Specificity and the Cognitive Hierarchy:
Value Orientations and the Acceptability
of Urban Wildlife Management Actions
Confluence Research and Consulting, Anchorage, Alaska, USA
JERRY J. VASKE AND MICHAEL J. MANFREDO
Colorado State University, Fort Collins, Colorado, USA
This article tests theory suggesting cognitions at the same level of specificity have
stronger associations than those at different levels. Using data from a survey of
Anchorage, AK, residents (n¼971, response rate¼59%), we explored relationships
between general wildlife value orientations and (1) the general acceptability of
hunting urban wildlife populations, and (2) specific wildlife management actions
(e.g., the acceptability of destroying a bear or moose after specific conflict situa-
tions). Consistent with previous research, patterns of basic wildlife beliefs aligned
along two distinct value orientations (protection–use and wildlife appreciation) that
differentially predicted management action acceptability. As hypothesized, general
wildlife value orientations had more influence on the acceptability of hunting to re-
duce wildlife populations than destroying an animal involved in specific conflict
situations. Findings suggested ways to improve measurement, ways to develop
broader models that include values-related variables, and the importance of
values-level information when addressing urban wildlife conflicts.
Keywords cognitive hierarchy, specificity, urban wildlife management, wildlife
Popular media commonly assert that values influence environmental attitudes and=or
behaviors, but empirical evidence showing direct predictive validity is sparse. Stern
(2000), for example, suggests that basic environmental beliefs have varying effects on
specific forms of ‘‘environmental activism’’ (e.g., signing environmental petitions) and
‘‘private-sphere environmentalism’’ (e.g., recycling). Similarly, pro-hunting ‘‘values’’
may not predict support for wildlife management actions such as hunts to reduce
deer populations in suburban neighborhoods or aerial wolf hunting. ‘‘Biocentric
values’’ may not coincide with opposition to wildlife control actions such as destroying
Received 02 September 2004; accepted 27 July 2005.
Address correspondence to Doug Whittaker, Confluence Research and Consulting, 6324
Red Tree Circle, Anchorage, AK 99507, USA. E-mail: firstname.lastname@example.org
Society and Natural Resources, 19:515–530
Copyright # 2006 Taylor & Francis Group, LLC
ISSN: 0894-1920 print=1521-0723 online
Social-psychological theories offer explanations for these disparities, suggesting
that attitudes, beliefs, and norms mediate the relationships between values and beha-
vior (Schwartz 1992; Stern et al. 1999). While other theorists have considered contex-
tual (Guagnano et al. 1995), personal capability (Stern et al. 1999), or habitual
(Dahlstrand and Biel 1997) influences on environmental behavior, substantial interest
remains in developing models that explain the conditions under which fundamental
values affect environmental behavior or evaluations (Stern 2000). Models with greater
predictive power are necessary to affirm that values do influence how people think and
behave related to environmental issues.
At the heart of these models, social-psychological theory distinguishes stable but
abstract values (Rokeach 1973; Homer and Kahle 1988) from more specific cogni-
tions (e.g., attitudes and norms) that evaluate objects or situations encountered in
daily life (Eagly and Chaiken 1993). These cognitions are best understood as part
of a ‘‘hierarchy’’ from general to specific (Homer and Kahle 1988). Specific belief,
attitudinal, or normative variables are more likely to predict behaviors than more
general measures like values (Fishbein and Ajzen 1975; Fiske and Taylor 1991).
One research tradition has applied this ‘‘cognitive hierarchy’’ to evaluations and
behavior associated with wildlife (Fulton et al. 1996; Tarrant et al. 1997; Manfredo
et al. 1997; Zinn et al. 1998) and forest planning issues (Vaske and Donnelly 1999;
Vaske et al. 2001). These researchers also suggest that ‘‘value orientations’’ (patterns
among basic beliefs) provide a link between abstract values and more specific
A central concept in this model is specificity or correspondence among the mea-
sured variables. When correspondence between variables is similar (in terms of
target, action, context, and time), correlations between variables are predicted to
be larger (Fishbein and Ajzen 1975; Manfredo et al. 1998; 2004). General ‘‘wildlife
value orientations,’’ for example, should predict the general acceptability of hunting
better than acceptability responses to specific conflict situations (e.g., destroying [the
action] a moose [the target] in a suburban neighborhood [the context] during the fall
of 2005 [time]). While correlation does not prove causality, the relative strength of
relationships can help assess the merits of including variables, or tests for mediation,
in larger models.
This study examined the specificity hypothesis in a cognitive hierarchy model.
Beginning with measures of general wildlife value orientations, we assessed relation-
ships between value orientations and (1) the general acceptability of bear or moose
hunting, and (2) the specific acceptability of destroying a bear or moose involved in
specific conflict situations. Our objectives were to identify methodological improve-
ments for measuring value orientations, assess support for the specificity hypothesis,
examine when values-level variables are important to include in a cognitive hierarchy
model, and discuss management implications of differential influences of value
orientations on the acceptability of wildlife management actions.
Cognitions and behaviors can be organized into a hierarchy leading from general
values to specific attitudes, norms, and behaviors (Rokeach 1973; Fishbein and
Ajzen 1975; Schwartz 1992; Homer and Kahle 1988; Stern et al. 1999). These ele-
ments build upon one another in what has been described as an inverted pyramid
(Figure 1) (Fulton et al. 1996; Vaske and Donnelly 1999).
516D. Whittaker et al.
Values represent the most basic social cognitions and differ from other elements in
the model because they transcend situations and issues (Rokeach 1973).For example, a
person who holds ‘‘honesty’’ as a value would be expected to be honest when complet-
ing IRS tax forms, conducting business deals, or interacting with friends.
Because values are abstract, linking them to more specific cognitions or behavior
is difficult, leading to the inclusion of value orientations in the model (Fulton et al.
1996; Vaske and Donnelly 1999; Bright et al. 2000; Vaske et al. 2001). Value orienta-
tions refer to patterns of ‘‘basic beliefs’’ about general objects, which give meaning to
more abstract values (Manfredo et al. 2004). For example, while values measure the
extent to which people identify with abstract concepts like altruism or honesty, value
orientations explore patterns of beliefs about broad classes of objects (e.g., wildlife,
forests), which are thought to link back to underlying values-level cognitions.
With this research tradition, one wildlife value orientation focuses on a single
‘‘protection-use’’ continuum that is measured by the degree of agreement with a
series of statements about ‘‘wildlife use,’’ ‘‘wildlife rights,’’ ‘‘hunting,’’ and ‘‘fishing’’
(Fulton et al. 1996). Example statements include ‘‘humans should manage wild
animal populations so that humans benefit’’ (wildlife use) and ‘‘hunting helps people
appreciate wildlife and natural processes’’ (hunting).
The wildlife ‘‘protection–use’’ orientation is similar to the biocentric–anthropo-
centric value orientation continuum (Shindler et al. 1993; Steel et al. 1994; Thompson
and Barton 1994; Vaske and Donnelly 1999; Vaske et al. 2001) and the ‘‘New Environ-
mental Paradigm’’(NEP;VanLiere andDunlap1980).Dependingonthe author,these
latent constructs (i.e., protection–use, biocentric–anthropocentric, NEP) have been
labeled general environmental values, value orientations, general attitudes, and even
‘‘worldviews.’’ Regardless of the label, each concept fundamentally measures an under-
lying general ‘‘environmental values’’ dimension.
A second, separate value orientation in wildlife-oriented hierarchy models
focuses on a ‘‘wildlife appreciation’’ continuum, measured by the degree of agreement
to statements about ‘‘wildlife benefits’’ in ‘‘recreation’’ and ‘‘residential settings,’’
Figure 1. The cognitive hierarchy. Source: Vaske and Donnelly (1999).
Specificity and the Cognitive Hierarchy 517
‘‘interest in wildlife learning,’’ and ‘‘wildlife existence’’ (Fulton et al. 1996). Although
protection–use and appreciation reflect two primary orthogonal orientations,
‘‘protection–use’’ has been a better predictor of specific attitudes and norms related
to management actions (e.g., destroying nuisance wildlife, hunting urban wildlife,
and wildlife trapping) than ‘‘appreciation’’ (Bright et al. 2000; Gliner et al. 2001;
Manfredo et al. 2003; 2004). In contrast, appreciation value orientations appear to
predict wildlife learning and wildlife viewing recreation behavior better than protec-
tion–use (Fulton et al. 1996).
More specific constructs in this cognitive hierarchy model include attitudes and
norms. Attitudes are positive or negative evaluations of some object, while norms are
judgments about what is appropriate in a specific situation (Wittmann et al. 1998;
Zinn et al. 1998), or standards that individuals use to evaluate whether behavior
or conditions should occur (Vaske and Whittaker 2004). While there are measure-
ment and conceptual distinctions between attitudes and norms, both are fundamen-
tally evaluative variables that can vary considerably by the specificity of their object.
For example, if the object is ‘‘wolves,’’ the evaluation reflects the mix of cognitions
that form a general attitude or norm. If the object is ‘‘wolf reintroduction in Color-
ado during 2005,’’ the evaluation reflects a narrower context and time frame, and
thus a more specific evaluation.
Both attitudinal and normative research traditions also formally recognize situa-
tional elements (Ajzen and Fishbein 1980). For example, attitudes and norms toward
human–bear conflict responses may differ depending on the action (e.g., destroy vs.
relocate the bear), target (e.g., bear), context (bear kills a person vs. bear charges a per-
son), and location (conflict occurred in a wildland area vs. a suburban neighborhood).
In the cognitive hierarchy model examined here, value orientations influence
both attitudes and norms (Vaske et al. 2001). Zinn et al. (1998), for example, found
that individuals on the protection end of the protection–use continuum were less
willing to have agencies destroy an animal (a norm) across three different wildlife
species (beavers, coyotes, mountain lions) and situation contexts (e.g., seeing the
animal in a residential area, human injury or death caused by wildlife). Respon-
dents on the use end of the continuum were more accepting of this management
Based on the cognitive hierarchy’s specificity principle and prior wildlife value orien-
tation research (Fulton et al. 1996; Bright et al. 2000; Gliner et al. 2001; Manfredo
et al. 2003; 2004), this article advances three hypotheses:
H1: General, protection–use and appreciation wildlife value orientations
will be related to (1) the general acceptability of hunting and (2) the
specific acceptability of lethal management actions after a specific
H2: General wildlife value orientations will have a greater influence on
the general acceptability of hunting than the specific acceptability
of killing an animal after specific conflict situations.
H3: Protection–use orientations will have stronger influences than
appreciation orientations on the acceptability of lethal wildlife
518 D. Whittaker et al.
Study Area and Context
Increasing suburban development generates conflict between humans and wildlife by
introducing more people into natural areas or ‘‘artificially’’ increasing wildlife popula-
tions that benefit from human-modified environments. When human communities
become intolerant of wildlife conflict levels, an ‘‘evaluative standard’’ (Manfredo et al.
and Purdy 1988) has been exceeded. To address these problems, wildlife managers typi-
callychooseamongseveral general prevention-orientedapproaches thatmodifyhuman
behaviors that may cause wildlife conflicts (e.g., develop bear proof garbage containers
to reduce food conditioning) or reduce overall wildlife population levels. If these
approaches fail, however, managers have specific options to (1) capture and relocate
the animal involved, (2) harass the animal from the area, or (3) destroy the animal.
Each of these actions has social and biological consequences and the potential
for controversy; public acceptability of these general and specific actions may limit
management options (Decker et al. 2004). For example, proposed moose hunts have
been stalled by a lack of public support (Donnelly and Vaske 1995) and ballot initia-
tives have changed hunting or trapping policies toward predator species (Manfredo
et al. 1997; Loker et al. 1998).
Anchorage, AK (population 260,000), faces several diverse human–wildlife con-
flict issues, particularly related to moose and bears. With extensive open space,
Anchorage is home to approximately 500 to 700 resident moose, 250 resident black
bears, and 60 brown bears (Alaska Department of Fish and Game [ADF&G] 2000).
Wildlife provide a variety of benefits to city residents, but they can create problems
(Whittaker and Manfredo 1997; Whittaker et al. 2001). For example, an average of 156
2000), they damage residential landscaping, and pose safety risks to residents and their
pets (ADF&G 1995). Similarly, bears frequently raid neighborhood food sources (e.g.,
garbage cans, bird feeders) and present a potential hazard on the city’s trail system,
along salmon streams, and during moose calving in the early summer (ADF&G 2000).
Several wildlife incidents in the mid-1990s (including human deaths from bears
and moose) led to concern about conflict levels in Anchorage, resulting in joint fed-
eral, state, and local research and planning efforts from 1996 to 2001. A survey of
residents on wildlife issues during the effort provided data for this study (Whittaker
and Manfredo 1997).
Data came from a 1996 mail survey of Anchorage residents (18 years of age and
older). The sample was drawn from registered voter lists (Whittaker and Manfredo
1997). Of the 1654 individuals contacted, 971 completed and returned surveys
(response rate¼59%). A telephone survey of nonrespondents (n¼107) allowed test-
ing for nonresponse bias on two participation, two attitudinal, and one demographic
variable; results suggested respondent=nonrespondent differences were not signifi-
cant and data were not weighted.
Following Fulton et al. (1996), wildlife value orientations were computed from a two-
step process. The first step used 28 belief variables to measure 8 wildlife dimensions:
Specificity and the Cognitive Hierarchy519
(1) wildlife use, (2) hunting, (3) wildlife rights, (4) wildlife welfare, (5) wildlife appreci-
ation in recreation settings, (6) appreciation in residential settings, (7) wildlife learn-
ing, and (8) wildlife existence and bequest value. Each of the scales contained two to
six items. Each variable was measured on a 7-point agree–disagree scale. The second
step combined these eight variables to measure two value orientations: (1) wildlife
protection–use, and (2) wildlife appreciation.
The survey asked two sets of questions about the acceptability of wildlife
management actions, each measured on a 7-point scale. The first measured the gen-
eral acceptability of public hunts to reduce moose or bear populations (i.e., general
population-level conflict prevention actions). The second measured the acceptability
of destroying individual animals in response to 14 distinct moose or bear encounter
situations. These situations ranged from the benign (e.g., a moose eating landscaping
in a residential neighborhood, a bear observed along a popular hiking trail) to the
severe (e.g., a moose tramples a person to death, a bear with a history of aggression
kills a person). There were 14 situations (5 moose, 9 bear) that represented a range
of encounter situations that occur in Anchorage. Situation descriptions included
information about the type, location, and timing of the hypothetical encounter.
A two-step confirmatory factor analysis (CFA) tested the wildlife value orientations
measurement model. A first-order factor analysis examined whether 28 items (observed
variables) loaded on 8 attitudinal dimensions (latent variables). A second-order
factor analysis examined how these eight scales loaded on two value orientation
dimensions. Both analyses were performed on the variance–covariance matrices using
LISREL 8.5. Four statistics (RMR, GFI, NFI, CFI) were used to assess each model’s
goodness of fit (Jo ¨reskog and So ¨rbom 1993). Cronbach’s alpha was also computed for
each belief and orientation scale.
Structural equation modeling explored relationships between the two value
orientations and action evaluations (both general and specific). Standardized path
coefficients and effect sizes (multiple Rs) are presented for each model. A single
descriptive statistic for each action (the percent of the sample reporting each action
as acceptable) is provided for context.
Wildlife Value Orientations
CFA supported the measurement models for general wildlife attitudes beliefs and
value orientations. In the first-order factor analysis, each of the 28 items loaded
on the 8 general attitude dimensions, with factor loadings ranging from .52 to .86
(Table 1). The goodness-of-fit indices suggested that the data fit the proposed models
(GFI, NFI, and CFI?.90, RMR<.05). Reliabilities for the 8 scales were all over .60
(7 of 8 Cronbach alphas>.72).
In the second-order factor analysis, the 8 attitude scales loaded on the 2 value
orientations with factor loadings ranging from .62 to .86 (Table 2). Acceptable fit
indices (GFI, NFI, and CFI?.93, RMR¼.079) were also observed. The reliabilities
for the protection–use and appreciation value orientations were .80 and .84, respect-
ively. Taken together, the CFA and reliability analyses supported the measurement
model for the two value orientations.
520 D. Whittaker et al.
Table 1. Items used to measure eight basic wildlife beliefs
Basic belief dimensions and items
comprising each scalea
Humans should manage wild animal populations so
that humans benefit.
It is acceptable for human use to cause the loss of
some individual animals as long as populations are
It is important for humans to manage the populations
of wild animals.
Hunting helps people enjoy the outdoors in a positive
Hunting is cruel and inhumane to animals.?
Hunting helps people appreciate wildlife and natural
Hunting makes people insensitive to suffering.?
I would be offended or upset if I saw an animal killed
or injured by a hunter.?
In general, I think hunting is a safe outdoor activity.
Although wildlife may have certain rights, most needs
that people have are more important than the rights
The needs of people are always more important than
any rights that wildlife may have.
The rights of people and the rights of wildlife are
People should not treat wildlife in ways that may
cause pain and suffering, regardless of how much
we may benefit.
If we can’t minimize the pain and suffering caused to
wildlife by an activity, then we should not allow
I enjoy watching wildlife when I take a trip outdoors.
Some of my most memorable outdoor experiences
occurred when I saw wildlife I didn’t expect to see.
The opportunity to see wildlife is one of the reasons I
take trips outdoors (like camping, hiking, or
I’m interested in making the area around my home
attractive to wildlife.
Specificity and the Cognitive Hierarchy 521
Table 1. Continued
Basic belief dimensions and items
comprising each scalea
I notice the moose, birds, and other wildlife around
me every day.
Having wildlife around my home is important to me.
An important part of my community is the wildlife I
It is important that we learn as much as we can about
I enjoy learning about wildlife.
It is important that all Alaskans have a chance to
learn about wildlife.
We should be sure future generations of Alaskans will
have an abundance of wildlife.
It is important to maintain wildlife so that future
generations can enjoy them.
It’s important to know that there are healthy
populations of wildlife in Alaska.
Whether or not I get out to see wildlife as much as I
like, it is important to know they exist in Alaska.
aResponses were on 7-point scales from ‘‘strongly agree’’ to ‘‘strongly disagree,’’ with a ‘‘no
opinion’’ mid-point. Starred items were reverse coded in the confirmatory factor analysis.
bFit statistics: v2¼1,443 with 319 degrees of freedom; RMR¼.047; GFI¼.90; NFI¼.91;
Table 2. Basic wildlife belief indices and value orientations
Wildlife value orientations and basic
beliefs comprising each scalea
Number of items in
basic belief scale
Protection–use value orientation
Appreciation value orientation
Wildlife existence=bequest value
aStarred belief dimensions were reverse coded in the confirmatory factor analysis.
bFit statistics: v2¼234.5 with 17 degrees of freedom; RMR¼.079; GFI¼.95; NFI¼.93;
522 D. Whittaker et al.
Relating Value Orientations and Action Evaluations
A series of structural equation models assessed relationships between the two
general wildlife value orientations and the acceptability of (1) public hunts in general
(Table 3), and (2) the specific acceptability of destroying a bear or moose after
specific conflict situations (Table 4). As hypothesized (H1), both value orientations
had significant effects on all the acceptability evaluations, with multiple R’s ranging
from .40 to .73. Multiple Rs of this magnitude reflect typical to substantial effects
(Vaske et al. 2002).
Consistent with hypothesis 2, the wildlife value orientations predicted the
acceptability of public hunts in general better than they predicted the acceptability
of destroying an animal after specific conflict situations (comparing Tables 3 and 4).
Relationships between value orientations and the acceptability of hunting in general
produced Rs ranging from .67 to .73, whereas relationships between the value orien-
tations and acceptability of destroying a specific animal ranged from R¼.46 to .52
for moose situations, and from R¼.40 to .57 for bear situations. Fisher Z pairwise
comparisons of the multiple Rs indicated that each of the general hunting equations
explained more of the variance than the situation-specific equations, even after
controlling for multiple tests (p<.001 in all comparisons).
Differential Effects of the Two Value Orientations
Consistent with H3, the protection–use orientations had a greater influence than
appreciation orientations on the acceptability of all of the general hunts (Table 3),
for three of the five moose situations (Table 4), and for eight of the nine bear situa-
tions (Table 4). For example, the influence of protection–use evaluations was much
stronger than appreciation in the general hunt models (b?.69 for protection–use,
b<.07 for appreciation). In addition, although the signs for the protection–
use paths in the hunt models were in the expected direction (i.e., people on the use
end of the protection–use scale were more likely to accept hunting), the signs for
the appreciation paths were not (i.e., people with higher appreciation were equal
or more likely to accept hunting actions).
Table 3. Influence of value orientations on the general management actions
Standardized regression coefficientsa
hunt is acceptableb
Public hunting to
reduce the number of:
aAll standardized regression coefficients were significant (p<.05).
bPercent reporting slightly, moderately, or highly acceptable on 7-point scale; when used as
continuous variable in models, higher acceptability score¼more acceptable.
cHigher protection–use scores¼more use oriented; lower scores¼more protection oriented.
dHigher appreciation scores¼higher appreciation orientation.
Specificity and the Cognitive Hierarchy523
Table 4. Influence of value orientations on the acceptability of destroying animals
Standardized regression coefficientsa
Specific conflict situation
Moose eating landscaping in
Moose blocking school
children in neighborhood.
Moose repeatedly charges
cross-country skiers on a
popular trail (but no
Moose charges and knocks
down hiker on park trail.
Moose tramples and kills a
person in neighborhood.
Brown bear repeatedly seen
on park trail.
Brown bear repeatedly seen
in a neighborhood.
Brown bear feeding on
moose carcass on park
Black bear repeatedly
getting into trash, etc. in
Bear kills several pets in
Bear charges=knocks down
hiker on trail, then leaves.
Bear kills a person on park
trail (no other details).
Bear kills person on a trail,
but bear is a sow with
Bear kills person on a trail,
but bear has history of
aAll standardized regression coefficients were significant (p<.05).
bPercent reporting slightly, moderately, or highly acceptable from 7-point scale; when used
as continuous variable in models, higher acceptability score¼more acceptable.
cHigher protection–use scores¼more use oriented; lower scores¼more protection oriented.
dHigher appreciation scores¼higher appreciation orientation.
524D. Whittaker et al.
For specific moose situations, both protection–use and appreciation orientations
had significant effects, although protection–use orientations were usually stronger
influences. Protection–use coefficients ranged from b¼.25 to .37, whereas appreci-
ation coefficients ranged from b¼.18 to .33. The highest protection–use influence
and the lowest appreciation influence were for the ‘‘moose kills a person’’ situation,
where a majority accepted destroying the moose. Path signs for both protection–use
and appreciation were in the expected directions for all of the moose models—people
on the protection end of the protection–use scale and with higher appreciation orien-
tations were less likely to accept destroying a moose.
For the specific bear situations, both protection–use and appreciation orienta-
tions were linked to acceptability, although protection–use was slightly stronger.
Protection–use coefficients ranged from b¼.25 to .46 and averaged .38, whereas
appreciation coefficients ranged from b¼.06 to .36 and averaged .22. There were
two notable patterns among coefficients in the different bear models: (1) Appreciation
predicted more than protection–use for only one model (bear feeding on a moose
carcass), and (2) appreciation paths were generally lower for more severe conflict
situations, with one exception (sow with cubs kills a human). Path signs for all models
were in the hypothesized directions—people on the protection end of the protection–
use continuum and with higher appreciation orientations were less likely to accept
destroying a bear.
Results suggest several conclusions about the cognitive hierarchy’s specificity prin-
ciple, wildlife value orientations, and their influence on the acceptability of urban
wildlife management actions. Discussion is organized by theoretical, measurement,
and management issues.
Results provide support for the cognitive hierarchy’s specificity principle. First, as
hypothesized, the findings showed links between elements in the cognitive hierarchy
(Homer and Kahle 1988; Fulton et al. 1996; Vaske and Donnelly 1999), and revealed
stronger relationships between general value orientations and general management
actions (e.g., hunting in general) than for specific actions (e.g., acceptability of
destroying an animal after specific human–wildlife conflict situations).
Support for the specificity principle has implications for developing broader
synthesized models that explain environmental attitudes or behavior (Stern 2000).
If one is trying to predict public acceptability of specific management alternatives,
value orientations might not be the best place to start. However, developing broader
theories about how people think about natural resource issues and how those influ-
ence a range of specific environmental attitudes and behaviors requires attention to
values-level information (Manfredo et al. 2004). Specific attitudes and norms are less
stable, context specific, and may not link strongly with fundamental beliefs. To
understand actual population shifts in values and general attitudes toward wildlife,
one needs to measure broader, more abstract, and more stable variables like value
orientations (Manfredo et al. 2003). A more complete understanding of public atti-
tudes and norms related to wildlife issues requires conceptualizing, measuring, classi-
fying, and ordering a full range of variables from the general to the specific.
Specificity and the Cognitive Hierarchy525
Second, findings indicate there are some strong associations between value
orientations and acceptability for this wildlife management topic (urban wildlife
conflict). Fulton et al. (1996) demonstrated similarly strong links between value orien-
tations and wildlife recreation participation, and further suggested value orientations
would predict attitudes or behaviors across several other wildlife issues. These data
join a growing list of studies that support the assertion for wildlife or other environ-
mental issues (Manfredo et al. 2004), suggesting wildlife value orientations deserve
consideration in these types of models.
The cognitive hierarchy model tested here is not the only version that organizes
cognitive variables from values to behavior, or that relies on the specificity principle.
Building on conceptual work linking environmentalism to values (Heberlein 1972;
Schwartz 1977), Stern et al. (1999) have developed a values–beliefs–norms (VBN)
activation model, Stern et al. suggest that an individual’s beliefs (e.g., awareness of
environment consequences, ascription of personal responsibility) mediate the relation-
shipbetweenvaluesandbehavior.The model alsoincorporatesabroadenvironmental
belief measure, the NEP (Dunlap and Van Liere 1978), which clearly taps underlying
dimensions of anthropocentric–biocentric and protection–use value orientations.
These different models offer competing conceptualizations, but there is consider-
able overlap and complimentary ideas. We support Stern’s (2000) suggestion that the
time is ripe for syntheses that are more complex but comprehensive. Successful
models linking values to behaviors likely will have to account for several layers of
specificity. Comparisons and subsequent filtering of various ‘‘environmental values,’’
‘‘value orientations,’’ ‘‘general attitudes,’’ or ‘‘worldview’’ variables will be important
for the sake of parsimony. Similarly, increased clarity about labeling and measure-
ment is needed in the literature.
Third, our findings indicate that the acceptability of wildlife actions is more
values-based for some actions than others, with the relative contributions of the
protection–use and appreciation orientations varying by action. For example,
protection–use orientations influenced the acceptability of public hunts more than
appreciation orientations. This may be related to an ‘‘action-severity’’ effect that
increases the influence of protection–use orientations when management actions
are likely to have large consequences for a wildlife population (e.g., hunting, lethal
Relative differences in the influence of the two value orientations were also
evident in the conflict situation results. Protection–use had greater influence than
appreciation in ‘‘severe’’ conflict situations (e.g., a bear kills a person), but the
differences narrowed when conflict situations were benign or involved viewing
opportunities (e.g., bear seen on a park trail). These findings may also suggest a
‘‘species-hazard’’ effect where ‘‘more dangerous’’ animals (e.g., bears in the more
severe situations, a moose that has killed a person) are considered differently from
‘‘less dangerous’’ animals. With the former, a protection–use orientation was domi-
nant; with the latter, appreciation orientations also apply. Similarly, a ‘‘viewing
opportunity’’ effect may help explain why hunting acceptability was strongly linked
with high appreciation orientations in situations that provided quality viewing
opportunities (as long as the situation seemed benign).
Although these ‘‘effects’’ are speculative, the relative influence of protection–use
and appreciation orientations suggests opportunities for future research. The search
for consistent relationships between general variables (e.g., value orientations) and
526 D. Whittaker et al.
issue-specific evaluations (e.g., acceptability of management actions) is likely to be
more successful when the types of actions or situations in which they apply are classi-
fied. Action severity, situation severity, and consequences on viewing opportunities
offer ways to distinguish the concepts.
Findings suggest that methods for measuring wildlife value orientations demonstrate
discriminant validity among the wildlife basic beliefs and reveal two underlying value
orientation dimensions. Consistent with similar research (Fulton et al. 1996; Bright
et al. 2000; Manfredo et al. 2003; 2004), these procedures offer a reliable and valid
approach to measuring wildlife value orientations.
Although the data fit the measurement model, there are opportunities for
improvement. The beliefs measured here may be an incomplete sample and the
inclusion of others may identify additional wildlife value orientations (Fulton et al.
1996). For example, beliefs about wildlife as a source of subsistence or the religious
importance of wildlife could underlie a wildlife spirituality orientation (Manfredo
and Fulton 1997; DeRuiter and Donnelly 2002). Research exploring cross-cultural
differences could also be illuminating (Manfredo and Dayer 2004).
Individual belief items can always be improved through additional testing, and
development of a smaller subset to measure value orientations would be more par-
simonious. Zinn et al. (1998) have made a start in this direction, but more systematic
work is needed. There is also a need to develop alternative measurement methods to
address construct validity issues.
Findings from this research have implications for planners addressing urban wildlife
conflicts. Although descriptive information can identify acceptable prevention
actions or conflict response policies, the results offer additional understanding of
those evaluations in at least three ways that are relevant to management.
First, results confirm that values influence the acceptability of some wildlife man-
agement actions and that protection–use and appreciation orientations are differen-
tially relevant to specific actions. This finding is important because most planning
processes require actions to link with broader goals in a plan. Agency goals are quali-
tative, abstract statements about what managers are trying to provide and they should
reflect values-level concepts. However, more specific management objectives tier down
from these broader goals and ultimately must be linked with actions. Research
illuminating these links can help planners articulate values and value orientations,
and describe how those may fit with various management strategies and programs.
We are sympathetic to managers who express opposition to ‘‘management by
polling,’’ because the purpose and promise of social research are decidedly different
from those of taking a vote (Manfredo et al. 2004). This research highlights why
choosing management actions based solely on their public acceptability is simplistic,
and that de facto values-level decisions come attached to those choices (even if man-
agers don’t acknowledge them). More sophisticated models that document links
among a diversity of general and specific cognitions clarify the need for plans to have
internally consistent series of goals, objectives, and actions that express a defensible
philosophy of natural resources management.
Specificity and the Cognitive Hierarchy 527
Second, descriptive studies can identify divisions related to public acceptability,
but information about the degree that they are values-based may indicate whether
controversies are likely to be resolved. Values are more resistant to change than
specific evaluations (Rokeach 1973; Stern 2000; Manfredo et al. 2003; 2004), so
resolving values-based issues is more difficult. This finding characterizes planning
efforts associated with a proposed moose hunt in Anchorage. Although a majority
of residents supported the hunt (51% supported, 34% opposed), there was consider-
able diversity in beliefs among those for and against (Whittaker et al. 2001). In
collaborative planning sessions, some disagreements among stakeholders were
specific to various hunt proposals, but the fundamental lack of consensus appeared
to reflect differences in core values about the appropriateness of using or protecting
wildlife, as predicted in these findings.
As managers search for ‘‘elegant solutions’’ to controversial natural resource
problems (i.e., actions that a diversity of stakeholders and a majority of the public
can accept), it is crucial to link decisions to articulated values positions (particularly
if those can be aligned with legislative and administrative mandates). Simple compro-
mises or ‘‘deals’’ without links to a foundation of more general values are more likely
to unravel, and may be open to challenge as ‘‘arbitrary and capricious,’’ potentially
feeding a controversy rather than resolving it. More deliberative decisions that tie a
hierarchy of cognitions to the action clarify which ‘‘side’’ is being chosen, and com-
municate the logic behind it. Controversy and poor management decisions are still
possible in these situations, but the effort may increase transparency and raise the
level of debate.
Finally, information about value orientations and acceptability can help
planners address issues when specific survey data are unavailable. It is not possible
to collect acceptability information for every possible wildlife conflict action, but
these data suggest that may not be necessary. Results show consistency among rela-
tionships between values and acceptability within classes of actions (e.g., hunting,
situations involving the death of a human), so planners may be able to extend infor-
mation about the known acceptability of some actions to those for other species
or situations. By understanding the structure of cognitions about wildlife issues,
planners have the ability to anticipate public reaction as new issues emerge.
A decade on from the wildlife conflicts that initiated this study and subsequent
planning, Anchorage continues to have wildlife conflicts with bears, moose, and other
species. Several actions in the wildlife plan have been implemented, while others
remain ‘‘good ideas’’ rather than imminent programs, but controversies erupt with
the occasional (and seemingly inevitable) highly publicized incident (e.g., a bear or
moose injures a person) or human=management response (e.g., the moose or bear is
or is not destroyed). With most of the ‘‘noise’’ in these controversies generated by
minorityadvocates atthe bothends of the protection–usevalue orientationcontinuum,
this type of research puts advocacy positions into perspective, and provides structured
arguments for more balanced policies that link to more broadly shared values.
Cliffs, NJ: Prentice Hall.
Alaska Department of Fish and Game. 1995. Moose management in the anchorage manage-
ment area: A discussion paper. Anchorage, AK: Division of Wildlife Conservation.
528D. Whittaker et al.
Alaska Department of Fish and Game. 2000. Living with wildlife in anchorage: A cooperative Download full-text
planning effort. Anchorage, AK: Division of Wildlife Conservation, January.
Bright, A. D., M. J. Manfredo, and D. C. Fulton. 2000. Segmenting the public: An application
of value orientations to wildlife planning in Colorado. Wildl. Society Bull. 28(1):218–226.
Dahlstrand, U. and A. Biel. 1997. Pro-environment habits: Propensity levels in behavioral
change. J. Appl. Social Psychol. 27:588–601.
Decker, D. J. and K. G. Purdy. 1988. Toward a concept of wildlife acceptance capacity in
wildlife management. Wildl. Society Bull. 16(1):53–57.
Decker, D. J., T. L. Brown, J. J. Vaske, and M. J. Manfredo. 2004. Human dimensions
of wildlife management. In Society and natural resources: A summary of knowledge,
eds. M. J. Manfredo, J. J. Vaske, B. L. Bruyere, D. R. Field, and P. Brown, 187–198.
Jefferson, MO: Modern Litho.
DeRuiter, D. S. and M. P. Donnelly. 2002. A qualitative approach to measuring determinants
of wildlife value orientations. Hum. Dimens. Wildl. 7(4):251–271.
Donnelly, M. P. and J. J. Vaske. 1995. Predicting attitudes toward a proposed moose hunt.
Society Nat. Resources 8(4):307–319.
Dunlap, R. and K. Van Liere. 1978. The new environmental paradigm. J. Environ. Educ.
Eagly, A. H. and S. Chaiken. 1993. The psychology of attitudes. New York: Harcourt Brace
Fishbein, M. and I. Ajzen. 1975. Belief, attitude, intention, and behavior: An introduction to
theory and research. Reading, MA: Addison-Wesley.
Fiske, S. T. and S. E. Taylor. 1991. Social cognition. New York: McGraw-Hill.
Fulton, D. C., M. J. Manfredo, and J. Lipscomb. 1996. Wildlife value orientations: A concep-
tual and measurement approach. Hum. Dimens. Wildl. 1(2):22–47.
Guagnano, G. A., P. C. Stern, and T. Dietz. 1995. Influences on attitude-behavior relation-
ships: A natural experiment with curbside recycling. Environ. Behav. 27(5):699–718.
Gliner, J. A., J. J. Vaske, and G. A. Morgan. 2001. Null hypothesis significance testing: Effect
size matters. Hum. Dimens. Wildl. 6(4):291–301.
Heberlein, T. A. 1972. The land ethic realized: Some social psychological explanations of
changing environmental attitudes. J. Social Issues 28(4):79–87.
Homer, P. M. and L. R. Kahle. 1988. A structural equation test of the value–attitude–behavior
hierarchy. J. Personality Social Psychol. 54(4):638–646.
Jo ¨reskog, K. G. and D. So ¨rbom. 1993. LISREL 8: Structural equation modeling with the SIM-
PLIS command language. Hillsdale, NJ: Lawrence Erlbaum Associates.
Loker, C. A., D. J. Decker, and L. C. Chase. 1998. Ballot initiatives—antithesis of human
dimensions approaches or catalyst for change? Hum. Dimens. Wildl. 3(2):8–20.
Manfredo, M. J. and A. A. Dayer. 2004. Concepts for exploring the social aspects of human–
wildlife conflict in a global context. Hum. Dimens. Wildl. 9(4):317–322.
Manfredo, M. J. and D. Fulton. 1997. A comparison of wildlife values in Belize and Colorado.
Hum. Dimens. Wildl. 2(2):62–63.
Manfredo, M. J., T. L. Teel, and A. D. Bright. 2003. Why are public values toward wildlife
changing? Hum. Dimens. Wildl. 8(4):287–306.
Manfredo, M. J., D. C. Fulton, and C. L. Pierce. 1997. Understanding voter behavior on wild-
life ballot initiatives: Colorado’s trapping amendment. Hum. Dimens. Wildl. 2(4):22–39.
Manfredo, M. J., T. L. Teel, and A. D. Bright. 2004. Applications of the concept of values and
attitudes in human dimensions of natural resources research. In Society and natural
resources: A summary of knowledge, eds. M. J. Manfredo, J. J. Vaske, B. L. Bruyere,
D. R. Field, and P. Brown, 271–282. Jefferson, MO: Modern Litho.
Manfredo, M. J., H. C. Zinn, L. Sikorowski, and J. Jones. 1998. Public acceptance of moun-
tain lion management: A case study of Denver, Colorado, and nearby foothills areas.
Wildl. Society Bull. 26(4):964–970.
Rokeach, M. 1973. The nature of human values. New York: Free Press.
Schwartz, S. H. 1977. Normative influences on altruism. In Advances in experimental social
psychology, ed. L. Berkowitz, vol. 10, 221–279. New York: Academic Press.
Schwartz, S. H. 1992. Universals in the content and structure of values: Theoretical advances
and empirical tests in 20 countries. In Advances in experimental social psychology,
ed. M. P. Zanna, vol. 25, 1–66. San Diego, CA: Academic Press.
Specificity and the Cognitive Hierarchy529