This article was downloaded by: [Fac Psicologia/Biblioteca]

On: 25 September 2012, At: 00:20

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

Human Dimensions of
Wildlife: An International
Journal
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/uhdw20

Development and
Validation of a
Specialization Index and
Testing of Specialization
Theory
Ronald J. Salz, David K. Loomis & Kelly L.
Finn
Version of record first published: 29 Oct 2010.

To cite this article: Ronald J. Salz, David K. Loomis & Kelly L. Finn (2001):
Development and Validation of a Specialization Index and Testing of
Specialization Theory, Human Dimensions of Wildlife: An International Journal,
6:4, 239-258
To link to this article: http://dx.doi.org/10.1080/108712001753473939

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/
terms-and-conditions
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,

reselling, loan, sub-licensing, systematic supply, or distribution in any

form to anyone is expressly forbidden.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

The publisher does not give any warranty express or implied or make
any representation that the contents will be complete or accurate
or up to date. The accuracy of any instructions, formulae, and drug
doses should be independently verified with primary sources. The
publisher shall not be liable for any loss, actions, claims, proceedings,
demand, or costs or damages whatsoever or howsoever caused arising
directly or indirectly in connection with or arising out of the use of this
material.

Human Dimensions of Wildlife, 6:239258, 2001

Copyright 2001 Taylor & Francis
1087-1209 /01 $12.00 + .00

Peer-Reviewed Articles

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization

Index and Testing of Specialization Theory
RONALD J. SALZ
DAVID K. LOOMIS
Human Dimensions Research Unit
University of Massachusetts-Amherst
Amherst, Massachusetts, USA

KELLY L. FINN
California Department of Transportation
San Diego, California, USA
Recreation specialization can be viewed as a continuum of behavior from
the general to the particular. Along this continuum, participants can be located into meaningful subgroups based on specific criteria. Previous studies have defined, measured, and segmented specialization groups in a variety of ways. The research reported here builds on the Ditton, Loomis, and
Choi reconceptualization of recreation specialization. A specialization index was developed to segment anglers into four groups based on their orientation, experiences, relationships, and commitment. Internal validation
analysis supported the use of this specialization index as a tool for angler
segmentation. Subsequent hypotheses tested for differences among specialization groups in frequency of participation, importance of activity and
nonactivity-specific elements, support for management regulations, and sidebets. Results provide strong support for the conceptual framework developed
by Ditton et al. These findings indicate a multidimensional index can be
used to segment anglers into discreet, meaningful specialization categories.
Keywords Recreation specialization, segmentation, specialization index,
anglers

Partial funding for this project was provided by the Cooperative State Research, Extension,
Education Service, U.S. Department of Agriculture, Massachusetts Agricultural Experiment Station
Project Number 782.
Address correspondence to David K. Loomis, Department of Natural Resources Conservation,
Human Dimensions Research Unit, Holdsworth National Resources Center, University of Massachusetts, Amherst, MA 01003-4210. E-mail: Loomis@forwild.umass.edu

239

240

R. J. Salz et al.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Introduction
Outdoor recreation participants generally display wide variation in their experiences, avidity, expertise, commitment, economic expenditures, and social interactions related to a particular activity. Connected to this variation are important
sociological and psychological differences affecting motivations, expectations,
desired outcomes, satisfaction levels, perceptions, and social norms. Outdoor recreation managers must recognize and accommodate these differences to provide
satisfactory experiences to a widely diverse clientele. Recreation specialization is
an area of study that attempts to describe this variation through segmentation of
participants into meaningful and identifiable subgroups. Bryan (1977) was the
first to conceptualize recreational specialization as a continuum of behavior from
the general to the particular, reflected by equipment and skills used and activity
setting preferences. The four levels of specialization he identified in a population
of trout anglers were occasional anglers, generalists, technique specialists, and
technique-setting specialists. Bryan (1977) suggested that more highly specialized anglers are part of a leisure social world with a shared sense of group identification derived from similar attitudes, beliefs, and experiences.
Recreation specialization studies following Bryan used a variety of classification techniques and variables to segment participants into specialization levels.
Some studies found that a single-item measure of specialization could be used to
segment participants. For example, Graefe (1980) noted that frequency of participation (i.e., avidity) was a useful surrogate for measuring angler specialization. He found that anglers who fished more frequently (i.e., were more specialized) had higher self-reported skill levels, participated in more diverse fishing
settings, and had a greater dependency on the resource. Ditton, Loomis, and Choi
(1992) also used avidity to segment recreational anglers into four specialization
levels. Similarly, Schreyer, Lime, and Williams (1984) used total number of river
runs as a means of classifying river users into six groups and found differences
between the groups in the type of prior river experience, motives for participation,
perceptions of conflict, and support for managerial regulations.
Other studies took a multidimensional approach to recreation specialization
by incorporating several variables into a specialization index. Chipman and Helfrich
(1988) concluded that investment, consumptive habits, and frequency of participation were important characteristics for determining specialization among anglers. Kauffman and Graefe (1984) used preferences for river characteristics to
segment canoeists into more-specialized and less-specialized groups. Fedler and
Ditton (1986) segmented anglers into levels of consumptive orientation based on
responses to statements regarding the importance of catching fish. Wellman,
Roggenbuck, and Smith (1982) used a specialization index based on equipment
investment, past experience, and centrality to lifestyle to segment anglers into
groups that reflect respondents attitudes toward depreciative behavior. Virden
and Schreyer (1988) constructed a specialization index to segment hikers based
on equipment and economic commitment, centrality to lifestyle, general experience, and past experience variables.

241

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization Index

Using a variety of segmentation methods, recreation specialization studies

showed that more specialized users differed from less-specialized users on numerous attributes. These included motives for participation (Kauffman & Graefe,
1984; Schreyer et al., 1984), importance of nonactivity-specific elements (Fedler
& Ditton, 1986), preferences for management strategies (Chipman & Helfrich,
1988; Hammitt & McDonald, 1983), perceptions about crowding (Vaske, Donnelly,
& Heberlein, 1978), environmental preferences (Kauffman & Graefe, 1984;
Schreyer et al. 1984; Virden & Schreyer, 1988), equipment ownership and use
(Chipman & Helfrich, 1988; Wellman et al., 1982), and centrality to lifestyle
(Virden & Schreyer, 1988; Wellman et al., 1982). In general, these studies provided support for Bryans specialization concept, and greatly advanced the general understanding of diversity among outdoor recreation participants.
However, the lack of any empirical testing of recreation specialization remained an issue. As pointed out by Ditton et al. (1992), any attempt to test Bryans
framework for specialization was problematic because it was tautological (circular) in its reasoning; specialization level, defined in terms of behaviors and preferences, was then used to predict specialized behaviors and experiential preferences.
As a result, recreation specialization as a concept could never be empirically tested
because specialization and its subsequent propositions were both defined and
measured in the same terms (Ditton et al., 1992).
Ditton et al. (1992) initiated development of a testable theory that links recreation specialization with elements of social worlds as described by Unruh (1979).
Unruh (1979) defined a social world as an internally recognizable constellation
of actors, organizations, events and practices which have coalesced into a perceived sphere of interest and involvement for participants. According to this perspective, members of the same social world hold similar attitudes, beliefs, and
motivations that create a sense of group identity. Unruh (1979) further suggested
that members within a social world could be ordered along a theoretical dimension of involvement level based on four key characteristics: orientation, experiences, relationships, and commitment. For each characteristic, Unruh (1979) describes four involvement levels that correspond to four trans-situational social
types: strangers, tourists, regulars, and insiders (Table 1).
Ditton et al. (1992) reconceptualized and redefined recreation specialization
TABLE 1 Characteristics and Types of Social World Participation (from Unruh,
1979)
Social types or subworlds
Characteristics

Strangers

Tourists

Regulars

Insiders

Orientation
Experiences
Relationships
Commitment

naivete
disorientation
superficiality
detachment

curiosity
orientation
transiency
entertainment

habituation
integration
familiarity
attachment

identity
creation
intimacy
recruitment

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

242

R. J. Salz et al.

as a process by which recreation social worlds and subworlds segment and intersect into new recreation subworlds, and the subsequent ordered arrangement of
these subworlds and their members along a continuum. Subworld types are arranged
by Ditton et al. (1992) on a continuum from least specialized to most specialized.
Ditton et al. (1992) developed eight recreation specialization propositions.
They tested three of these, using frequency of participation to segment anglers
into four specialization levels. Their results provided empirical support for specialization by showing that the four groups differed as predicted in their resource
dependency, level of mediated interaction, and the importance they attach to activity-specific and nonactivity-specific elements within a recreational activity.
Highly specialized anglers were found to have a higher resource dependency than
did less specialized anglers. The highly specialized groups placed more importance on catching big, distinctive, or trophy fish, whereas the less specialized
anglers appear to be less interested in the rare event aspect of the fishing experience. They found that anglers who were more specialized had a greater involvement in various types of mediated means of communication than did less specialized anglers. Finally, Ditton et al. (1992) found that as level of specialization
increased, the importance attached to catch-related angling motivations (e.g., catching fish of preferred size, number, or species) decreased relative to noncatch-related
angling motivations (e.g., to be outdoors, to relax, to be with friends, etc.).
Although their single dimension (i.e., frequency of participation) approach
to angler segmentation proved successful, Ditton et al. (1992) recognized that
other variables can and should be used as a means of classifying individuals into
specialization subgroups. A single variable (such as avidity) cannot adequately
measure these distinct dimensions of specialization and may result in high
misclassification rates. In this paper, we suggest that the testing of recreation specialization theory, and its application, is advanced when using a multivariable approach to segmentation that incorporates orientation, experiences, relationships,
and commitment.

Study Objectives
The first purpose of this research was to develop and validate a multivariable
specialization index based on a social world view of recreation specialization.
The second purpose of this research was to use this index to test recreation specialization theory by re-examining one of the propositions tested by Ditton et al.
(1992), examining two other propositions that have not yet been tested, and developing and testing a new proposition. The proposition to be retested states: as level
of specialization in a given recreation activity increases, the importance of activityspecific elements of the experience will decrease relative to nonactivity-specific
elements of the experience (Proposition Eight in Ditton et al., 1992). Ditton et al.
(1992) found that more-specialized anglers placed less importance on activityspecific elements, such as catching fish, and more importance on the nonactivity-

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization Index

243

specific elements of the fishing experience, such as enjoying nature, relaxing,

being with friends or family, and so forth.
The second proposition states that participants who are more specialized would
indicate greater support for management rules and regulatory procedures, as well
as for social norms that identify and often dictate acceptable behavior, than would
less-specialized participants (Proposition Four in Ditton et al., 1992). Temporary
or seasonal closures due to overfishing, for example, would have a greater impact
for more-specialized individuals than for less-specialized individuals. Therefore,
by voluntarily accepting rules and social norms associated with the activity, participants help to ensure its continuation (Ditton et al., 1992). The third proposition states that more-specialized anglers have higher levels of side-bets than do
less-specialized anglers. Side-bets denote when something of value (time, money,
social relations) is invested in the activity with the condition that to discontinue
the activity could result in a loss of the investment (Alluto, Hrebiniak, & Alonso,
1973; Becker, 1960). More-specialized individuals are proposed to have a greater
financial and emotional investment in a given activity than less-specialized individuals (Proposition Two in Ditton et al., 1992).
The new proposition that we propose here states that as level of specialization in a given recreation activity increases, frequency of participation in that
activity will increase. We base this proposition on the results of previous research.
Graefe (1980) found avidity to be a surrogate measure for specialization level.
Schreyer et al. (1984) similarly used number of river runs to segment river users
into subgroups and found significant differences between these subgroups .
Chipman and Helfrich (1988) successfully used frequency of participation as one
element of determining specialization level. Finally, Ditton et al. (1992) used avidity
to segment a population of anglers into specialization subgroups, and found significant differences between the subgroups. We would view this as Proposition
Nine as added to the eight previously stated by Ditton et al. (1992).

Hypotheses
Based on the previous propositions, the following hypotheses were generated.
Ha1(a): High-specialization anglers will attach less importance to activity-specific elements of the fishing experience than will low-specialization
anglers.
Ha1(b): High-specialization anglers will attach more importance to nonactivityspecific elements of the fishing experience than will low-specialization
anglers.
Ha2:
High-specialization anglers will have a greater support for various management tools and regulations than will low-specialization anglers.
Ha3:
High-specialization anglers will have generated a greater value of sidebets than will low-specialization anglers.
Ha4:
High-specialization anglers will have a greater frequency of participation than will low-specialization anglers.

244

R. J. Salz et al.

Methods

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Specialization Index
In developing our specialization index, we chose to pursue an a priori approach
that builds on theory, and that uses theory to generate the index items. Our specialization index items, therefore, were derived from the four characteristics (orientation, experiences, relationships, and commitment) used by Unruh (1979) to
place participants in a particular subworld (or in our case a particular specialization level). For each characteristic, Unruh described four subworld types of participants: strangers, tourists, regulars, and insiders (Table 1). Based on these descriptions, we developed four survey questions (i.e., corresponding to the four
characteristics), each containing four possible response options (i.e., corresponding to four specialization levels). Question response options, consisting of statements describing a participants connection to an activity relative to that particular characteristic, were ordered from least specialized (response option = 1) to
most specialized (response option = 4) along a 4-point scale (Table 2). It was
expected that for each item, the least-specialized participants would select response option 1, and the most-specialized participants would select response option 4.
The sum of the four responses (e.g., least specialized: 1 + 1 + 1 + 1 = 4,
highly specialized: 4 + 4 + 4 + 4 = 16) was then used to locate anglers along the
recreation specialization continuum. The actual process of developing and testing
the specialization index used for segmentation of anglers into specialization levels is described in the Results section.
Data Collection
Data were collected by way of a mail survey administered to a random sample of
licensed Massachusetts anglers. The basic survey design and implementation followed accepted principles based on Salant and Dillman (1994). A personalized
advance-notice letter was sent to all members of the sample announcing they had
been selected to participate in the survey and that they would be receiving the
questionnaire in the mail within the following week. One week later a set of survey materials was mailed to all members of the sample. These materials included
the questionnaire, a cover letter describing the intent of the survey, and a selfaddressed stamped envelope for returning the completed survey. Two weeks after
mailing the advance notice letter, a thank you/reminder postcard was mailed to all
members of the sample. This follow-up served to thank those who had already
completed and returned their questionnaire, and to request a response from those
who had not. Five weeks after mailing the advance notice letter, a second set of
survey materials was sent to those who had not yet responded. This second survey
package was identical to the first, except that the personalized cover letter was
revised to further encourage the subject to complete and return their survey.

Development and Validation of a Specialization Index

245

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

TABLE 2 Recreation Specialization Index Survey Questions and Response

Options
Q. Please indicate your general orientation to the sport of fishing.
1) I am an outsider. I am uncomfortable when I go fishing, and dont really feel
like I am part of the fishing scene.
2) I am an observer or irregular participant. Sometimes it is fun, entertaining, or
rewarding to go fishing.
3) I am a habitual and regular participant in the sport of fishing.
4) I am an insider to the sport. Fishing is an important part of who I am.
Q. Please indicate how you would best describe yourself during a fishing experience.
1) I am often uncertain. I am unsure about what I can or cannot do while fishing,
or how to do it.
2) I have some understanding of fishing, but I am still in the process of learning
more about fishing. I am becoming more familiar and comfortable with fishing.
3) I have become comfortable with the sport. I have regular, routine and predictable experiences. I have a good understanding of what I can do while fishing,
and how to do it.
4) I am a facilitator in the sport. I encourage, teach and enhance opportunities for
others who are interested in fishing.
Q. Please indicate how you would best describe your relationships with other
anglers.
1) Superficial. I really dont know any other anglers.
2) Very limited. I know some other anglers by sight and sometimes talk with
them, but I dont know their names.
3) One of familiarity. I know the names of other anglers, and often speak with
them.
4) Close. I have personal and close relationships with other anglers. These friendships often revolve around fishing.
Q. Please indicate how you would best describe your commitment to fishing.
1) Almost nonexistent. I am basically indifferent about going fishing.
2) Moderate commitment. I will continue to go fishing as long as it is entertaining and provides the benefits I want.
3) Fairly strong commitment. I have a sense of being a member of the activity,
and it is likely that I will continue to fish for a long time.
4) Very strong commitment. I am totally committed to fishing. I encourage others to go fishing and seek to ensure the activity continues into the future.

246

R. J. Salz et al.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Testing Specialization Theory

One-way ANOVA tests were used to test for mean differences between specialization groups. A significance level of .10 was used to test the null hypotheses.
This level of confidence reflects a balance between a higher probability of committing a Type I error (rejecting a true null hypothesis) and consequently decreasing the probability of committing a Type II error (failure to reject the null when it
is false). Gregorie and Driver (1979) suggest this as being a more appropriate
level (than 0.01 or 0.05), so later studies would not mistakenly consider some of
the insignificant differences as being unimportant, when in fact they might have
been due to the commission of a Type II error.

Results
Response Rate
A total of 1,411 questionnaires (54.6%) were returned in usable form (Table 3).
There were 312 questionnaires returned as undeliverable by the U.S. Postal Service, 3 were returned because the addressee was deceased, and 29 returned by
respondents were unusable. The remainder were nonresponses.
Index Development and Internal Validation
Frequency distributions were calculated for each of the four index items (Figure
1). On a scale of responses from 1 (least specialized) to 4 (highly specialized), the modal response for all four items was 3. The proportion of responses
in the least-specialized category (i.e., response = 1) was 2% or less for orientation, experience, and commitment. The proportion of least-specialized responses (response = 1) was considerably greater for relationships (7.3%), although this was still small compared to the proportion for the other three response

TABLE 3 Status of Sport Angler Questionnaire Response

Type of response

Initial sample
Mortality
Deceased (3)
Nondeliverable (312)
Not-usable upon return (29)
Effective sample
Nonresponse

2,930
344

2,586
1,175

100.0
45.4

Usable returned surveys

1,411

54.6

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization Index

247

FIGURE 1. Distribution of angler selections of response options according to the

four index items.
options. Nearly 60% of respondents chose 3 for the question regarding experience. The other variables were more evenly distributed across responses, except
for the previously noted lack of 1 responses (Figure 1).
Bivariate relationships among the items considered for inclusion in the index were then examined to determine the degree to which the items were related
(Babbie, 1995). Correlation coefficients for the six pair-wise comparisons ranged
from 0.41 to 0.60 and were all statistically significant (Table 4). This middle range
suggests that no two items were so similar as to warrant exclusion from the index
to avoid redundancy. Therefore, although significant positive relationships were
found for all pair-wise comparisons, each item measures a somewhat different
aspect of recreation specialization. The two lowest correlation coefficients involved the variable relationships (0.41 and 0.43), whereas the highest correlation was between orientation and commitment (0.60).
Another way to analyze bivariate relationships is to examine the percent of
occurrences when two variables differ from each other by more than a particular
amount. For each of our four variables, possible responses ranged from 1 (least
specialized) to 4 (highly specialized). For all pair-wise comparisons, less than
9% of all respondents had responses for any two variables that differed by more
than one (Table 4). This further supports the strong positive relationships between
all items. Most of the cases where an anglers responses for two variables did
differ by more than one involved the variable relationships. Pair-wise comparisons not involving the variable relationships differed by more than one for only
about 3% of respondents.
Index item reliability was tested using Cronbachs coefficient alpha
(Cronbach, 1951). The reliability of the final multiple-item index was measured
with an internal consistency coefficient (Cronbachs alpha) of 0.78. Alpha values

248

R. J. Salz et al.

TABLE 4 Bivariate Relationships Among Index Items

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Index item pair

Relationships and Experience
Relationships and Orientation
Relationships and Commitment
Experiences and Orientation
Experiences and Commitment
Orientation and Commitment

Correlation
coefficient

% of responses differing
by more than one

0.41
0.43
0.49
0.48
0.50
0.60

8.2%
8.9%
7.8%
3.0%
3.0%
3.0%

when a particular item was deleted were 0.68 for commitment, 0.74 for experience, 0.70 for orientation, and 0.76 for relationships. This further supported the
inclusion of all four recreation specialization social world characteristics (i.e.,
orientation, commitment, experience, and relationships) in our index.
Based on our results from the bivariate comparisons and Cronbachs alpha,
we decided to include all four items in creating our recreation specialization index. A composite specialization rank was calculated by summing the responses to
the four items for each respondent (Figure 2). Composite scores ranged from 4
through 16. Respondents were segmented into specialization groups based on
their cumulative item score as follows:
If cumulative score = 46 Index Level = 1 (least specialized)
If cumulative score = 710 Index Level = 2 (moderately specialized)
If cumulative score = 1113 Index Level = 3 (very specialized)
If cumulative score = 1416 Index Level = 4 (highly specialized)

FIGURE 2. Distribution of anglers according to cumulative index score.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization Index

249

Again, we pursued an a priori process in developing the index. This also

applied to determining which item scores should correspond to which specialization levels. We chose to make the score brackets as equal in size as possible (levels 1, 3, and 4 all had a range of 3 in their score, whereas level 2 had a range of 4).
The number of anglers classified into each specialization level is the result of this
process, rather than the opposite in which some preconceived distribution of anglers is forced into a manipulated set of index brackets.
This process resulted in the least specialized angler group (Level = 1) accounting for only 1.2% (n = 16) of all respondents (Figure 3). Moderately specialized anglers (Level = 2) accounted for 32.5% (n = 440), very specialized
anglers (Level = 3) accounted for 42.3% (n = 572), and highly specialized anglers (Level = 4) accounted for 24.0% (n = 325) of all respondents.
Internal index validation is conducted to demonstrate that an index successfully measures what it is intended to measure (Babbie, 1995). A method of internal validation called item analysis was conducted to examine the extent to which
our composite index is related to (or predicts responses to) the four items (i.e.,
relationships, commitment, experience, and orientation) that comprise it. Item
analyses using direct comparisons were possible because both index scores (i.e.,
index level) and item scores were based on equivalent 4-point scales ranging from
least to highly specialized. The index score was identical to the item score for
orientation 72% of the time, commitment 74% of the time, experiences
66% of the time, and relationships 60% of the time. For all items, the absolute
difference between index score and item score exceeded one for less than 3% of
respondents. These results support the internal validity of our specialization index.
As a final test, correlations were computed between specialization index level
and more traditional measures of specialization such as avidity (freshwater days
fished in past 12 months) and total years fished. The correlation between special-

FIGURE 3. Distribution of anglers according to specialization level.

250

R. J. Salz et al.

ization index level and freshwater days fished in past 12 months was 0.38, whereas
the correlation between specialization index level and years fished was 0.18. Both
were highly significant (p < 0.0001), indicating that our specialization index correlates with these unidimensional specialization indicators. However, both correlations were also fairly low, suggesting that important differences between our
index and these unidimensional indicators do exist.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Testing Recreation Specialization Theory

As mentioned before, our segmentation of respondents resulted in only 16 individuals (1.2%) being classified into the least-specialized level. Because this is an
inadequate sample size for our analyses, this group was subsequently dropped for
hypothesis testing. Therefore, hypotheses were tested using only three specialization levels: Moderately specialized (M), Very specialized (V), and Highly specialized (H); levels 2, 3, and 4, respectively.
Hypothesis One
Seven items were used to measure the importance of activity-specific elements of
the fishing experience. Results show significant differences for five of these seven
measures (Table 5). However, one of these items was contrary to specialization
theory because more-specialized anglers rated the item experience of the catch
as more important than did less-specialized anglers. For two other items, there
was no significant difference among specialization levels. Still, four out of the
five items with significant differences were ordered as predicted by specialization
theory. Based on these results, the null hypothesis that there are no differences
according to level of specialization on activity-specific measures of the fishing
experience was rejected, and we accept hypothesis Ha1(a) as stated, but propose
that more investigation is needed regarding the items that did not behave as predicted by specialization theory.
Ten items were used to measure the importance of nonactivity-specific elements of the fishing experience. Results show significant differences for 9 out of
the 10 items according to level of specialization (Table 6). It was predicted that
more-specialized anglers would place greater importance on nonactivity-specific
activities than would less-specialized anglers. Because the results are as predicted,
the null hypothesis is rejected and Ha1(b) is accepted as stated.
Hypothesis Two
Eleven items were used to measure support or opposition to various management
regulations. The null hypothesis, which states that there are no differences between anglers in their support and opposition to management rules, was rejected
because significant differences were found for ten of the eleven items (Table 7).

251

Development and Validation of a Specialization Index

TABLE 5 One-way ANOVA Tests for Mean Differences in Importance of
Activity-specific Items According to Specialization Level
Level of specialization

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Items*

For the experience of

3.500** 3.818
the catch
For the sport of fishing,
3.556
3.904
not to obtain food to eat
Im just as happy if I
4.110
4.181
release the fish I catch
I am just as happy if I dont
4.053
4.158
keep the fish I catch
A fishing trip can be
3.792
3.834
successful even if no
fish are caught
When I go fishing, Im just
3.095
3.034
as happy if I dont catch
a fish
To obtain fish for eating, and 1.502
1.480
not for sport

4.128

30.29

0.000

4.183

26.11

0.000

4.370

7.57

0.001

4.329

7.55

0.001

4.031

5.90

0.003

3.111

0.67

0.510

1.547

0.55

0.578

*For items 1, 2, and 7 mean scores were based on responses to the following categories;
1 = Not at all important, 2 = Slightly important, 3 = Moderately important, 4 = Very
important, 5 = Extremely important. For all other items, mean scores were based on
responses to the following categories; 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4
= Agree, 5 = Strongly agree.
**Means underscored by same line are not significantly different (.10) using Tukeys
test.

The prediction that more-specialized anglers would indicate a greater support for
management rules than would less-specialized anglers was supported on 9 of the
10 significant items. The mean values for one item (restricted fishing area) were
directly opposite of that predicted. Because 9 of the 10 significant items were
ordered as predicted, Ha2 is accepted as stated.
Hypothesis Three
Four items relating to the cost of replacing fishing equipment were used to measure side-bets. It was predicted that more-specialized anglers would generate a
greater value in side-bets than would less-specialized anglers. Significant differences supporting this prediction were found according to specialization level for
all four items (Table 8). Therefore, we reject the null hypothesis. Because the
mean differences are as predicted, we accept Ha3 as stated.

252

R. J. Salz et al.

TABLE 6 One-way ANOVA Tests for Mean Differences in Importance of

Nonactivity-specific Items According to Specialization Level
Level of specialization

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Items*
To experience adventure
and excitement
To be close to the water
For relaxation
To be with friends
To experience natural
surroundings
To experience new and
different things
To get away from the
demands of other people
To be outdoors
To get away from
the regular routine
For family recreation

3.405**

3.732

4.009

28.77

0.000

3.366
4.218
3.107
4.134

3.576
4.345
3.206
4.248

3.973
4.559
3.559
4.453

21.20
16.48
13.21
12.82

0.000
0.000
0.000
0.000

2.842

2.939

3.279

12.35

0.000

3.409

3.474

3.842

10.49

0.000

4.177
3.800

4.236
3.912

4.450
4.159

10.44
9.78

0.000
0.000

3.279

3.136

3.256

1.72

0.179

*Mean scores were based on responses to the following categories; 1 = Not at all important, 2 = Slightly important, 3 = Moderately important, 4 = Very important, 5 = Extremely
important.
**Means underscored by same line are not significantly different (.10) using Tukeys
test.

Hypothesis Four
Results showed significant differences on angler frequency of participation according to level of specialization (Table 9). The null hypothesis is therefore rejected as stated. Highly-specialized anglers had significantly higher rates of participation than did moderately-specialized anglers, who in turn had significantly
higher rates of participation than did lower-specialized anglers. Because this is
consistent with what was predicted, Ha4 is accepted as stated.

Discussion
Our results provide strong support for the theory of recreation specialization as
reconceptualized by Ditton et al. (1992), and for use of the specialization index
developed here. Results from our hypotheses tests were as predicted for an overwhelming majority of the items we investigated. Our study also strongly supports
the inclusion of all four characteristics of social worlds (commitment, orienta-

Development and Validation of a Specialization Index

253

TABLE 7 One-way ANOVA Tests for Mean Differences in Support and

Opposition of Management Regulation Items According to Specialization Level
Level of specialization

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Items*
Creel limit
No stocking allowed
Maximum size
Stock non-native fish
Minimum size limit
Restricted fishing area
Mandatory catch
and release
Stock native fish
Slot limit
Voluntary catch and
release
Prohibit use of certain
gear

4.109**
3.559
3.284
3.009
4.108
3.434
3.122

4.293
3.673
3.547
3.282
4.211
3.260
3.162

4.463
3.935
3.733
3.343
4.433
3.069
3.439

14.79
13.70
13.64
11.13
8.62
7.23
6.67

0.000
0.000
0.000
0.000
0.000
0.001
0.001

4.219
3.117
3.874 a

4.336
3.190
4.028 b

4.403
3.388
4.022 a,b

6.03
5.86
2.83

0.003
0.003
0.059

3.612

3.545

3.581

0.43

0.653

*Mean scores were based on responses to the following categories; 1 = Strongly oppose,
2 = Oppose, 3 = Neutral, 4 = Support, 5 = Strongly support.
**Means underscored by same line or same superscript are not significantly different
(.10) using Tukeys test.

tions, experience, and relationships) as related and reliable measures of recreation

specialization.
Specialization Index Development
There are several possible explanations for the fact that the least specialized
subworld made up such a small proportion of our sample (only 1.2%). First, we
should not rule out the possibility that this group may, in fact, be much smaller in
size than the other groups. This would be the case if the learning curve from least
specialized to moderately specialized requires a relatively short time period.
Because our survey was administered to those people who had purchased licenses
during the previous year, anglers who were least specialized at the time of license purchase had at least a full fishing season to increase their specialization
level prior to receiving our survey. Another possible explanation is that our sample
did not tap into those groups of anglers that make up the majority of the least
specialized group. For example, children (under 17 years old), out-of-state anglers, and 3-day license holders were not part of our survey population. One might
reasonably expect these anglers to be among the least specialized. We consider

254

R. J. Salz et al.

TABLE 8 One-way ANOVA Tests for Mean Differences in the Cost of Replacing
Fishing Equipment with Similar Equipment Between Specialization Level

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Items*
Replace reels
Replace tackle
Replace rods
Replace electronic
equipment

Level of specialization
M
V
H
$119.33*
114.80
138.31
262.00

$229.49
282.28
284.52
436.65

$455.80
579.84
555.28
580.42

90.00
78.65
38.18
6.95

0.000
0.000
0.000
0.001

*Means underscored by same line are not significantly different (.10) using Tukeys test.

these explanations to be the most likely reasons for the small size of the least
specialized group.
Nonresponse bias could also be a possible explanation if the probability of
an angler returning our survey was positively correlated to the anglers specialization level. However, in a study of nonresponse bias on angler surveys, Fisher
(1996) found that species preferences and scores from summated Likert scales
were independent of response probabilities. Finally, the choice of words we used
for the least specialized response options could explain the low percent of respondents selecting those options. Anglers may have felt embarrassed to identify
themselves with words such as outsider, uncomfortable, unsure or uncertain, all of which may have strong negative connotations. Our results suggest
that least specialized subworlds may be more difficult to sample for a variety of
reasons. A special sample design may be needed in certain situations to adequately
address this group.
Our results showed that although all four social world characteristics (relationships, orientation, experience, and commitment) should be included in the
index, the relationships dimension behaved somewhat differently from the other
three. Specifically, some anglers scored least specialized for relationships but
were in the middle-to-high range of specialization for the other three dimensions.
This suggests that for the activity of freshwater fishing, having personal relationTABLE 9 One-way ANOVA Tests for Mean Differences in Frequency of
Participation According to Specialization Level
Level of specialization
Items*
Mean total days
fished

15.566*

36.656

56.609

105.54

0.000

*Means underscored by same line are not significantly different (.10) using Tukeys test.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Development and Validation of a Specialization Index

255

ships with other anglers may not be as important of a component when advancing
to higher specialization levels as previously thought. Although interaction and
communication relate to social world boundaries (Unruh, 1980), in todays world
these can be readily achieved through mediated channels instead of personal contact. Some highly specialized anglers may rely on journals, magazines, cable television, and the Internet to acquire and exchange information about fishing. If so,
our question measuring relationships, which focuses only on personal contacts,
may have to be expanded to include a wider range of interactive and communicative possibilities.
The characteristics included in our index were derived directly from the social worlds literature. Still, the question of which specific measures should be
used to define specific characteristics of a specialization index is open to interpretation (Kuentzel & McDonald, 1992). For example, commitment to an activity
has been measured as the number of related magazines one subscribes to (Bloch,
Black, & Lichtenstein, 1989), the level of activity involvement (Williams &
Huffman, 1986), the centrality of the activity to ones lifestyle (Chipman &
Helfrich, 1988), the number of side-bets invested in, and an affective attachment to the activity (Buchanan, 1985). Similarly, one could come up with multiple ways to define and measure orientation, experience, and relationships
related to a particular activity.
Specialization dimensions can also be measured using either behavioral or
cognitive measures. One of the main features of social world involvement is voluntary identification, meaning one chooses to become a member of a social world
rather than it being a requirement (Unruh, 1980). The necessity of voluntary identification suggests a strong cognitive component to entry into a social world and
movement between subworlds within that social world. This cognitive component is reflected in the questions we used in this study to measure specialization
dimensions. For example, rather than measure commitment through other variables as described above, anglers were asked directly to choose the statements
that best describe their involvement in the sport.
Approaching specialization from a social worlds perspective may add subjectivity to the index because words like commitment, insider, and orientation can mean different things to different people. However, this subjectivity
does not necessarily bias the segmentation process, but rather, it may redefine
specialization in a new way. The assumption that a specialization index derived
from objective measures (i.e., gear used, days fished, magazines purchased) is
preferable to one that uses more subjective, cognitive measures should not automatically be made. The decision of which index to use should, perhaps, be based
on the goals of the particular study and the research purpose or management application it is intended for. The general lack of consistency in measuring specialization in the outdoor recreation literature supports this contention. For future
study, it would be interesting to compare participant segmentation using our index with previous specialization indices using the same survey population.

256

R. J. Salz et al.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

Testing Recreation Specialization Theory

Results indicated that more-specialized anglers were more interested in a qualitative experience, whereas less-specialized anglers had a more simplistic view of
fishing that did not consider other intrinsic elements of the experience to be quite
as important. This supports specialization theory and further reconfirms the results of Ditton et al. (1992).
Frequency of participation was also shown to increase as specialization levels increase. Our results are consistent with previous work and justify the addition
of our proposed Proposition Nine. Individuals are likely to increase their frequency of participation when they feel some sort of attachment to an activity. As
specialization level increases, alternative activities will be rejected as the commitment to participating in the primary activity increases (Buchanan, 1985; Unruh,
1979).
It appears that more-specialized anglers are more receptive to management
regulations than are less-specialized anglers. The support for management regulations was shown to increase as specialization increases. The former group is
more likely to be impacted than the latter group if fishing activities were discontinued; therefore, as predicted from specialization theory, the former would be
more supportive of rules and regulations issued from fisheries management
agencies.
Finally, as predicted, side-bets anglers appropriated for fishing equipment
were shown to increase as level of specialization increased. Because of a greater
involvement within the activity, more-specialized anglers will commit greater financial costs towards fishing than will less-specialized anglers.
Management Implications
There is potential here for fisheries managers to gain an understanding of group
differences on a variety of issues to efficiently improve services already provided.
By developing and promoting services based on some aggregation of anglers, the
interests of many anglers are ignored. Managers may then be confronted with a
fairness issue, where some anglers perceive that resources are allocated unfairly.
Segmentation by specialization recognizes that different groups have different
attributes that require different marketing schemes. Through a better understanding of the angling constituency, managers can avoid making resource allocation
decisions that may result in the loss of credibility for the fisheries agency (Ditton,
1996; Loomis & Ditton, 1993). The results of this study provide strong support
for the use of a multidimensional index as a means of classifying participants into
homogeneous groups, based on the recreation specialization theory developed by
Ditton et al. (1992). Such insight to anglers can also be used to effectively evaluate current management objectives and services.

Development and Validation of a Specialization Index

257

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

References
Alluto, J. A., Hrebiniak, L. G., & Alonso, R. C. (1973). On operationalizing the concept
of commitment. Social Forces, 51, 448454.
Babbie, E. (1995). The practice of social research. Belmont, CA: Wadsworth Publishing
Company.
Becker, H. S. (1960). Notes on the concept of commitment. American Journal of Sociology, 66, 3240.
Bloch, P. H., Black, W. C., & Lichtenstein, D. (1989). Involvement with the equipment
component of sport: Links to recreational commitment. Leisure Sciences, 11, 187
200.
Bryan, H. (1977). Leisure value systems and recreational specialization: The case of trout
fishermen. Journal of Leisure Research, 9(3), 174187.
Buchanan, T. (1985). Commitment and leisure behavior: A theoretical perspective. Leisure Sciences, 7, 401420.
Chipman, B. D., & Helfrich, L. A. (1988). Recreational specialization and motivations of
Virginia river anglers. North American Journal of Fisheries Management, 8, 390
398.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika,
16, 297334.
Ditton, R. B. (1996). Understanding the diversity among largemouth bass anglers. American Fisheries Society Symposium, 16, 135144.
Ditton, R. B., Loomis, D. K., & Choi, S. (1992). Recreation specialization:
Reconceptualization from a social worlds perspective. Journal of Leisure Research,
24(1), 3551.
Fedler, A. J., & Ditton, R. B. (1986). A framework for understanding the consumptive
orientation of recreational fishermen. Environmental Management, 10(2), 221227.
Fisher, M. R. (1996). Estimating the effect of nonresponse bias on angler surveys. Transaction of the American Fisheries Society, 125, 118126.
Graefe, A. R. (1980). The relationship between level of participation and selected aspects of specialization in recreational fishing. Doctoral dissertation. Texas A&M
University, College Station.
Gregorie, T. G., & Driver, B. L. (1987). Type II errors in leisure research. Journal of
Leisure Research, 19(4), 261272.
Hammitt, W. E., & McDonald, C. D. (1983). Past on-site experience and its relationship
to managing river recreation resources. Forest Science, 29, 262266.
Kauffman, R. B., & Graefe, A. R. (1984). Canoeing specialization, expected rewards, and
resource related attitudes. In J. S. Popodic, D. I. Butterfield, D. H. Anderson, & M.
R. Popodic (Eds.), National River Recreation Symposium Proceedings (pp. 629
641). Baton Rouge, LA: Louisiana State University.
Kuentzel, W. F., & McDonald, C. D. (1992). Differential effects of past experience, commitment, and lifestyle dimensions of river use specialization. Journal of Leisure
Research, 24, 269287.
Loomis, D. K., & Ditton, R. B. (1993). Distributive justice in fisheries management.
Fisheries, 18, 1418.
Salant, P., & Dillman, D. A. (1994). How to conduct your own survey. New York: Wiley.

Downloaded by [Fac Psicologia/Biblioteca] at 00:20 25 September 2012

258

R. J. Salz et al.

Schreyer, R. M., Lime, D. W., & Williams, D. R. (1984). Characterizing the influence of
past experience on recreation behavior. Journal of Leisure Research, 16, 3450.
Unruh, D. R. (1979). Characteristics and types of participation in social worlds. Symbolic
Interaction, 2, 115130.
Unruh, D. R. (1980). The nature of social worlds. Pacific Sociological Review, 23(3),
271296.
Vaske, J. J., Donnelly, M. P., & Heberlein, T. A. (1978). Perceptions of crowding and
resource quality by early and more recent visitors. Leisure Sciences, 3, 367381.
Virden, R. J., & Schreyer, R. M. (1988). Recreation specialization as an indicator of
environmental preference. Environment and Behavior, 20, 721739.
Wellman, J. D., Roggenbuck, J. W., & Smith, A. C. (1982). Recreation specialization and
norms of depreciative behavior among canoeists. Journal of Leisure Research, 14,
323340.
Williams, D. R., & Huffman, M. G. (1986). Recreation specialization as a factor in
backcountry trail choice. In R. Lucas (Ed.) Proceeding of the National Wilderness
Research Conference Current Research, (General Technical Report INT-211, pp.
339344), Ogden, UT: USDA Forest Service Intermountain Research Station.