership especially powerful [23, 30, 31]. Often, trust in state wildlife management agencies is linked to political affiliation; conservatives generally report lower trust in government programs even while holding traditionalist, utilitarian views of wildlife use [32]. Messaging can also resonate with other aspects of identity besides group membership. For instance, identity theory posits that personal values and morals exert substantial influence on self-concepts, ultimately driving behaviors and beliefs [33, 34]. Examples of such values include family, tradition, and spirituality [35]. Message designers can use these and other elements of identity to tailor communications to audiences to increase their efficacy.

The source of a message is just as important as its contents and relatability to recipients. Audiences must accept the credibility of a message’s author or distributor; otherwise, they may perceive information provided in the message as biased or incorrect, redoubling adherence to prior beliefs [21, 36]. Given the increasing trend of political polarization in the United States [37], perceptions of credibility are frequently linked to political affiliation. Audiences place greater trust in information from figures whose political beliefs align with their own [25, 36]. Notably, scientific information about contentious topics is subject to varying acceptance depending on the political alignment of both source and recipient, though this can be mitigated using data from apolitical or uncontested sources [22]. Prior research has suggested that wildlife and conservation agency personnel are generally viewed as credible providers of information about hunting, though it remains unclear whether this holds true for information about trophy hunting [4].

Existing trophy hunting communications, which are currently neither well-designed nor widely distributed, do not represent successful outreach efforts. Some outreach professionals even discourage public discussion of the practice. One popular guide to hunting communications labels trophy hunting a “perfect villain” used by anti-hunting activists to negatively influence broader attitudes toward hunting, encouraging its readers to steer debates away from trophy hunting to more palatable motivations (e.g., subsistence hunting) [1]. This argument positions trophy hunting as too objectionable to even be mentioned in public discussions, limiting conversations about the practice to intra-community discourse. To this point, research suggests that communications from hunting operations primarily target their customer base without attempting to appeal to a broader audience. A review of online marketing by safari hunting outfits found that such websites emphasize the value of the hunt to the hunter rather than issues of morality or sustainability; descriptions of pricing plans and species of interest are prioritized over promotion of ethical standards or conservation practices [38]. Trophy hunters, meanwhile, often discuss hunting in terms of achievement or accomplishment, which can be off-putting to the general public [4, 39].

By ignoring non-hunters’ concerns about trophy hunting in outreach materials, hunt operators and trophy hunters limit their ability to connect with audiences not already familiar with the practice. In doing so, they risk losing what little public support exists for trophy hunting as the social license (i.e., unofficial sanction) for the practice diminishes [3, 5, 40]. Thus, the problem of inadequate public-facing communication about trophy hunting persists. To address the issues with current outreach efforts and improve public attitudes concerning this controversial practice, new communication strategies are required. In particular, priming and message framing represent promising options, especially when used to tailor messages to specific target audiences.

Priming is a psychological process in which people’s responses to a given situation or piece of information are influenced by prior exposure to other information [41, 42]. In other words, priming describes how recent experiences subconsciously affect present reactions. In the context of trophy hunting and wildlife management, audiences could be primed to connect messages they receive with their place identity, relationships to nature, or social identity groups (such as political affiliation, hunter status, or conservation organization membership) in order to increase message acceptance [29, 31, 32, 43].

As with identity-focused priming, the frame of a message can substantially affect how audiences receive it. A message’s “frame” is the combined effect of its contents, context, organization, structure, and medium, all contributing to the overall presentation of the information contained in the message [44, 45]. Framing describes how people (re)conceptualize an issue based on contextualized information [46]. The same information can provoke different responses according to how it is framed for audiences. For example, messaging about invasive species mitigation invoking legal consequences of negligence is more likely to promote proper watercraft sterilization behaviors than messages emphasizing cultural norms surrounding sterilization [45]. Regarding the efficacy of different frames in improving hunting communications, research focused on developing pro-hunting messages has found receptive audiences for communications with different themes, including the economic and ecological benefits of such activities [4].

Research goals

To address the growing need for effective communication about trophy hunting, this research intended to explore the effects of different message designs on public attitudes toward the practice. Therefore, we had two main objectives: to investigate how different combinations of identity priming items and message frames affect respondents’ attitudes toward trophy hunting and to compare respondents’ willingness to trust information about trophy hunting from different hypothetical sources. We hypothesized that messages emphasizing potential benefits of trophy hunting would increase respondents’ support for the practice as compared with a neutral control message (H1), and we anticipated that wildlife conservation-focused messages would have a greater positive effect than messages centered around the socioeconomic benefits of trophy hunting given the prevalence of the popular belief that trophy hunting poses a substantial threat to wildlife species [1, 15, 16] (H2). We also expected that priming for personal value identity or social identity would affect respondents’ post-message support for trophy hunting (H3). Finally, we predicted that respondents would be more trusting of trophy hunting communications from government agencies or conservation organizations than those from hunting associations (H4). Overall, this study aimed to evaluate how different aspects of message design can influence public attitudes regarding trophy hunting; we successfully achieved this goal and provided suggestions for improving future public-facing trophy hunting communications based on our findings.

Survey and message design

We followed the recommendations of Dillman et al. [47] while developing our online survey instrument. After indicating their written consent to participate and completing an introductory section regarding attitudes towards hunting in general, respondents transitioned to the trophy hunting-specific message testing experiment, which employed a pretest-posttest design [48]. First, each respondent answered the same pretest question, indicating their current level of approval for “legal, regulated” trophy hunting using a five-point Likert scale (1 = “strongly disapprove” to 5 = “strongly approve”) with an additional “I don’t know” option. This question was based on prior research by Responsive Management and the National Shooting Sports Foundation [5]. Next, respondents were randomly assigned to one of nine treatment groups, where each treatment had two components: an identity-focused priming item (or lack thereof, as a control) and one of three sets of message content (including a control message) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Description of message treatments.

https://doi.org/10.1371/journal.pone.0312949.t001

Given our interest in the potential roles of social identity theory [24] and identity theory [33, 34] in determining approval for trophy hunting, we included questions that primed for both types of identities (independently) in this experiment. To maximize the effects of these primes, respondents answered their designated priming item just prior to viewing their assigned messaging [41, 42]. Respondents assigned to treatments subject to the value identity prime (Treatments 3, 6, and 9; Table 1), corresponding to identity theory [33, 34], were asked to consider a list of personal values adapted from the Chronic Pain Values Inventory [35], which consisted of family/friendship, conservation/stewardship, economic/financial success, growth and learning, and recreation/fun. Respondents were then instructed to rank these items according to the level of importance they would assign to each in their own lives and the lives of others. Respondents assigned to Treatments 1, 2, 4, 5, 7, and 8 (Table 1) received this question as part of the demographics section at the end of the survey. Respondents assigned to the “social identity” treatments (i.e., Treatments 2, 5, and 8; Table 1), corresponding to social identity theory [24], were asked to report their political alignment; options included “strongly conservative” (1), “moderately conservative” (2), “centrist” (3), “moderately liberal” (4), and “strongly liberal” (5). Respondents assigned to Treatments 1, 3, 4, 6, 7, and 9 received this question as part of the demographics section at the end of the survey.

After the identity priming item (or absence of one), respondents viewed one of three messages corresponding to their assigned frame (Table in S1 Appendix). The control message featured a single paragraph providing descriptive information about trophy hunting, including a basic definition of the term and a brief summary of typical trophy hunting operations [7, 49]. The two experimental messages contained the same descriptive information as the control message, and additional information related to the treatment frames was appended to the definitional paragraph. Prior research suggested that these messages should focus on the ecological and socioeconomic benefits of managed trophy hunting, as audiences have proven receptive to these two frames in similar contexts [4, 44]. Treatments 4–6 focused on the wildlife conservation benefits associated with managed trophy hunting [9, 50]. Treatments 7–9 emphasized the socioeconomic benefits that trophy hunting can provide to local communities and society at large [12, 49, 51].

To determine whether the messages resonated as intended, we asked respondents to indicate their agreement with a series of seven induction statements regarding their assigned message using a five-point Likert scale with an additional “I don’t know” option [52]. We then asked respondents the post-test question, which rephrased the pre-test question about approval for “legal, regulated” trophy hunting to reflect our interest in respondents’ attitudes following their exposure to the messages. Additionally, we employed a post-only design [48] where respondents were asked to indicate the likelihood of their trust in the message that they just read if it came from various hypothetical sources using a three-point scale (1 = “less likely,” 2 = “neither more nor less likely,” and 3 = “more likely”) with an additional “I don’t know” option. Respondents also answered questions about various demographic characteristics (i.e., age, gender, race/ethnicity, region, setting, household income, and highest level of education).

We cognitively pretested the survey from December 2022 to January 2023 and made revisions based on participants’ comments. The final survey instrument (text in S2 Appendix) and protocol were approved by the University of Arkansas at Monticello Institutional Review Board (#FANR1002).

Survey distribution

The Qualtrics survey platform hosted the survey, and respondents were randomly selected from an opt-in panel maintained by Qualtrics. Respondents were required to reside in the United States and be over the age of eighteen to be eligible for the survey. Qualtrics used respondents’ self-reported demographic characteristics (namely, gender, age, race/ethnicity, region, and highest level of education) as quotas to ensure the sample was representative of the U.S. population as reported in the 2020 U.S. Census [53]. As a note, the racial and ethnic categories used in this study were adapted from the most recent U.S. Census [53]. We acknowledge the limitations of this methodology, including the possible disparities between our sample and the overall U.S. population, as well as the potential bias introduced by the online distribution of our survey, as differences in attitudes regarding trophy hunting may exist between American residents with and without reliable internet access [47]. Responses were collected between January and February of 2023. During this period, Qualtrics solicited responses from the online panel, adjusting the targeted distribution of the survey as the pre-established demographic quotas were met. Throughout this process, the survey company provided updates to the researchers and allowed for periodic inspection of data quality. Upon completion of data collection, we performed a final quality check on all responses, including removal of some responses due to suspected “straight-lining” [54].

Data analysis

We used R statistical software version 4.2.2. to analyze our data [55]. For both the pre- and post-test questions regarding approval for trophy hunting and the induction checks, all “I don’t know” responses were converted to neutral responses prior to analysis to allow for paired analyses [56]. For the question about trust in various hypothetical message sources, all “I don’t know” responses were removed prior to analysis. Shapiro-Wilk tests and Levene’s tests (R package “car” [57]) indicated that our data violated assumptions of normality and homogeneity of variance. However, given the size of our sample and the general robustness of parametric tests such as ANOVA to violations of their prerequisite assumptions when applied to large-scale Likert type survey data, we elected to use parametric analyses to answer our research questions [58, 59]. For purposes of comparison, we included results from nonparametric alternatives for each parametric test we conducted in S3 Appendix.

We used paired t-tests to compare respondents’ attitudes regarding trophy hunting before and after viewing their assigned message [60]. We assessed the effect size of pre- and post-message differences via Cohen’s d, which we determined with the “effsize” package in R [61]. Effect size thresholds were as follows: |d| < 0.2 = "negligible,” 0.2 ≤ |d| < 0.5 = "small,” 0.5 ≤ |d| < 0.8 = "medium,” and |d| ≥ 0.8 = "large" [62]. We used one-way analysis of variance (ANOVA) tests to determine statistically significant differences in attitudes toward trophy hunting based on treatment group. We measured effect size using eta squared (η²), interpreted as follows: η² < 0.01 = “negligible,” 0.01 ≤ η² < 0.06 = “small,” 0.06 ≤ η² < 0.14 = “medium”, and η² ≥ 0.14 = “large” [63, 64]. One-way ANOVA was also used to determine significant differences in the induction checks based on treatment group, trust in the message based on hypothetical message source, and attitudes toward trophy hunting based on political alignment and personal values. When the ANOVA test results were significant, we performed multiple pairwise comparisons with a Bonferroni correction to assess which specific differences were statistically significant [60].

To determine the effects of different message treatments on respondents with opposing political affiliations or different personal values, we sorted the responses to relevant questions into binary response bins (e.g., liberal versus conservative, individuals who highly value conservation versus individuals who place a low value on conservation, etc.). For political affiliation, we collapsed the 5-point Likert scale into two options: conservative versus liberal, excluding centrists from the analysis [32]. Similarly, for personal values, a rank of 1 or 2 for a given value area indicated "highly valued,” and 4 or 5 was "low value,” while ranks of 3 (“medium value”) were excluded. For this analysis, we focused on respondents who highly valued “conservation/stewardship” and “economic/financial success,” as these values were directly related to the two experimental message frames. While the political identity groups were mutually exclusive, it was possible for a respondent to be in both high-rank personal value groups, as the two values are not opposite ends of the same spectrum. We used paired t-tests to compare pre- and post-message attitudes toward trophy hunting within each group. We also used one-way ANOVA tests and multiple pairwise comparisons with a Bonferroni correction to examine treatment effect on the pre- and post-message attitudes among respondents in each group [60]. Data and code used for all analyses are found in S4 and S5 Appendices.

Sample demographics

Our final sample size was 2000 responses, and the demographics of the sample were consistent with those of the 2020 U.S. Census [53]. Slightly more than half of our respondents identified as women, and the mean age of our sample was 46.73 years old (Table 2). Nearly 60% of respondents identified as White (Non-Hispanic) alone; the rest belonged to one or more other racial or ethnic groups (Table 2). Over 35% reported that their highest level of formal education was at least a bachelor’s degree, and less than half had a total annual household income of $60,000 or more (Table 2). Over three-quarters of respondents resided in urban or suburban areas based on their own self-reporting (Table 2). Nearly 40% identified as either strongly or moderately conservative while almost a third identified as either strongly or moderately liberal (Table 3). Regarding personal values, 45% of respondents ranked “family/friendship” as the most important of the provided options. Just over one-fifth indicated that “economic/financial success” was the most important, and approximately 15% listed “conservation/stewardship” as being most important. “Growth and learning” ranked as the most important value for slightly over 12% of respondents, while 7% considered “recreation/fun” to be most important (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of sample demographics.

https://doi.org/10.1371/journal.pone.0312949.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Descriptive statistics and ANOVA results showing differences in approval for trophy hunting across identity groups.

https://doi.org/10.1371/journal.pone.0312949.t003

Induction check

Overall, respondents tended to disagree that their assigned message aligned with their personal views, and they generally indicated that they wanted to “argue back” with what they read (Table 4). Respondents were relatively neutral regarding both the statement that the message they read addressed their concerns about trophy hunting (indicating that the messages did not adequately respond to such concerns) and the statement that the message provided factual evidence about the benefits of trophy hunting (suggesting that the message contents were generally seen as at least somewhat factual, though not overwhelmingly so) (Table 4). Overall, respondents also tended to be fairly neutral when asked whether their assigned message made them believe that trophy hunting could be good for conservation or for local people, indicating that the messages did not provoke strong feelings about the positive impacts of trophy hunting (Table 4). Finally, on average, respondents neither agreed nor disagreed that the message they read was well-reasoned (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. ANOVA results showing differences in response to induction checks by treatment.

https://doi.org/10.1371/journal.pone.0312949.t004

One-way ANOVA tests and post-hoc multiple pairwise comparisons with a Bonferroni correction, indicated that the various treatments produced significant differences in the messages’ resonance with respondents. The responses of those who saw the conservation message (Treatments 4–6) generally differed from those of their counterparts who viewed the control message (Treatments 1–3) (Table 4). For example, while respondents who saw the conservation message (Treatments 4–6) tended to perceive the message as more in line with their personal views than those who saw the control message (Treatments 1–3) (F(8, 1990) = 10.56, p < 0.001, η² = 0.04), mean agreement between Treatment 4 and Treatment 2 did not significantly differ (Table 4). Mean agreement that the messages provided factual evidence of the benefits of trophy hunting also generally adhered to this pattern (F(8, 1989) = 5.98, p < 0.001, η² = 0.02), as did responses to statements about the messages being well-reasoned (F(8, 1987) = 8.40, p < 0.001, η² = 0.03) and able to convince respondents that trophy hunting could be good for conservation (F(8, 1990) = 11.52, p < 0.001, η² = 0.04) (Table 4). Respondents who viewed either the conservation or socioeconomic benefits messages (Treatments 4–9) tended to be slightly more willing to agree that the message convinced them about potential socioeconomic benefits of trophy hunting than those who saw the control message (Treatments 1–3) (F(8, 1988) = 11.66, p < 0.001, η² = 0.04) (Table 4). Other induction checks did not follow the common pattern: the only significant differences regarding respondents’ desire to “argue back” with the messages were that those in Treatment 2 were, on average, significantly more likely to agree with this sentiment than those who received Treatment 4 or Treatment 5 (F(8, 1990) = 2.99, p = 0.003, η² = 0.01) (Table 4). Finally, while the nine treatments produced some significant differences in mean agreement with the statement about addressing concerns (F(8, 1990) = 4.37, p < 0.001, η² = 0.02), there were no clear patterns based on either priming item or message version (Table 4).

Message treatment effect on trophy hunting attitudes

Among all respondents, mean approval for trophy hunting prior to viewing any message was low ( (SD) = 2.41 (1.27)). The total number of respondents randomly assigned to each treatment group ranged from 203 to 231, and a one-way ANOVA test found no significant differences in average pre-treatment support for trophy hunting based on treatment group (Table 5). Supporting H1, all treatments including either the conservation or the socioeconomic benefits message significantly increased mean approval for trophy hunting, though the effect size varied (Table 5). A one-way ANOVA and post-hoc multiple pairwise comparisons with a Bonferroni correction revealed some significant differences in post-treatment mean approval for trophy hunting between treatments (Table 5). However, inconsistent with H2 and H3, we found no significant differences in post-message approval for trophy hunting between respondents who viewed the conservation benefits message and those who viewed the socioeconomic benefits message, nor did we observe any patterns in post-message attitudes toward the practice based on the priming items (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. ANOVA and paired t-test results showing treatment effect on pre- and post-message approval for trophy hunting.

https://doi.org/10.1371/journal.pone.0312949.t005

Based on the results of a one-way ANOVA test and a post-hoc multiple pairwise comparison with a Bonferroni correction, we found that conservative respondents were more supportive of trophy hunting than their liberal counterparts, but those identifying as moderately conservative did not significantly differ from centrists (Table 3). After dividing the sample into liberals and conservatives, we found that both groups generally disapproved of trophy hunting before viewing any message (Table 5). The number of conservative respondents randomly assigned to each treatment group ranged from 80 to 97 while treatment group size for liberal respondents ranged from 60 to 92 (Table 5). One-way ANOVA tests indicated that pre-test mean approval for trophy hunting did not significantly differ by treatment for conservatives or liberals (Table 5). For conservative respondents, all treatments aside from Treatment 1 and Treatment 2 significantly improved attitudes towards trophy hunting (Table 5). Similarly, paired t-tests found that all treatments including either the conservation or socioeconomic benefits message (Treatments 4–9) significantly increased mean approval for trophy hunting among liberal respondents while those including the control message (Treatments 1–3) produced no change (Table 5). Although a one-way ANOVA test indicated that conservatives’ post-test approval differed significantly by treatment, post-hoc multiple pairwise comparisons with a Bonferroni correction did not find any such differences; meanwhile, the only significant difference in liberals’ post-test attitudes was that those in Treatment 7 tended to be more approving than those in Treatment 3 (Table 5).

One-way ANOVA tests and post-hoc multiple pairwise comparisons with a Bonferroni correction indicated that the only significant differences in attitudes regarding trophy hunting based on the relative importance of respondents’ personal values were that those who deemed “conservation/stewardship” the least important value from among the provided options tended to be more supportive of trophy hunting than those who reported it as being of greater importance, and those who ranked “recreation/fun” as being of the highest importance were more approving of the practice than those who considered it to be either their lowest or second-lowest priority (Table 3). Respondents who highly valued “conservation/stewardship” (approximately 33% of the sample) were largely unsupportive of trophy hunting before viewing any messages ( (SD) = 2.28 (1.25)), and pre-test approval for the practice among those who highly valued “economic/financial success” (just under 43% of respondents) was likewise low ( (SD) = 2.49 (1.28)) (Table 5). The number of respondents who highly valued “conservation/stewardship” who were randomly assigned to each treatment group ranged from 65 to 84 while treatment group size for respondents who highly valued “economic/financial success” ranged from 85 to 109 (Table 5). One-way ANOVA tests found that pre-test mean approval for trophy hunting did not significantly differ by treatment for either group (Table 5). For both groups, paired t-tests showed that treatments including either the conservation or socioeconomic benefits message significantly increased mean approval for trophy hunting while those including the control message had no significant effect on attitudes toward the practice (Table 5). Among those who highly valued “conservation/stewardship,” Treatment 4 had the greatest effect, while for respondents who highly valued “economic/financial success”, Treatment 6 and Treatment 7 provoked the most substantial changes (Table 5). A one-way ANOVA and post-hoc multiple pairwise comparison with a Bonferroni correction found that among respondents who highly valued “conservation/stewardship”, Treatment 1 resulted in a significantly lower post-test mean approval level for trophy hunting than treatments including the conservation message (Treatments 4–6) and Treatment 7, and Treatment 3 had a significantly lower post-test mean approval than Treatment 4 (Table 5). Meanwhile, the only significant difference in post-test attitudes toward trophy hunting among respondents who highly valued “economic/financial success” was that those in Treatment 6 tended to be more approving than those in Treatment 3 (Table 5).

Trust in message sources

Consistent with H4, a one-way ANOVA and post-hoc multiple pairwise comparison with a Bonferroni correction indicated that respondents found Safari Club International (SCI) to be the least trustworthy provider of information among the available sources (Table 6). This was the only entity for which respondents reported a decreased likelihood of trust in hypothetical messaging (i.e., mean likelihood < 2.0) (Table 6). Notably, respondents tended to view information from the South African Department of Environmental Affairs (SADEA) as neither more nor less likely to be trusted, yet average trust placed in messages from the U.S. Fish & Wildlife Service (USFWS) was significantly higher (Table 6). The highest level of trust was reserved for USFWS or conservation-focused organizations such as the World Wildlife Fund (WWF) or The Nature Conservancy (TNC) (Table 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Descriptive statistics and ANOVA results showing differences in mean likelihood of trust in message sources.

https://doi.org/10.1371/journal.pone.0312949.t006