Effect of semantics in the study of tolerance for wolves

As conservation scholars increasingly recognize the critical role of human thought and behavior in determining the persistence of biodiversity across the globe, a growing line of inquiry regarding the validity and comparability of previous applications of core psychological concepts has emerged. Specifically, inconsistent measurement and use of terms, such as attitudes and acceptance, reveal important questions about previous approaches. Given that these concepts differ by definition, yet have been used interchangeably, we explored what drives differences in people's responses when each concept is operationalized in the context of a contested wildlife species, the gray wolf (Canis lupus). To do so, we used data from a 2014 survey of U.S. residents (n = 1287) to test how measures of six concepts (i.e., acceptance, attitudes, benefits, risks, [prior] behavior, and behavioral intentions) often employed in the conservation social sciences were related with a broad set of possible explanatory variables. Despite moderate to strong correlations between all concepts measured (| Pearson's r | = 0.39–0.65, p < 0.001), results revealed considerable variation in their respective relationships with identical explanatory variables. Specifically, although wildlife value orientation (i.e., domination or mutualism) operated fairly consistently across cognitive and behavioral concepts, the relationship between the six concepts and other factors, such as social trust, identification with various interest groups (i.e., hunter, farmer, or rancher, environmentalist, and animal rights advocate), and political ideology (i.e., liberal vs. conservative), varied considerably. Our findings underscore that differences exist in these measures, which could have serious implications for conservationists integrating social science findings in their decision‐making processes if they are unaware of the theoretical underpinnings of and distinctions between core psychological concepts.


INTRODUCTION
As conservation scholars increasingly recognize the role of human thought and behavior in determining the persistence of biodiversity globally (Baruch-Mordo et al., 2009;Ghoddousi et al., 2021;Manfredo et al., 2019;Teel et al., 2018), questions concerning the systematic study of human perceptions have emerged. A growing line of inquiry regarding both the construct validity (Drost, 2011) and comparability of the core psychological concepts commonly found throughout the literature reveals important questions about previous approaches (Brenner & Metcalf, 2020;Browne-Nuñez & Jonker, 2008;Bruskotter et al., 2015;Glikman et al., 2021;Knox et al., 2021;Whitehouse-Tedd et al., 2021). From attitudes and normative beliefs to behavioral intentions and more--the operationalization of these concepts has varied nearly as widely as the range of species assessed. Inconsistency in measurement and the use of these concepts hinders researchers' ability to make meaningful comparisons across data sets and identify trends over time. Further, the accurate measurement of such concepts is imperative to not only the scientific rigor of conservation research, but also the conservation actions that such efforts often seek to inform (Merkle et al., 2019). Thus, we responded to urgent calls to investigate the consistency and robustness of psychological measures frequently employed Kyle et al., 2020).

Use and misuse of core psychological concepts
Many discrepancies in prior research may result from a lack of clarification, or agreement, regarding cognitive and behavioral concepts. For example, attitude--derived from the late Latin aptus (i.e., fit or fitness)--is defined by Eagly and Chaiken (1993:1) as "a psychological tendency…expressed by evaluating a particular entity with some degree of favor or disfavor." The concept of attitudes is commonly applied in conservation social science research, yet measurement of the term varies considerably, and the reliability and validity of related psychometric scales are infrequently assessed (Whitehouse-Tedd et al., 2021). For example, Whitehouse-Tedd et al. (2021) identified at least four different scales (summated ratings, checklists, semantic differentials, and equal appearing intervals) across approximately 100 publications focused on human-wildlife interactions, half of which investigated the reliability and fewer of which investigated validity specifically. Despite the definitional precision offered by social scientists, attitudes have been referenced interchangeably with several other core psychological constructs, such as acceptance.
Acceptance, from Anglo-French, reflects an agreement to abide by the act of another and is often characterized in psychological literature as a normative belief or appraisal about what is considered appropriate for oneself and others (e.g., in terms of behavior, idea, or situation). In the context of wildlife, acceptance is typically assessed via the use of the wildlife acceptance capacity (WAC) measure--which evaluates the willingness of an individual (or collectively, a society) to "carry" a particular species (Decker & Purdy, 1988;Gigliotti et al., 2000). Equating the general favorability of a species (attitude) with distinct preferences for the size of future populations in a given area (e.g., WAC) may contribute to ill-informed management decisions. In fact, in the context of certain species (e.g., wolves), evidence suggests that some people may prefer local populations to decrease despite holding positive attitudes toward the species overall (Ericsson et al., 2008;von Essen & Allen, 2020). Given the well-established status quo bias--a cognitive heuristic whereby a desire for things to remain the same results in an aversion to change (Samuelson & Zeckhauser, 1988)--it is possible that preferences about current population sizes may be largely independent of attitudes.
In addition to how attitudes and acceptance are defined above, scholars (e.g., Bath & Buchanan, 1989;Ericsson & Heberlein, 2003;Treves et al., 2013) have also measured perceived risks and benefits of species as if such ideas could approximate attitudes. Originating from the Italian word risco (i.e., danger) and the Latin expression bene facere (i.e., do good to), these perceptions were once believed to be the result of a rational weighing of the negative and positive impacts of a given hazard (e.g., wild animal); however, they are often shaped by far more than technical assessments of an event's probability and severity (Slovic, 1987;Wilson et al., 2019). Among nonexperts specifically, subjective judgments of risks and benefits--whether ecological, economic, or otherwise--are understood to be the instinctual and intuitive products of one's affective (e.g., emotional) response to a hazard (Slovic & Peters, 2006), their trust in the managing authority (Siegrist, 2021), and more. Thus, risk perceptions, attitudes, and acceptability are related (Lischka et al., 2008;Riley & Decker, 2000;Carter et al., 2012;Slagle et al., 2013;Zajac et al., 2012) but distinct constructs, necessitating rigorous and systematic evaluations of their use.
Combining concepts into a multidimensional measure that accounts for cognitive antecedents of behavior and actual behavior (or intentions) may further help conservationists understand public sentiment about challenging conservation issues, such as human-wildlife conflict (Brenner & Metcalf, 2020;Treves, 2012;Bhatia et al., 2020;Bruskotter & Fulton, 2012;Nilsson et al., 2020). For example, in conservation, the term tolerance--stemming from the Latin verb tolerare (i.e., to endure)--typically represents an unwillingness to engage in negative behaviors toward (or acceptance of) a biodiversity target (e.g., a carnivore) that one dislikes. In contrast, a negative attitude and low acceptability of--and possibly negative behavior toward--a species would reflect intolerance. Bhatia et al. (2020) go further by identifying five types of human responses from a literature review in relation to how attitudes and behaviors (or inaction) interact: manifested intolerance (negative attitudes acted upon), latent intolerance (negative attitudes not acted upon), neutral (ambivalent or no attitudes and no resultant behaviors), appreciation (positive attitudes not acted upon), and stewardship (positive attitudes acted upon).
Behaviors acted upon--whether in support of or opposition to the biodiversity target--might include a range of activities that are political (e.g., writing to one's congressperson), social (e.g., posting to social media), or physical (e.g., illegally killing a member of the species). When an absence of past behaviors is expected (e.g., due to contextual factors, such as geographical proximity or biographical availability), measures of behavioral intentions may be used to estimate one's selfperceived likelihood of engaging in a particular behavior if given the opportunity in the future (Fishbein & Ajzen, 1975). Thus, the inclusion of multiple behavioral measures (e.g., prior behavior and intentions) provides a means of assessing behavior at multiple scales, allowing researchers to maximize the utility of their results.
Despite attempts by scholars to differentiate the meaning of such core constructs, the operationalization of key terms in applied contexts, such as conservation, remains inconsistent, unstandardized, and in some cases untested in terms of reliability and validity (Whitehouse-Tedd et al., 2021). Thus, conservationists could benefit from a more thorough evaluation of the distinctions and nuances of the concepts commonly invoked in applied social science research to ensure better alignment with research objectives (i.e., validity), to improve the reliability of measures when operationalized and comparability of results across spatial and temporal context, and to obtain information from different constituent groups or the public that can inform conservation decision-making.

Case study of wolves
Inconsistent application of core social science concepts may be especially problematic in the context of a wide-ranging species, such as the gray wolf (Canis lupus). As populations of this contested carnivore continue to expand throughout the United States and Europe (Chapron et al., 2014;Mech, 2017), the need to better understand and explain public perceptions and behavior toward wolves has intensified. However, just as the operationalization of public perceptions and behaviors toward wolves has varied, so too have the factors believed to influence them. From demographic characteristics (e.g., age, sex, education, and income) (Williams et al., 2002), place of residence (e.g., rural vs. urban) (Williams et al., 2002), trust Slagle & Bruskotter, 2019;Volski et al., 2021), and sociopolitical identity (e.g., organizational affiliation) (Hamilton et al., 2020;van Eeden et al., 2019van Eeden et al., , 2021 to individual values (e.g., wildlife value orientations) (Dietsch et al., 2016;Teel & Manfredo, 2010), occupation (e.g., farming and ranching) (Williams et al., 2002), and engagement in outdoor recreation (e.g., hunting) (Ericsson & Heberlein, 2003;Hogberg et al., 2016;Treves & Martin, 2011), a range of cognitive, interpersonal, and contextual factors are explanatory. Although the specific setting of each study in which such constructs are applied undoubtedly contributes to the considerable variation observed (Zimmermann et al., 2020), such variation may also be a product of the inconsistent ways in which these concepts are measured.
We examined the relationship between key indicators of tolerance (i.e., attitudes, acceptance, benefits, risks, prior behavior, and behavioral intentions toward a species) and a consistent set of independent variables (i.e., demographics, residency, political and social identity, trust, and wildlife value orientations). We then investigated which of the four cognitive concepts tested (i.e., attitudes, acceptance, benefits, and risks) best explained the propensity to act in either support of or opposition to wolves while controlling for the same independent variables to demonstrate how operationalization of these concepts can inform conservation decision-making.

METHODS
The study population included U.S. residents 18 years and older (n = 1287). Responses were collected by the GfK Group through a Qualtrics survey administered in 2014. Participants were recruited via address-based sampling and randomly selected for participation through the GfK (currently Ipsos) Knowledge Panel. Data collection methods were approved by The Ohio State University's Office of Responsible Research Practices (study 2013E0553) and detailed in Slagle et al. (2017). Because we sought to examine the relationship between different variables, rather than represent the views of any particular portion of the U.S. population (Appendix S1), all analyses were performed on unweighted data. Data quality was assessed via a variety of criteria, including scale reliability tests (Cronbach's alpha), collinearity statistics (variance inflation factor), and more (details in Appendices S2 & S3).
To understand the relationship between all six operationalizations of key concepts (i.e., attitudes, acceptance, benefits, risks, prior behavior, and behavioral intentions) in relation to wolves, we first calculated descriptive statistics (i.e., frequencies, means, and standard deviations) of each measure and used Pearson's correlation coefficients to detect significant associations between the measures. Following that, we constructed six separate linear regression models to test each of these measures against an identical set of independent variables (outlined below). Finally, we constructed two logistic regression models that examined engagement in prior stewardship behaviors and prior intolerant behaviors. We controlled for the same set of independent variables, in addition to participants' attitudes, acceptance, benefits, and risks associated with wolves.

Dependent variables
Attitudes toward wolves were quantified using a semantic differential scale consisting of four response items (Appendix S2). These items, each of which was measured on a 7-point bipolar scale ranging from 1 (a negative attitude toward wolves) to 7 (a positive attitude toward wolves), were averaged to reflect a respondent's overall attitude toward the species.
Acceptance of wolves was calculated via participants' WAC, which reflected their preferences for future wolf populations nationwide. Responses were measured on a 5-point bipolar scale ranging from 1 (wolf populations should be decreased greatly) to 5 (wolf populations should be increased greatly). The original survey also contained a sixth option of "I don't know," which was treated as missing in these analyses.
Perceived benefits of wolves were captured via participants' level of agreement with the following five statements regarding the expansion of wolf populations: Allowing wolf populations to expand into other areas (outside of the areas they currently occupy) would keep deer, elk, and moose populations in balance; increase tourism in areas where wolves have moved into; preserve the wolf as a wildlife species; return the natural environment back to the way it was; and help control coyote populations. Responses to each item were measured on a 7-point bipolar scale ranging from 1 (strongly disagree) to 7 (strongly agree) and averaged to reflect respondents' perceived benefit of the species.
To assess perceived risks associated with wolves, participants rated their level of agreement with the following five statements regarding the expansion of wolf populations: allowing wolf populations to expand into other areas (outside of the areas they currently occupy) would result in large numbers of wolf attacks on livestock; ranchers losing money; wolf attacks on humans; wolves wandering into residential areas; and fewer deer, elk, and moose available to hunters. Responses to each item were measured on a 7-point bipolar scale ranging from 1 (strongly disagree) to 7 (strongly agree) and averaged to reflect respondents' perceived risk of the species.
To quantify prior wolf conservation behavior, participants reported their engagement in six stewardship and six intolerant actions during the past 6 months. Specifically, participants were asked whether they had performed any of the following behaviors in support of or opposition to wolf recovery: donated to an organization; spoke with family or friends; posted to Facebook or Twitter; wrote, phoned, or emailed their congressperson; wrote a letter to their newspaper; or contacted a wildlife manager or management agency. For our linear regression analyses, prior intolerant behaviors were subtracted from prior stewardship behaviors to yield a single composite measure of participants' prior wolf conservation-related behavior (i.e., a positive number indicated stewardship, whereas a negative number indicated intolerance). For our logistic regression analyses, responses to the six items representing stewardship were dichotomized into a binary measure indicating whether participants had engaged in at least one or more of the supportive behaviors in the past 6 months (0, no; 1, yes). Similarly, responses to the six intolerant items were dichotomized into a binary measure indicating whether participants had engaged in at least one or more of the intolerant behaviors in the past 6 months (0, no; 1, yes).
To capture wolf conservation behavioral intentions, participants indicated their estimated likelihood to engage in five stewardship and six intolerant actions. Specifically, participants were asked how likely they were to perform any of the following behaviors in support of or opposition to wolf recovery: contribute to an organization; post to Facebook or Twitter; write their congressperson; sign a petition; and contact a wildlife manager or management agency. The sixth intolerant action assessed how likely participants were to shoot a wolf if they saw one. Responses were measured on a 5-point bipolar scale ranging from 1 (very unlikely) to 7 (very likely). Intolerant behavioral intentions were subtracted from stewardship intentions so that a negative number indicated intentions for intolerance and a positive number indicated intentions for stewardship.

Independent variables
We examined several independent variables in each regression model, including demographics, residency, political and social identity, trust, and wildlife value orientations. Participants' self-reported age, sex (male or female), education, and household income were assessed. Participants indicated their highest degree earned on a categorical scale from 1 (less than high school) to 4 (bachelor's degree or higher). Responses for household income were measured on a categorical scale from 1 (less than $5000) to 19 ($175,000 or more).
Given the significant role of residency (e.g., rurality, listing region) established in previous studies, whether or not participants lived in a metropolitan statistical area (MSA) (0, no; 1, yes) was examined, as was participants' residency in the Northern Rocky Mountains (NRM)--where wolves were reintroduced in the mid-1990s (Smith & Bangs, 2009) and removed from ESA (Endangered Species Act) protections in 2011; the Western Great Lakes (WGL)--where wolves have always been present and have been listed under the ESA since the early 1970s; or the rest of the country (RoC)--where the species' status has varied and no recovery plan exists.
Participants' political ideology was quantified via the following question: "When it comes to politics, please indicate which of the following you consider yourself ?" Responses ranged on a 7-point bipolar scale from 1 (extremely liberal) to 7 (extremely conservative).
Participants reported the extent to which they identified as a hunter, farmer, or rancher, environmentalist, and animal rights advocate. Responses were measured on a 5-point unipolar scale ranging from 1 (not at all) to 5 (very strongly). Participants could identify with more than one interest group; therefore, response categories are not discrete.
Participants rated the extent to which they agreed with the following four statements concerning the federal agency charged with the management and conservation of species listed as threatened or endangered under the ESA (i.e., the U.S. Fish and Wildlife Service [USFWS]): "I feel that the U.S. Fish and Wildlife Service…" shares similar values as me; takes similar actions as I would; is trustworthy in their management of wildlife in the United States; and is capable in their management of wildlife in the United States. Responses were measured on a 5-point bipolar scale ranging from 1 (strongly disagree) to 5 (strongly agree).
To assess participants' core beliefs about wildlife, an abbreviated form of wildlife value orientations (WVOs), consisting of a set of seven domination and mutualism-based items, was examined (Appendix S2). Responses were measured on a 5point bipolar scale ranging from 1 (strong disagreement) to 5 (strong agreement). Responses were averaged to reflect either participants' overall domination or mutualist orientations.

Linear regression
Attitudes toward wolves were not significantly associated with age, sex, income, education, political ideology, residency in an MSA, the NRM, or the WGL (Table 1). Attitudes toward wolves were positively associated with identification as an environmentalist or animal rights advocate, trust in the USFWS, and mutualism and negatively associated with identification as a farmer or rancher and domination. Collectively, these factors explained ∼24% of the variance in attitudes toward wolves.
Identification as an environmentalist, trust in the USFWS, and mutualism were associated with a preference for wolf populations to be increased nationwide (Table 1). Identification as ideologically conservative was significantly associated with a preference for wolf populations to be decreased nationwide, as was higher educational attainment and identification as a hunter or farmer or rancher. Overall, these factors explained ∼27% of the variance in the acceptance of wolves.
Identification as an environmentalist or an animal rights advocate, trust in the USFWS, and mutualism were positively associated with perceived benefits, whereas identification as ideologically conservative, a hunter or farmer or rancher were negatively associated with perceived benefits (Table 1). These factors explained ∼29% of the variance in the perceived benefits of wolves.
Identification as ideologically conservative was positively associated with perceived risks of wolves, as was identification as a hunter or farmer or rancher (Table 1). In contrast, identification as an animal rights advocate and mutualism were negatively associated with perceived risks. Together, these factors explained 22% of the variance in perceived risks of wolves.
Identification as an environmentalist or animal rights advocate was positively associated with prior wolf conservation behaviors, indicating that these individuals were more likely to engage in stewardship behavior (Table 1). Identification as ideologically conservative, a hunter or farmer or rancher, higher educational attainment, and domination were negatively associated with prior wolf conservation, demonstrating that these participants were less likely to engage in stewardship behaviors. Jointly, these factors explained ∼20% of the variance in prior wolf conservation behavior.
Identification as an environmentalist or animal rights advocate, trust in the USFWS, and mutualism were positively  associated with behavioral intentions, indicating that these individuals were more likely to report an intention to engage in actions supporting wolves in the future. In contrast, identification as ideologically conservative, a hunter or farmer or rancher, higher educational attainment, and domination were negatively associated with behavioral intentions. Collectively, these factors explained ∼36% of the variance in wolf conservation behavioral intentions.

Logistic regression
Our examination of the likelihood that participants had previously engaged in at least one or more actions supporting wolf recovery indicated that the more positive attitude participants had toward wolves, the more likely they were to have engaged in stewardship behavior or behaviors in the past 6 months ( Table 2). Increased identification as an environmentalist or animal rights advocate was also associated with an increased likelihood of engagement in one or more behaviors supporting wolf recovery. In contrast, the less trust participants expressed in the USFWS and the fewer risks they perceived associated with wolves' expansion, the more likely they were to have engaged in stewardship behavior. Overall, the model correctly classified ∼87% of cases. Our assessment of the likelihood of prior engagement in at least one action opposing wolf recovery during the past 6 months showed that a greater domination wildlife value orientation and identification as a hunter were associated with an increased likelihood of intolerant behaviors (Table 2). In contrast, greater trust in the USFWS, a more positive attitude toward wolves, greater acceptance of the species, and greater perceived benefits of their population's expansion were associated with a reduction in the likelihood of engaging in actions opposing the species. Collectively, the model correctly classified ∼92% of cases.

DISCUSSION
The success of conservation efforts, particularly for contested species, such as the gray wolf, depends not only on the biological landscapes that species inhabit, but also on sociopolitical landscapes Ghoddousi et al., 2021;Manfredo et al., 2021;Treves & Karanth, 2003). Thus, calls for the meaningful integration of conservation social science research in management decision-making processes have proliferated (Bennett et al., 2017;Ghoddousi et al., 2021;Manfredo et al., 2019). However, inconsistent applications of core concepts related to human thought and behavior have raised concerns regarding the construct validity and comparability of previous efforts (Bruskotter et al., 2015;Knox et al., 2021;Whitehouse-Tedd et al., 2021). Ultimately demonstrating that the operational measures commonly used to gauge public perceptions and behaviors toward wolves do indeed yield different results across a sample of the same individuals, our study highlights a fundamental need for the intentional selection and iterative evaluation of measures employed in future conservation social science research. Despite moderate to strong correlations between all six of the psychological concepts tested (i.e., attitudes, acceptance, benefits, risks, prior behavior, and behavioral intentions), linear and logistic regression analyses revealed considerable variation in their respective relationships with an identical set of explanatory variables. Therefore, our findings offer important implications for survey development across a wide range of contexts to precisely capture the core concepts of interest and thus best meet the needs of the specific agency or organization of operation and the constituents whom they represent.
For example, although wildlife scholars tracking longitudinal trends in the public's general perception of wolves (Bruskotter et al., 2007;George et al., 2016;Treves et al., 2013) and various methods of their control (Slagle et al., 2017) may choose to measure attitudes given the availability of existing data needed for such comparisons, managers of a wildlife management authority in a democratized context, where public input is valued, may be most interested in gauging the public's response to future plans to increase wolf populations. As such, data collection efforts in this context may be best structured to assess residents' acceptance of the species (Behr et al., 2017;Niemiec et al., 2020). Despite relatively similar effects from values, trust, and social identity, political ideology was one of the strongest indicators of WAC--yet had a negligible association with attitudes. Because political ideology is often resisted, or even excluded from agency-based questionnaires, our findings suggest that supplementing analyses of WAC with secondary data regarding political ideology may be an important strategy.
Given the political polarization regarding a host of management decisions (Nie, 2004), a failure to integrate such information may adversely affect the accuracy of results obtained and thwart efforts to mitigate social conflict among diverse stakeholders due to a lack of understanding of such conflict's underlying forces (Hamilton et al., 2020;van Eden et al., 2021). Of course, considering that the terms used to define political ideologies (e.g., liberalism and conservatism) vary by country (Waytz et al., 2019), scholars must account for differences in semantics when measuring and interpreting the impact of such beliefs across broader contexts. Furthermore, because several other indicators of one's political identity exist (e.g., party affiliation and voting practices), future studies may benefit from the use of multidimensional measures to capture the complexity of one's political beliefs (and behaviors).
Conservation practitioners may also seek to identify groups with whom they could collaborate or partner with for future community-driven initiatives (Handy et al., 2014), such as a citizen science project. To accomplish this, the organization might benefit from acquiring a better understanding of residents' prior engagement in behaviors affecting the conservation of wolves or their estimated likelihood to do so in the future. Therefore, developing a study focused on factors influencing the propensity to act may be ideal. Results indicate that while most respondents in our sample were not active stewards of wolves, they were tolerant of the species (as indicated by low engagement in intolerant behaviors). This finding is consistent with the proposition that "most people, most of the time, will not be motivated to act" (i.e., inaction is typically the default [Bruskotter & Fulton, 2012:99;Slagle et al., 2022]). Drawing insight from the theory of planned behavior (Ajzen, 1991) and its myriad applications to wildlife conservation (reviewed in Miller [2017]), the lack of intolerant behavior may reflect a perceived inability to (or disinterest in) performing such behaviors or a social context that regards intolerant behaviors as non-normative.
Researchers have long used surveys to identify acceptance and attitudes--both of which can be indicators of one's general tolerance for a species. We found value in using multiple indicators of tolerance within the same study and operationalizing tolerance as both supportive and oppositional behaviors. Specifically, our results showed how supportive and oppositional behaviors--what Bruskotter and Fulton (2012) called "stewardship" and "intolerance"--were explained by different suites of variables. These differences challenge efforts to expand tolerance through communications. To wit, a variety of prior studies stress the importance of perceived benefits as a mechanism for increasing tolerance (e.g., Carter et al., 2012;Kansky et al., 2014), and experimental evidence supports the idea that communicating benefits can increase tolerance. However, these studies typically rely on WAC as a measure of tolerance. Our behavioral analyses showed that, even when controlling for independent effects of attitudes and WAC, perceiving the benefits of wolves was negatively associated with prior engagement in intolerant behaviors. Thus, consistent with prior research, our findings emphasize the importance of perceived benefits as a means of reducing intolerance. However, perceiving more benefits did not increase supportive behaviors when other factors were controlled. Rather, a reduced perception of risk was associated with increased support. This suggests that while benefits are important (for reducing intolerance), communicating about risk in cases where supportive actions are necessary for conservation is also needed. To do so effectively, messaging may place the significance or severity of said risk in relation to other potential risks, while simultaneously increasing one's perception of control by detailing specific behaviors that can be taken to achieve desired outcomes (Gore et al., 2009). Of course, it is essential that all relevant information--regarding both risks and benefits--be balanced in a way that is socially responsible, given the context of interest (Slagle et al., 2013).
Research on acceptance of risk (i.e., hazard acceptance) shows that having higher trust in the agency charged with managing risks associated with wildlife, such as carnivores, tends to increase the acceptance of that species. We found that increased trust was negatively associated with both supportive and oppositional behaviors; trusting individuals may feel less compelled to act on behalf of wildlife (Wald et al., 2019) or to engage at all (Li, 2015). That is, trust built upon shared values and perceived capability tempers a need for personal action--for or against wildlife. Thus, in cases where supportive advocacy is beneficial for the conservation status of a species, a focus on increasing the trust of the managing agency might actually reduce people's engagement in needed supportive actions (Slovic, 1993). The lesson is simple-how one communicates about carnivores depends upon the goals of one's conservation actions and the intended audience.
Recognizing that one's interpretation of new information is automatically filtered through existing schemas (or mental classification systems), it is important to consider the effect of other factors in shaping what people think and do--such as social identity (Carlson et al., 2020). For example, people with substantial knowledge and experience with wildlife may discount any discussion of the benefits derived from wolves' presence on the landscape due to conflicting preexisting beliefs. Thus, given people's propensity to reduce such dissonance (Festinger, 1962)--coupled with a tendency for benefits to be inversely correlated with risks in people's minds (Alhakami & Slovic, 1994), as evidenced in our own sample--practitioners may need to highlight the negative outcomes (i.e., risks) associated with the species' absence to boost support for conservation inter-ventions. This strategy may be critical given that self-identified hunters in our study were significantly more likely to have engaged in at least one or more actions opposing wolf recovery--a particularly salient finding given that those who hunt are the only group who reliably have the motivation, means, and opportunity to physically express their intolerance by killing a wolf (Bruskotter & Fulton, 2012;Ghoddousi et al., 2021;Treves, 2012).
Beyond directly affecting wolves, interest groups (e.g., farmers and ranchers) could affect conservation outcomes for species indirectly through political pressure or other means and across longer time spans. We sought to optimize the recall accuracy of participants during the past 6 months because "memory decreases over time, even when the event is relatively important and distinctive" (Schwarz & Oyserman, 2001:136). Thus, responses obtained do not necessarily reflect the entirety of participants' prior engagement in intolerant or stewardship behaviors. We encourage adaptation of the present sampling strategy to capture a wider range of behaviors for specific groups of interest and, where possible, to assess those behaviors across multiple points in time (i.e., employ a longitudinal approach). We also acknowledge that other mechanisms may shape beliefs and behaviors beyond those examined herein-including one's level of support for specific management actions (e.g., lethal control) (Manfredo et al., 2021) and the affective components of such judgments (Vaske et al., 2021). Similarly, despite controlling for a wide range of cognitive, demographic, and place-based characteristics--many of which have been quantified throughout the literature in equally diverse ways--the effect of some factors identified in other models (e.g., wildlife tolerance model [Kansky et al., 2016]) remains untested. Therefore, the standardization and incorporation of additional factors in future approaches may help explain variation in responses not presently accounted for.
By examining the relationship between what people think about wolves (i.e., cognitions), how they act toward the species (i.e., prior behavior), and how they say they will act in the future (i.e., behavioral intentions), our study offers a critical contribution to current understandings of the comparability of psychological concepts in the literature. With growing questions regarding the validity of measures used to capture a variety of conservation-related thoughts and behaviors (e.g., environmental concern [Cruz & Manata, 2020]), it is imperative that efforts to clarify the relationships between existing social science concepts continue to be pursued, especially in the context of survey development. The significant variation we found suggests such clarity in the operationalization of constructs is needed worldwide--in different cultural contexts, different languages, with different research methods, and across different species and spectrums of human-wildlife interaction. As we and others have demonstrated (e.g., Knox et al., 2021), different factors shape the operationalization of key social science constructs, which has critical implications for management and policy recommendations when social science results are incorporated into decision-making processes.