INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.

U n iversity M icrofilm s In tern ational
A B ell & H ow ell Inform ation C o m p a n y
3 0 0 N orth Z e e b R oad . A n n Arbor. Ml 4 8 1 0 6 - 1 3 4 6 U S A
3 1 3 /7 6 1 -4 7 0 0

8 0 0 /5 2 1 -0 6 0 0

O rder N u m b e r 9 5 1 2 0 2 0

T h e im p ortan ce o f an cillary a ttrib u te s in d e stin a tio n choice
an d len g th o f stay: C ross-cou n try skiers a t a sm all n o rth ern
M ich igan com m u n ity

Bishop, Glen Robert, Ph.D.
Michigan State University, 1994

UMI

300 N. Zeeb Rd.
Ann Arbor, MI 48106

THE IMPORTANCE OF ANCILLARY ATTRIBUTES IN DESTINATION
CHOICE AND LENGTH OF STAY: CROSS-COUNTRY SKIERS AT A
SMALL NORTHERN MICHIGAN COMMUNITY
By
Glen Robert Bishop

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
D epartm ent of Park and Recreation Resources
1994

ABSTRACT
THE IMPORTANCE OF ANCILLARY ATTRIBUTES IN DESTINATION
CHOICE AND LENGTH OF STAY: CROSS-COUNTRY SKIERS AT A
SMALL NORTHERN MICHIGAN COMMUNITY
by
Glen Robert Bishop

Ancillary attributes are of secondary importance in achieving the main
objective of an activity. They facilitate the activity, but are not directly
related to the principle activity of an outdoor recreation trip. Those ancillary
attributes related to basic needs and present in the community were expected
to be rated higher in importance in choosing a destination than those ancillary
attributes not related to basic needs and not present in the community. In
addition, variations in length of stay and in willingness to pay higher trail fees
were expected to be associated with importance ratings of some ancillary
attributes. The importance of ancillary attributes may be undervalued in
winter recreation activities such as cross-country skiing.
Cross-country skiers responded to a questionnaire distributed at cross­
country ski trails during the winter of 1991-92 near Munising, Michigan. The
questionnaire contained a list of 21 ancillary attributes and 1 cross-country ski
trail attribute. Importance ratings of individual attributes were compared to
the average rating of all attributes with paired t-tests. Regression models were
then developed to examine variations in length of stay and willingness to pay
trail fees. These models were then forced into stepwise regression analyses
with the importance ratings of the 21 ancillary attributes.

Attributes receiving higher importance ratings where those directly
related to the main activity of the trip (cross-country skiing) or to sustenance.
The importance ratings of some ancillary attributes were significantly
associated with variations in length of stay or willingness to pay trail fees.
Often, the cross-country ski trail is provided by a government agency while
the ancillary attributes are provided by local businesses. Both types of
organizations may benefit as more customers are draw n by the combination of
goods and services provided by both sectors than would have been attracted
to the destination by the efforts of only one sector. Particular ancillary
attributes may encourage longer stays, higher spending, and the selection of a
community as a cross-country ski destination.

TABLE OF CONTENTS

CHAPTER I: INTRODUCTION ...........................................................................
Ancillary Attributes .....................................................................................
Purpose of Study ...........................................................................................
Importance of Study .....................................................................................

1
6
7
7

CHAPTER II: LITERATURE REVIEW ...............................................................
Theoretical Framework .............................................................................
Satisfaction ......................................................................
Regression Models .......................................................................................

10
11
16
22

CHAPTER III: METHODOLOGY .....................................................................
Statement of the Problem ...........................................................................
Hypotheses ...................................................................................................
Assumptions .................................................................................................
Variables .......................................................................................................
Methods .........................................................................................................
Questionnaire Design .....................................................................
Sampling Strategy ...........................................................................
Data Analysis ...................................................................................
Ancillary Attribute Importance ......................................
Differences Among Response Groups ..........................
Relative Importance ...........................................................
Limitations .......................................................................................

38
38
39
39
40
48
48
49
51
51
53
53
60

CHAPTER IV: RESULTS .....................................................................................
Responses to Open Format Questions .....................................................
Hign Ancillary Attribute Importance Ratings .....................................
Ancillary Attribute Ratings, All Respondents ........................
Ancillary Attribute Ratings, Local Munising Respondents ..
Ancillary Attribute Ratings, Tourist Respondents ..................
Differences Among Cross-country Skier Groups ..............................
Ancillary Attribute Importance Relationship to Length of Stay . . . .

63
64
67
67
71
75
78
83

CHAPTER V: SUMMARY AND CONCLUSIONS ........................................... 101
Applications ....................................................................................................I l l
Community Applications ............................................................... 112
Resource M anagement Applications ........................................... 115
Future Research ..................
116
Summary ..........................................................................................................123
iv

LITERATURE CITED
APPENDIX:

QUESTIONNAIRES

LIST OF TABLES

Table 1. Responses to open-ended question by mail questionnaire
respondents ...................................................................................................

65

Table 2. Frequency destination considerations mentioned, mail
questionnaire respondents .........................................................................

66

Table 3.

Mean importance ratings, all respondents .......................

69

Table 4.

Mean importance ratings, local respondents

..................................

73

Table 5.

Mean importance ratings, tourist respondents ................................

76

Table 6.

Comparison of ancillary importance ratings by groups

..............

80

Table 7.

Pearson correlation matrix of independent variables ...................

85

Table 8.

Length of stay base model regression ............

90

Table 9.

Length of stay stepwise regression with ancillary attributes .......

91

Table 10. Trail fee base model regression

.......................................................

95

Table 11. Trail fee stepwise regression with ancillary attributes ..................

96

Table 12. Significant differences between distance zones

vi

.................................100

LIST OF FIGURES

Figure 1. Location of Munising, Michigan............................................................

3

Figure 2. Closed ended question and open ended question .........................

42

Figure 3. Constraint measures ............................................................................

45

vii

THE IMPORTANCE OF ANCILLARY ATTRIBUTES IN DESTINATION
CHOICE AND LENGTH OF STAY: CROSS-COUNTRY SKIERS AT A
SMALL NORTHERN MICHIGAN COMMUNITY
CHAPTER I: INTRODUCTION
Cross-country skiers in Michigan and the northwestern Great Lakes
Region in general have many high quality destinations to choose from with
superb ski trails. From this relatively large choice set of potential
destinations, cross-country skiers usually seriously consider only two to six
possibilities (Woodside and Sherrell 1977; Woodside and Carr 1988).
In marketing these high quality destination places, each destination
community can be considered to be a whole product in and of itself. In
addition, each destination community can be subdivided into separate
individual service and attraction products which together make up the
destination product. The customer in choosing a destination makes choices
among alternative destinations, at least in part, based upon the separate
individual service and attraction products contained in the destination
communities being considered. These individual service and attraction
products can be termed attributes. The exact combination of attributes or
separate attractions and services within the destination chosen by the tourist
creates a unique travel experience or package (Ashworth and Voogd 1990;
Doswell and Gamble 1981), or more specifically in this study a unique cross­
country ski experience.

1

2
This is not to say that the same destination is chosen every time under
similar circumstances. People are likely to have several regular destinations
(Fesenmaier 1985). The goal of marketing a destination should be to become a
member of the group of destinations that receive regular consideration by a
large num ber of cross-country skiers. To accomplish this, the destination
should have a basic package of attributes which are im portant to various
cross-country skier market segments. This study examines which individual
products within a larger destination product were of importance to cross­
country skiers visiting Munising, Michigan during the winter of 1991-1992.
The study further examines the importance of these individual products in
understanding variations in length-of-stay and willingness to pay trail fees
(Bell and Leeworthy 1990; Silberman 1985; Usyal, McDonald, and O'leary
1988; Walsh and Davitt 1983).
Munising is a small community located on the southern shore of Lake
Superior in Alger County, Michigan (Figure 1). Cross-country ski trails are
provided by the U.S. Forest Service and the National Park Service. These
agencies belong to different departm ents of the federal government and have
substantially different management philosophies. The differences in
administrative philosophy may result in the provision of different cross­
country ski experiences by the two agencies. These differing experiences may
attract skiers who prefer differing combinations of attributes.
Much recreation research has focused on the importance of setting to
the recreational experience or outcome and to the motivations for
participating in outdoor recreation activities (i.e. McCool, Stankey, and Clark
1985; Driver and others 1987; Virden and Knopf 1989). Recreabonists prefer to
visit those areas which have attributes or characteristics which they consider

3

Munising

Figure 1. Location of Munising, Michigan.

4
to be im portant in achieving desired experiences or benefits. Usually, the
recreator considers more than one attribute when selecting an area to visit.
The relative attractiveness of an area can be measured by adding the amount
of each attribute for each area weighted by the importance of that attribute
and comparing the sums for each area under consideration. This approach
has been given several titles in the literature including multi-attribute theory,
rational choice theory, and expected utility theory (Hogarth 1987). Several
related theories, expectancy theory, the theory of reasoned behavior, and the
theory of planned action (Ajzen and Driver 1991) are somewhat different in
that they examine attitudes and intentions and their relation to behavior.
Essentially the same approach of combining attitudes is used to arrive at
behavioral intentions. People undertake actions or select destinations based
on the relative attractiveness of the behavior across a variety of motivations or
needs weighted by the likelihood of occurrence, or performance.
While some studies have examined attribute importance in the context
of general tourism (Goodrich 1978; Bronner and de Hoog 1985; Crompton,
Fakeye, and Lue 1992; Bonn, Furr, and Uysal 1992; Kretchman and Eagles
1990; Um and Crompton 1991), no studies of specific outdoor recreation
activities have included importance ratings of support services as
independent variables in examining length of stay or num ber of trips.
Tourism studies usually use attribute importance ratings either by
themselves or in conjunction with measures of perceptions of destinations to
compare potential destinations or to segment respondents into potential
market segments.
Current cross-country skier research using the multi-attribute
approach focuses primarily on the trail setting and is related to the recreation

5
opportunity spectrum concept (Clark and Stankey 1979). Few studies have
examined non-trail attributes which may be important in attracting cross­
country skiers to particular destinations. No studies have linked non-trail
attributes to length of stay other than to use aggregate spending on such
attributes as a constraint. Because skiers cannot spend 24 hours a day skiing,
it is logical to suppose that destination attributes other than those directly
associated with the cross-country ski trail may be important in attracting
skiers. These other attributes may aid the cross-country skier in fulfilling
motivations and attaining beneficial outcomes. The combination of non-trail
attributes and trail attributes may act synergistically (Poon 1989) and result in
higher value to cross-country skiers than trail attribute and non-trail
attributes considered separately. A product composed of a combination of
several goods may possess characteristics different from those of the goods if
considered individually (Lancaster 1966).
In rural areas of roughly even natural resource distribution, natural
resource characteristics may have little effect on tourism spending (Roehl,
Fesenmaier, and Fesenmaier 1993). Pearce (1982) speculates that natural
environments may be particularly valuable for self-actualizing experiences
while man-made environments fulfill physiological needs. In addition, these
ancillary attributes may be easier for destination managers (either local
business persons or government agency managers) to control than trail
conditions. Trail conditions may largely depend on weather during a
particular year, government agency funding, or the efforts of local advocates
of cross-country skiing.
Moreover, because of the large num ber of excellent ski trails in
northern Michigan, one could suppose that decisions about which trail or

6
local area to select for a weekend of skiing may be largely based on the
presence of key ancillary attributes. Previous studies have focused on such
trail attributes as difficulty and the opportunity to observe wildlife and
picturesque scenery. The contribution of these trail attributes towards
experience outcomes or benefits has been frequently examined. More
recently, a few studies have looked at the attributes of the cross-country ski
resort (Brayley 1991). However, no studies have examined which small town
attributes are important to the cross-country skier in choosing a destination.
And, no studies have tried to place such ancillary attributes in context with
other variables such as distance from home, perceived constraints,
satisfaction, experience, skill level, and trail quality which have been
discussed as influencing destination choice and length of stay in the
recreation literature.

Ancillary Attributes
All products are m ade up of varying inputs of services and physical
goods. There are no products which are entirely physical in nature nor are
there any products which are entirely service in nature. Key physical
attributes of a service product can be thought of as facilitating products
(Sasser, Olsen, and Wyckoff 1978). A collection of related product attributes
can be thought of as a service package for products which are primarily
service in nature, as most tourism end products are (Normann 1991). Cross­
country skiing in the case of small rural communities consists of a physical
product, the ski trail, and a bundle of services, trail grooming, lodging,
restaurant meals, and entertainment.

7
Product attributes can generally be divided into two groups, core and
ancillary or peripheral (Normann 1991). A core attribute is one that clearly is
the main focus of activity or interest, a cross-country ski trail for example.
Ancillary or peripheral attributes are part of the package but are of lesser
importance, the external appearance of a hotel for example. When there is
little differentiation among core services or product attributes offered by a
num ber of different providers, then competition for customers may shift to
ancillary attributes or services (Normann 1991). Since there are many cross­
country ski trails in northern Michigan of roughly the same quality,
competition for cross-country skiers may be taking place on the presence or
absence of key ancillary attributes.

Purpose of Study
The purpose of this study is to examine which ancillary attributes are
of greater importance to cross-country skiers in choosing a destination and to
examine ancillary attribute importance in explaining variations in length of
stay in combination with distance from home, perceived constraints,
satisfaction, past overnight ski trips, skill level, and trail quality.

Importance of Study
Instead of asking ". . . at what level tourism development detracts from
the wilderness experience" (Fridgen 1984, p. 23), this study examines which
types of developm ent contribute to the rural, small-town, cross-country ski
experience. No other studies have examined the importance of the attributes
to be found in small rural communities to cross-country skiers. Several other

8
studies have examined trail attributes, motel or resort attributes, and
motivations and experience outcomes.
The ancillary attributes of a destination often result in additional
visitor activities which take place during the same trip as the main activity.
For some members of the cross-country ski group, these additional activities,
socializing at a tavern for example, may become the main activity. Activities
and the attributes they are based in a trip to a specific destination are termed
activity packages (McCool 1978). The activities considered as part of a
package in any one publication are often those which take place in a wild
land setting, or those which take place in a built environment, but rarely are
the activities of the two settings combined.

The activity package concept can

incorporate a combination of attributes which make up the entire trip
experience, including those attributes of direct importance to performing the
main trip activity and those which facilitate the main activity by providing
comfort, entertainment, and sustenance while the recreator is disengaged
from the main activity (Gunn 1994).
Witt and Wright (1992) recommend examining valence and
instrumentality of the attributes to tourists across destinations along with
expectations of actually visiting the destination. While the specific situation
of the present study makes such an arrangem ent unworkable, study results
represent a valuable first step in that direction for northern rural
communities.
Results from this study are of interest to the cross-country ski industry
and especially to small communities in Michigan's Upper Peninsula
attem pting to increase their cross-country skier market share. Knowing which
attributes are of importance to cross-country skiers will allow more effective

9
m arketing campaigns aimed at changing beliefs about these attributes,
changing the relative importance of attributes, adding new important
attributes at destinations currently lacking them, and changing beliefs about
the "ideal" cross-country ski destination (Hawkins, Best, and Coney 1992).
This study could be used as a part of a much larger study to measure
tourism potential and suitability on either a regional or statewide basis. Often
such studies select variables or weight them based on literature reviews or
expert opinions (Pearce 1982). Variables and weights should be selected
based upon empirical evidence of what tourists consider to be of importance.
The availability of attributes im portant to cross-country skiing and other
activities could be measured, m apped, and summed for each activity. Such
an endeavor could be useful in guiding investment in tourism infrastructure
or avoiding unnecessary or dam aging development. Blank and Gunn (1965)
developed such a plan for Michigan's Upper Peninsula in the early 1960s. In
the intervening 30 years, much change has taken place in the region and in the
activities which recreationists pursue. Advocates of this approach to tourism
development have focussed prim arily on warm weather activities, making a
study of winter tourism potential especially valuable.

CHAPTER II: LITERATURE REVIEW
Several areas of research contribute to the foundation of this study.
Multi-attribute theory has been the basis of much research in recreation and
tourism. Attributes have te e n used to analyze attitudes, motivation,
decision-making or choice, preferences, and satisfaction. The attributes used
most often in recreation research are those which directly influence on-site
recreation experiences (such as trail, park, or campground attributes), while
tourism research often focuses on attributes of destination cities, regions, and
service centers. Few tourism studies, other than those which consider hiking
or camping, examine the roles of attributes in specific outdoor recreation
activities or highly specialized activities such as cross-country skiing.
Few studies combine ancillary attributes (those attributes not directly
related to the prim ary purpose of the trip) with other variables which have
been shown to be associated with length of stay. Those that have linked
attributes to other variables either use general items such as "hotel quality" or
rather narrow items such as "hotel appearance." Ancillary attributes are
sometimes considered in economic studies as a portion of such variables as
on-site price.
Attribute importance or preference has been examined on an attribute
by attribute basis. Attribute ratings have been combined into underlying
dimensions and frequently these dimensions have been rotated into
orthogonal variables through factor analysis. The resulting factors, or
sometimes even the individual attributes, have then been used to segment the
10

11
market either through cluster or discriminant analysis (i. e. Crask 1981; Shih
1986, Taylor 1986; Teye 1989; Davis and Sternquist 1987; Ritchie and Zins 1978;
Crompton, Fakeye, and Chi-Chuan Lue 1992; Rao, Thomas, and Javalgi 1992).
Attributes have also been combined and manipulated in factorial designs in
experimental choice research (Louviere and Timmermans 1992). Typically in
such research, each attribute is divided into several levels. The different
levels of the attributes under study are then combined into descriptions of
several possible destinations from which the experimental subjects choose.
The most im portant attributes and attribute levels are then revealed through
statistical analysis of the num ber of times each hypothetical destination
description was chosen. Attributes have also been evaluated through
examination of numbers of actual visitors and variations in travel cost in the
economic literature (Mendelshohn 1987; Caulkins, Bishop, and Bouwes 1986,
Morey 1981).

Theoretical Framework
As Levy (1979) has discussed, variations in leisure behavior are a result
of the interaction of differences among persons and environments. This study
focuses prim arily on the variations in attributes which comprise the unique
environment of a small northern Michigan community and the differences in
the importance of these attributes to people. In combination, these attributes
form a complete and somewhat unique cross-country ski destination. Cross­
country skiers who desire those attributes present in the Munising
environment would more likely be found in Munising than those who desire
an environment offering other amenities.

12
The theoretical background for this study is based on economic utility
theory or the rational economic model. Utility theory states that a rational
individual will choose the option which has the most value for that particular
consumer (Hogarth 1987). However, the individual is unlikely to be able to
know about or compare all possible products of interest on all dimensions of
interest and select the alternative that maximizes expected utility or value. To
reduce the complexity of the decision task, people are likely to look for
certain levels or degrees of quality on several key attributes. When an
alternative is encountered which meets or exceeds each of the specified levels
of the key attributes, it is chosen. As the search continues, the decision-maker
may change the looked for levels of attribute qualities to reflect what is
encountered in the market.
Often, a choice m ust be made among several alternatives. Decision
makers usually employ one or a combination of several different strategies or
models in choosing among alternatives (Hogarth 1987). Whatever the strategy
used, it can be considered a form of the basic linear compensatory model.
Each salient dimension of an alternative is measured; the measure is weighted
to reflect the importance of the dimension in making the decision; the
alternative having the greatest combined value of all weighted dimension
measures is chosen. Variations of this modei exist. The additive difference
model evaluates the differences between two alternatives on a dimension by
dimension basis. The differences are then totaled and compared. With the
ideal point model, each alternative is compared with what the decision maker
w ould consider the perfect alternative which may or may not exist.
There are several non-compensatory models (Hogarth 1987), but they
can be considered special cases of the compensatory model with extreme

13
weights applied to some dimensions (or portions of their measurements), a
time order comparison factor, or an importance order comparison factor used
to determine the order of attribute comparisons.
The conjunctive model sets cutoff points on the dimensions, and any
alternative with a dim ension not exceeding the cutoff point is dropped from
consideration. The disjunctive model allows for a low score on one
dimension provided some other dimension receives a very high score. The
lexicographic model compares alternatives on a dimension by dimension
basis starting with the most im portant dimension, continuing dimension by
dimension in order of importance until a difference is found among the
alternatives. The elimination-by-aspects model eliminates alternatives based
upon the which alternative lacks a dimension. The dimension used for
comparison is chosen probabilisticly on a dimension by dimension basis from
a pool of dimensions of interest.
In the above strategies, the key decision points are the choosing of the
dimensions, or in the terminology of this study the attributes, to be used in
evaluating the alternatives and the weighting of those alternatives. The
attributes chosen and their assigned weights largely determine which
alternative is ultimately chosen.
If the outcomes of choosing among alternatives are uncertain, as they
usually are in choosing a travel destination, then expected utility or
expectancy theory ( H ogarth 1987, Witt and Wright 1992) can be used to
evaluate the uncertainty involved in the choice along with the quality and
importance of the combination of key attributes. Expectancy introduces a
probability that a specific outcome will occur.

14
Each product can be thought of as a bundle of attributes from which
utility is derived (Lancaster 1966), just as the ownership of property can be
conceptualized as the ow nership of a bundle of rights (Barlowe 1986).
Whether or not the product is property with a bundle of rights or a vacation
destination with a bundle of related service and attraction attributes, the
consumer bases his or her purchase decision upon the benefits contained
within the bundle. Competing destinations can be thought of as competing
products or competing bundles of attributes. Only when a consumer has
made a purchase decision to visit a particular destination, can attributes
which fill a similar function, or niche, within the destination be thought of as
competing products (Goodall 1991).

Until the customer has made a

destination selection, all the attributes of a particular destination are in effect
on the same team and part of the same product. As in sports, some members
of the team (shops, restaurants, hotels, etc.) may be more valuable in scoring,
attracting customers, or, more specifically, attracting cross-country skiers
than others. Attributes may be experiences, activities, facilities, or
environmental factors. The destination which the tourist believes to have a
higher degree of the attributes that the tourist considers important is the
destination most likely to be chosen, given the tourist has roughly equal
capabilities and risks in visiting each destination.
According to some authors (i.e. Dann 1977), the decision to travel is
based mostly upon so-called push factors (need for a break from work,
relaxation, social interaction, ego enhancement). However, in this study, the
decision to travel and the decision to participate in a specific activity (cross­
country skiing) had already been made by the subjects so that an examination
of push factors (Dann 1977) is beyond the scope of this study. Hodgson (1983)

15
indicates that many of the push factors, the decision to go, how long, budget,
traveling companions, and type and purpose of trip maybe beyond the
influence of strategic m arketing programs. And, since the destination had
already been chosen by the respondents, measurement of the beliefs of the
respondents that destinations have varying amounts of pull factors would
have been difficult to measure without a large amount of bias caused by
closeness of the decision and the actual visit to Munising, Michigan.
This study focuses on the types of pull factors which lead to the
selection of the destination out of a wide range of possible destinations after
the push factors have resulted in a basic level of motivation to travel
(Manfredo 1989). Those ancillary attributes deemed more im portant could
have a greater role to play in assuring the cross-country skier the behavioral
outcome sought on the cross-country ski trip. Recreation Opportunity
Spectrum literature (i.e. Driver and others 1987, Virden and Knopf 1989,)
indicates that these pull attributes may differ for people seeking different
experiences. Kaltenborn and Emmelin (1993) found that five groups of
visitors to Svalbard, a relatively undeveloped group of islands in the
Norwegian arctic, had similar desirability ratings for setting attributes such as
remoteness and naturalness, but differed on such management attributes as
transportation, hut access, and information. The groups also differed on their
travel patterns.

Experiences offered by National Park Service and U.S. Forest

Service trails in Munising may be different leading to skiers frequenting each
agency and who desire different ancillary attributes.
When both groups are considered together, some attributes could be
expected to be more im portant than others. Pearce (1982) states that natural
environments may be of particular value in providing self-actualizing

16
experiences while the service centers or cities provide for physiological
needs. Physiological needs such as hunger and thirst are of primary concern
to the tourist and are accompanied by needs for identity, status, and security
(Teare 1990). Attributes which either provide satiation of physiological needs
or directly contribute to self-actualizing experiences could be expected to be
rated as being more im portant by tourists than other attributes which do not
have a strong need link, but perhaps serve as fillers between strong need link
activities. Mayo and Jarvis (1981) note that change from everyday routine
during travel may lessen the motivation to satiate lower-order needs. Those
needs moved into the background would most likely be mid-range needs
capable of being fulfilled at home and not critical to short term well-being,
leaving physiological needs, which need satiation on a short-cycle periodical
basis, and higher order needs at a higher state of awareness, resulting in
stronger motivations for fulfillment (Atkinson and Raynor 1974). Positive or
satisfactory experiences result from the fulfillment of physiological, love and
belongingness, and self-actualizing needs. Negative or unsatisfactory
experiences result from a failure to attain lower order needs such as
physiological and safety needs (Pearce and Caltabiano 1983).

Satisfaction
Satisfaction is related to ancillary attribute importance. A community
w ith a num ber of im portant attributes could interest customers in visiting.
Satisfaction with those attributes could then influence length of stay or the
num ber of return visits. Including both importance and satisfaction measures
in a consumer behavior model are important because importance indicates
expected ancillary attributes and satisfaction indicates how well the

17
community fulfilled those expectations. Lawler (1973) in his discussion of
fulfillment, discrepancy, and equity theories of satisfaction indicates that
when a customer receives more of the things he feels he should receive,
greater satisfaction results. According to Dorfman (1979) overall satisfaction
is most dependent on the perception of the conditions deemed valuable, on
the differences between perceived and preferred states, and on the differences
between perceived and expected conditions. Or more simply put

..

benefits provided by a destination must match [the] benefits sought by the
market . . . " (Woodside 1982). Therefore, those skiers who rated the ancillary
facilities which Munising offers as being more important could be expected
to be more satisfied than those skiers who rated as more important the
attributes which Munising is lacking. Those skiers who are satisfied could be
expected to stay longer than cross-country skiers who are not. The physical
environment can signal the intended market as to the types of services and
products available (Bitner 1992) . Those skiers who did not receive the signal
or misinterpreted it are apt to be dissatisfied if they are looking for different
products and services.
Although the respondents for this study were merely requested to
indicate how im portant a selection of ancillary attributes were in selecting a
destination community for the activity of cross-skiing, importance in this
study could be interpreted as asking the cross-country skier if he or she felt
this benefit or product should be included in the destination service package
for consumption during the cross-country ski vacation. The questionnaire
did not specifically ask respondents to compare the services received in
Munising to other similar products, but respondents perhaps compared them
anyway as a matter of habit. Data collected on site tend to yield a large

18
num ber of satisfied respondents because dissatisfied respondents tend to
leave or not come in the first place. People tend to convince themselves they
are satisfied regardless of their experience to justify to themselves the time,
money, and trouble they took to come to destinations such as Munising
(Rollins 1985; Rollins and Chambers 1990; Heberlein 1977).
Dorfman (1979) determ ined that the most im portant determinant of
satisfaction in camping is the presence or absence of annoying conditions.
Lawler (1973) also notes that facets with which workers are dissatisfied are
likely to be perceived as being more important by the workers than those job
facets with which they are satisfied. People notice service which is below
their expectations much more than they notice satisfactory service (Normann
1991). Dissatisfaction may be more highly correlated with importance than
satisfaction (Lawler 1973). Thus people who rated gourmet restaurant as
extremely im portant are likely to be more dissatisfied than those skiers who
rated family restaurant as extremely important are satisfied because Munising
comes closer to supplying family restaurants than gourmet restaurants.
Satisfaction has been conceptualized as being composed of several
underlying performance dimensions (Pizam 1978, Lawler 1973) which must
be identified and measured and their relative importance determined. Pizam
(1978) measured 32 items developed from a review of consumer satisfaction
and destination attractiveness literature, interviews with local tourism
experts, and open-ended interviews with tourists. Eight factors were derived
from the 32 items: beach opportunities, cost, hospitality, eating and drinking
facilities, accommodation facilities, campground facilities, environment, and
extent of commercialization. However, some of the items lumped together
included some very different attributes (restaurants, cafeterias, and bars), and

19
in other cases one item represented a whole class of attributes which could
contain a substantial amount of variation (hotels/ motels) for example.
In a similar fashion, Noe (1987) compared instrumental and expressive
measures of satisfaction in two regression analyses with two different general,
overall satisfaction measures as the dependent variables. The expressive
measures were goal related and the instrumental measures were items likely
to facilitate the attainment of the goal. Participants and spectators at a river
raft responded to 23 items. The responses were then factor-analyzed using
principal components into five factors. The expressive factor, which included
such items as partying, music festivities, and crowd watching, explained most
of the variance in the general, overall satisfaction measures. The other four
factors which included items which were more controllable by managers such
as satisfaction with police, park rangers, rest rooms, food, beverages, traffic,
and access were not nearly as influential on overall satisfaction. Noe (1987)
interpreted this result to indicate specific instrumental items were considered
a means to the overall expressive end of partying, considered to be a main
function of the event. However, the overall measures of satisfaction were
slanted towards the river raft race activities with both overall measures of
satisfaction specifically mentioning the raft race or river raft race activities.
The river race study did not distinguish between respondents who were
tourist and those who were local participants. Given the questionnaire items
used in the study, Noe (1987) likely expected mostly to have respondents
from the local area. A sample of purely tourists would perhaps be more
sensitive to instrumental attributes such as food and beverage.
Overall satisfaction is often discussed as being determined by some
combination of separate satisfactions derived from different facets of the

20
experience (Lawler 1973). Therefore, overall community satisfaction can be
used as a measure of the performance of all the community's ancillary
attributes in general. Overall community satisfaction in this study would be
expected to be significantly determined by the component satisfactions
related to lodging, restaurants, and shopping. Trail satisfaction was also
measured and could also be expected to influence overall community
satisfaction but to a lesser degree than attributes to be actually found in the
community. Respondents may not have considered ski trails to be a part of
the community because a num ber of the trails are located several miles from
lodging establishments and the main part of town. However, based on the
results of Noe (1987), trail satisfaction, a more goal oriented satisfaction
measure than overall community satisfaction, could perhaps be expected to
have a stronger relationship to length of stay than overall community
satisfaction.
Overall community satisfaction may be more than the sum of its parts
when the experience is vacation related and satisfaction is measured in a
general fashion. Some people may be determined to enjoy themselves on
vacation. Enjoyment may be difficult for the vacationer if she or he is not
highly satisfied. Such vacationers may report high levels of satisfaction as
part of their effort to enjoy themselves. More insight may be gained in an
examination of dissatisfaction during vacation experiences than satisfaction.
Episodes of poor service may have a stronger effect on overall satisfaction
than satisfactoiy service, or outstanding service. Or, as the results in the study
by Noe (1987) may indicate, a goal oriented measure of satisfaction (trail
quality satisfaction in the case of the present study) may have more influence
than satisfaction with facilitating or ancillary attributes.

21
Visitor satisfaction in some situations may be highly associated with
visitor density or crowding. Visitor density or a crowding variable has been
included in some recreation valuation studies along with various other
measurable quality attributes (Englin and Mendelsohn 1991; McConnell 1977).
The value and importance of the quality attributes included in the models is
usually determ ined by comparing the number of visits to a num ber of similar
sites which differ on the quality variables in question. Some authors such as
McConnell (1977) examine differences in perceived quality through direct
interviews to determine differences in individual willingness to pay. The
potential of diminishing marginal benefits and ultimately total benefits
caused by dissatisfaction related to the presence of additional visitors at
recreation sites has been much discussed both in the recreation literature and
the economic literature (i.e. Shelby 1980; Rollins and Chambers 1990).
Crowding, in the case of Munising, was thought by the U.S. Forest Service not
to be a problem. However, the lack of a suitable num ber of patrons can also
be a problem for some recreation activities or for some types of participants
(McConnell 1977). This potential problem for Munising cross-country skiers
was not addressed in the questionnaire from which the data for this study
were derived. However, comments volunteered by many of the respondents
indicate that cross-country skiers in the Munising area favor uncrowded
conditions. Several respondents mentioned they were fleeing crowded
conditions elsewhere when they discovered Munising as a destination at
which to undertake cross-country skiing.

22
Regression Models
Regression models have been used at least since the 1960s to predict
visitation to downhill ski areas and explain variations in visitation and length
of stay, often in a travel cost framework. Echelberger and Shafer (1970) related
total visitor-days at 26 ski areas in northern New England and New York to
facilities, management practices, and distance to metropolitan centers during
the winters of 1964-65 and 1965-66. During a year of low snowfall (1964-65),
total visitor-days were found to be related to advertising program.
Intermediate trails, driving time, rolled and packed trails, and num ber of
instructors were found to be related to visitor-days during a year of plentiful
snowfall (1965-66). A model taking into account both years indicated that
advertising and accessibility were the most significant factors in explaining
variations in visitation (Echelberger and Shafer 1970).
Miles of novice, intermediate, and expert trails, lift capacity, slope
exposure, snow making capacity, days of operation, advertising budget,
driving time, num ber of instructors, percent of groomed slopes, percent of
rolled slopes, percent of advertising budget for broadcast media and
magazines, and percent of advertising budget for brochures and leaflets were
also included in the analysis. All variables were entered into a factor analysis,
then the variable with the highest loading on each resulting factor was
incorporated in a regression analysis. The regression models developed for
the low snow, plentiful snow, and combined years explained 71% to 89% of
the variation in visitor-days. In addition, Echelberger and Shafer (1970)
wanted to include in the analysis, but did not because of measurement
problems, num ber of local ski rentals, num ber of private dwellings for rent,
num ber of liquor licenses, sleeping capacity of surrounding hotels, motels,

23
and tourist homes, and dining capacity of nearby restaurants. These variables
are ancillary attributes and not directly related to the down hill ski slope.
Echelberger and Shafer (1970) recommended collecting data over a num ber of
years to take into account climatic and other variations. They noted the
difficulty of obtaining accurate data from commercial recreation facilities.
Johnston and Eisner (1972) used data collected by Herrington (1967) to
analyze downhill-skier day-visits at a large num ber of California resorts
during the 1963-64 season. Single equation multiple regression models were
used to develop models of the effects of the cost of a day lift ticket, total lift
capacity, length of season, competing lift capacity within 30 minutes driving
time, competition from other sites, and location characteristics on the annual
num ber of day-visits. Three out of many models were considered "most
satisfactory" (Herrington, 1967, p.46). Total lift capacity, capacity at nearby
sites, and length of season were positively related and cost of a day lift ticket
was negatively related to the num ber of day-visitors.
Johnston and Eisner (1972) speculated that the addition of population,
distance, and socioeconomic characteristics, would improve the models. The
authors also strongly recommend using site specific variables rather than the
location shift dum m y variables which they incorporated into their models.
They also suggested that visitation may be strongly dependent on visitation in
prior years. If these results are applicable to cross country-skiing, then it
would seem that destinations with more trails and more kilometers of trails
(more capacity) would have a larger num ber of skiers.
Eisner (1971) in a two stage effort to develop a model to predict
visitation to new California downhill ski resorts included distance in a
regression model as the m inim um driving time from the center of each

24
county to each ski resort in the study. Gross county skier outflow was
regressed with an urbanization variable and the product of population and
availability, to predict the num ber of skiers from each county. Availability
was calculated by dividing uphill capacity by distance and summing for all
resorts.
In the second stage of the model, ski use allocation was calculated as
the percentage of skiers from one county which went to a particular resort.
Distance for this part of the model was the minimum travel time in minutes
from the most populous city in a county to each resort. There was no
explanation as to why this different definition of distance was used.
Predicted allocation was based on uphill capacity, open slope and distance.
Other site characteristics considered were vertical rise, length, num ber of lifts,
rental equipment, night skiing, instructors, price, jump, sledding, skating,
swimming, dancing, gambling, bar, meals, snacks, rooms, cabin, dorms. It
was not clear why these variables were dropped. Predicted outflows were
then combined with predicted allocations to predict attendance at each resort
in the study. The model is somewhat circular in that distance and lift
capacity play large roles in both the allocation and outflow sides of the
model. It seems that, once the population of each county and the percentage
of skiers in each county are known, each ski resort could predict its share of
skiers based on distance without the first stage of the model.
Mak and Moncur (1980) tested (using a tobit model) Rugg's (1973)
theory, which was based on the work of Lancaster (1966), that destination
choice and length of stay depend on the attributes of individual destinations.
According to Rugg (1973), utility is derived from dwelling at the destination
for a period of time which allows the consumption of destination attributes.

25
However time on-site also has an opportunity cost in that it could be used in
some other leisure activity or in work. These two aspects of time, as product
generating utility (time used in enjoying recreation experience) and as an
input (raw material to be combined with other inputs to create the product)
continue to create a problem in recreation dem and estimation (McConnell
1992; Bockstael, Strand, and Hanem ann 1992).
The Tobit model developed by Mak and Moncur (1980) was tested
using data gathered from U.S. mainland visitors to Hawaii in 1974. However,
only the amount of rain and the concentration of rooms were used as
individual destination attributes. Rain was found to reduce visits and room
concentration was found to attract visits. Room price had a negative effect on
both destination choice and length of stay. Age was the only personal
variable with an effect on travel choice, but only for first time visitors. Other
personal variables included in the study were education, marital status, and
party size. In addition to the individual destination attributes of room
concentration and rain, money and time constraints also impacted length of
stay and destination choice, with time being more im portant once travelers
had chosen Hawaii as a destination.
Silberman (1985) examined the effects of 12 variables on length of-stay
at Virginia Beach using an approach he attributes to Walsh and Davitt (1983).
Silberman (1985) developed a model using stepwise regression and then with
the same data estimated dem and curves. He first calculated a dem and curve
for average total variable cost per day and length of stay holding the
remaining variables constant at their mean levels. He then dem onstrated the
effects of shifting the dem and curve by varying each of the remaining
variables individually while holding the other variables constant.

26
Direct cost per day, distance traveled, annual household income, effect
of the recession, more than one trip, staying at a campground, advance
planning, planning to visit again, participating in sports, advertising, classy
image, and rundow n image were significantly related to length of stay.
However, length of stay, in the case of Virginia Beach, was not sensitive to
changes in cost per day, income, proportion of visitors affected by the
recession, making more than one trip, number of months of advance planning,
proportion of visitors planning to visit again, proportion of visitors
participating in sports, classy resort image, and rundow n image. Length of
stay was more sensitive to distance traveled, increases in the proportion of
visitors staying in campgrounds, and proportions of visitors learning about
Virginia Beach through advertising.
In formulating the dem and curve, Silberman took the somewhat
unusual step of using total cost per day as the price rather than distance
traveled or travel cost while holding constant the effects of all other
independent variables. This would tend to confuse the effects of travel cost
which may in fact increase length of stay while costs on site may work to
decrease length of stay. Those traveling farther would tend to stay longer to
recoup their investments and those staying in low cost accommodations such
as cam pgrounds could perhaps afford to stay longer. Silberman also took a
somewhat unique approach in allowing respondent ratings of various aspects
of the image of Virginia beach into the stepwise regression analysis. Usually,
site characteristics in regression dem and studies are merely combined in an
on-site price or cost variable, if prices exist, or measured in a strictly objective
m anner based on physical properties, dissolved oxygen for example in water
quality studies (Bockstael, Hanemann, and Kling 1987; Parsons and

27
Needelman 1992; Parsons and Kealy 1992). In hedonic travel cost models,
variations in travel distance to a variety of recreation sites are used to value
site characteristics which usually have been measured in an objective fashion,
square meters of exposed rock for example (Brown and Mendelsohn 1984;
Englin and M endelsohn 1991; M endelsohn 1987). Importance and preference
ratings have been used in a num ber of other ways including preferences for
various landscapes and recreation facilities (Lime 1971; Lucas 1980; Allton
and Lieber 1983; Peterson, Dwyer, and Darragh 1983; Lucas 1985; Watson,
Williams, Roggenbuck, and Daigle 1992; Love and Watson 1992). However,
these ratings usually were not then inserted into regression equations to arrive
at dem and or consumers surplus measures.
Uysal, McDonald, and O'Leary (1988) developed a dem and model for
cross-country skiing using length-of-stay as a measure of quantity for the
dependent variable . Data were draw n from the 1982-83 Nationwide
Recreation Survey. Length of stay was hypothesized to be a function of cost,
distance traveled, skiing trips in the past year, num ber of persons in party,
number of persons encountered at the site, man-made structures, prominence
of non-recreational activities. Distance, num ber of trips, cost, and man-made
structures were found to have the greatest effect on length of stay. The
coefficients for these variables had negative signs indicating that as these
variables increased, length of stay decreased. Num ber of persons in party
(positive correlation) and num ber of trips (negative correlation) had the
highest correlations with length of stay.
Distance is often correlated positively with length of stay in that people
who travel farther m ust stay longer to recoup their investment. Num ber of
trips could also be hypothesized to correlate positively with length of stay in

28
that those people with greater numbers of trips would be more involved with
the sport of cross-country skiing and would thus not only travel more
frequently to cross-country ski but would also stay longer while on those
more frequent trips. An explanation offered by Uysal, McDonald, and
O'Leary (1988) was that cross-country skiing is an activity which frequently
takes place close to home during an allotment of only a portion of the day.
Time not spent traveling is spent on the site skiing. The type of trip, whether
overnight or a local trip, may be an important additional variable in
examining length of stay of cross-country skiers. Furthermore, Wanhill (1990)
states that "stay tourists" should be kept separate from "day" tourists in
analyses because only those tourists who stay overnight consume the entire
product.
Morey (1981) developed a constrained utility model to examine the
variations in the num ber of days Colorado skiers spent at a num ber of
Colorado ski resorts on day trips. Costs, skier ability, and site characteristics
were im portant in determ ining a resort's share of the market. The model
developed was found to predict skier choice better than a logit model.
Results were consistent with the theory of constrained utility maximizing
behavior. Effects of changing prices and characteristics on market share were
estimated. It was assumed that budget allocation was independent of income,
prices, and nonskiing activities, and, that marginal utility of skiing at a site
decreases with increasing trips to that site. Site characteristics included were
acres of ski terrain, acres designed for the skier's ability, vertical transport
capacity, average annual snowfall, and an index of other physical attributes,
assumed to be zero. Because participants in the study were mostly single,

29
day-trip skiers, characteristics such as child-care facilities, lodges, and
nightlife were assumed to be unimportant.
Walsh and Davitt (1983) provide a good rational for using length of
stay in dem and functions and dem and curves. The skier faces two decisions:
how many trips to make and how many days to spend on each trip. At resorts
located a great distance from the home residences of their customers, the
average num ber of trips dem anded is usually relatively low, usually around
one. The num ber of trips, because of travel time and cost, is greatly
constrained. The num ber of days to stay at the resort given that a trip will be
made is the principle decision. Variables significantly associated with length
of stay at Aspen, Colorado during the winter of 1977-78 were cost, household
income, distance, substitution, party size, skiing ability, state population, and
whether or not the skiers were on a package plan. Other variables were
examined, some of which may have been significant if they had been entered
into the stepwise regression in a different order.
Bell and Leeworthy (1990) in a study of dem and for Florida beach days
recommended in response to problems noted about the travel cost method of
estimating recreation values (Smith and Kopp 1980) that tourists and residents
be treated differently as they face different decision situations caused
primarily by distance. The travel cost method assumes that recreation trips
are single purpose, result in the same amount of on-site time, and that the
same mode of travel was used by all travelers (Smith and Kopp 1980). The
travel cost method ". . . treats tourists and residents . . .[as] utility clones" (Bell
and Leeworthy 1990). Bell and Leeworthy (1990) argue that consumption of
recreation services should be measured as num ber of trips times average trip
length for those trips. As noted by Walsh and Davitt (1983) the decision faced

30
by distant tourists is not how many trips to make but how long to stay as time
and travel cost constraints limit the num ber of trips severely. Bell and
Leeworthy (1990) discussed modifying this approach somewhat by
multiplying average trip length with num ber of trips. In their analysis
however, they used the num ber of days on the beach (the prime attraction in
their study) m ultiplied by the num ber of trips. This approach theoretically
would better model the trade off between the num ber of trips and trip length
for visitors traveling a w ider range of distances.
Length of stay at the destination would be a better variable to use than
trip length as the measure of the am ount of recreation dem anded in such an
analysis. If trip length is considered to be the duration rather than the
distance of the trip, then a portion of the trip length may be considered a cost
rather than a benefit. Travel to the destination, depending on the individual
consumer and several environmental factors, could be either cost or benefit.
The travel cost models for analyzing dem and frequently include such
variables as travel cost and travel time as part of the price paid to enjoy an
amount of recreation usually measured by the num ber of trips to a
destination. By using trip length as the measure of benefit and also using
travel cost and travel time, which are often highly correlated with trip length,
as price variables, very similar variables are used as both the price of
recreation and am ount of recreation dem anded. Only a portion of the trip
length should be used to represent the amount of recreation received. The
most likely portion of trip length to be viewed as a benefit by the consumer is
the portion of the trip actually spent at the destination, the portion of the trip
from which the consumer is most likely to derive pleasure. Other portions of
trip length (duration), notably the portions devoted to travel to the

31
destination and travel back home from the destination are more likely to be
viewed as costs and part of the price paid to experience leisure at the
destination.

Even a small percentage of the time spent at the destination may

be viewed as a cost, especially if some time must be spent on family
obligations with disliked people or on a disliked activity. Certainly, the time
on-site has an opportunity cost (McConnel 1992).
The models proposed by Bell and Leeworthy (1990), and Walsh and
Davitt (1983) anticipate the thinking of many small communities at large
distances from what could logically be considered their primary markets.
One of the original U.S. Forest Service objectives in sponsoring the study
from which the data for this analysis were obtained was to explore how to get
cross-country skiers to stay longer in the community (Chase 1991). This
objective assumes that most cross-country skiers will make a small num ber of
trips to Munising during the cross country ski season. The objectives of the
U.S. Forest Service were clearly not on the valuation of the cross-country ski
experience offered on their own and nearby trails and not on the value of the
resource as is the focus of many travel-cost studies (Hof and King 1992), but
on stimulating dem and. The U.S. Forest Service believed that dem and
during the winter of 1990-91 was relatively low. The interest was on
increasing cross-country skier days, and not on the prediction of an exact
num ber of cross-country skier days.
Hof and King (1992) have commented that the Bell and Leeworthy
(1990) model is really not a travel cost model but a new method of valuation
based on on-site costs. This is actually somewhat incorrect. Distance, which
as in the case of Munising, would be highly correlated with travel cost, is still
an im portant factor, perhaps even more im portant than on-site costs. It is the

32
interplay between on-site costs and travel costs (or distance) which
determines the ratio of the num ber of trips and length of stay (Shaw 1991).
Higher travel cost (or distance) works to increase length of stay while on-site
cost should work to decrease length of stay.
Shaw (1991) criticizes Bell and Leeworthy (1990) for not including more
destination characteristics in their model and handling the effects of crowding
in an overly simplistic fashion. In the current study, crowding, based on the
reports of U.S. Forest Service personnel, was not thought to be a significant
problem. Several respondents wrote comments on their questionnaire that
one reason they came to Munising was to escape the crowds at some of their
former destination communities in Wisconsin. As Shelby (1980) points out,
crowding is not always significantly related to satisfaction. However, it is
through its effects on satisfaction that crowding is thought to effect demand.
Shaw (1991) was also critical of Bell and Lee Worthy (1990) for not
including additional personal characteristics in the model. The lack of
individual characteristics in the Bell and Leeworthy (1990) model assumed
that all visitors were homogeneous, which usually is not the case. In the
model proposed in the current study, skill and previous overnight trips are
used as personal variables. Cross-country skiers, especially tourist skiers at a
small market destination village such as Munising, are likely to have
relatively homogenous socio-economic characteristics since small destination
communities are unlikely to have the diversity of tourist services which
would attract a wide variety of cross-country skiers. The distances and
economics of skiing at such communities are likely to act as filters, reducing
the variability in socio-economic characteristics of the cross-country skiers
who visit.

33
Destination characteristics usually included in travel cost models are
very activity specific (i.e. Morey 1981; McConnell 1977; Bockstael, Hanemann,
and Kling 1987). Bell and Leeworthy (1990) accounted for destination
characteristics with two variables, actual price on site per day and the number
of days spent in non beach activities. Price on site per day was a composite of
spending at the destination for hotel, restaurants, fees, and other expenses.
Bell and Leeworthy hypothesized that as price on site increased, length
of stay would decrease. Even though their results indicated support for this
hypothesis, it would not be too difficult to imagine the opposite situation
occurring where price on site may be positively correlated with increased
length of stay. Perhaps in that case, a portion of the value usually credited to
the recreation site in travel cost studies would be better attributed to
supporting services found in destination communities.

Increased on-site

price may indicate increased spending. Increased spending may indicate that
tourists are more involved in activities supported by the local establishments
in the community. More involvement may indicate increased enjoyment of
the destination. The increased enjoyment caused by taking part in more
activities as evidenced by increased spending may increase stay length.
Higher prices may also indicate higher quality and more diverse product
availability and draw a more upscale clientele. People like to have a good
time and will stay longer at places where good times are being enjoyed.
However, usually the thinking is that as price on site declines, stay
length should increase, and as price increases, stay length should decrease.
The results of Bell and Leeworthy (1990) indicated that travel cost per trip
were strongly positively related to beach days. The higher the price to get to
the beach, the more days were spent at the beach. Travel cost is not a sunk

34
cost but a variable cost which provides a choice between num ber of trips and
beach days per trip.
The model Bell and Leeworthy (1990) introduced based on their
discussion of theory used beach days per year. The variable they use in their
dem and equation for which they discussed results was apparently beach days
per trip. Its only beach days per trip that logically should be positively
correlated with travel cost per trip. Beach days per year would cause some
confusion because of the trade off effects between num ber of trips and travel
cost. Visitors from a wide range of distances could have similar beach days
per year by combining differing numbers of trips and beach days per trip.
Destinations which have relatively few customers living at such a distance
where multiple trips are viable should be most interested in increasing
dem and through increasing stay length. Stay length may be a better indicator
of dem and than num ber of trips for destination communities at distances
where multiple trips by a single tourist during any one season are unusual or
where tourists are likely to visit a variety of destination communities in
successive trips, seldom repeating destinations in any one season.
Multiple short-trip customers are likely to spend less per trip, not only
because of their shorter stays (requiring fewer lodging nights and restaurant
meals), but also they are not as dependent on the host community for a variety
of other goods and services which they can either bring from or enjoy upon
their return to home. Those customers who find certain destination ancillary
attributes im portant are likely to spend more money while at the destination
and stay longer at the destination.
Although Fletcher, Adamowicz and Graham-Tomasi (1990) state that ".
. . perceptions of attributes is of particular importance in recreation dem and

35
models", and suggest that psychometric approaches be incorporated in single
site recreation dem and models, few dem and studies actually incorporate
such measures in their models. Some have recently appeared in the tourism
literature. Silbermann's article (1985), discussed above, used perceptions of
Virginia Beach's image based on seven image characteristics in modeling
length of stay in a two-stage least squares model. Dadgostar and Isotalo (1992)
examined

u c S u R u iiG n ii i' ia g c a t id ii5 ic i c x i i u i i a i i i p iu ij'ic iV u in b c r Oi Ciciy 5 S p e n t

in leisure and recreation activities in near-home city destinations during the
previous 12 months. The importance of overall cost of the trip was measured
on a 1 to 7 scale as opposed to average dollar cost per day. This choice of
m easurement was defended on the basis of the substitutability of activities
within a single visit and the m ultipurpose nature of most trips. Since the
trips were often m ultipurpose, the sorting out of costs according to purpose
of activity would be difficult, especially over a 12 month period.
Substitutability of activities w ould be less of a concern because income was
incorporated in the model as a constraint. Lower priced activities would
tend to result in lower on-site costs and possibly more tourist days overall for
each income level. Tourists were defined as those who spent at least 50% of
the time at the city destinations engaged in leisure and recreational activities.
Ancillary attribute importance to cross-country skiers has not yet been
adequately analyzed, and cross-country skier ancillary attribute importance
ratings have not been adequately linked to other constructs such as
satisfaction and dem and concepts such as length-of-stay. As Getz (1993)
recently stated in an article on tourist shopping villages:

36
A working hypothesis . . . is that heritage resources are in
themselves the initial attraction, but tourist-oriented shopping
and other services provide the draw for repeat visits. Retail
surveys are . . . needed to test tourist preferences for types of
goods and services, especially the assum ption that arts and
crafts, 'country' goods and souvenirs are preferred.
In the case of cross-country skiing, it is the ski trail that is the initial attraction,
but what are the services that provide a draw for repeat visits or make a single
visit of increased duration?
Kelly and Godbey (1992) discuss the idea that for some people
consumption of goods has become synonymous with leisure. That for many
people, leisure has become largely viewed as an opportunity to consume
experiences produced by the actions of other people (as in viewing athletic
events), as opposed to consuming experiences based on self- action. If
consumption based on the actions of others is an im portant part of the cross­
country ski holiday, then ancillary attributes will be important.
It is the ancillary attributes that provide the setting for consumption
based on the actions of others, not the ski trail. An individual skier does not
consume the trail, but merely uses a constantly changing 10 foot section of the
trail during a progressing period of time. The trail will still exist for others to
use as an input in a self-created experience. The trail provides the setting for
an individual activity. But food, drink, entertainment, clothing, skis, and
lodging rely more heavily on others both to create the services and products
consumed and to appreciate the consumption. These products come in a
variety of styles and prices whose consumption provides identification to the
consumer as being part of a class or segment (Veblen 1899).

The trail comes

in basically one size, color, and style, and is available to anyone at low or no

37
cost. The ancillary attributes provide the bulk of any differentiation among
consumers of cross-country ski trail experiences.
The ancillary attribute importance ratings could be expected to vary
among local and tourist skiers, and among Forest Service, and Park Service,
respondents if the skiers prefer different types of cross-country ski
experiences and these preferences are reflected in their choice of trail
operated by different agencies. Tourist and local cross-country skiers have
different needs for ancillary services because the local skier can provide
many services for themselves at home which the tourist skier must purchase
from commercial establishments.

Ancillary attribute importance ratings

should make a contribution in explaining the variance in the number of days
spent in Munising to cross-country ski. Those ancillary attributes which can
be found in Munising, fulfill basic needs, or are related to cross-country
skiing should be rated more highly in importance.

CHAPTER III: METHODOLOGY
Statement of the Problem
O utdoor recreation behavior is often conceptualized as dependent
upon facilities or attributes provided by government and the time, money,
and interests of the participants. Increased understanding of participation
may result from a more thorough examination of attributes which are not
directly related to the main purpose of the trip, ancillary attributes.
Ancillary attributes (often provided by commercial establishments) may be
especially im portant to cross-country skiers because of the winter conditions
and physical exertion which characterize this sport. The problem is to
determ ine which ancillary attributes are im portant to cross-country skiers.
This problem is examined through two methods: (1) comparing the ancillary
attribute importance ratings of individual attributes with an average rating of
22 attributes and (2) determ ining which ancillary attribute importance ratings
have a significant association with length of stay or willingness to pay trail
fees in stepwise regression models. Those ancillary attributes associated with
basic needs and present in the community could be expected to receive
higher importance ratings than other ancillary attributes. Possible differences
in the way U.S. Forest Service and National Park Service respondents and
tourists and local respondents rated the importance of ancillary attributes are
also examined. Such differences could indicate a need for separate regression
models for those groups.

38

39
Hypotheses
HI: Mid price lodging, family restaurants, and quality of cross­
country ski trails will be rated as more important in choosing a cross-country
skiing destination community than an average of 22 attribute importance
ratings.
H2: Differences in ancillary attribute importance ratings will occur
between local and tourist cross-country skiers and between U.S. Forest
Service and National Park Service respondents.
H3: Ancillary attribute importance variables will improve a base
regression model with distance, constraints, trail satisfaction, overall
community satisfaction, previous overnight trips, skill level, spending, and
trail quality importance as independent variables and length of stay or
willingness to pay trail fees as a dependent variable when analyzed in a
stepwise regression procedure.

A ssum ptions
Cross-country skiers who visited Munising during the season of 199192 had some prior knowledge of the services and products available in
Munising and based their decision to visit Munising in part on that
knowledge. Cross-country skiers visited Munising, rather than an alternative
destination, in part because of the ancillary attributes present in the
community. This analysis examines which ancillary attributes are important
after making the assumption that some ancillary attributes are of importance.
Respondents were able to correctly ascribe varying levels of
importance to various destination community attributes and were able to
record accurately these importance levels on the questionnaire.

40
Respondents use a linear additive compensatory model or a variation
thereof in selecting cross-country ski destinations in which they consider
destination community attributes in addition to cross-country ski trail
attributes. This analysis does not examine decision mechanisms but assumes
respondents evaluate community characteristics and select the community
with the largest num ber of im portant attributes. The presence of one attribute
may compensate for an attribute which the community lacks.

Variables
U.S. Forest Service respondents are those cross-country skiers who
received a questionnaire at one of the five cross-country ski trails operated
under the auspices of the U.S. Forest Service in the vicinity of Munising,
Michigan.
National Park Service respondents are those cross-country skiers who
received a questionnaire at the Munising Trail which is managed by Pictured
Rocks National Lakeshore.
Mail respondents are those who inquired about cross-country skiing
opportunities at the Upper Peninsula Travel and Recreation Association and
appeared on a mailing list supplied to the Travel, Tourism, and Recreation
Resource Center at Michigan State University. These respondents were sent a
two-page questionnaire.
Local cross-country skiers are those skiers who reported that they did
not spend the night away from home while on the trip during which they
received the questionnaire. Some of these respondents may have traveled a
considerable distance to spend the day cross-country skiing at trails in the
Munising area before returning home to spend the evening. Local cross­

41
country skiers are im portant to include in the analysis because participation
is often measured as num ber of trips. Only recently has the recreation
economic literature explicitly discussed the different decisions faced by local
recreation participants and those traveling a considerable distance. Tourists
may focus their decision on how long to stay while local participants may
focus their decision on w hether or not to make an additional trip during a
specific time period. These differences in planning may be reflected in the
importance ratings of some attributes.
Nonlocal or tourist skiers are those cross-country skiers who reported
that they spent at least one night away from home while on the trip during
which they received the questionnaire. Frequently, other definitions of
tourists by other authors have also included a distance from home criteria.
However, for the purposes of this study, the criteria of interest was to
differentiate the dependence of the tourist on community attributes as
compared to a local respondent. Staying overnight away from home was
thought to be a more efficient indicator of dependence on the local
community than a combination of distance and staying overnight.
Ancillary attribute importance ratings were measured by asking
respondents to rate, on a 1 (not important) to 7 (extremely important) scale,
the importance of 22 items in choosing a destination (Figure 2).

The list was

developed based on literature, consultation with local Munising, Michigan
offices of the U.S. Forest Service, and staff and faculty at Michigan State
University. Ancillary attributes are those goods and services which facilitate
the main activity of the trip but are secondary in nature and not directly
connected to the trip purpose. The Ancillary attributes with which Munising
is well supplied is based on the somewhat subjective personal opinion of the

42

H ow im p o rta n t are the availability of th follow ing features to you w h en choosing a cross­
country skiing d estination com m unity? Place a check m ark in the blank w hich reflects how
y o u feel.
Low price m otel
Not Im portant________________________Extremely important
Mid price m o te l................................ Not Im portant________________________Extremely important
Ski lodge or ski reso rt...................... Not Im portant________________________Extremely important
Basic kitchen (sink and microwave) Not Im portant_______________________ Extremely
Child care ......................................... Not Im portant________________________Extremely
Swimming pool.................................. Not Im portant________________________Extremely
Sauna ................................................Not Im portant________________________Extremely
Ventilated waxing room ...................Not Im portant________________________Extremely
L au n d ry............................................ Not Im portant________________________Extremely

important
important
important
important
important
important

Fast food ...........................................Not Im portant________________________Extremely
Family restaurant............................. Not Im portant________________________Extremely
Gourmet restau ran t..........................Not Im portant________________________Extremely
Night club ......................................... Not Important
Extremely
Family oriented entertainm ent
Not Im portant_______________________ Extremely
Cultural attraction........................... Not Im portant________________________Extremely
B a r .....................................................Not Im portant________________________Extremely

important
important
important
important
important
important
important

Gift shops ......................................... Not Im portant________________________Extremely
Clothing sh o p s.................................. Not Im portant________________________Extremely
Ski s h o p s ...........................................Not Im portant________________________Extremely
Local art and craft sho p s................. Not Im portant________________________Extremely

important
important
important
important

Quality of cross-country ski trails . Not Im portant_______________________ Extremely important
Overall community........................... Not Im portant________________________Extremely important

A.

B.

What considerations
come to mind when deciding on
a X-C destination

List each of these considerations
in order of importance

1.

1.

2.

2.

3.

3.

4.

4.

Figure 2. Closed Ended Question and Open Ended Question.

43
author which was formed during several visits to Munising just before and
during the 1991-92 cross-country ski season. Munising is well supplied with
family restaurants and m edium priced lodging. These attributes were
expected to receive higher importance ratings as cross-country skiers desiring
other accommodations and services would be more likely to choose other
destinations.
Satisfaction consisted of asking Forest Service and National Park
Service Respondents how satisfied they were with lodging, restaurants,
shopping, cross-country ski trails, and overall community on a 1 to 7 scale
with 1 being not satisfied and 7 being extremely satisfied. Overall community
satisfaction and trail satisfaction were then included in the regression analyses
below. Overall community satisfaction and trail satisfaction were expected to
have a significant influence on length of stay. Less satisfied respondents were
expected to have shorter stays as they would be more likely to leave earlier
than more satisfied respondents. In addition, very well satisfied respondents
could be expected to stretch their stay to include as many days as possible
before obligations at home or work force them to return to their home
com m unities.
Length of stay was measured by asking respondents how many days
they would be within 15 miles of Munising on this trip. This measure
precluded the inclusion of local respondents in the regression analyses.
Distance was m easured by asking Forest Service and National Park
Service Respondents approximately how many miles Munising is from where
they live. Often in the economic literature distance is used to calculate travel
cost a n d /o r the opportunity cost of travel time or is used as a proxy for those
variables (Donnelly and others 1990; Bockstael, Hanemann, and Kling 1987;

44
Silberman 1985; Bockstael, Strand, and Hanemann 1992). Since travel to
Munising for cross-country skiing is almost exclusively by car with only a
very occasional chartered bus, variations in distance are good approximations
of variations in travel time and travel costs among respondents. The
conversion of distance to actual costs measured in dollars would be more
critical if the purpose of this study was to measure consumer surplus
associated with cross-country skiing in the Munising area.
Constraints are factors which reduce, modify, or eliminate
participation in a specific recreation activity (Crawford and Godbey 1987).
Constraints were measured by asking Forest Service and Park Service
respondents to rate on a 1 to 7 scale whether they agreed or disagreed with
each of a list 17 reasons people do not ski as often as they would like to. The
list of constraints was developed in consultation with U.S. Forest Service and
National Park Service personnel, literature, and staff and faculty at Michigan
State University and is presented in Figure 3. The list included intrapersonal
items (poor physical condition), interpersonal items (lack of companions),
and structural items (too many snowmobiles) (Crawford and Godbey 1987;
Raymore and others 1993). Recent literature on leisure constraints has
discussed a model consisting of a hierarchy of types of constraints.
Intrapersonal constraints are those that are within the potential participant
and are thought to dim inish interest in participating in a given activity.
Interpersonal constraints arise when the potential participant has difficulty
finding suitable partners w ith which to participate, and structural constraints
consist of difficulties arising out of society or the surrounding environment
which reduce participation.

Participation occurs when participants are able

to negotiate through all potential constraints which they face (Jackson,

45

T here are m any reasons people d o not ski as often as they w ould like to. If you strongly
agree w ith one of the reasons below w hich m ay reduce the num ber of tim es people cross­
country ski in th e M u n isin g area, check the blank closest to agree. If you strongly disagree
m ark th e blank closest to disagree. If you feel som ew here in betw een, m ark the blank
betw een agree and disagree w hich m ost closely reflects how you feel about the reason. If
y ou neither agree n o r disagree, m ark the m iddle blank.
It takes too much time to travel to M u nising............... disagree________________________________
Concern about snow on the roads ...............................disagree________________________________
Concern about being unable to return
home on time because of snow ...................... disagree________________________________
Fees charged at Forest Service ski areas
are too expensive .......................................... disagree________________________________

agree
agree

Munising is too expensive............................................ disagree________________________________
. __________________
Obligations at home or w o r k ....................................... disagree__________
Members of family or group would
rather go elsew here....................................... disagree__________
Parking lots are not plowed .........................................disagree_______________________________
Concern about medical care .........................................disagree________________________________
Concern about vehicle burglary .................................. disagree________________________________
I don't feel welcome in the community..........................disagree________________________________
Lack of com panions...................................................... disagree________________________

agree
agree
agree
agree
agree
agree
agree
agree

Concern about becoming lost ....................................... disagree________________________________
Poor personal physical condition................................ disagree____________
Too many snowmobiles around M unising...................disagree________________________________
Not enough to do in Munising ..................................... disagree______________________
Lack of ski r e n ta l.......................................................... disagree____________________________

agree
agree
agree
agree
agree

Figure 3. Constraint Measures.

agree
agree

46
Crawford, and Godbey 1993). The constraint ratings were then totaled for
each respondent for use in regression models examining the number of days
respondents expected to spend in Munising during their trip. Constraints
used in the economic literature usually consist of income and time (Fletcher,
Adamowicz, and Graham-Tomasi 1990). However, a variety of other
constraining factors have been examined in the recreation and leisure
literature (Jackson, Crawford, and Godbey 1993; Jackson and D unn 1991;
Jackson 1993). Some of those constraints are included in this analysis as well
as some which may be particular to the Munising situation. Income was not
measured adequately to be included in the analysis as a separate variable.
However, the list of constraints included some items reflecting feelings about
expense and travel time required to cross-country ski at Munising (Figure 3).
N um ber of overnight trips was measured by asking Forest Service and
Park Service respondents how many overnight trips they had made for the
purpose of cross-country skiing during the previous season. Previous
overnight trips was included as a possible indicator of familiarity with
substitute destinations.
Skill level was measured by asking U.S. Forest Service and National
Park Service respondents to indicate their skill level on a 1 (beginner) to 5
(expert) scale. This variable and num ber of overnight trips during the
previous season, while not only serving to provide some personal
differentiation among respondents as recommended by Shaw (1991), were
originally included as a result of examining recreation literature concerning
activity involvement and specialization (Dimanche, Havitz, and H ow ard 1991;
Bryan 1977; Hollenhorst 1990). This literature suggests that participants of
differing skill and participation levels may have different preferences and

47
behaviors. Those cross-country skiers who are more skilled and travel more
frequently overnight to cross-country ski would be thought to be more
involved in the activity. Those cross-country skiers who are more involved
may have different ancillary attribute preferences which may be reflected in
their importance ratings which may affect the num ber of days of skiing in
Munising either positively or negatively. Highly involved skiers may prefer
fancier accommodations than are available in the Munising area. On the other
hand they are more likely to stay longer because they like skiing regardless of
the ancillary attributes available. McConnell, Strand, and Bockstael (1990)
discuss the long use of such "habit" variables as skill and previous experience
in recreation economic dem and studies and the ability of such variables to
increase the explanatory power of recreational dem and models. Past
behavior can be very helpful in explaining current behavior. Both skill level,
which usually increases over time, and previous overnight trips provide an
index of past behavior. Preferences are formed and shaped, and information
is gathered and stored during previous trips.
Spending in Munising was measured by asking the Forest Service and
Park Service respondents how much they and their party spent on grocery
food and beverages, restaurant food and beverages, vehicle related items,
lodging, ski equipment, clothing, and all other items on the day they received
the questionnaire. The amounts for each item listed above were totaled to use
in regression equations. Similar measures termed on site price, or on site
spending appear in the economic literature (Bell and Leeworthy 1990) and are
usually thought to reduce length of stay.

48
M ethods
Munising, Michigan was selected for the study because the local
Munising, Michigan offices of the U.S. Forest Service and the National Park
Service which contributed funds for the initial data collection, analysis, and
report were primarily interested in skiers visiting that area. Because of the
difficulty in finding a sampling frame containing a large num ber of cross­
country skiers who were familiar with cross-country skiing in Munising, it
was decided to contact people directly who were currently cross-country
skiing in the Munising area.
Because of budget and personnel limitations, stationing personnel in
Munising at the various trails for any length of time was impossible. The
Forest Service agreed that at two of the trails, the questionnaires would be
placed in boxes at the trail heads. At two other trails at which store owners
cooperated with the Forest Service in operating the trails, the store owners
would administer the questionnaires. At a fifth trail, concessionaires
operated a warming hut on weekends and adm inistered the questionnaire.
D uring the week, the questionnaire was m ade available in a box placed
outside the warming hut and near the trail head. At the trail managed by the
National Park Service, questionnaires were placed in a box near the trail head
with a sign asking skiers to participate in the study.

The questionnaires distributed on-site were primarily designed to
reflect the objectives of the U.S. Forest Service which was the primary source
of funds for the study. Questions were based on the need to gather
information to help the U.S. Forest Service make management decisions. The

49
design and content of the U.S. Forest Service questionnaire was developed in
consultation with the Forest Service and faculty and staff at the Travel,
Tourism, and Recreation Resource Center and the Department of Park and
Recreation Resources at Michigan State University. Staff at Pictured Rocks
National Lakeshore were consulted in designing the questionnaire
distributed there.
The questionnaires were not pretested because of time constraints. As a
further check on the validity on the ancillary attribute items used in the
questionnaire, a group of cross-country skiers who had inquired at the Upper
Peninsula Travel and Tourism Association office during the 1991-92 ski
season were sent a shortened version of the questionnaire which contained an
open format question in which respondents were asked to list the most
im portant considerations when deciding on a cross-country ski destination.
These responses were compared to the closed format list of ancillary items.
Figure 2 displays the close ended ancillary attributes analyzed in the study
and the open ended question format.

Questionnaires were distributed at the 5 Munising area U.S. Forest
Service cross-country ski trails and a modified version of the questionnaire
was distributed at the M unising Trail, operated by the National Park Service.
Quantities assigned to each Forest Service trail were based upon U.S. Forest
Service use estimates from the 1990-91 ski season. The trails were supplied in
such a manner as to have respondents represented from the Christmas holiday
until the end of the season w ith snow melt at the end of March. National Park
Service questionnaires were placed in early January. The goal was to recover

50
approximately 50% of the questionnaires taken by skiers. The return rate
realized was approximately 43%.
Approximately half of the questionnaires were distributed at the Valley
Spur trail because, based on U.S. Forest Service use estimates, it was by far the
busiest trail of those m anaged by the U.S. Forest Service. On weekends,
questionnaires were passed out to every fourth person who entered a
warming hut to purchase a cross-country ski permit. In addition, the
operators of the warming hut distributed questionnaires to those who showed
a great deal of interest in skiing and to those who collected most of the other
brochures available at the counter. During the week, when the warming hut
was closed, the questionnaires were placed in a wooden box labeled
registration which was attached to a sign board just in front and to the right of
the warm ing hut. A letter from the district ranger on the sign board invited
skiers to participate in the study. Late in the season, the distribution rate was
doubled to obtain more responses.
At the McKeever Hills and the Christmas cross-country ski trials,
convenience store operators distributed a questionnaire to every fourth
person who purchased a permit. The store operator at Mckeever Hills
distributed questionnaires to every fourth skier who entered his store
regardless of whether or not the skier purchased a permit. Late in the season
the distribution rate was doubled to obtain more responses.
At Hiawatha and Buckhorn cross-country ski trails, the questionnaires
were placed in a wooden box attached to a signboard. Posted to the sign
board was a letter from the district ranger inviting skiers to participate in the
study. These boxes were replenished regularly. In February, the U.S. Forest

51
Service placed a laminated copy of the first page of the questionnaire on the
outside of the boxes to further encourage responses.
The Pictured Rocks National Lakeshore began distributing a modified
version of the questionnaire in mid-January by using a box placed at the
Munising Trail similar to the one used by the Forest Service. A sign asking
people to participate was in clear view. Distribution of questionnaires
continued until the end of the cross-country ski season at the end of March.
Although the National Park Service trail received more use than any of the
U.S. Forest Service trails, the Park Service distributed fewer questionnaires
because of lower funding levels.
All respondents were asked to either mail the questionnaire to the
Travel, Tourism, and Recreation Resource Center at Michigan State
University in a postage-paid envelope which was attached to the last page of
the questionnaire or to drop the questionnaire off at the joint National Park
Service-U.S. Forest Service visitor center in Munising.

To assess the completeness of the close-ended ancillary attribute
response scales, a group of cross-country skiers interested in skiing in the
Upper Peninsula of Michigan was contacted by mail. Respondents to the
mail questionnaire were asked what four considerations come to mind when
selecting a cross-country ski destination (Figure 2). Respondents were then
asked to list these considerations in order of importance. Respondents were
instructed to answer these two questions along with the rest of the questions
appearing on the front of the questionnaire before answering the questions on

52
the back of the single page. The back of the questionnaire included the
closed-format question. Respondents were instructed not to return to the first
page after turning to the second and that most people would probably leave
some answer spaces blank. Questionnaires distributed at Munising area trails
did not have an open-format question.
As stated in the first hypothesis, mid price lodging, family restaurants,
and quality of cross-country ski trails were expected to be rated as higher in
im portant than an average of all 22 importance ratings. For each respondent,
a mean importance rating was calculated from the ratings of the

2 2

importance items. This new variable was then compared with each of the 22
separate importance rating items in 22 paired t-tests. Manfredo (1989) and
Saleh and Ryan (1992) followed similar procedures. The paired t-test
assumed that the

2

measures were not independent and were related in that

the same individual had given both scores. The paired t-test summed the
individual differences between the two variables in question and determined
if the summ ed differences were significantly different from zero (Mendenhall
1979). If the variable was significantly different from zero and the mean
difference had a positive sign, the variable was classified as significantly more
im portant than average in making destination decisions. As stated in the first
hypothesis, ancillary attributes with which Munising was well supplied were
expected to be rated more highly than variables with which Munising was not
well supplied. Specifically, family restaurants, mid price lodging, and
quality of cross-country ski trails were expected to be rated more highly than
the average variable.

53

For the second hypothesis, a two-way analysis of variance procedure
was used to compare ancillary attribute importance ratings between local and
tourist respondents and between U.S. Forest Service, and National Park
Service respondents. It was hypothesized that the groups would exhibit some
differences in their importance ratings of the ancillary attributes.

As stated in the third hypothesis, the importance ratings of ancillary
attributes were expected to make a statistically significant contribution to
explaining variations in length of stay. Distance, perceived constraints,
overall community satisfaction, trail quality satisfaction, num ber of overnight
cross-country skiing trips during the past season, skill level, party spending
in Munising during the day the questionnaire was received, and trail quality
importance ratings were included in a base model of independent variables.
All of these variables have been mentioned to varying degrees in the literature
as having an influence on the decision to participate in an activity, the
selection of a destination, or the length of stay at a destination. Distance has
been used as an approximation of travel cost, travel time, or both in dem and
studies or as a major component in cost or price estimation. Distance rather
than travel cost or travel time is especially useful where nearly all
participants use the same mode of transportation to a common destination.
Distance and travel cost and travel time are usually highly intercorrelated.
Where valuation of the recreation resource is needed, distance can readily by
transformed into monetary figures by using a conversion factor which reflects

54
the average cost per mile of travel to the destination. Distance can also
usually be converted into time cost.
To measure the significance of ancillary attribute importance ratings, a
regression model was developed which was similar to those that have been
used with travel cost analysis. Such models can be used to either forecast or
predict dem and or to examine relationships among variables (Witt and Witt
1992). The dependent variable was length of stay during the trip which
respondents picked up a questionnaire. Those variables with a significant
standardized regression coefficient w ould have a significant effect in
determining the amount of time a cross-country skier spends in Munising
(Witt and Witt 1992). The size of the standardized regression coefficient
indicates the relative importance of a variable in the equation. The larger the
coefficient, the stronger association of the independent variable with the
dependent variable.
This analysis was performed only for tourist skiers. As noted by Bell
and Leeworthy (1990), there is evidence that local and tourist visitors face
much different decisions and probably should be kept separate in analyses.
Length of stay for local skiers could be measured and analyzed using the
num ber of hours during which they skied during the trip. This variable is
available from the data set being used for this study. However, the result
would not be comparable to length of stay measured as days at Munising or
days away from home for tourist cross-country skiers. Ancillary attribute
importance variables would be expected to have less of an effect on local
cross-country skier length of stay than for tourist cross-country skier length
of stay as local skiers would have access to similar services and facilities in
their own homes and would not necessarily rely on the community's

55
restaurants to acquire meals, lodging establishments for sleeping, or bars for
refreshments. The literature has indicated that including both local and
nonlocal skiers in the same length of stay regression can confuse the results.
Uysal, McDonald, and OTeary (1988) found that skiers who travel shorter
distances stay longer. Logic would indicate that skiers who travel farther
should stay longer to recoup their larger monetary and travel time
investment. Uysal, McDonald, and O'leary (1988) explain this problem in
their study by acknowledging that data were included from both resident and
tourist respondents. Bell and Leeworthy (1990) note tourists and local
respondents are different and should be treated separately.
The regression model used in this study (below) is a linear multivariate
model based on the travel cost model. Statistics for the model were
calculated with SYSTAT for the Macintosh. The use of this computer
program is described in its user's manual (Wilkinson 1989).
The basic parts of the travel cost model as discussed by Fletcher,
Adamowicz, and Graham-Tomasi (1990) consist of the variation in distance
and therefore costs experienced by travelers from various distances to a
recreation site. As distance increases, so does cost, thus reducing the number
of visits. The travel costs are analogous to a price for access. The higher the
price the fewer visits which occur at that price. The relationship between
price and visits allow the formulation of a dem and curve. Researchers using
travel cost models to analyze dem and usually assume equal lengths of stay
and chose data which meet this assumption. The num ber of visits is
constrained by the desire for other goods and visits to other recreation sites,
money budget constraint and time budget constraint, and other restrictions
(Fletcher, Adamowicz, and Graham-Tomasi 1990).

56
Travel cost models usually include income and often a measure of the
cost of time. Neither were available from the data set used for this analysis.
Some measure of the availability or use of substitute sites is also often
included in travel cost models, but such a variable was also unavailable from
the data set used for this analysis. The am ount of one-day spending in
Munising was included in the model because most travel cost studies include
an on-site price variable which usually has a negative effect on length of stay
or number of trips (depending on what is being examined). The higher the
cost the fewer days or trips customers should be able to afford. Spending in
Munising should to some degree reflect income constraints and also the
availability of substitute destinations. Higher costs usually reduce the
quantity dem anded. Spending should be lower for those with lower incomes
and lower for those who could travel to substitute sites with lower costs.
Spending may be greater for those respondents with greater time values.
Such respondents may be able to afford higher on-site costs and be more
willing to undergo them to ensure a return on their more valuable investment
of time.
Satisfaction was included in the model because of the potential for
dissatisfied visitors to shorten their stay and for highly satisfied customers to
perhaps lengthen their stay. Satisfaction with services is highly dependent on
the manner in which service is provided and not just whether or not services
exist (Teare 1990; Arnould, Price, Tierney 1993). The presence of ancillary
attributes may be related to initiation of the trip with satisfaction occurring
with the consumption of destination attributes. Satisfaction may have a
strong influence on trip duration. Two measures of satisfaction were
included in the regression model, overall community satisfaction and trail

57
satisfaction. It was anticipated that trail satisfaction, being more directly tied
to the main trip activity of skiing, would have a greater association with
length of stay. Overall community satisfaction was included to provide a
performance measure of the ancillary attributes. Ideally satisfaction with each
ancillary attribute included in the importance measures would exist, but
these were condensed into a few broad general satisfaction measures to
reduce the length of the questionnaire.
Constraints may have the effect of shortening length of stay for highly
satisfied people who perhaps would like to stay an extra day or two but
cannot because of obligations at home or work or a lack of financial resources.
Constraints may also work to lengthen visit duration for those visitors who
are highly dissatisfied. Dissatisfied customers may not be able to switch
destinations easily once they have arrived because of time shortage or lack of
knowledge about other potential destinations. Returning home may also not
be an option for dissatisfied customers because that would amount to the
wastage of what for some may be a substantial sunk investment in the trip.
This may be especially true for those cross-country skiers who, because of
other commitments, have a limited num ber of days during the cross-country
ski season during which they can ski and even fewer days during which they
can arrange a trip lasting several days.
Skill level and previous overnight trips represent personal variables
which may shift the num ber of visits and days for any particular distance or
travel cost. Skill level and previous overnight trips can be particularly
valuable as they reflect a degree of experience and learning about the specific
recreation activities to which they are related. In addition, "Consumers with
extensive prior experience develop . . . clearly defined expectations and

58
selection criteria" (Teare 1990:244). These expectations and selection criteria
may simplify into a search for a particular brand or any one of a group of
brands that have proved satisfactory in the past. A visitor to Munising may
book lodging at the Comfort Inn, as opposed to the Day's Inn or Best Western
because based on other stays at Comfort Inn motels, the customer knows that
this brand performs well on most of the visitor s selection criteria.
Trail quality importance was included to serve as a means of
comparison for ancillary attribute importance ratings. Trail quality
importance would be expected to be more strongly associated with length of
stay than any of the ancillary attributes as it is more closely connected to the
overall goal of the tourists' trips, cross-country skiing.
Although psychometric measures are rarely included in travel cost
models, Fletcher, Adamowicz, and Graham-Tomasi (1990) support their
inclusion. Constraints, ancillary attribute importance, and satisfaction were
all measured by asking respondents to indicate how they felt on a one to
seven scale. The resulting equation then can be stated as follows:
Y= A

+ B iX i + B2 X2 + B3 X3 + B4 X4 + B5 X5 + B6 X6 +B 7 X7 + BgXg + B9 X9

Y= Num ber of days within 15 miles of Munising on this trip
A=

Constant

Xj= Distance from respondents residence in miles as reported by respondents
X2 = Perceived Constraints (Developed from the addition of responses to
items in question

1 0

)

X3 = Satisfaction with cross-country ski trail quality
X4 = Overall Community satisfaction
X5 = Num ber of overnight cross-country ski trips during last season
X6=

Skill level.

59
X7 = On site cost for one day per party
X8 = Importance of trail quality
X9 =Ancillary attribute importance ratings (21 items were entered into a
stepwise regression analysis as separate variables)

Because of limitations caused by a relatively small num ber of
respondents for this type of analysis, not all

21

ancillary attribute importance

variables could be included in the model simultaneously. A separate model
was calculated for a base model of key variables which included the first
eight variables listed above. Then that model was used in a stepwise
regression analysis with the addition of the ancillary importance ratings. The
only difference among the models was the ancillary attribute importance
variables used. The other variables remained the same for each regression
equation.
Respondents were also asked to state the maximum am ount they would
pay to keep using the cross-country ski trail where they received the
questionnaire, w ith the understanding that the trail may be closed if the
am ount they were willing to pay was not high enough. For the purposes of
this study, the am ount skiers were willing to pay to keep using the trail was
considered another measure of dem and and was substituted as the dependent
variable in a regression of the base model and a stepwise regression including
all of the ancillary attribute importance ratings. This measure was expected
to provide some confirmation of the results achieved through the regression
analyses with length of stay as the dependent variable.
Principle component analysis was also considered as a means of
incorporating the ancillary attribute importance ratings into the analysis.

60
However, the variables did not appear to be strongly intercorrelated and the
resulting factors did not account for a large portion of the variance.

Lim itations
The usefulness of this study is constrained because of the methods of
data collection. Data were collected at cross-country ski trails in the
Munising, Michigan area or by mail from people who had inquired at the
Upper Peninsula Travel and Recreation Association about cross-country
skiing opportunities in the Upper Peninsula of Michigan.
Many of the respondents were self-selected by their willingness to take
a questionnaire from a box placed near a trail head. This was true for the trail
operated by the National Park Service and at the U.S. Forest Service Valley
Spur trail when the warm ing hut was closed. Questionnaires were also
distributed at the U.S. Forest Service lesser used trails by the box method.
The hut at the Valley Spur Trail was usually open on weekends when most
use occurred. When the hut was open, every fourth skier who entered the
warm ing hut received a questionnaire. Because of these distribution
methods, the sample was not randomly drawn. Skiers responding to the
questionnaire are probably more interested in cross-country skiing, the
Munising area, or both than are cross-country skiers in general. Also, the
probability of any one particular skier participating in the study increased
the longer that one particular skier stayed in the Munising area, resulting in
length of stay bias.
Because all of the U.S. Forest Service trails were located within 20 miles
of the National Park Service trail, some cross-country skiers likely visited
trails managed by both agencies. However, they were more likely to obtain a

61
questionnaire at those trails they visited more often. Cross-country skiers
were classified as a U.S. Forest Service respondent or a National Park Service
respondent based upon where they obtained their questionnaire. There were
likely some skiers who preferred or usually skied at trails operated by the
other agency than the one for which they obtained a questionnaire. There may
also have been some skiers who sent in more than one questionnaire or sent in
questionnaires for each agency.
Because most of the respondents were contacted at Munising,
Michigan, only limited inferences can be draw n about skiers who choose
other destinations. However, the mail questionnaire respondents to some
degree make up for this problem in spite of the fact that the mail respondents
were already interested in the Upper Peninsula and so may be more similar to
Munising cross-country skiers than different from Munising cross-country
skiers. Some mail respondents were familiar with Munising or had visited
Munising during the course of the cross-country ski season.
Data collection began during the last week in December at U.S. Forest
Service trails and in mid January at the National Park Service trail, missing
early season skiers, and continued until to the end of March. Although there
is no evidence that early season skiers differ from mid to late season skiers,
that possibility remains unexplored.
In the analyses examining the contribution of ancillary attribute
importance ratings to explaining variance in length of stay, two variables
often considered critical for such analyses where not available, income and
the opportunity cost of travel time. However, some authors discuss the
possibility that such omissions may not be quite as serious in some situations
where there is little variation in respondents' incomes and most visitors use

62
the same mode of transportation. Despite the above limitations, the analysis
provides some very interesting results, and some avenues for additional
investigations which could prove fruitful.

CHAPTER IV: RESULTS
This chapter presents the results of the data analyses described in the
previous chapter. After presenting responses to the open-ended question, the
discussion is organized in the order the hypotheses were presented in chapter
III. The results show that ancillary attributes vary in their importance to
cross-country skiers, with those ancillary attributes strongly represented in
Munising and serving basic needs, or connected to the main purpose of the
trip (cross-country skiing) being most important. Local and tourist cross­
country skiers rated the importance of several ancillary attributes differently.
In addition, differences also occurred in the way U.S. Forest Service, and
National Park Service respondents rated the importance of some ancillary
attributes. Some ancillary attribute importance ratings were significantly
related to length of stay in a regression model.
In general, respondents were middle-aged and well-educated with
above average incomes (Bishop, Forsberg, and Holecek 1992). Only 34% of
tourist respondents were first time visitors. The mean U.S. Forest Service
tourist respondent traveled 307.9 miles away from home to come to Munising,
had 16.4 years of education, and was 41.9 years old. Nearly all U.S. Forest
Service tourist respondents were traveling with family or friends. However,
11.5% were traveling by themselves. The mean National Park Service tourist
respondent traveled 386.6 miles away from home, had 15.9 years of education
and was 40.5 years old. Most were traveling with family or friends.

63

64
However, 9.7% of the National Park Service tourist respondents were
traveling by themselves.

Responses to Open Format Questions
Only respondents to the mail questionnaire are included in this section
as questionnaires distributed in the Munising area did not have an open
format question. There was a substantial amount of overlap in the placement
of some considerations in the order of importance (Table 1). One respondent
would place a consideration as being most important while another
respondent would place the same consideration as being least important.
Respondents listed a wide variety of considerations, with several responses
being unique. Even though a consideration may have been listed as third or
fourth in importance, each consideration should still be considered important
because it came to mind with relatively little prompting and was important
enough to be written dow n by the respondent. Several considerations appear
in each section of the list because they were ordered differently by different
people.
The responses in Table 1, in order to give the reader an understanding
of the array of responses, are only slightly condensed versions of the actual
answers written on the questionnaire. These answers have been further
condensed into categories with response frequencies in Table 2. The number
of responses recorded for each category can be used to infer those attributes
which were im portant to the most respondents.
For the most part, the free format responses mirrored the closed format
list with the exception of trail quality. The close-ended list simply asked for
the importance of trail quality in general. Open-ended responses described

Table 1. Responses to open-ended question by mail questionnaire respondents^

Trail quality
Snow
Trail difficulty
Quality and diversity of
trails
Well groomed trails
Hills
Good skiing
Difficulty
Trails
Lighted area for night
skiing
Variety
Facilities
Pretty
Isolation
Terrain
Wilderness
Lodging
Reasonable prices
Rated
cottages
Distance from home
Country-side or area
Scenery
No snowmobiles
Location
Cost
Easy access
How well organized (this
respondent appeared to be thinking
in terms of a group tour)
Vacation abilities
Peace and Quiet

Trail quality
Groomed
Trail Maintenance
Snow
Facilities
Wooded
Hilly
Reasonable use fees
Length
Accessibility
Numbers of
Usage of
Different skill levels
Marked trails
Difficulty
Beauty
Variety
Lodging
Reasonable prices
A great B&B
Close to trails
Skiing out the door
Condominium style
Pool
Whirlpool
Moderate priced
Restaurants
Reasonable prices
Quaint
No crowds
Distance from home
Cost
Near large city
Natural beauty
Equipment rental
Activities for children
Local attractions

Trail quality
Difficulty
Miles of trails
Distance from cottage
Accessibility
Groomed
Low traffic
Scenery variety
Facilities nearby
Snow
Challenging terrain
Striding only __ no
skating
Total miles
Number of km
Variety in hills, scenery
Lay out
Average snow fall
Heavy snow
Lodging
Cabins
Lodge
Good
Unusual accommodations
Cost
Restaurants
Good
Great
Distance from home
Cost
Physical Characteristics
Scenic
Shopping
Entertainment
Free-way Access
Not crowded
Time Available

Trail Quality
Difficulty
Facilities
Skiing challenge
Access with dogs
Groomed
Rest areas
Good snow base
Trail fee
Low number of skiers
Lodging
Sauna
Good
Motels
Comfortable
Restaurants
Good food
Family
Rental equipment
Community
Other things to do
No crowds
No snowmobiles
Crowded
Distance from home
Surroundings
Reasonable Prices
Scenery
Community activities
Atmosphere
Accommodations for kids on
trails and in lodging
Cost
Weather
Accessibility

66

Table 2
Frequency destination considerations m entioned, mail questionnaire

O rder listed
A ttribute
category

First

Second

Third

Fourth

W eighted
total 3
Total

Snow

18b

04

04

02

9.4

28

Trails

18

32

21

08

2 1 .8

79

Scenic beauty

06

0 2

02

05

3.9

15

Distance from
home

15

05

06

02

8.9

28

Crow ding

03

0 2

02

04

2 .6

11

Lodging

06

13

14

06

9.7

39

Snow-mobiles

01

0 0

00

01

0.5

0 2

Cost

01

04

01

03

2 .1

09

Restaurants

00

01

09

03

2.4

13

Access

01

01

02

02

1.3

06

Shopping

00

0 0

01

00

0 .2

01

O ther

03

05

05

07

4.4

2 0

Total

72

69

67

43

67.2

251

aWeighted total, num ber of response multiplied by 4 for first, 3 for second, 2
for third, and 1 for fourth, added together then divided by 10. bNum ber of
mail respondents who mentioned attribute category as a reason for choosing a
cross-country ski destination.

67
trail quality in detail including snow, terrain, and facilities. Trail quality was
included in the close-ended format primarily as a means of comparison for
the other items. Other studies have extensively examined aspects of trail
quality (Ballman 1980; Rosenthal and Driver 1980; McLaughlin and Paradice
1980; Rauhauser 1979; Nelson 1988).

High Ancillary Attribute Importance Ratings
In the first hypothesis, importance ratings were expected to be higher
for those ancillary attributes which directly affected basic hum an needs,
directly affected the main purpose of the trip (cross-country skiing), and were
available in Munising at the time of data collection.

Mid price lodging,

family restaurants, and quality of cross-country ski trails were expected to
receive importance ratings higher than an average of all ancillary attribute
ratings. The remaining ancillary attributes were either not well represented in
Munising or were not as closely connected to basic sustenance needs or to the
cross-country ski experience.

An overall average attribute rating was calculated for each respondent
by adding the rating given each attribute and then dividing the sum by

2 2

(the

num ber of attributes). The average rating was compared with the importance
rating for each attribute using a paired t-test. For the paired t-test, the average
attribute rating was subtracted from the particular attribute rating,
respondent by respondent. These results were summed across all
respondents. A t statistic and a probability value were calculated based on
the sum of the above differences, the number of respondents, and the

68

variability in their responses. The probability value indicated the likelihood
that the resulting sum was so small that essentially there was no difference in
the two measures, the average attribute rating and the particular attribute
rating (Wilkinson 1989). A small value of p (0.05 or less) indicates that the
probability of the attribute rating being average is quite small and that the
ancillary attribute should be considered either to have been rated as higher or
lower in importance than the average ancillary attribute rating. Since, the
interest is in those ancillary attributes which were rated more highly, the twotailed p value calculated with the use of SYSTAT for Macintosh was divided
by two for the probability of a one-tailed test, or the probability that a value
on the high side of average is average. The probability that an ancillary
attribute rating is greater than average would be given by

l-p /2 .

For this analysis, a probability level of 0.05 (for a one-tailed test) was
chosen as the cut off point for a value being significantly different from 0. A
probability level of 0.05 indicates that there is a 5% or less probability that a
difference between the average of all attribute ratings and one particular
ancillary attribute rating would be considered to be greater than

0

when it is

not. The hypothesis for this analysis is only concerned with those attributes
which are more important than average. Table 3 reports the t statistic and its
relationship to the cutoff point for each ancillary attribute. The reader is
urged to interpret all p values in this study with caution. Such values assume
that the data upon which they are based were draw n from a randomly
selected sample of respondents, which was not the case with this study.
Low price motel, mid price motel, basic kitchen, sauna, family
restaurant, trail quality, and overall community can be considered more
im portant than the average attribute (Table 3). These results give some

69
Table 3

Ancillary Attribute

M

SD

M Pa

SDDb

Low price motel

3.92

2.31

0.78

2.32

5.69*

Mid price motel

4.30

2 .0 1

1.15

1.72

11.35*

Ski lodge

2.56

1 .8 8

-0.67

1.54

-7.36

Basic kitchen

3.34

2.17

0 .2 0

1.95

1.70*

Child care

1.62

1.43

-1.50

1.33

-19.16

Swimming pool

2.94

2.05

-0.18

1.71

-1.77

Sauna

3.52

2.30

0.36

1 .8 8

Waxing room

1.97

1.53

-1.16

1.36

-14.48

L aundry

2.06

1.74

-1.07

1.41

-12.90

Fast food

2.81

1.92

-0.32

1.71

-3.20

Family restaurant

4.90

1.81

1.72

1.59

18.43*

G ourm et
restaurant

2.58

1.92

-0.56

1.61

-5.86

Night Club

2.04

1.73

-1.13

1.43

-13.44

Family
entertainm ent

2.75

2.04

-0.40

1.64

-4.16

Cultural attraction

3.08

2 .1 1

-0.13

1.60

-1.38

Bar

2.74

2 .0 1

-0.43

1.82

-4.00

Gift shops

2.18

1.63

-0.97

1.28

-12.82

Clothing shops

2.27

1.62
-0.87
1.21
(continued on next page)

-12.28

t

3.20*

70
Table 3 (Continued)

Ancillary Attribute

M

SD

M Pa

SDP.b

Ski Shops

3.28

1.85

0.12

1.49

1.40

Art and craft shops

3.15

1.96

-0.04

1.54

-0.40

Quality of ski trails

6.65

0.82

3.48

1.09

54.39*

Overall
com m unity

4.80

1.66

1.62

1.39

19.68*

t

N ote. Mean importance ratings ( l=not im portant and 7= extremely
important) and paired t tests on differences between individual
attribute importance ratings and the average importance rating of all
22 items for all respondents (N=288). aMD = mean difference.
bSDD=standard deviation of the differences. *p <. .05, positive one­
tailed.

71
support to the first hypothesis with the exception of low price motel, basic
kitchen, sauna and overall community being of greater importance than
average. The relatively high ratings of these attributes were not predicted in
the first hypothesis. However, it is possible that some respondents may have
considered Munising m id price lodging to be low price. Basic kitchen is
related to basic sustenance needs, perhaps more so than restaurants. For some
respondents, a sauna may be an integral part of the cross country ski
experience. Overall community may be an expression of a synthesis of basic
needs that must be met by the destination community and may include such
issues as safety and cleanliness. The relatively high importance ratings given
to mid price motel, family restaurant, and trail quality support the first
hypothesis.

These respondents are cross-country skiers who received a
questionnaire at either the Munising Trail, managed by Pictured Rocks
National Lakeshore or at one of the trails managed by the Munising district of
the U.S. Forest Service and who indicated that they did not spend or intend to
spend any nights away from home while on this particular cross-country
skiing trip. Most local respondents live either in Munising, Marquette,
Escanaba, or one of the other towns or villages within 40 to 50 miles of
Munising. Local residents are likely to have different needs than skiers who
stay overnight during their holiday. These needs may be reflected in the
importance they ascribe to various ancillary attributes. However, the
question was stated in such a way as to apply to selecting a destination in
general. Some local cross-country skiers who did not stay overnight on the

72
particular trip for which they filled out the questionnaire may stay overnight
on other trips at other destinations. Their importance ratings may reflect this
behavior. Such ratings would be of interest to managers of establishments in
the Munising area assuming that these individuals represent potential
customers either from Munising or surrounding communities.
Information about this group of skiers can still be valuable in that it
suggests what cross-country skiers from other towns and villages in the
Upper Peninsula may be looking for when deciding upon a cross-country
skiing destination community. Many of the local skiers were in fact from
surrounding communities and not the immediate Munising area and
contributed to the local Munising economy during their visit. Cross-country
skiers who resided in Munising were likely to be similar to other Upper
Peninsula residents.
Mid price motel, sauna, family restaurant, quality of cross-country ski
trails, and overall community were statistically rated higher in importance
than average (Table 4) by local cross-country skiers. These results give some
support to the first hypothesis with the exception of sauna and overall
community being of greater importance than average. The relatively high
ratings of these ancillary attributes were not predicted in the first hypothesis.
The relatively high rating of sauna may reflect the Scandinavian heritage of
many of the people who live in the Upper Peninsula of Michigan. The
relatively high rating of overall community may emphasize the critical
importance of packages of (rather than solitary) ancillary attributes to cross­
country skiers. The relatively high ratings of mid price motel, family
restaurant, and quality of cross-country ski trails are in keeping with the
predictions in the first hypothesis. The lodging importance ratings may

73
Table 4

Ancillary Attribute

M

SD

M Pa

SDDb

Low price motel

3.31

2.34

0.04

2.14

0.18

Mid price motel

3.89

2.15

0.64

1.76

3.47*

Ski lodge

2.71

2.05

-0.60

1.58

-3.66

Basic kitchen

3.16

2.15

-0.09

1.63

-0.54

Child care

1.99

1.79

-1.24

1.63

-7.31

Swimming pool

3.16

2.16

-0 . 0 2

1.71

-0.08

Sauna

4.03

2.45

0.84

1.78

Waxing room

2 .0 2

1.62

-1 . 2 1

1.32

-8.78

L aundry

2.48

2.08

-0.79

1.57

-4.80

Fast food

3.06

2.04

-0.19

1.63

- 1 .1 1

Family restaurant

4.62

1.92

1.30

1.47

G ourm et
restaurant

2.74

1.97

-0.47

1.54

-2.93

Night club

2.28

1.93

-0.99

1.57

-6.08

Family
entertainm ent

3.22

2.13

-0.04

1.62

- 0 .2 2

Cultural attraction

3.47

2.29

0.16

1.59

0.96

Bar

2.83

2.16

-0.45

1.97

-2.18

Gift shops

2.30

1.71

-0.93

1.34

-6.62

Clothing shops

2.45

-0.78
1.30
1.74
(continued on next page)

-5.73

t

4.54*

8.48*

74
Table 4 (Continued)

Ancillary Attribute

M

sn

M Pa

SDD>

Ski Shops

3.28

1.92

0.04

1.47

0.26

Art and craft shops

3.20

2.06

-0.05

1.54

-0.30

Quality ski trails

6.60

0.98

3.34

1 .2 1

4.73

1.89

1.47

1.59

t

26.54*

Overall community
8.96*

N ote. Mean importance ratings ( l=not im portant and 7= extremely
important) and paired t tests on differences between individual
attribute importance ratings and the average importance rating of all
22 items for local Munising respondents (N=92). aMD = mean
difference. t>SDD = standard deviation of the differences. *p <, .05,
positive one-tailed.

75
indicate that many local cross-country skiers answered this question for
overnight trips they may have already taken or may take in the future.

These respondents received a questionnaire at one of the Munising area
trails managed by either the National Park Service or the U.S. Forest Service
and stayed at least one night away from home while on their cross-country ski
trip. Some of these skiers came from locations (Escanaba and M arquette for
example) which also supplied local skiers (those that did not stay overnight
away from home) to the Munising trails. The majority of tourist cross-country
skiers came from lower Michigan and Wisconsin, with a handful from Illinois,
Indiana, and Ohio, and a very few from other states such as Tennessee, Texas,
and Virginia.
Low price motel, mid price motel, basic kitchen, family restaurant, ski
shops, quality of cross-country ski trails, and overall community were rated
statistically higher in importance than the average importance rating of the
attributes used in this study (Table 5).

2 2

These results give some support to the

first hypothesis with the exception of low price motel, basic kitchen, ski shops
and overall community being of greater importance than average. The
relatively high importance ratings of these attribute were not predicted in the
first hypothesis. Some respondents may have considered Munising mid price
lodging to be low price lodging, or some of the respondents may have been
willing to put up with less than what they would have considered to be ideal
lodging to take advantage of good cross-country skiing conditions.
Kitchenettes were available in some lodging establishments in the Munising
area. The obvious relationship of kitchenetts to food supports the intent of

76
Table 5

Ancillary A ttribute

M

SJQ

M Pa

SDDb

Low price motel

4.48

2 .2 1

1.42

2.32

6.81*

Mid price motel

4.41

1.99

1.34

1.74

8.57*

Ski lodge

2.24

1 .6 6

-0.94

1.41

Basic kitchen

3.47

2.23

0.40

2.16

Child care

1.46

1.18

-1.59

1 .1 0

-16.16

Swimming pool

2.60

1.97

-0.43

1.70

-2.80

Sauna

3.34

2.25

0.18

1.92

1.03

Waxing room

1.96

1.50

-1 . 1 1

1.32

-9.30

Laundry

1.91

1.57

-1.14

1.31

-9.66

Fast food

2.61

1.79

-0.42

1.73

-2.71

Family restaurant

5.16

1.71

2 .1 2

1.55

15.19*

G ourm et
restaurant

2.52

2 .0 0

-0.61

1.61

-4.24

Night Club

1 .8 6

1.57

-1.26

1.32

-1 0 . 6 6

Family
entertainm ent

2.40

1 .8 6

-0 . 6 6

1.56

-4.72

Cultural attraction

2.77

2 .0 1

-0.35

1.52

-2.55

Bar

2.71

2 .0 2

-0.39

1.83

-2.36

Gift shops

1.95

1.49

-1.14

1 .2 1

Clothing shops

2.05

1.55
-1.06
1 .1 0
(continued on next page)

t

-7.41
2.07*

-10.50
-10.76

77
Table 5 (continued)

Ancillary Attribute

M

SD

M Pa

SDDb

Ski Shops

3.51

1.87

0.34

1.54

Art and craft shops

3.06

1.98

-0.04

1.56

-0.29

Quality of trails

6.67

0.78

3.58

1.05

37.83*

Overall community

4.84

1.48

1.76

1.26

15.48*

t
2.44*

N ote. Mean importance ratings ( l=not important and 7= extremely
important) and paired t tests on differences between
individual attribute importance ratings and the average importance
rating of all 22 items for U.S. Forest Service and National Park Service
tourist respondents (N=124). aMD = mean difference.
t>SDD=standard deviation of the differences. *p <. .05, positive one­
tailed.

78
the first hypothesis. The relatively high rating of ski shop is different from the
way local skiers responded to this item. This perhaps reflects not so much a
desire to purchase a pair of skis, but the need to have available a local source
of cross-country ski information. Tourists may also feel that a ski shop is a
method of providing trip insurance. Should a cross-country skier need to
repair or replace equipment or other paraphernalia, the skier would be able
to do so without having to travel a great distance. The relatively high rating
of overall community may emphasize the critical importance of packages of
(rather than solitary) ancillary attributes to cross-country skiers.

Differences Among Cross-country Skier Groups

Data analyzed in this study potentially represent four groups of cross­
country skiers: tourist cross-country skiers, local cross-country skiers, U.S.
Forest Service cross-country skiers, and National Park Service cross-country
skiers. Because these skiers were contacted at different places (five U.S. Forest
Service trails and one National Park Service trail) using somewhat different
questionnaires, the ancillary attribute importance ratings may vary among the
four groups, as indicated in the second hypothesis. The literature describing
the recreation opportunity spectrum would suggest that the differences in
importance ratings may be associated with desires for different recreational
experiences among the three groups. While the recreation opportunity
spectrum concept has not been extended to the service / accommodation
industry from its original wild land recreation application, it is logical to
suppose that groups taking advantage of different recreation opportunities
may also desire different services when not on the cross-country ski trail.

79
Tourist cross-country skiers could also be expected to have different
importance ratings than local respondents. Tourists who are away from home
m ust rely more fully on services found in destination communities for
sustenance than local cross-country skiers who can more easily either return
home or bring food and drink with them.
Importance ratings of the 22 ancillary attributes used in this study were
compared across U.S. Forest Service, National Park Service, tourist, and local
respondents using the two-way analysis of variance procedure in the SYSTAT
for Macintosh statistical package (SYSTAT 1992). This analysis compared
tourist and local respondents while statistically holding agency constant and
also compared respondents contacted at trails managed by the two agencies
while holding constant whether or not they were tourist or local cross-country
skiers. Tourists rated low price motel and family restaurant higher in
importance (p ^ 0.05) while local respondents rated ski lodge, child care,
swimming pool, sauna, laundry, night club, family entertainment, and
cultural attraction higher in importance ( p <. 0.05, Table 6 ). The higher
importance ratings by local cross-country skiers may have resulted from
consideration of w hat they would like to see more of in Munising, what they
think would be im portant to tourist cross-country skiers, or what they would
look for themselves when traveling to other communities for cross-country
skiing. Tourist cross-country skiers rated higher in importance than local
cross-country skiers, those attributes (low price motel and family restaurant)
which were of key importance to their current trip. Although low price
lodging was lacking in Munising, there was a good supply of mid price
lodging which some respondents may have considered to be low price

80
Table 6

.groups
Ancillary
A ttribute

TMa

L Mb

Fc

FMd

PMe

Low price motel

4.56

3.84

15.20*

3.82

4.12

0.75

Mid price motel

4.36

3.83

3.58

4.20

3.99

0.48

Ski lodge

2 .1 1

2.59

3.89*

2.60

2 .1 0

3.24

Basic Kitchen

3.50

3.19

1.14

3.29

3.40

0 .1 2

Child Care

1.44

1.98

7.41*

1.74

1 .6 8

0.06

Swimming pool

2.58

3.14

4.21*

2.91

2.81

0 .1 1

Sauna

3.20

3.90

5.11*

3.82

3.27

2.39

Waxing room

2 .0 0

2.03

0 .1 0

2 .0 0

2 .0 2

0.04

Laundry

1 .8 6

2.43

5.57*

2.24

2.05

0.45

Fast food

2.54

2.97

3.26

2.93

2.55

1 .6 8

Family restaurant

5.10

4.54

4.86*

4.97

4.63

1.52

G ourm et
restaurant

2.38

2.61

0.72

2.76

2.23

3.13

Night club

1 .6 8

2 .1 1

3.50

2.24

1.55

6.80*

Family
entertainm ent

2.32

3.14

9.79*

2.90

2.56

1.24

Cultural attraction

2 .6 6

3.56

6

.1 1 *

2.78

1.94

3.23

Bar

2.43

2.56

0 .2 2

3.04

1.94

12.74*

Gift shops

1.87

2.23

2.78

2.19

1.92

1.27

Clothing shops

2 .0 0

2.40

3.39

2.30

2 .1 1

0.54

(continued on next page)

Ff

81

Table 6 (Continued)

A ncillary
A ttribute

TMa

LM b

Fc

FMd

PMe

Ff

Ski shops

3.51

3.28

0.85

3.39

3.41

0 .0 0

Art and craft shops

3.00

3.15

0.30

3.18

2.97

0.46

Quality of trails

6.54

6.49

0 .2 2

6.76

6.26

16.34*

O verall
com m unity

4.77

4.66

0 .2 2

4.85

4.58

1.16

N ote. Two-way analysis of variance comparing ancillary attribute
importance ratings between tourist and local respondents and between U.S.
Forest Service and National Park Service Respondents. aTM is the adjusted
least square mean for tourist respondents. bLM is the adjusted least square
mean for local respondents. CF is the F statistic for the comparison of tourist
and local adjusted least square means. dpM is the adjusted least square mean
for U.S. Forest Service respondents. ePM js the adjusted least square mean for
National Park Service respondents. fp is the F statistic for the comparison of
U.S. Forest Service and National Park Service adjusted least square means.
‘Difference between groups significant at p <. 0.05.

82
lodging. Other respondents may have decided to put up with less than ideal
lodging to take advantage of good skiing conditions in the Munising area.
U.S. Forest Service respondents rated night club, bar and quality of
cross-country ski trails higher in importance than National Park Service
respondents (p s 0.05, Table 6 ). U.S. Forest Service skiers may have been more
socially motivated than National Park Service skiers. However, there are no
measures available from the data set to adequately explore this conjecture.
Higher importance ratings for trail quality may be a reflection of the $3.00 fee
charged at U.S. Forest Service trails. The fee requirement may have increased
trail quality expectations which resulted in higher quality importance
rankings by U.S. Forest Service respondents.
The differences in importance ratings between local and tourist
respondents support the second hypothesis that differences in ancillary
attribute importance ratings occur among subgroups. The differences in
importance ratings between U.S. Forest Service and National Park Service
cross-country skiers give less support for the second hypothesis as some
differences between the two groups would be expected to be significant just
by chance (it would be surprising not to find any differences). However,
because the ancillary attributes (night club and bar) which were rated
significantly higher by U.S. Forest Service respondents (p ^ 0.05) could be
considered to represent ancillary attributes which are similar in nature, there
is a stronger likelihood that these differences are actual differences and did
not occur by chance. These differences may only have been ephemeral,
perhaps only existing during the season these data were collected or among
the skiers who happened to respond to the questionnaire. Additional
research would be necessary to determine if these differences are persistent

83
and what if any deeper meaning they may hold. Location of U.S. Forest
Service trails contrasted to the location of the National Park Service trail and
management strategies may have attracted skiers with different motivations.
Three of the U.S. Forest Service trails were located near either a convenience
store or a lodge (which was patronized by a great many snowmobilers and a
much smaller quantity of cross-country skiers). However, the most popular
U.S. Forest Service trail (Valley Spur) did not have such an establishment
nearby. A fee was charged at all of the U.S. Forest Service trails, with the
exception of one which was ungroomed. The National Park Service did not
require a fee to use the trail (Munising Trail) under its management. This trail
was regarded by some skiers as being narrower and more scenic than many of
the U.S. Forest Service trails. The National Park Service trail may have
attracted cross-country skiers who were more motivated by nature
appreciation opportunities than socialization opportunities, but there is no
means of investigating this conjecture further without additional data
collection.

Ancillary Attribute Importance Relationship to Length of Stay
In the results so far examined, there has been evidence that ancillary
attributes are im portant to cross-country skiers and that importance varies
according to whether or not the cross-country skier is a tourist or a resident of
the local area. As stated in the final hypothesis, ancillary attribute importance
ratings were expected to make a significant contribution in explaining the
variance in length of stay, when combined in a model with distance,
perceived constraints, trail satisfaction, overall community satisfaction,
num ber of previous overnight cross-country skiing trips, skill level,

84
spending, and quality of cross-country ski trails. All of these variables have
been mentioned to varying degrees in the literature as having an influence on
participation, the num ber of trips to a particular outdoor recreation site, or
the length of stay at a destination.
To determine the significance of each variable, a regression model was
developed. Those independent variables with a statistically significant
standardized regression coefficient had a significant association with the
amount of time cross-country skiers spent in Munising. The size of the
standardized regression coefficient indicates the relative importance of the
variable in the equation. The larger the coefficient, the larger the association
of the independent variable with the dependent variable. A regression of the
basic or core model was performed to compare with the model resulting
when ancillary attribute importance ratings were added to the analysis. The
basic or core model consisted of distance, total constraints, trail satisfaction,
satisfaction with the overall community, trips previous season, skill level,
spending in Munising, and trail quality. Stepwise regression was used for the
analysis which included the ancillary attribute importance ratings because
the num ber of variables was two large to include all of them in a single
standard regression model w ithout causing overspecification problems.
Several of the variables are somewhat correlated with each other, but
not to the degree that multicolinearity would be a problem (Table 7). Because
a stepwise procedure was used to generate the models which included
ancillary attribute importance ratings, p values should be viewed with
caution. Stepwise Regression tends to result in different p values from what
they would have been if the model were calculated with a standard regression

85
Table 7

1
1
2

3
4
5
6

7
8

9
10
11
12

13
14
15
16
17
18
19
2 0
21
22

2

3

4

5

6

7

8

9

10

1

0.028
0.161
0.161
-0.045
-0.195
-0.068
-0.071
-0.138
0.063
0.064
-0.097
0.038
0.107
-0.183
-0.034
-0.067
0.207
-0.096
0.074
-0 . 0 1 2
-0.039

0 .0 1
23
24 -0.048
25 -0.061
26 0.062
27 0.048
28 0.058
29 -0.006
30 -0.05

1

-0.07
-0.169
-0.08
-0.285
-0.093
-0.052
-0.026
0.133
0.171
0.166
0.085
0.149
-0.019
-0.088
0.045
0.291
0.266
0.164
0.062
0.191
0.167
0.255
0.171
0.182
0.203

1

0.165
1
-0.108 0.197
1
1
0.014 0.051 0.388
0.075 0.136 -0.037 0.105
1
0.029 0.081 0.081 0.187 0.056
-0.106 0.056 0.008 0 . 0 0 2 -0.08
0.189 -0.125
0.128 0.077
0 .2 2
-0.03 0.138 0 . 1 2 2 0.128 -0.035
-0.218 -0 . 0 1 0.009 -0.064 0.006
-0.015 -0.096 -0.188 0.152 0.005
0 .0 2
0 .1 2
0.061 -0.044 0.029
0.16
-0.192 0 . 0 1 1 -0.189 -0.127
-0.142 -0.035 -0.177 0.145 0.209
-0.255 -0.041 0.113 0.134 -0.041
-0.19 0.128
-0.066 0.164 -0 . 1 1
-0.056
0.09 -0.119 -0.243 0 . 0 0 1
-0.068 -0.006 -0.025 0.087 0.038
-0.028 -0.057 -0 . 1 2 2 -0.019 0 . 0 2 2
0.04
-0.105 -0.039 0.032 -0.091
-0.035 -0.088 -0.123 -0.228 0.081
-0 . 1 1 -0.195 -0.118
-0 . 1 0.067
-0.03
0 .0 2
-0.097 -0.066 0.217
-0.017 0.046 -0.244 -0.24 0.172
-0.072 0.093 -0.163 -0.219 0.189
0.025 0.131 -0.044 -0.064 0.171
0 .1 2
0.259 -0.03 -0.241 -0.288 -0.267 0.136
0.274 -0.125 0.135 -0.199 -0.398 0.168
(continued on next page)

1

0.109
-0 . 0 0 1
0.183
0.03
0.142
0.037
0.046
0.015
0.047
-0 . 0 2 2
-0 . 0 2
0.281
0.055
-0.068
-0.067
0 .0 2 1

-0.028
-0.061
-0.052
0.143
-0.008
0.142

1

-0.051
0.003
0 .1 2

-0.143
0.072
-0.031
-0 . 0 2 1
0

0.064
0.178
-0 . 0 1 1
0.072
0.135
0.092
0.068
0.148
0.104
0.114
0.017
0.048
0.166

1

0.258
-0.23
0.262
-0.005
-0.281
-0 . 1 0 2
-0.123
-0.025
0 .0 2 2

0.3
-0.293
0.015
0.042
0 .0 1

0.026
-0.317
-0.275
-0.133
-0.09
-0.174

86
Table 7 (Continued)
Peason correlation matrix of independent variables

11
11

12

13

14

15

16

17

18

19

1

0.029
1
13 0.154 -0.018
1
14 0.032 0.288 0.254
1
15
0.26
0 .1 2
0.048 0.227
1
16 0.218 0.091 0.279 0.246 0.535
1
17 0.251 0.472 0.052 0.374 0.286 0.359
18 0.244 0.215 0.149 0.335 0.192 0.233
19 -0.03 0.191 -0.236 0.161 0.086 -0.181
2 0
0.511 -0.085 0.335 0.057 0.059 0.086
21
0.226 0.456 -0.034 0.244
0.43 0.333
2 2
0.159 0.103 -0.154 0.139 0.251 0.147
23 0.177 0.138 0.127
0.43 0.304 0 . 2 1 1
24
0 .2 2
0.3 0.159 0.303 0.356 0.277
25 0.099 0.184 0.055 0.178 0.123 0.263
26 0.031 0.297 0.019 0.262
0.33 0.231
27 0.183
0.36
0.07 0.253 0.375 0.215
28 0.134
0.24 0.033 0.186
0.25 0.168
29 0.144 0.181 0.044
0.24 0.355 0.243
30
0 .2
0.208 -0.075
0.08 0.321 0.107
(continued on next page)
12

1

0.414
0.044
0.059
0.476
0.086
0.116
0.273
0.003
0.024
0.143
0.183
0.076
0 .1

1

0.158
0 .2 0 2

0.31
0.225
0.24
0.381
0.168
0.355
0.491
0.332
0.372
0.293

1

0.141
0.008
0.189
0.214
0 .1 1 2

0.179
0.237
0.235
0.164
0.109
0.271

87
Table 7 (Continued)
Peason correlation matrix of independent variables
2 0
2 0
21
2 2

23
24
25
26
27
28
29
30

1
0.098
-0.065
0.151
0.146
-0 . 0 1 1
-0.024
0.164
0.326
0.105
0.153

21

2 2

23

24

25

26

27

28

1

0.297
0.111
0.458
0.14
0.373
0.469
0.269
0.432
0 .2

1

0.36
0.386
0.547
0.258
0.284
0.028
0.314
0.315

1

0.447
0.187
0.423
0.407
0.156
0.421
0.313

1

0.287
0.404
0.48
0.359
0.592
0.329

1

0.258
0.181
0 .0 2 1

0.276
0.289

1

0.812
0.349
0.57
0.358

1

0.502
0.576
0.431

1

0.35
0.33

Note. l=distance, 2=constraints, 3=trail satisfaction, 4=overall community
satisfaction, 5=trips past season, 6 =skill level, 7=spending, 8 =trail quality,
9=permit price, 10=low price motel, ll= m id price motel, 12=ski lodge,
13=basic kitchen, 14=child care, 15=swimming pool, 16=sauna,
17=ventillated waxing room, 18=laundry, 19=fastfood restaurant,
20=family restaurant, 21=gourmet restaurant, 22=night club, 23=family
entertainment, 24=cultural attraction, 25=bar, 26=gift ship, 27= clothing
shop, 28=ski ship, 29=arts and crafts shop, 30=overall community

88
procedure (SYSTAT 1992). However, the results do indicate which ancillary
attribute importance rating variables are useful in examining length of stay.
This analysis was performed only for tourist cross-country skiers.
Length of stay, by definition, for local cross-country skiers was one day. This
lack of variation would make inclusion of local respondents meaningless.
Also, the analysis reported above indicated that local and nonlocal
respondents rated the importance of several ancillary attributes differently.
Other authors have indicated that including both local and nonlocal skiers in
the same length of stay regression analysis can lead to results which do not
make much sense.

Stay
Four models are presented below. First a base model of basic variables
which have been discussed as influencing length of stay, participation, or
dem and in the recreation or recreation economics literature was developed.
This base model is as follows:
Length of stay = Constant + BjDistance + ZLTotal constraints + f^T rail
satisfaction + B4 overall community satisfaction + BsNumber of overnight
trips during the past season + BgSkill level + B;Spending in Munising +
BgQuality trail importance.
The same base model was then forced into a stepwise regression procedure
with the remaining

21

ancillary attribute importance variables to determine

which of these variables could make significant additions to the model.
Length of stay was then replaced as the dependent variable with willingness
to pay trail fees to determine how the model would change if a different

89
measure of dem and was used as the dependent variable. In addition the
current price at the time of data collection to use the trail ($3.00 for U.S. Forest
Service respondents and $0.00 for National Park Service respondent) was
added to the model because previous analysis (Bishop, Forsberg, and
Flolecek 1992) had shown this variable to have had a significant impact on the
amount cross-country skiers would be willing to pay in the future to use the
trails. The new base or core model with the willingness to pay trail fees as the
dependent variable was developed to see if a somewhat different measure of
dem and for cross-country skiing at Munising would effect the model. The
new base or core model was then included in a stepwise regression model
with the ancillary attribute importance ratings.
The analysis of the base model indicated that it does explain a
significant (p s 0.05) am ount of the variance with an adjusted R2 of 0.24 which
indicates that 24% of the variance in length of stay was explained by the
model (Table 8 ). Distance and the amount of spending in Munising the day
the questionnaire was received were significant independent variables (p ^
0.05). Overnight trips during the previous season (p <. 0.10) and skill level (p
^0.10) were almost significant in explaining length of stay variance. As
would be expected from the findings of Noe (1987), satisfaction with trails had
a greater association with length of stay than overall community satisfaction,
but both measures were insignificant. Quality of cross-country ski trail
importance also had little effect in the base model.
The base model was then forced into a stepwise regression procedure
with the ancillary attribute importance rating variables. The resulting model
had an adjusted R2 of 0.378 which indicated that an additional 13.8% of the
variance in length of stay was explained (Table 9). Distance, overnight cross-

90
Table 8

Variable

Ba

SEb

SBc

Constant

-1.92

2 .1 0

0 .0 0

Distance

0 .0 0

0 .0 0

0.36

Constraints

0 .0 1

0 .0 1

Trail satisfaction

0 .1 1

O verall
com m unity
satisfaction

Tol.d

te

pf

-0.91

0.36

0.91

4.26

0 .0 0

0.09

0 .8 8

1.03

0.30

0.14

0.07

0.91

0.81

0.42

0.03

0.13

0 .0 2

0.83

0.24

0.81

Trips previous
season

0.08

0.05

0.15

0.79

1.65

0 .1 0

Skill level

0.32

0.19

0.16

0.76

1 .6 8

0 .1 0

Spending

0 .0 1

0 .0 0

0.35

0.92

4.09

0 .0 0

Trail quality

0 .1 2

0.25

0.04

0.96

0.47

0.64

N ote. Analysis with length of stay in Munising, Michigan as the dependent
variable included U.S.Forest Service and National Park Service tourist
respondents (N = 116, adjusted

R2

= 0.24, F = 5.55, p model <; 0.05).

aB = R e g ressio n

coefficient. t>SE=Standard error. cSB=Standardized regression coefficient.
<FTol.=Tolerance. et=t statistic. fp=individual variable probability.

91
Table 9

Variable

Ba

SEb

SBc

Constant

-2.15

1.99

0 .0 0

Distance

0 .0 0

0 .0 0

0.32

0.82

3.90

0 .0 0

Constraints

0.03

0 .0 2

0.16

0.73

1.85

0.07

Trail satisfaction

0 .1 2

0.13

0.07

0.89

0.97

0.34

Com m unity
satisfaction

0.05

0.13

0.03

0.74

0.39

0.70

Trips previous
season

0 .1 1

0.05

0.19

0.71

2.15

0.03

Skill level

0.26

0.19

0.13

0.65

1.38

0.17

Spending

0 .0 1

0 .0 0

0.34

0.91

4.32

0 .0 0

Trail quality

0 .2 0

0.24

0.07

0.91

0.84

0.40

-0.17

0.08

-0.18

0.77

-2.07

0.04

0 .2 1

0.08

0.24

0.70

2.72

0 .0 1

-0.32

0.13

-0 . 2 0

0.75

-2.38

0 .0 2

Laundry

0 .2 1

0 .1 2

0.17

0.64

1.83

0.07

Fast food

-0 . 2 1

0.09

-0 . 2 0

0.76

-2.28

0 .0 2

Night club

0.24

0 .1 1

0.19

0.77

2.25

0.03

Cultural attraction

-0.19

0.09

-0.19

0 .6 6

-2.08

0.04

Mid price motel
Basic kitchen
Child care

Tol.d

te

Pf

-1.08

0.28

N ote. Analysis with length of stay in Munising, Michigan as the dependent
variable included U.S.Forest Service and National Park Service tourist
(continued on next page)

Table 9 (continued)

respondents (N = 112, adjusted R2 = 0.38, F = 5.50, p m od el ^ 0.05).
aB=Regression coefficient.

bSE=Standard error.

cSB=Standardized

regression coefficient. dTol.=Tolerance. et=t statistic.
variable probability.

fp=individual

93
country ski trips during the previous season, and Munising spending were
significant ( p <, 0.05) base model variables. In addition, the constraints
variable was nearly significant ( p <. 0.10). Mid price lodging, Basic kitchen,
child care, laundry, fast food, night club, and, cultural attraction were the
ancillary attribute variables which were added to the base model. Mid price
lodging, child care, fast food, and cultural attraction importance ratings had
negative associations with length of stay. Basic kitchen, laundry, and night
club importance ratings had positive associations with length of stay. Trail
satisfaction continued to be of more importance than overall community
satisfaction, but both variables were insignificant. Trail quality importance
again was insignificant. Basic kitchen and laundry would be expected to be
im portant to cross-country skiers planning to stay longer as they would allow
lower expenses and the bringing of fewer clothes. There is no single
explanation as to why mid price lodging, child care, fast food, and cultural
attraction importance ratings would be associated with shorter lengths of stay.
Those respondents with children may be more anxious to return home
because of obligations such as school associated with children. They may also
have less discretionary income. Fast food customers may be less well
established in their careers, younger, or perhaps cannot afford to stay as long.
Cultural attraction importance may indicate a desire for some meaningful
activity other than cross-country skiing. Munising did not offer a clearly
identifiable cultural attraction during the season which these data were
collected.
The same analysis was then repeated with the amount of trail fee
respondents would pay rather than risk closure of the trail as the dependent
variable. Price of a trail perm it ($3.00 for Forest Service respondents and $0.00

94
for National Park Service respondents) was added to the analysis. The basic
or core model continued to be significant (p <. 0.05) with an adjusted

R2

of

0.134 (Table 10). Price was the only significant ( p <; 0.05) variable in the core
model. However, overnight cross-country ski trips during the previous
season ( p -

0

.1 0 ) and the importance of trail quality ( p =

0

.1 2 ) were nearly

significant.
The inclusion of the ancillary attribute importance ratings in a stepwise
regression analysis with the revised trail fee core model resulted in a
substantially different model from the length of stay analysis. The model was
significant ( p ^ 0.05) with an adjusted
additional

R2

of 0.336 which indicates that an

. % of the variance in willingness to pay trail fees was explained

2 0 2

(Table 11). Satisfaction with the overall community, past season overnight
trips, and price were the only significant (p s 0.05) base model variables of the
nine forced into the analysis. The num ber of overnight cross-country ski trips
during the previous season was the only core variable which was significant
in both the length of stay and the trail fee stepwise regression models.
However previous season overnight trips had a positive association with
length of stay and a negative association with trail fees. Cross-country ski trail
quality importance continued to be insignificant.
Overall community satisfaction had a positive association with the
willingness to pay trail fees while trail satisfaction was insignificant.
Common sense might lead to the expectation that trail satisfaction would have
a stronger relationship with trail fees than overall community satisfaction
since most people would expect that higher trail satisfaction would be more
likely to correspond with a willingness to pay higher trail fees, but such was
not the case in this analysis. Perhaps the lack of variation in the relatively

95
Table 10

Ba

Variable

SEb

S Be

Tol.d

te

pi

-0.40

0.69

Constant

0 .8 8

2 .2 0

0 .0 0

Distance

0 .0 0

0 .0 0

0.04

0 .8 8

0.38

0.71

-0 . 0 2

0 .0 1

0.04

0.89

-1.36

0.18

0 .0 1

0.14

0 .0 1

0.92

0.08

0.94

O verall
com m unity
satisfaction

0.08

0.13

0.06

0.83

0.64

0.53

Trips previous
season

-0.09

0.05

-0.16

0.79

-1.67

0 .1 0

Skill level

-0 . 1 0

0 .2 0

0 .0 0

0.76

-0.05

0.96

Constraints
Trail satisfaction

Spending

0 .0 0

0 .0 0

0.03

0.92

0.34

0.74

Trail quality

0.41

0.26

0.14

0.94

1.56

0 .1 2

Price

0.50

0.13

0.35

0.94

3.90

0 .0 0

N ote. Analysis with contingent trail fee as the dependent variable included
U.S.Forest Service and National Park Service tourist respondents (N = 116,
adjusted

R2

= 0.13, F =2.97, p

m o d el

^ 0.05). aB=Regression coefficient.

bSE=Standard error. cSB=Standardized regression coefficient.
rfTol.=Tolerance. et=t statistic. i p in d iv id u a l variable probability.

96
Table 11

Variable

Ba

SEb

SBc

Constant

-0.69

2 .0 2

0 .0 0

Distance

0 .0 0

0 .0 0

0.06

Constraints

-0 . 0 2

0 .0 2

Trail satisfaction

0 .1 0

C om m unity
satisfaction

te

pf

-0.34

0.73

0.82

0.73

0.47

-0.15

0.73

-1.69

0.09

0.13

0.06

0.85

0.77

0.44

0.27

0.13

0 .2 0

0.69

2 .1 1

0.04

Trips previous
season

-0 . 1 0

0.05

-0.19

0.75

-2.09

0.04

Skill level

-0.08

0.19

-0.04

0 .6 8

-0.44

0 .6 6

Spending

0 .0 0

0 .0 0

-0 . 0 1

0.87

-0 . 1 2

0.91

Trail quality

0.33

0.23

0 .1 2

0.93

1.44

0.15

Price

0.52

0 .1 2

0.37

0 .8 8

4.44

0 .0 0

Ski lodge

0.16

0 .1 1

0.14

0 .6 8

1.45

0.15

Waxing room

0.32

0.13

0.26

0.56

2.54

0 .0 1

Laundry

-0.37

0.13

-0.31

0.52

-2 . 8 6

0 .0 0

Fast food

-0.14

0.09

-0.14

0.80

-1.58

0 .1 2

Cultural attraction

0.16

0 .1 0

0.17

0.54

1.65

0 .1 0

Arts and crafts
shop

0 .2 1

0 .1 0

0 .2 2

0.50

2.04

0.04

Tol.d

N ote. Analysis with contingent trail fee as the dependent variable included
U.S.Forest Service and National Park Service tourist respondents (N = 111,
(continued on next page)

97
Table 11 (continued)

adjusted R2 = 0.34, F = 4.70, p

m od el

^ 0.05). aB=Regression coefficient.

t>SE=Standard error. cSB=Standardized regression coefficient.
4Tol.=Tolerance. et=t statistic. fp in d iv id u a l variable probability.

98
high importance ratings given trail quality caused its insignificance in the
model. Overall community satisfaction may be an indication of greater
income, more spending, and the ability to pay higher trail fees. However,
spending in Munising and its association with trail fees was insignificant.
Overall community satisfaction may be an indication of brand (Munising)
loyalty. Respondents with high community satisfaction perhaps would rather
pay a little more in trail fees than go elsewhere for their cross-country skiing
just as people who prefer a nationally advertised brand of groceries pay a
little more than they would for the store brand. Respondents with high
overall community satisfaction may have a special connection to Munising
(perhaps relatives or property in the area) which would make switching to
another cross-country ski destination community more difficult. The
negative association of the num ber of overnight cross-country ski trips taken
during the past season and trail fees may be the result of better knowledge of
alternative trails and destinations and the trail fees charged at these
alternatives. For example, at many Michigan State Forest cross-country ski
trails payment is only requested in the form of a voluntary contribution.
Of the 21 ancillary attribute importance variables in the trail fee step­
wise regression, ski lodge, ventilated waxing room, laundry, fast food,
cultural attraction, and arts and crafts shop were added to the base model.
Ski lodge, waxing room, cultural attraction, and arts and crafts shop
importance ratings were associated with higher trail fees while laundry and
fast food were associated with lower trail fees. Waxing room, laundry, and
arts and crafts shop were of more influence as indicated by their larger
standardized coefficients. The ancillary attribute importance ratings which
were significant in the trail fee analysis may have been proxies for income.

99

Those attributes positively associated with trail fees were those which
possibly would add to the cost of the visit and therefore could possibly be
more im portant to those w ith higher incomes. Those attributes negatively
associated with trail fees were those which could possibly reduce the cost of
the visit and therefore could possibly be more im portant to those with lower
incomes. The m anner in which income was measured in the questionnaire
precluded its inclusion in the analysis.
To further explore attribute importance rating variable validity,
distance zones were examined as a kind of proxy for attribute preferences.
Perhaps respondents from certain regions, the Upper Peninsula of Michigan
for example, would be different in their importance ratings from other
respondents. Respondents were grouped into four distance or proximity
zones: those residing up to 99 miles away from Munising, those residing
between 100 and 299 miles away from Munising, those residing 300 to 499
miles away from Munising, and those residing at least 500 miles away from
Munising. Because of the small num ber of respondents in the first group (11
respondents) and the last group (15 respondents) a Kruskal-Wallis one-way
analysis of variance procedure (SYSTAT 1992) was used to determine if there
were any significant differences in the 22 importance ratings among
respondents from the various distance zones. Significant differences (p s 0.05)
occurred in the importance ratings of mid price lodging, swimming pool,
sauna, and laundry. An analysis of differences in the importance ratings of
these four ancillary attributes between the distance zones compared two at a
time did not suggest any plausible explanations other than chance for these
differences (Table 12).

100

Table 12
Significant differences between distance zones

Ancillary attribute

1 to 2

Mid price lodging

✓

Swimming pool

✓

Sauna
L aundry

1 to 3

1 to 4

2 to 3

2 -lo A

✓

✓

3 to 4

✓
✓

✓
✓

✓

✓

N ote. Respondents within 99 miles of home were in zone 1. Respondents
within 100 to 299 miles of home were in zone 2. Respondents within 300 to 499
miles of home were in zone 3. Respondents more than 499 miles away from
home were in zone 4. ✓ indicates importance ratings differed significantly
between the two groups ( p ^ 0.05) as determ ined by Kruskal-Wallis analysis of
variance (SYSTAT 1992).

CHAPTER V: SUMMARY AND CONCLUSIONS
This study examined the importance of ancillary attributes to cross­
country skiers visiting a small, rural, northern Michigan community.
Ancillary attributes are those attributes which are secondary to the main
purpose of the trip but facilitate the main trip activity, hotel accommodations
or fast food restaurants for example in the case of cross-country skiing.
Respondents rated the importance of 21 ancillary attributes and the
importance of quality ski trails on a 1 (not important) to 7 (very important)
scale. Previous studies of outdoor recreation activities, and specifically
cross-country skiing, have examined the importance of attributes of the
outdoor recreation site, usually a wild land setting. A few studies of dem and
for down hill skiing have incorporated resort characteristics and one study
(Brayley 1991) compared resort characteristics important to cross-country
skiers to those im portant to snowmobilers. Other studies which have
examined the importance of attributes have done so for a specific or several
specific tourism destinations in general and not for a specific outdoor
recreation activity.

Studies of specific outdoor recreation activities have

focused on attributes to be found at the outdoor recreation site and have
ignored the role of supporting facilities and services, such as lodging
establishments, restaurants, entertainment establishments, and shopping
establishments, in augm enting outdoor recreational experiences (Lime 1971;
Lucas 1980; Allton and Lieber 1983; Peterson, Dwyer, Darragh 1983; Lucas

101

102
1985; Watson, Williams, Roggenbuck, and Daigle 1992; Love and Watson 1992;
Roggenbuck, Williams, and Watson 1993).
The results of this study indicate that importance ratings of some
ancillary attributes may be useful in comparing various groups of cross­
country skiers and in explaining variations in length of stay. Ancillary
attribute importance ratings were higher for those variables that were
available at the destination community, fulfilled basic shelter and sustenance
needs, or were strongly connected with the main purpose of the recreational
trip. These results indicate that cross-country skiers base their destination
selection at least in part on the presence of key ancillary attributes. The
results of this study tend to support each of the three hypotheses presented in
chapter III.
Only nine of the 21 ancillary attributes measured were found to be
significant in either length of stay or willingness to pay models.

Evidence of

the synergetic relationship between the natural, wild land, recreation setting
of cross-country skiing and the support services provided by destination
communities would have been strengthened if additional ancillary
importance ratings had m ade significant contributions in the models. Several
ancillary attributes were added to both the length of stay model and the trail
fee model in stepwise regression procedures, but none were significant (p ^
0.05) across both models. Mid price lodging, basic kitchen, child care, fast
food, night club, and cultural attraction were significant in the length of stay
regression model, while ventilated waxing room, laundry, and arts and crafts
shop were significant in the trail fee regression model. Laundry was a part of
the length of stay model but was not significant, and ski lodge, fast food, and

103
cultural attraction were part of the trail fee model but did not have a
significant p value.
Attribute importance, in the trail fee model, may have served as a
substitute for a measure of income. Those ancillary attributes with a positive
influence on willingness to pay trail fees were those that probably would be
available at higher prices (ski lodge, ventilated waxing room, cultural
attraction, and arts and crafts shop). This may indicate the Munising cross­
country ski market could be divided into budget and luxury segments.
Overall community satisfaction had a significant positive association
with trail fees. This indicates that those who reported greater overall
satisfaction with the community were willing to pay higher trail fees.
However, satisfaction with trails did not have a significant effect on the
willingness to pay trail fees even though cross-country ski trails would
appear to have a more direct link with the main activity of the trip, cross­
country skiing. Perhaps those skiers who report greater overall community
satisfaction have developed a degree of brand loyalty and are willing to pay
more to keep skiing in the Munising area or perhaps they have stronger
connections to the Munising area either through the ownership of property or
the presence of relatives or friends.

Lack of variation in the trail satisfaction

variable may be the reason for its lack of association with trail fees. Nearly all
respondents rated satisfaction with trails highly.
The num ber of overnight trips taken during the previous cross-country
ski season had a negative association with willingness to pay trail fees.
Perhaps these skiers have a greater knowledge of alternative ski destinations
where trail fees are lower or nonexistent. At many Michigan State Forest
cross-country ski trails, for example, payment is voluntary. The data for this

104
study were collected during a winter when snow was not plentiful in lower
Michigan and farther to the south. During winters with plentiful snow to the
south of Munising, price sensitivity towards both trail fee and travel cost may
cause some cross-country skiers, especially those who have traveled over
night to a num ber of different locations, to use substitute cross-country ski
destinations which have lower costs. Munising's major competitive
advantage, identified by local U.S. Forest Service personnel, is that the
Munising area frequently has abundant snow when competing locations to
the south have none or little. During winters when snowfall is plentiful in
more southern locations, Munising's major competitive advantage may
disappear for large portions of the cross-country ski season, making lessexpensive substitute destinations more attractive.
A relatively large am ount of variance in willingness to pay trail use
fees and in length of stay remained unexplained by the models used in the
current study. This may reflect the reluctance of many individuals to pay any
fees for whatever purpose to a government agency which they have already
supported through taxes. Some other wording of this question which would
more clearly identify the benefits which would be obtained with fee payment
and would identify some organization other than a government agency as the
recipient of the money would perhaps be an improvement. However, an
additional three of four dollars per person per day may not be a large enough
expense, when compared to the several hundred dollars being spent in total
on a cross-country ski excursion, to be of concern to most respondents.
Personal income and time may also have been very limiting, both in the time
spent in Munising and the am ounts respondents would be willing to pay for
trail fees. The addition of these two variables perhaps would account for a

105
greater portion of the variance in both models. However, both time and
income were at least partially accounted for by the distance from home
variable and also as a portion of the constraints variable. Constraints were
nearly significant in both models and distance was a significant variable in the
length of stay model, but not in the trail fee model.
Although most studies of recreation dem and use the num ber of trips
individuals make to the recreation site as the dependent variable, this study,
as did Bell and Leeworthy (1990) and Silberman (1985) among others, used
length of stay as the dependent variable. The focus of Bell and Leeworthy's
(1990) study was beach attendance. They did not include ancillary attributes
in their dem and model except to include on site spending as a constraint.
Silberman included some attributes which could be considered ancillary in a
measure of image, but his study was of tourism in general to a specific
destination and not of a specific outdoor recreation activity. The current
study examined a specific outdoor recreation activity and the contribution
that certain ancillary attribute importance ratings can make in explaining the
variance in length of stay. Somewhat contrary to what could be considered
common wisdom among natural resource managers, trail quality importance
and trail satisfaction had little effect on length of stay or willingness to pay
trail fees. Several ancillary attribute importance rating variables were
significant in one or the other of the models and overall community
satisfaction was significant in the trail fee model.
Significance differences in two ancillary attribute im portant ratings,
bar and night club, occurred between U.S. Forest Service and National Park
Service respondents. Also, there was a significant difference in the
importance rating of quality of ski trails. All of these attributes were rated

106
significantly higher in importance by U.S. Forest Service respondents. U.S.
Forest Service skiers may be more interested in such social opportunities as
interacting with family and friends afforded by cross-country skiing than
National Park Service respondents. Perhaps National Park service
respondents would be more interested in opportunities to observe scenery
and nature. However, such issues could not be adequately examined with the
data available. U.S. Forest Service respondents may have accorded more
importance to the quality of the cross-country ski trail because of the trail fee
charged at U.S. Forest Service trails.
In accordance with need theories of motivation, those ancillary
attributes considered more im portant than average were those directly
related to cross-country skiing, which was the main trip activity for all but a
handful of the respondents, and the basic needs of sustenance and shelter.
One exception to the above generalization was the relative importance of
sauna for some for local skiers. Taking a sauna at first glance appears to be
unrelated to either a basic need or the cross-country ski experience. However,
for some groups of cross-country skiers, saunas may be closely linked with
cross-country skiing activity. Some skiers may consider a sauna after a cross­
country ski run as a key part of the cross-country ski experience. The sauna
may be likened to a rich desert after the main meal of several hours of cross­
country skiing.
The results of this study indicate that cross-country skiers view the
quality of cross-country ski trails as the most important attribute in choosing
a destination community. However, trail quality importance ratings were not
a significant variable in explaining variance in length of stay in Munising or in

107
explaining variance in willingness to pay trail fees. Other ancillary attributes
had a greater impact on both of these dem and related variables.
The cross-country skier must decide if the potential destination has the
type of trails desired. While the current study did not examine how cross­
country skiers may define trail quality, other studies have determined the
qualities that cross-country skiers value in a cross-countiy ski trail (Nelson
1988; Smith 1980; Rosenthal, Driver, and Rauhauser 1980; McLaughlin and
Paradice 1980). If a community receives adequate snowfall, quality cross­
country ski trails may be relatively easy to provide. If several competing
destinations have trails with the desired qualities, ancillary attributes may
prove to be the deciding factors on which a destination community is chosen.
Cross-country skiing is not an activity the recreationist chooses lightly.
Some expense is involved in obtaining the necessary equipment, usually a
minim um of $100.00. Additional expenses of money and time are necessary to
travel to a destination community and stay at least a couple of days. The
cross-country skier usually doesn't arrive at a destination such as Munising,
spend the night, and then decide to go for a ski. Cross-country skiers, as
indicated by this study, usually arrive with their equipment in hand with
cross-country skiing in mind as the main activity of the trip. Travelers during
other seasons may be touring for a variety of reasons, choosing activities as the
mood strikes, or as they are motivated to do so by advertising or sights and
sounds along the way (Atkinson and Raynor 1974). Many summer travelers
may not even have a specific destination in mind, but just head up north, back
east, out west, or dow n south. Nearly all of the respondents contacted at
Munising listed cross-country skiing as the primary purpose of the trip.
Given the central importance of skiing to the entire journey for nearly all the

108
respondents, it is highly likely that the first test of a destination is the
availability of quality trails. Quality to a cross-country skier may mean that
the trail is groomed and passes through scenic terrain with a number of hills.
Once the availability of that attribute has been established, the skier probably
looks for suitable supporting ancillary attributes.
Logic, and much of the recreation economics literature would indicate
that on-site costs, often related to ancillary attributes, would have a negative
effect on length of stay, but the results in the present study indicate that
higher spending results in longer stays. Perhaps the inclusion of income and
time costs in the model w ould change this result. According to Bockstael,
McConnell and Strand (1991), the omission of time opportunity cost should
increase the absolute value of money costs. The inclusion of income and a
time value variable would perhaps cause on site spending to have a negative
effect on length of stay, or at least lessen its positive effect.
Ancillary attributes in economic studies are usually represented only
as a composite on site price, or in the hedonic travel cost model, as a specific,
objective, physical quantity m easure which is then valued according to how
much further tourists travel to reach a destination with a specific set of
attributes. However, the present study has demonstrated that ancillary
attribute importance measures can be valuable additions in models
examining dem and variation. Outdoor recreation sites provided by the
government are calculated to have a consumer's surplus, a value beyond
what the consumer pays to obtain the use of that site. Presently, when
recreation sites are valued using travel cost models or contingent valuation
studies, all of the consumer's surplus is taken as the value of the recreation
site. However, it seems, given the results of the current study, that at least

109
some ancillary or support services provided by private business
establishments are used by the recreationist in the production of the
recreation experience. Thus, attributing the entire consumer's surplus
associated with a visit to a public recreation area will result in an over
estimate of the value of the recreation area. The association of increased
spending in the community with longer stays and higher overall community
satisfaction with higher trail fees would seem to indicate, at least in the case of
cross-country skiing, that a portion of the recreation value of a public
recreation site is derived from nearby supporting services provided by the
private sector. The private sector in turn benefits from additional customers
which are draw n to the community specifically by the public sector
recreation developments. Both types of organizations benefit from goods and
services provided by the other. Ultimately it is the customer who has a better
recreation experience and receives a better return on the recreation
investment of time, money, and equipment because of joint efforts by public
and private sectors.
Government provided attributes (packaged in the form of recreation
sites) can be thought of as inputs to the production of recreation experiences
by the household (Cicchetti, Fisher and Smith 1976; Walsh 1986; Deyak and
Smith 1978). These experiences tend to be valued more highly than the costs
of obtaining the government provided inputs. Hence the inputs are said to
have consumer's surplus. Ancillary attributes such as hotels and restaurants,
have traditionally been valued only at their market price in recreation
dem and studies. Theses costs are then treated as constraints or are added into
the travel cost variable in the model. However, they can also be thought of as

110
inputs to creating the recreation experience which results in consumer's
surplus.
If ancillary attributes help to create the recreation experience which
results in consumer's surplus, then some of the consumer's surplus
traditionally used to value the government provided outdoor recreation site
should be credited to the private sector ancillary attributes. The purchase of
hotel nights, restaurant meals, travel, and access to the recreation site all serve
as inputs to the production of a recreation experience.
As suggested by Fletcher, Adamowicz, and Graham-Tomasi (1990), if
utility is generated by staying in a hotel, then only the minimum lodging
price should be counted as travel cost.

Only by spending time at the

destination collecting (consuming) both private sector and government sector
inputs (Mak and Moncur 1980; Rugg 1973) can the consumer create a
recreation experience. If the hotel is not separable from the trip product and
part of the trip enjoyment comes from eating out, staying in motels, and
shopping, then the attributes providing these services should be included in
the dem and system used to evaluate an outdoor recreation site (Fletcher,
Adamowicz, and Graham-Tomasi 1990). If ancillary attributes are im portant
in creating recreation experiences, then planning concepts such as the
recreation opportunity spectrum could perhaps be expanded to include
secondary service facilities.
Different recreation activities and different participation styles could
be expected to depend on different combinations of ancillary attribute inputs
which would have different contributions to consumer's surplus. Willingness
to pay is not only a m easure of the value of the outdoor recreation site but is
also a measure of the value of all inputs to outdoor recreation experience

Ill

production. As Getz (1993) indicates, "Retail surveys are . . . needed to test
tourist preferences for types of goods and services. . . . "
Without some measure of what the visitor perceives as the important
qualities of the outdoor recreation site and supporting destination
community, to determine exactly what the visitor values in a particular
destination is difficult. Even when the trip is a single purpose, single
destination event, to ascribe all the consumer's surplus to the outdoor
recreation site ignores the contributions of other inputs which are often found
in supporting destination communities.
To determine the importance of ancillary and primary attributes,
customers need to be asked what attributes are of importance. Observation of
behavior or experimental choice situations can then be used to confirm the
importance level of specific attributes. Such observation may reveal that
some respondents may be unaw are of some of the attributes which influence
their decision. Respondents may also ascribe too much or too little
importance to other attributes. Variations in willingness to pay to experience
various combinations of ancillary and primary attributes could be used to
determ ine the values of the individual attributes in creating recreation
experiences.

A pplications
This research has identified several ancillary attributes which were
rated by skiers as being more im portant in selecting a cross-country ski
destination and which were related to longer length of stay in the Munising,
Michigan area. Additional attributes were identified which were related to a
willingness to pay higher trail fees. Knowing which attributes are of

112
importance to cross-country skiers will allow more effective marketing
campaigns aimed at changing beliefs about these attributes, changing the
relative importance of attributes, adding new im portant attributes at
destinations currently lacking them, and changing beliefs about the "ideal"
cross-country ski destination (Hawkins, Best, and Coney 1992).

Co m m u n ity A p plications
Munising, as well as other communities with characteristics similar to
those of Munising, could find the results of this research useful.
Organizations and businesses in the Munising area could emphasize in their
promotional efforts those attributes present in Munising which were rated as
being more im portant in choosing a cross-country ski destination.
Given the much higher average rating of trail quality importance and
the detail with which this attribute was described by the mail questionnaire
respondents in the open-ended questions, any promotional efforts should
emphasize the availability and quality of a num ber of superior cross-country
ski trails in the vicinity of the community and the special features of these
trails. However, promotion is only one portion of any tourism program.
Additional effort perhaps could be made to further develop the cross­
country ski tourism product in Munising. The cross-country ski tourism
product consists of trails, lodging establishments, restaurants, shops,
emergency services, infrastructure and other services and products which
tourists use (Gunn 1994). Sometimes the cross-country ski product is thought
only to be provision of the groomed trail. However, this research indicates
ancillary attributes are a vital part of the product. Some importance ratings
for ancillary attributes were associated with longer stays, others with higher

113
trail fees. Promotion of trail quality is necessary to interest customers in
coming to Munising. Quality ancillary attributes are then necessary in order
to maintain and increase the customer base. Providing key ancillary
attributes is necessary to assure return visits and longer stays. Neither is
sufficient by itself to assure continued success.
Product development would involve providing additional services
and individual products that cross-country skiers would find useful or
enjoyable. For example Munising restaurants, as restaurants have in other
regions, could add additional so-called "healthy" menu items to their menus.
Lodging establishments could set aside certain floors or sets of rooms
especially for cross-country skiers or others who desire quiet stays as
opposed to other groups of w inter sports enthusiasts who may prefer or
generate more noise. Special events could be organized, especially during
usually less crowded periods. The U.S. Forest Service in cooperation with
local businesses and other organizations organized during the 1992-1993 and
the 1993-1994 ski seasons an event in which hors d'oeuvres were served at
several stations along one of the cross-country ski trails followed by a banquet
for all participants. Such product development may be especially valuable
during seasons when snow is plentiful and Munising is faced with many
potential competing destinations. The results indicate that experienced cross­
country ski tourists may be especially price sensitive. During winters with
good snow cover to the south, Munising perhaps should consider lower trail
fees, making trail fee payment voluntary, increased promotion, and special
packages and events to continue to draw cross-country skiers who might
otherwise stop at destinations closer to home.

114
However, in order to know what products to develop, a customer
information system must be developed to determine what products and
services customers and potential customers desire. Information can be
developed from surveys, focus groups, test marketing new ideas, or informal
conversations with visitors (Covey 1990). Important insights could be gained
by using each method. Tastes change. Only through systematic, diligent
effort can customer changes be identified, acted upon and new products
developed. One trail in the Munising area could be selected at which to
sample weekend cross-country ski visitors with a short questionnaire to gage
how well the community and the cross-country ski trails are meeting the
needs of the skiers. Through the collection of information over several
seasons, marketing efforts could be fine-tuned to appeal to the types of cross­
country skiers most likely to visit Munising. Changes in the ages, life stages,
and incomes of skiers could be identified and necessary changes in product
planned. Results of the regression analysis reported above indicate that
providing more high end services, such as those found in ski lodges, may
attract higher spending cross-country skiers while providing basic kitchens
and entertainment may encourage cross-country skiers to stay longer.
Through emphasizing quality cross-country ski trails in promotional
efforts and assisting resource management in providing high quality trails,
business from cross-country skiers could for a time increase. Providing
desired quality ancillary services while maintaining quality cross-country ski
trails would encourage a long term increase in business. The cross-country
ski trails in the Munising area give the local business organizations a concise,
concrete product to promote. The provision of additional services which

115
exceed customer expectations provide the foundation for long term business
growth.

The results of this research indicate that the provision of quality cross­
country ski trails is highly important. National Park Service and the U.S.
Forest Service currently provide the land for the trails in the area and
organize resources to see that they are groomed on a regular basis during the
cross-country ski season. W ithout these efforts, cross-country skiers would
probably not come to Munising in any appreciable numbers. In addition, the
U.S. Forest Service has taken the lead in efforts to organize the community to
more effectively promote and provide for cross-country skiing. The results of
this research underscore the importance of these efforts. W ithout the
provision of key ancillary attributes by the community at large, efforts to
promote cross-country skiing by the National Park Service and U.S. Forest
Service through the provision of groomed trails would only reach those
skiers who live nearby. If only skiers from the local area were to use the cross­
country ski trails, the value of the trails would be greatly reduced, possibly to
the point that there would be no justification for keeping them open. Cross­
country ski trails in the proximity of a destination community, at least in the
case of Munising, result in more business and greater value for businesses
located in the local area.
Although this research did not closely examine the attributes of a
quality cross-country ski trail, this research does emphasize the importance of
providing quality trails in attracting cross-country ski tourists to the local
community. To detect changes in what quality means in cross-country skiing,

116
the U.S. Forest Service could assist the community in instigating a customer
information system. Cross-country skiers are most easily contacted at the
trails. A suggestion box could provide valuable insights to the kinds of trail
features desired by skiers and also the kinds of community attributes cross­
country skiers would like to have available. More sophisticated means of
information collection could confirm these insights or detect other trends.
This research indicates that while quality trails are im portant in attracting
skiers, the characteristics of the destination community influence length of
stay and spending. Resource managers could emphasize the importance of
trails in attracting customers in efforts to increase community support for and
cooperation in cross-country ski trail management.
The results of this research indicated that there is a synergistic
relationship between resource management agencies and local communities
providing tourist services. Thus, there is a need for resource management
agencies and local communities to work together to provide satisfying cross­
country ski experiences.

Future Research
Additional research is needed to clarify the importance of ancillary
attributes to visitors who engage in specific, specialized recreation activities.
Ancillary attributes can and do influence dem and for wild land recreation
facilities and wild land recreation facilities influence dem and for lodging,
restaurant, entertainment, and shopping establishments. However, there is
currently little explicit recognition of this in the recreation, recreation
economics, or tourism literature.

117
Additional research is needed to more closely identify variations in
ancillary attribute importance and the relationship of ancillary attributes to
recreation experiences in wild land settings. Specifically, the relationship
among rural recreation site attributes and management practices, recreation
motivations, and support community characteristics needs further
exploration. The results of the current study indicate that ancillary attributes
found in support communities are an important part of the recreation
product. Through the combination of key ancillary attributes, destination
products can be designed to serve selected segments of the cross-country ski
market. Packages may be developed to specifically serve high spending, low
spending, long stay, or short stay markets or markets with various
combinations of these characteristics. Motivations of cross-country skiers
may be reflected in desired characteristics of the ski trail and in the desired
support services. Skiers primarily motivated by fitness may desire trail
features which provide a strenuous workout with nearby support services
which include athletic training facilities. Skiers primarily motivated by
socializing may enjoy warm ing huts along the trail complete with
refreshments and communities with a variety of bars, night clubs, and family
entertainment. Skiers motivated by the opportunity to observe nature may
enjoy interpretive signs along the trail and naturalist talks at night.
Future research could be improved by including additional variables
and measures. As noted above, the inclusion of motivational variables in
future research would allow closer linking of ancillary and primary
attributes. Direct measures of primary attributes identified as im portant by
respondent groups would allow the identification of the ways in which
prim ary attributes offered in the trail setting differ according to managing

118
agency. The current study only measured differences in primary attributes
indirectly through the identification of the agency managing the trail where
respondents were contacted. Inclusion of income, time, and substitute
destination measures would allow further exploration of how ancillary and
primary attributes influence dem and for specific outdoor recreation activity
facilities and their supporting communities. Documentation of income and
time effects in the dem and models for such destination communities should
be undertaken to isolate the effects of ancillary attributes in generating
dem and.
However, McConnell (1985,) and Bockstael, McConnell, and Strand
(1991) note that in many site dem and functions, the income effect is
insignificant because higher incomes lead to the visitation of other sites and
lower incomes preclude visitation at all. Income may be more associated
with participation and non participation than the num ber of trips, or perhaps
in the case of the present study, length of stay (Bockstael, McConnell and
Strand 1991). This issue has not been addressed for cross-country skiing in a
small town destination community.
Additional research could compare slightly different combinations of
ancillary attributes present in nearby, similar communities. A study of the
variation in customer characteristics among these communities could explore
w hether or not small communities providing recreation experiences can be
substituted one for the other. If customers of each community have similar
motivations and socio-demographic characteristics, then the products and
experiences offered in and around the communities could be considered
substitutable. These communities together, as a group, may be thought of as a
single destination by their customers (Fridgen 1987). If small communities are

119
substitutable for one another, then a substitutability variable could be
included in valuation models for nearby wild land recreation facilities.
The findings of the current study could serve as the basis for additional
investigations. Ancillary attributes found to be of importance could be
combined into several differing descriptions of hypothetical destination
communities. Respondents would be asked to chose or rate these
descriptions according to their preference for visiting each hypothetical
community (Louviere and Timmermans 1990). Statistical analysis could then
determine the importance of each attribute in a number of selection decisions.
This approach would allow a more controlled manipulation of the attributes
and a comparison of product types without the respondent having had to
visit a num ber of actual destinations.
Many cross-country skiers who came to Munising during the winter of
1991-92 considered this destination an undiscovered gem (comments written
on questionnaire). M arketing efforts may change Munising from a destination
which provides a sense of discovery to a destination that seems crowded to
some cross-country skiers. Several skiers commented that in past seasons they
skied at several Wisconsin destinations but now prefer to come to Munising
to escape crowds or to escape increasing commercialization. A study of the
attributes currently available in the Munising area offers an opportunity to
document possible changes in visitor characteristics as Munising becomes a
more widely known destination. Changes in visitor characteristics may occur
as attributes change. The addition of a ski loge or an upscale hotel could
bring substantially different visitors to the Munising area. A change in visitor
characteristics coupled with a change in community attributes would give
additional credence to the idea that ancillary attributes are important. A

120
comparison of Munising visitors to cross-country ski resort customers may
forecast how customers characteristics may change should such an
establishment locate in the Munising area.
Munising skiers must travel varying distances to arrive at their
destination. Usually in travel cost dem and studies, all travelers are examined
in the same model. Recent economic literature has discussed and
implemented the separation of local participants from tourists in the
development of dem and models because these two groups face inherently
different decisions about recreation participation. The key determinant is
w hether the visitor spends the night away from home. This measure is
sometimes coupled with a standard travel distance of 100 miles.

Additional

analysis could examine at what distance local skiers begin to behave more like
tourists in their preference for or use of ancillary attributes. Perhaps tourist
cross-country skiers could be further separated into middle distance and long
distance travelers. Scott, Schewe, and Frederick (1978) in a study of tourists
to Massachusetts provided some evidence that travelers consider different
attributes in their travel decisions according to the distance their destination
is from their home. Of more practical importance is whether or not
differences in attribute importance warrant the development of separate
models for travelers of different distances. The identification of the break
points for distance groups could provide a relatively untapped means of
market segmentation and a means of targeting promotion and advertising
activities. Analysis examining relationships between distance zones and
ancillary attribute importance ratings using the Munising data suggested
such relationships may exist. As indicated in Table 12, ratings for mid price
lodging, swimming pool, sauna, and laundry, appear to differ across distance

121
zones. Mid price lodging, swimming pool, and sauna appear to be rated
higher in importance by visitors from within 100 miles and visitors from more
than 500 miles away. Laundry is more im portant to travelers from greater
than 500 miles away. Analysis of a larger data set with more respondents from
both shorter and longer distances may be necessary to explore this issue
further.
Destination decisions may be made based on any num ber of criteria
from trail quality to visiting someplace new.

These criteria may change

depending on the home locale, the distance traveled, previous skiing
experience, and the needs of the travel group in general. Specific reasons for
choosing specific destinations need to be recorded, perhaps by the subjects
themselves, at the time decisions are made. In depth interviews (Henderson
1991) would perhaps reveal different reasons for visiting different sites for
those skiers who travel to a variety of destinations during the cross-country
ski season. Cross-country skiers may not be entirely mindful of the reasons a
destination is selected and may later justify their destination selection to
reflect those reasons which have come to be accepted in their social group or
appear on a standardized questionnaire.
Information about the temporal aspects of decision making would also
be helpful. Trip decisions for a given period (a ski season for example) may
be made separately, on a trip by trip basis, or almost simultaneously with the
total num ber of trips for a given period (season) decided upon before the
season begins. If the latter case is true, then the probability of taking a
particular trip would be highly connected to the other trips. Constraints
upon trip length would not only involve the cost of one trip, but all the trips

122
planned for the given season. The time period for which the num ber of trips is
decided may be a week, a month, or several months.
Parallel to the issue of whether trip decisions are dependent on each
other is the question of w hether decisions about trip and trip length are made
together or at different times, Even if the trip length is decided upon at the
same time as the num ber of trips, the trip length decision can be modified if
conditions do not match expectations as in the case of rain and campers,
anglers who are not having much luck, and skiers who do not find good
snow. Modifications of trip length may also have an effect (increasing or
decreasing the num ber or increasing or decreasing the length) on the
remaining, as yet untaken, trips, assuming trip decisions for a period occur at
the same time. If trip decisions are made on a relatively separate trip by trip
basis, modification of one trip length may still have such effects. A traveler
may think that conditions were so bad they will never return, or, conversely, a
traveler may decide to return sooner to take advantage of better weather
conditions.
Future studies should include measures of motivation.

Motivation

measures would have been useful in this study to further explore
hypothetical differences in sociability between U.S. Forest Service and
National Park Service respondents. Motivation may influence the kinds of
ancillary attributes im portant to particular groups of skiers and the manner in
which available attributes are interpreted in the mind of the skier. Measures
of beliefs about quantities of attributes available would help with problems
caused by different interpretations of the way ancillary attributes are
described and provide a sounder basis for comparing visitor groups and
different destination communities. For example, skiers may differ in how

123
they define, low, mid, and high levels of lodging. One respondent may
consider a motel as a mid-range establishment while another respondent may
consider the same establishment to be low price.

Sum m ary
High quality cross-country ski trails are numerous in the Upper
Midwest. It has been the contention of this study that because high quality
trails are numerous, destination communities wishing to attract cross-country
skiers can do so by providing key ancillary attributes. Cross-country skiers
visiting Munising, Michigan rated the importance of ancillary attributes in
choosing a destination. Cross-country skiers tended to rate attributes
associated with shelter, food, and the main trip activity more highly than
other attributes. Cross-country skiers also differed in their importance ratings
of some attributes according to the agency which managed the trail where
they obtained their questionnaire. Importance ratings for some ancillary
attributes were significant in explaining the variance in the number of days
spent in Munising. The importance ratings for some other ancillary attributes
were significant in a regression analysis of willingness to pay trail fees. Trail
quality importance and trail satisfaction were not significantly related to
either length of stay or willingness to pay trail fees. But overall community
satisfaction was significant in a regression model examining willingness to
pay trail fees.
Future research should include measures of income and time as
constraints and a measure of substitute sites. Measures of motivation would
also be valuable in assessing the strengths of ancillary attribute preferences.
Similar and dissimilar cities and their visitors could be compared with each

124
other to further analyze the role of ancillary attributes in destination selection.
Measures of attribute quantity and quality from the point of view of the
cross-country skier would be of value in analyzing the importance of
ancillary attributes.
Demand for cross-country skiing at a specific location is at least
partially dependent on the presence of key ancillary facilities, as well as on
trail quality. The results of the current study indicate that the presence of
certain community attributes, which appear to be secondary or ancillary in
importance, can add to or subtract from the duration of visits and can attract
visitors willing to pay higher or lower trail use fees. Ancillary attributes may
be more im portant in examining length of stay and willingness to pay trail
fees than trail quality. High quality trails are the most im portant attribute a
cross-country ski destination can offer, but secondary attributes may
substantially affect the overall volume of business. A portion of the
valuation, consumer's surplus, usually credited to public recreation sites
perhaps results from key ancillary attributes of nearby supporting
communities. A portion of the consumer's surplus which has previously
been counted as a portion of the value of a publicly provided recreation site
perhaps should be counted as value contributed by services provided by
destination communities located near public recreation sites. Through joint
efforts, public and commercial organizations appear to be able to offer a
product superior to that which would be available from single sector
endeavors.

LITERATURE CITED

LITERATURE CITED
Ajzen, Icek and B. L. Driver. 1991. Prediction of Leisure Participation from
behavioral, normative, and control beliefs: An application of the Theory of
Planned Behavior. Leisure Sciences 13(3):185-204.
Allton, David J. and Stanley R. Lieber. 1983. Attributes of Chicago trail areas.
Arnould, Eric J.; Linda L. Price; Patrick Tierney. 1993. Managing hedonic
service encounters: The view from the river. Paper presented at the 1993
National Recreation and Parks Association Congress Research Symposium,
San Jose, California.
Ashworth, Gregory and Henk Voogd. 1990. Can places by sold for tourism? In
Tourism Places, ed. G. Ashworth ana B. Goodall. London:
Routledge.
Atkinson, John W. and Joel O. Raynor. 1974. M otivation and Achievement.
Washington, D.C.: V. H. Winston & Sons, Inc.
Ballman, Gary. 1980. Operationalizing the cross-countiy skiing opportunity
spectrum. N orth American Symposium on Dispersed Winter Recreation,
February 27-29, 1980, University of Minnesota, St. Paul, Minnesota,
Proceedings. Office of Special Programs — Educational Series 2-3,
Agricultural Extension Service, University of Minnesota.
Barlowe, Raleigh. 1986. Land Resource Economics: The Economics of Real
Estate, fourth edition. Englewood Cliffs, N.J.: Prentice-Hall.
Bell and Leeworthy. 1990. Recreational dem and by tourists for saltwater beach
Bishop, Glen R.; Peter K. Forsberg; and Donald F. Holecek. 1992. 1991-1992
Munising Area Nordic Ski Survey. East Lansing, Michigan: Michigan Travel
Tourism, and Recreation REsource Center, Michigan State University.
Bitner, Mary Jo. 1992. Servicescapes: The impact of physical surroundings on
customers and employees. Journal of Marketing 56(April 1992):57-71.
Blank, Uel and Clare A. Gunn. 1965. Guidelines for Tourism-Recreation in
Michigan's U pper Peninsula. Cooperative Extension Service. East Lansing:
Michigan State University.

125

Bockstael, Nancy E., W. Michael Hanemann, and Catherine L. Kling. 1987.
Estimating the value of water quality improvements in a recreational demand
fram ew ork Water Resources Research 23(5):951-960.
Bockstael, Nancy E., Kenneth E. McConnell, and Ivar Strand. 1991. Recreation.
Chapter 8 in Measuring_the Dem and for Environmental Quality, eds. John B.
Braden & Charles D. Kolstad. Amsterdam: Esevier Science Publishers B. V.
(North-Holland).
Bockstael, Nancy E., Ivar E. Strand and W. Michael Hanemann. 1992. Time
and the recreational dem and model, Chapter 13. In Environm ental
Economics: A Reader, ed. Anil M arkandya and Julie Richardson. New York:
St M artin's Press.
Bonn, Mark A., H. Leslie Furr, and Muzaffer Uysal. 1992. Seasonal variation of
coastal resort visitors: Hilton Head Island. Journal of Travel Research 31(1):5056.
Brayley, Russell E. 1991. Beer drinking smokers and granola crunchers: The
challenge of physically and socially accommodating incompatible winter
resort market segments. Paper presented at the 1991 RCRA Research
Symposium, November 1-3, 1991, Bend, Oregon.
Bronner, Fred and Robert de Hoog. 1985. A recipe for mixing decision
ingredients. European Research 13(3):109-115.
Brown, G ardner Jr. and Robet Mendelsohn. 1984. The hedonic travel cost
Bryan, Hobson. 1977. Leisure value systems and recreational specialization:
The case of trout fishermen. Journal of Leisure Research 9(3):174-187.
Caulkins, Peter P.; Richard C. Bishop; and Nicolas W. Bouwes, Sr. 1986. The
travel cost model for lake recreation: A comparison of two methods for
incorporating site quality and substitution effects. American Journal of
‘ ■ ’
' : 68(2):291-297.
Chase, Terresa. 1991. Letter to Peter Forsberg, Travel, Tourism, and Recreation
Resource Center, Michigan State University, East Lansing, Michigan.
Cicchetti, C. J., A.C. Fisher, and V. K. Smith. 1976. An econometric evaluation
of a generalized consumer surplus measure: The Mineral King controversy.
F.conometrica 44(Nov.):1259-1276.
Clark, Roger N. and George H. Stankey. 1979. The Recreation O pportunity
™
, Management, and Research. General
orest Service, Pacific Northwest Forest
and Range Experiment Station.

P

127
Crask, Melvin R. 1981. Segmenting the vacationer market: Identifying the
vacation preferences, demographics, and magazine readership of each group.
Journal of Travel Research 20(Fall):29-34.
Crawford, Duane W. and Geoffrey Godbey. 1987. Reconceptualizing barriers
to family leisure. Leisur£_Scieiices 9(2):119-127.
Crompton, John L., Paul C. Fakeye, and Chi-Chuan Lue. Positioning: The
example of the lower Rio Grande Valley in the winter long stay destination
Dadgostar, Bahram, and Robert M. Isotalo. 1992. Factors affecting time spent
by near-home tourists in city destinations. Journal of Travel Research 31(2):3439.
Dann, G. 1977. Anomie, ego-enhancement and tourism. Annals of Tourism
Research 4:184-94.
Davis, Bonnie D. and Brenda Sternquist. 1987. Appealing to the elusive
tourist: an attribute cluster strategy. Journal of Travel Research 25(4):25-31.
Deyak, T. A., and V. K. Smith. 1978. Congestion and participation in outdoor
recreation: A household production function approach. Journal of
~ '
' ' "
' ' '
5(March):63-80.
Dimanche, Havitz, and How ard. 1991. Testing the involvement profile (IP)
scale in the context of select recreational touristic activities. Journal of Leisure
Donnelly, Dennis M., Cindy Sorg-Swanson, John B. Loomis, and Louis J.
Nelson. 1990. Net economic value of hunting and fishing in Idaho. In Social
Science and Natural Resource Recreation M anagement, ed. Joanne Vining.
Boulder, Colorado: Westview Press.
Dorfman, Peter W. 1979. M easurement and meaning of recreation satisfaction:
A case study in camping. Environment and Behavior 11(4):483-510.
Doswell, Roger and Paul R. Gamble. 1981.
~
~
London: Hutchinson.
Driver, B. L., Periy J. Brown, George H. Stankey, and Timothy G. Gregoire.
1987. The ROS planning system: Evolution, Basic concepts, and research
needed. Leisure Sciences 9:201-212.
Echelberger, Herbert E. and Elwood L. Shafer, Jr. 1970. Snow + (X) = Use of
ski slopes. Journal of Marketing Research 7(August):388-392.
Eisner, Gary H. 1971. A regression method for estimating the level of use and
market area of a proposed larg e ski resort. ’
3(3):160-167.

128
Englin, Jeffrey and Robert Mendelsohn. 1991. A hedonic travel cost analysis
for valuation of multiple components of site quality: The recreation value of
forest management. Journal of Environmental Economics and Management
21:275-290.
Fesenmaier, Daniel R. 1985. Modeling variation in destination patronage for
outdoor recreation activity. Journal oLT ravel Research 24(2): 17-23.
Fletcher, Jerald J., Viktor L. Adamowicz, and Theodore Graham-Tomasi. 1990.
The travel cost model of recreation demand: Theoretical and empirical
issues. Leisure Sciences 12:119-147.
Fridgen, Joseph D. 1984. Environmental Psychology and tourism. Annals of
Fridgen, J. 1987. Use of cognitive maps to determine perceived tourism
regions. Leisure Sciences 9:101-117.
Getz, Donald. 1993. Tourist shopping villages - development and planning
strategies. Tourism M anagement 14(l):15-26.
Goodall, Brian. 1991. Chapter 4, Opportunity sets as analytical, marketing
instruments: A destination area view. In Marketing Tourism Places, Gregory
Ashworth and Brian Goodall, editors. London: Routledge.
Goodrich, Jonathan N. 1978. The relationship between preferences for and
perceptions of vacation destinations: Application of a choice model. Journal
of Travel Research 17(2):8-13.
Gunn, Clare A. 1994.
Francis.

Third Edition. New York: Taylor &

Hawkins, Del I., Roger J. Best, and Kenneth A. Coney.
Coney. 1992.
Behavior: Imp.................
Implications for Marketing Strategy, fifth edition. Homewood,
Illinois: Irwin.
Heberlein, T. A. 1977. Density, crowding and satisfaction. In
Service General Technical Report NC-28.
Henderson, Karla A. 1991.
l. State College, PA: Venture
Publishing.
Herrington, R. B. 1967.
Ogden, Utah: Intermountain"Forest and Range Experiment Station, U. S.
Forest Service Research Paper INT-34. 90p.
Hodgson, Peter. 1983. Research into the complex nature of the holiday choice
rocess. In Proceedings of the Seminar on the Importance of Research in the
ourist Industry, Helsinki, ESOMAR, Amsterdam, 1983, June 8-11:17-35.

¥

129
Hof, John G. and David A. King. 1992. Recreational demand by tourists for
saltwater beach days: Comment. Journal
M anagem ent 22:281-291.
Hogarth, Robin M. 1987. Judgement and Choice:
New York: John Wiley & Sons.
Hollenhorst, Steve. 1990. What makes a recreation specialist? The case of
rockclimbing, chapter 6. In Social Science and N atural Resource Recreation
M anagement, ed. Joanne Vining. Boulder, Colorado: Westview Press.
Jackson, Edgar L. 1993. Recognizing patterns of leisure constraints: Results
from alternative analyses. Journal of Leisure Research 25(2):129-149.
Jackson, Edgar L.; Duane W. Crawford; and Geoffrey Godbey. 1993.
Negotiation of leisure constraints. Leisure Sciences 15(l):l-lf.
Jackson, Edgar L. and Elaine Dunn. 1991. Is constrained leisure an internally
homogeneous concept? Leisure Sciences 13:167-184.
Johnston, Warren E. and Gary H. Eisner. 1972. Variability in use among ski
areas: A statistical study of the California market reg io n ./
Research 4(Winter):43-49.
Kaltenborn, Bjorn P. and Lars Emmelin. 1993. Tourism in the high north:
M anagement challenges and recreation opportunity spectrum planning in
Svalbard, Norway. Environmental M anagement 17 ( 1 ):41 -59.
Kelly, John R. and Geoffrey Godbey. 1992. Sociology of Leisure. State College,
Pennsylvania: Venture Publishing, Inc.
Kretchman, Janice A. and Paul Eagles. 1990. An analysis of the motives of
ecotourists in comparison to the general Canadian population. Loisir et
Lancaster, Kelvin J. 1966. A new approach to consumer theory. Journal of
......... ....
.74:132-157.
Lawler, Edw ard E., III. 1973.
California: Brooks/Cole Publishing Company.

Monterey,

Levy, Joseph. 1979. A paradigm for conceptualizing leisure behavior:
Towards a person-environment interaction analysis.
' ll(l):48-60.
Lime, David W. 1971.

use in
i. USDA, Forest Service, Research Paper NC-60.

130
Louviere, Jordan and Harry Timmermans. 1990. Preference analysis,choice
modelling and dem and forecasting. In Proceedings of the National Outdoor
Recreation Trends Symposium III, Indianapolis,Indiana, ed. Joseph T.
O'Leary, Daniel R. Fesenmaier, Tommy Brown, Dan Stynes and Bev Driver.
Louviere, Jordan J. and Harry J. P. Timmermans. 1992. Testing the external
validity of hierarchical conjoint analysis models of recreational destination
choice. Leisure Sciences 14(3):179-194.
Love, Timothy G. and Alan E. Watson. 1992. Effects of the Gates Park Fire on
Recreation Choices. U.S.D.A.,Forest Service. Imternountain Research Station,
Research Note INT-402.
Lucas, Robert C. 1980.
references in nine wilderness and other roadless areas. USDA, Forest
ervice, Intermountain Forest and Range Experiment Station Research Paper
INT-253.

g

Service, Intermountain Researclh^tation, Research Paper INT-345.
Mak, James and James E. T. Moncur. The choice of jc
lengths of stay: A micro analysis. The Revk

destinations and
Studies 10(2):38-48.

Manfredo, Michael J. 1989. A test of assumptions inherent in attribute-specific
Mayo, Edward J., Jr. and Lance P. Jarvis. 1981. The PsychologyuiLLeisure
T ravel. Boston: CBI Publishing Company, Inc.
McConnell, K. E. 1977. Congestion and willingness to pay: A study of beach
use. Land Economics 53(2):185-195.
McConnell, K. E. 1985. The economics of outdoor recreation, Chapter 15. In
Kneese and James L. Sweeney. Amsterdam: North-FIolland.
McConnell, K. E. 1992. On-site time in the dem and for recreation. A m erican
McConnell, Kenneth E.; Ivar E. Strand, and Nancy E. Bockstael. 1990. Habit
formation and the dem and for recreation: Issues and a case study. In
Advances in Micro-F.conomics, Volume 5:217-235. Greenwich, Connecticut:
JAI Press Inc.
McCool, Stephen F. 1978. Recreation activity Packages at water-based
resources. Leisure Sciences 1(2):163-173.

131
McCool, Stephen F.; George H.Stankey; and Roger N. Clark. 1985. Choosing
recreation settings: processes, findings and research directions. In
Proceedings-symposium on recreation choice behavior; 1984 March 22-23;
Missoula, MT. Gen. Tech. Rep. INT-184. Ogden, UT: U.S. Department of
Agriculture, Forest Service, Intermountain Research Station: 1-8.
McLaughlin and W. E. J. Paradice. 1980. Using visitor preference information
to guide dispersed winter recreation m anagem ent for cross-country skiing
ana snowmobiling. North American Symposium on Dispersed Winter
Recreation, February 27-29, 1980, University of Minnesota, St. Paul,
Minnesota, Proceedings. Office of Special Programs — Educational Series 2-3,
Agricultural Extension Service, University of Minnesota.
Mendelsohn, Robert. 1987. Modeling the dem and for outdoor recreation.
Mendenhall, William. 1979. Introduction to Probability and Statistics, Fifth
Edition. North Scituate, Massachusetts: Duxbury Press.
Morey, Edward R. 1981. The dem and for site-specific recreational activities: A
characteristics approach. Journal of Environmental Economics and
; 8:345-371.
Nelson, Charles M. 1988. Michigan State Forest Pathway Cross Country
Skiers- 1988. East Lansing, Michigan: Departm ent of Park and Recreation
Resources, Michigan State University.
Noe, Francis P. 1987. M easurement specification and leisure satisfaction.
Leisure Sciences 9:163-172.
Service Business, 2nd. ed. C h ich ester/^ est Sussex,
Sons Ltd.

E n g lan d :

John Wiley &

Parsons, George R. and Maiy Jo Kealy. 1992. Randomly draw n opportunity
sets in a random utility model of lake recreation. Land Economics 68(11:93106.
Parsons, George R. and Michael S. Needelman. 1992. Site aggregation in a
random utility model of recreation. I .and Economics 68(4):418-33.
Pearce, Philip L. 1982. The Social Psychology of TourisLBehayiour.
International Series in Experimental Social Psychology, Vol. 3. New York:
Pergamon Press.
Pearce, Philip L. and Marie L. Caltabiano. 1983. Inferring travel motivation
from travelers' experiences. Journal of Travel Research 22(2): 16-20.
Peterson, George L.; John F. Dwyer; and Alexander J. Darragh. 1983. A
behavioral urban recreation site choice model. Leisure Sciences 6(1):61-81.

132
Pizam, Abraham. 1978. Dimentions of tourist satisfaction with a destination
Poon A. 1989. Competitive strategies for a 'new tourism', Chapter 5. In
C M ^C ooper. London: Belhaven Press.
Rao, S. R., Edward G. Thomas, and Rajshekhar G. Javalgi. 1992. Activity
preferences and trip-planning behavior of the U. S. outbound pleasure travel
market. Journal of Travel Research 30(3):3-12.
Rauhauser, Donna Scott. 1979. Cross-country skier preferences for the
recreational setting. Unpublished Master of Arts Thesis.
The
Boulder: Department
of Geography, University of Colorado.
Raymore, Leslie; Geoffrey Godbey; Duane Crawford; and Alexander von Eye.
1993. Nature and process of leisure constraints: An empirical test. Leisure
15(3):99
“‘ 113.
Ritchie, J. R. and Michael Zins. 1978. Culture as determinant of the
attractiveness of a tourism region. Annals of Tourism Research 5(2):252-267.
Roehl, Wesley S., Julie Fesenmaier, and Daniel R. Fesenmaier. 1993. Highway
accessibility and regional tourist expenditures. Journal of Travel Research
31(3):58-63.
Roggenbuck, J.j W., D. R. Williams, and A. E. Watson. 1993. Defining
acceptable conditions in wilderness. Environmental M anagement 1/(2):187197.
Rollins, Richard B. 1985. Measuring recreation satisfaction within a national
ark setting: The west coast trail area of Pacific Rim National Park. Ph.D.
'issertation. Seattle: University of Washington.

E

Rollins, Rick and Doug Chambers. 1990. Camper satisfaction with Canadian
Park Service campgrounds, chapter 7. In Social Science and Natural Resouri
Recreation M anagement, ed. Joanne Vining. Boulder, Colorado: Westview
Press.
Rosenthal, Donald H., B. L. Driver, and Donna Rauhauser. 1980. Skiing
environments preferred by Colorado ski-tourers. North American
Symposium on Dispersed Winter Recreation, February 27-29, 1980, University
of Minnesota, St. Paul, Minnesota, Proceedings. Office of Special Programs —
Educational Series 2-3, Agricultural Extension Service, University of
Minnesota.
Rugg, Donald. 1973. The choice of journey destination: A theoretical and
empirical analysis. Review of Economics and Statistics 50:64-72.
Saleh, Farouk and Chris Ryan. 1992. Client perceptions of hotels. T ourism
M anagem ent: 163-168.

Scott, Douglas R., Charles D. Schere, and Donald G. Frederick. 1978. A multibrand/m ulti-attribute model of tourist state choice.
Research 17(l):23-29.
Shaw, W. Douglas. 1991. Recreational dem and by tourists for saltwater beach
days: Comment. '
' M anagement
20:284-289.
Shelby, Bo. 1980. Crowding models for backcountry recreation. Land
Shih, David. 1986. VALS as a tool of tourism market research: The
Pennsylvania experience. Journal of Travel Research 24(4):2-11.
Silberman, Jonathan. 1985. A dem and function for length of stay: The evidence
from Virginia Beach. Journal of Travel Research 23(4):16-23.
Smith, Charles K. 1980. Ski touring in the Twin Cities area. North American
Symposium on Dispersed Winter Recreation, February 27-29, 1980, University
of Minnesota, St. Paul, Minnesota, Proceedings. Office of Special Programs —
Educational Series 2-3, Agricultural Extension Service, University of
Minnesota.
Smith, V. K. and R. J. Kopp. 1980. The spatial limits of the travel cost
recreational demand model. T.and Economics 56:64-72.
SYSTAT, Inc. 1992. SYSTAT: Statistics. Version 5.2 Edition. Evanston, IL:
SYSTAT, Inc.
Taylor, Gordon D. 1986. M ulti-dimensional segmentation of the Canadian
pleasure travel market. Tourism M anagem ent7(3):146-153.
Teare, Richard. 1990. An exploration of the consumer decision process for
hospitality services, Chapter 15. In M anaging and Marketing Services in the
1990s. Edited by Richard Teare, Luiz Moutinho, and Neil Morgan. London:
Cassell Educational Limited.
Teye, Victor B. 1989. Marketing an emerging international tourist destination:
The case of Arizona. Journal of Travel Research 27(4):23-28.
Um, Seoho and John L. Crompton. 1991. Development of pleasure travel
attitude dimensions. Annals of Tourism Research 18(3’):500-504.
Uysal, Musaffer, Cary D. McDonald, and Joseph T. O'Leary. 1988. Length of
stay: A macro analysis for cross-country skiing trips. Journal of Travel
Research 26(Winter):29-31.

134
Virden, Randy J. and Richard C. Knopf. 1989. Activities, experiences, and
environmental settings: A case study of recreation opportunity spectrum
relationships. Leisure Sciences 11(3):159-176.
and Costs State College, Pennsylvania: Venture Publishing, Inc.
Walsh, Richard G. and Gordon J. Davitt. 1983. A dem and function for length
of stay on ski trips to Aspen. Journal of Travel Research 21 (4):23-29.
Wanhill, Stephen. 1990. Tourism statistics. Chapter 12 in Managing and
Marketing Services in the 1990s. ed. Richard Teare, Luiz Moutinho, and Neil
Morgan. London: Cassell Educational Limited.
Watson, Alan E.; Daniel R. Williams; Joseph W. Roggenbuck; and John J.
Daigle. 1992. Visitor characteristics and preferences for three national forest
wilderness in the south. USDA, Forest Service, Intermountain Research
Station. Research Paper INT-455.
Wilkinson, Leland. 1989.
SYSTAT, Inc.

3. Evanston, IL:

Witt, Stephen F. and Christine A. Witt. 1992.
Demand in Tourism . London: Academic Press, Harcourt Brace Jovanovich,
Publishers.
Witt, Christine A. and Peter L. Wright. 1992. Chapter 3, Tourist motivation:
Life after Maslow. In Choice and Demand in Tourism, Peter Johnson and
Barry Thomas, editors. London: Mansell Publishing Limited.
Woodside, Arch G. 1982. Positioning a province using travel research. Journal
Woodside, Arch G. and Jefrey A. Carr. 1988. Consumer decision making and
competitive m arketing strategies applications for tourism planning.Journal
Woodside, Arch G. and Dan Sherrell. 1977. Traveler evoked, inept, and inert
sets of vacation destinations. Journal of Travel Research 16(Winter):14-18.

APPENDIX:

QUESTIONNAIRES

135

D ear Skier:
In an effort to better m anage cross-country ski trails in the M unising area, the U.S. Forest Service, in cooperation
w ith the Travel, T ourism , and Recreation R esource C enter at M ichigan State U niversity (TTRRC), is asking a
select sam ple of skiers to complete this questionnaire. Your answ ers to the questions w ill be kept confidential
and w ill only be reported in total along w ith o th e r respondents to th e survey. W hen you have com pleted the
questionnaire, please either m ail it to u s using th e postage p aid envelope or drop it off at the U. S. Forest
Service office, 400 E. M unising Street, M unising.
The U.S. Forest Service w ill use the results to b e tter provide q uality w in ter recreation opportunities. Your
cooperation is vital to the success of this project as on ly a lim ited nu m b er of people can be contacted for their
opinions. Please take a few m inutes som etim e today to complete this questionnaire and return it.
Sincerely,

CX*JU
Peter Forsberg, T.T.R.R.C.

D ick A nderson, U.S.F.S.

P le ase check the trail at which this questionnaire w a s h a n d ed out:
O H iaw atha O V alley Spur
□ C hristm as
□ M cK eever Hills □

1.

P le a s e check th e type

2.

P le ase circle the am ount you paid for this p a ss.
$0 $3.00 $5.0 0 $10.00
$20.00 $30.00 $35.00 $50.00
W hat w a s th e d a te th e p a s s w a s p u rc h a se d ? __________________

3.

T he U.S. F orest Service o p e ra te s five cross-country ski trails in th e M unising a rea. B e c a u se of high
m aintenance c o sts and lack of u se , th e F orest S ervice in the future m ay h ave to cut back. P lease ch eck the
option below you would prefer.

□

4.

C losing trails

of p a s s

□

w hich

you p u rc h a se d .

Done

R educing groom ing quality

Oannual

B uckhorn

day

□

R aising fe e s

Qnone

Q S to p Groom ing

The U.S. Forest Service is considering closing so m e trails b e c a u se of lack of u se and th e high c o sts of trail
m aintenance. O ne m ea n s of offsetling the c o s ts of m aintaining trails would b e to ch arg e higher tee s. How
m uch would you b e willing to p ay? If the am ount is not high enough, the Forest Service would likely close this
and the o ther trails it o p e rates. W hat is the h ig h e st to ta l fee you w o u ld p a y p e r p e rs o n fo r a day o l skiing at
th is tr a il in o rd e r to keep using th is tr a il a t its current level o f gro o m in g ? Circle the d o lla r am ount below that
you w o u ld p a y to keep using th is t r a il.

$ 0 .0 0
5.

$3.00

$4.00

$5.00

$6.00

$7.00

$8.00

$9.00

$10.00

$11.00

$12,00

$13.00

W hat is the highest total fee you w ould b e willing to pay p e r p e r s o n p e r day to u se th is trail for a day of
skiing if it w ere u n - g r o o m e d ?
$ 0 .0 0

6.

$1.00

$1.00

$3.00

$4.00

$5.00

$6.00

$7.00

$8.00

$9.00

$10.00

$11.00

$12.00

$13.00

W ould you con tin u e to u s e Forest S erv ice trails if th e y w ere no lon ger g ro o m ed and th ere w ere no fe e s ?
(ch eck o n e)
O Y es
Q no

1

136

7.

If th e F orest S ervice d e cid ed to c lo se so m e of its cross-country ski trails, w hich one trail should b e kept
open.
□ H ia w a th a

8.

9.

O v a lle y S pur

□ C h ris tm a s

O M cK eever Hills

□ B u ck h o m

T he U.S. F orest S en/ice would like to attract m ore cross-country skiers to th e Munising
area. For eac h set
of possible actions, ch eck th e o n e in e a c h se t which you feel is the m ost attractive. (One X per line)
a.

Add more trails

Add more loops to current trails

b.

Offer special ski packages

c.

Add lodge to lodge trail

d.

Provide warming huts

e.

Provide trails which require more skill

f.

Increase signs on trails

g.

Increase trail variety

Add more skating loops

Sponsor cross-country ski festival
Develop a ski lodge

Provide direct access

Provide rest shelters along trails
P ro v id e

Offer ski lessons

easier trails

Increase signs on highway
Provide better quality grooming

to tra il from lodging

Provide outhouses

along trail

Groom loops for skating
Increase

p ro m o tio n and advertising

Provide side

by

side

g room ing

W hich letter from a b o v e is the m ost attractive action?_____

10. T here a re m any re a so n s p eople do not ski a s often a s they w ould like to. If you strongly a g ree with one of the
re a s o n s below w hich m ay re d u c e th e n um ber of tim es p eo p le c ro ss-country ski in th e M u n isin g a re a ,
check the blank clo sest to a g ree . If you strongly disagree m ark the blank closest to disagree. If you feel
so m ew h ere in b etw een , m ark th e blank betw een agree and d isa g re e which m ost closely reflects how you feel
about the re aso n . If you neither a g re e nor disagree, mark th e middle blank.
It takes too much time to travel to Munising
. . . . disaqree
Concern about snow on the ro a d s ...................... .........disagree
Concern about being unable to return
home on time because of sn o w ...........
Fees charged at Forest Service ski areas
are too expensive...................................
Munising is too expensive..................................... .........disaoree
Obligations at home or w ork................................ . .. . disaoree
Members of family or group would
rather go elsew here............................... . . . . disagree
Parking lots are not plow ed................................... . . . . disaqree
Concern about medical c a r e ................................. . . . . disaoree
Concern about vehicle burglary........................... .........disaqree
I don't feel welcome in the community............... .........disaqree
Lack of com panions................................................ . . . .disagree

____
__________. agree
____ ____________________agree
____________________agree
____________________agree
____________________agree
aqree

Concern about becoming lo st............................... . . . . disaoree
Poor personal physical condition......................... . . . . disaoree
Too many snowmobiles around Munising...................... disagree____________________________________
Not enough to do in Munising........................................... disagree
Lack of ski re n ta l.............................................................. disagree____________________________________

2

aaree
agree

aaree
aqree
aaree
aqree
aaree
aqree
agree
agree
agree

137

11. C heck e a c h of th e five U.S. F orest S ervice cross-country ski trails in th e M unising a re a that you skied
be fo re N ovem ber 1991.
□ H ia w a th a

Q v a lle y S pur

□ C h ris tm a s

OM cK eever Hills

Q B u ck h o m

12. How m any tim es did you cross-country ski last s e a s o n (N ovem ber 1990 through M arch 1 9 9 1 )? ______
13. How m any overnight trips during which you cross-country skied did you m ake during th e p a st s e a s o n
(N ovem ber 1990 through M arch 1 9 9 1 )? ______
14. How m any trips do you expect to m ake to cross-country ski in th e Munising a re a this s e a s o n (N ovem ber 1991
through M arch 1 9 9 2 )? ______
15. O ther than Munising, w hat o n e com m unity would you visit to cross-country ski? C heck none if applicable.
□
16.

P le a s e circle your skill level: 1=beginner, 5= expert
be ginner
1

17.

N one

2

interm ediate
3
4

e x p e rt
5

W hat n e w sp a p e r do you re ad m ost frequently?

18. Will you sp e n d the night aw ay from hom e while on this trip?

Q

no

Q Y es

19. In the Munising a rea , w here a re you staying?
□ y o u r second home

Q your own home

□ friend's home or friend's second home

□ relative's home or relative's second home

□ motel/hotel

20.

Is the place you are staying within 15 m iles of Munising?

21.

W hat is the main reason you a re in the Munising a rea ?
□ liv e here

22.

□visiting friends or relatives

□ recreation

O no

Q ybs

□ b u sin ess

□ o th e r

P le ase circle e a c h winter recreation activity in which you participate.
snowmobiling

ice fishing

ice climbing

camping

sledding

downhill skiing

snowshoeing

cross-country skiing

23. P le a s e circle th e m ain activity in which you will participate

hunting

qq

Other_____
Other

th is t r ip .

snowmobiling

ice fishing

ice climbing

camping

sledding

downhill skiing

snowshoeing

cross-country skiing

3

ice skating

hunting

ice skating
Other

138

Low price m otel........................................Not important
^

S T

,

! .................................................. N ot lm P ° n a n t

^

re m e V .i m por,ant

Ski lodge or ski re so rt............................. Not jmporlant------------------------------- ------------------------Extremely important
-------------------------------Extremely important
^ h n d ^ r Cr,h0n <S'nk a n d m i0 ro w a v e )' • ' N°> important
imoortant-----------------Xhlld c a r e .................................................. Not
Swimming pool.............................................. importam--------------------------------

c „.

, .

Extremely important
Extremely important

w ?-"I!/ V ‘
.......................................... Not important------------------------------- ----------------------- Extremely important
ventilated waxing room ......................
Not importam-------------------------------------------------------- Extremely important
Laundry......................................................Not important----------------------------------------------------------Extremely important
Extremely important
Fast food . . . .
., ..
_
.......................................... Not important
..
.
Family restaurant...................................... Not lmportant
™staPfant...................................Not important
Night C lub...................................................... imp0[tant
r-amily oriented entertainm ent................. Not important
Cultural attraction...................................... No! important
a r .............................................................. Not important
c

Extremely important
Extremely imponant
Extremely important
---- ‘ ----------------------------------------- Extremely important
Extremely imponant
~ "
'— Extremely imponant
Extremely important

irttportant_
Clnthh0PSh ................................................ ^ ot
Clothing s h o p s .............................................. i m p o r t a n t --------------------------------------Extremely important
, S, 0 p s .....................................................Not important
'--------------------------------------- Extremely important
Local art and craft s h o p s .......................... Not i m p o r t a n t ------------------------------------------------ Extremely important
Extremely important
Quality of cross-country ski trails
Not important
Overall community.................................... ... jmportant-------------------------------------------------------Extremely important
Extremely important

2 5 ' th e b l S S ^ f l e ^ s h o w you*,e l f " 9 ^

~ ^

,RneriS ,aUran,s.................................................No, satisfied
kpd9ln.9 ..................................................... No,satisfied
Shopping
Notsatisfied------------------------------------------------- n----Quality of cross-country ski trails
Not satisfied
Overall community................................... ; No,satlsfled

26. How m any d a y s will you ski during this trip? .
su rv ey ? _______

_

Place a

c h e c k mark in

Extremely satisfied
Extremely satisfied
Extremely satisfied
Extremely satisfied
Extremely satisfied

How m any hours did you ski the d a y you picked up this

27. How m any d a y s will you b e within 15 m iles of Munising on this trip? _
28. How m any nights will you b e within 15 m iles of Munising o n this trip?

4

139

29. W e would also like to know how much yo u a n d y o u r p arty spent in the Upper Peninsula and in the
Munising area on this trip from 12:01 a .m . until m id nig ht on the day you received this questionnaire.
Rem em ber to include c a s h , check, and c re d it card p u rc h a s e s for the following items. If you did not
buy an item during the 24 hour period, please record a zero in the blank. For purposes of this study, the
Munising area is within 15 miles of Munising and includes the towns of Munising, Christmas, Au train,
Shingleton, and Chatham . Do n o t In c lu d e M u n ls ln a a re a p u rc h a s e s In th e a m o u n ts re c o rd e d

t P i l h e U p p e r P e n in s u la .
U pper
P e n in s u la

M u nisin g
A rea

G rocery food & b e v e ra g e s

$

$

R e stau ra n t food a n d b e v e ra a e s

$

$

Vehicle related item s (aas. etc.I

$

$

L odging

$

$

Ski equipm ent

$

$

Clothing

$

$

All o th er item s

$

$

30. How m any Deoole. includina vourself. a re vou reDortina exD enditures tor?
31. How m any people, includina vourself. are travelina with vou?
32. Are the people in your party : O F am ily Q F r ie n d s

^ R e la tiv e s

□ club or organization

33. How m any d a y s h ave vou b e en awav from hom e on vour trip?
34. Approxim ately how m any m iles is Munising from w here you live?

Miles

35. C heck the m ost im portant source from which you learned about skiing opportunities in the Munising area.
Q F riends or relatives

Q T ravel brochure

QM agazine/newspaper article

Q c h am b e r of Commerce

Q p o re st Service employee

Q F orest Service publication

QHotel/Motel employee

Q

d NR

QDNR publication

Q P revious Trip

□

aaa

QMichigan Travel Bureau

QMichigan Welcome Center

Ql_ocal ski shop

QOut-of-town ski shop

employee

QlJpper Peninsula Travel and
Recreation Association

5

Q o th e r

140

36.

W hat is th e highest level of education you h a v e co m pleted?
1 2 3 4 5 6 7 8
elem en tary

37.

[Circle y e a rs com pleted]

13 14 15 16
college

9 10 11 12
high school

17 18 19 20
graduate school

W here is y our p erm anent re sid en c e?
City:.

County:.

S tate:.

Zip code:.

38. M edian h o usehold incom e in M ichigan is $ 3 1 ,6 4 5 . In 1990, w a s your total h o usehold incom e
taxes: (C heck One)
O

M o re

Q

L ess

□

be fo re

A pproxim ately th e sa m e

3 9 . How old a re you?

T h a n k y o u fo r y o u r a s s i s t a n c e . P le a s e d ro p off th e co m p leted q u estio n n aire at th e F o rest Service Office at
400 E. Munising S treet, M unising, or mail th e c om pleted q uestionnaire in th e p o sta g e-p a id envelope provided. If
th e envelope is m issing, p le a se se n d th e co m p le te d questio n n aire to:
Dr. P e ter F orsberg
Travel, Tourism , a n d R ecreation R e so u rc e C e n te r
172 N atural R e so u rc e s Building
M ichigan S ta te University
E ast L ansing, Ml 48824-1222
P le a se write in the sp a c e below any c o m m en ts you would like to m ake regarding cross-country skiing in the
Munising area.

6

141
Dear Skier;
The Travel, Tourism , an d Recreation Resource Center (TTRRC) at M ichigan State University is conducting a
study to assist public agencies in better managem ent of M unising area ski trails. The Center is asking a select
sam ple of skiers to complete this questionnaire. Your answers to the questions w ill be kept confidential and
w ill only be reported in total along w ith other respondents to the survey. W hen you have completed the
questionnaire, please eith er m ail it to us using the postage paid envelope or drop it off at the joint Park
Service/Forest Service visitor center in M unising.
Public agencies w ill use the results to better provide quality w inter recreation opportunities. Your cooperation
is vital to the success of this project as only a limited num ber of people can be contacted for their opinions.
Please take a few m inutes som etim e today to complete this questionnaire and return it.
Sincerely,

Glen Bishop, TTRRC

1.

If in the future the National Park Sen/ice h a s to cut back its cross-country skiing program b eca u se of high
m ain ten an ce c o sts, which of the options below would you would prefer.

O C losing trails

2.

Q R educing groom ing quality

Q C harging fe e s

Q S top G room ing

O ne m e a n s of offsetting the c o sts of maintaining trails would b e to charge higher fees. How m uch would you
be willing to p ay? If th e am ount is not high enough, this trail may be closed. What is the highest total tee you
w ould p a y p e r p e rs o n fo r a d a y o f skiing a t th is tr a il in o rd e r to keep using th is tr a il a t its current level o f
gro o m in g ? Circle the d o lla r a m o u n t below that you would p a y to keep using th is tr a il.

$ 0 .0 0
3.

$3.00

$4.00

$5.00

$6.00 $7.00 $8.00

$9.00 $10.00

$11.00

W hat is th e highest total fee you would be willing to pay p e r p e rs o n per day to
skiing if it w e re u n - g r o o m e d ?
$ 0 .0 0

4.

$1.00

$1.00

$3.00

$4.00

$5.00

$6.00

$7.00 $8.00

W ould you continue to u se National Park Service trails
fe es?
(check one)
□ Y es

$9.00 $10.00

if they w ere no
O

$11.00

$12.00

$13.00

u se th is trail for a day of

$12.00

$13.00

longer groom ed and there w ere no

no

5. C heck e a c h cross-country ski trail that you skied before N ovem ber 1991.
Q Munising □ G rand Marias

□ School Forest

□ M cK eever Hills

□ B uckhorn

1

Q H iaw atha

Q v alley Spur

Q c h ristm a s

142

6.

Som e people would like to attract m ore cross-country skiers to th e Munising a rea . For e a c h se t of possible
actions, check th e o n e in e a c h set which you feel is th e m ost attractive. (One X p e r line)
a.

Add more trails

b.

Offer special ski packages

c.

Provide warming huts

d.

Provide trails which require more skill

e.

Increase signs on trails

f.

Increase trail variety

Add more loops to current trails

Offer naturalist guided ski trips

Sponsor cross-country ski festival
Provide rest shelters along trails
Provide easier trails

Increase signs on highway
Provide better quality grooming

Offer ski lessons

Provide outhouses along trail
Improve scenic quality
Increase publications
Provide side by side grooming

7.

Which letter from above is the m ost attractive action?.

8.

There a re m any re a so n s people do not ski a s often a s they would like to. If you strongly a g re e with one of the
re aso n s below w hich m ay re d u ce the num ber of tim es p eople cross-country ski In t h e M u n isin g a re a ,
check the blank clo sest to agree. If you strongly d isa g ree m ark the blank closest to d isa g ree . If you feel
som ew here in betw een, m ark the blank betw een a g re e and disag ree which m ost closely reflects how you feel
about the reason. If you neither a g re e nor disagree, m ark the middle blank.
tt takes too much time to travel to Munising..................disagree____________________________________ agree
Concern about snow on the ro ad s..................................disagree____________________________________ agree
Concern about being unable to return
home on time because of sn o w ___________disagree____________________________________ agree
Fees charged at Forest Service ski areas
are too expensive............................................. disagree____________________________________ agree
Munising is too expensive............................................... disagree____________________________________ agree
Obligations at home or w ork........................................ disagree_____________________________________agree
Members of family or group would
rather go elsew here
....................... disagree_____________________________________agree
Parking lots are not plowed............................................. disagree____________________________________ agree
Concern about medical c a r e ........................................... disagree____________________________________ agree
Concern about vehicle burglary...................................... disagree____________________________________ agree
I don't feel welcome in the community........................... disagree____________________________________ agree
Lack of com panions___________________________ disagree_____________________________________ agree
Concern about becoming lo st............................ , . . . . disagree_____________________________________ agree
Poor personal physical condition................................ disagree_____________________________________ agree
Too many snowmobiles around Munising............d i s a g r e e ____________________________________ agree
Not enough to do in Munising..........................................disagree______________
agree
Lack of ski rental.............................................................. disagree__________________________________ agree

9.

How m any tim es did you cross-country ski last s e a s o n (N ovem ber 1990 through M arch 1991)?

2

143

10. How m any overnight trips did you m ake in the p a st s e a s o n (N ovem ber 1990 through March 1991) during

which you c ro ss-country sk ied ? ______

11 . How m any trips do you ex p ect to m ake to cross-country ski in th e Munising a rea this se a so n (N ovem ber 1991
through M arch 1 9 9 2 )? ______

12. O ther th an M unising, w hat o n e com m unity would you visit to cross-country ski? C heck none it applicable.

________________________________________________

Q

None

13. Please circle your skill level: 1 =beginner, 5=expert
beginner
1

2

intermediate
3
4

expert
5

14. W hat n e w sp a p e r do you re a d m ost frequently?

____________________________________________

QNo

15. Will you spend the night away from home while on this trip?

Q Yes

16. In th e Munising a rea , w here a re you staying?

Qyour second home

Qyour own home

Grelative's home or relative's second home

Qfriend's home or friend's second home
Q motel/hotel

17. Is the place you a re staying within 15 miles of Munising? Q

no

Q Y es

18. W hat is the main reaso n you a re in the Munising a re a ?

Qlive here Gvisrting friends or relatives Qrecreation Qbusiness Qother
19. P le ase circle e a c h winter recreation activity in which you participate,

snowmobiling

ice fishing

sledding

downhill skiing snowshoeing

20. P le a s e circle the m a in activity

ice climbing

in which

you

snowmobiling

ice fishing

ice climbing

sledding

downhill skiing snowshoeing

camping

hunting

cross-country skiing
will participate a n
camping

Other____

th is trip

hunting

cross-country skiing

3

ice skating

ice skating
Other____

144

Low price m otel...................................... Not important
, - . . . .
Mid-price m otel....................................... Not important-------------------------------------------------------- Extremely important
Extremely important
Ski lodge or ski resort............................ Not important
--------------------------------------------------------Extremely important
Basic kitchen (sink and microwave). . . Not important
e
? hild « " » ...................................................Not important------------------------------------------------------c f 8me,ly impor,ant
Swimming pool..........................................Not important------------------------------------------------------- ExUemely important
B au n a.........................................................Not important------------------------------------------------------- Earemely important
ventilated waxing room
Not important
txtremely important
iaundry ......................................................Notimportant-------------------------------------------------------- Extremely important
----------------------Extremely important
Fam i°0d ..................................................Not important
____
P
Family restaurant...................................
Not
important---------------------------------------- Extremely important
Gourmet restaurant..................................... importan, --------------------------------------------------------Extremely important
£ '9h,1C lub.................................................Not important--------------------------------------------------------Extremely important
Family oriented entertainment
Not important------------------------------------------------------- Extremely important
Cultural attraction................................... Not
important---------------------------------------- Extremely important
B a r.............................................................. Not important
Extremely important
------------------------------Extremely important

Extremely important
Ctothing s h o p s..........................................Not important'
Extremely important
om
Sklsh°Ps
sirups..................................................Not
................................................... Notin
important '
Extremely important
Local art and craft sh o p s........................ Not important_____________________________________ Extremely important
c u th ! ln PSh ....................................................... Not imP °riant

Quality of cross-country ski trails
Not important____________________________________ Extremely important
Overall community...................................Not important
____________________________
Extremely important

22. Flow sgtlgfiefl are you with the following types of serv ices In th g M u n isin g A re a ? P lace a check m ark in
the blank which reflects how you feel.
R estaurants..............................................Not satisfied
Lodging.................................................... Not satisfied
Shopping................................................. Not satisfied_______ .
Quality of cross-country ski trails
Not satisfied______________________________ 3
Overall community...................................Not satisfied

Extremely satisfied
Extremely satisfied
Extremely satisfied
Extremely satisfied
Extremely satisfied

23. Flow m any day s will you ski during this trip? ______ How m any hours did you ski the day you picked up this
su rvey? _______
24
25.

How m any days will you be within 15 miles of Munising on this trip? _______
How m any nights will you b e within 15 miles of Munising on this trip? _____

4

145
26. W e w ould also like to know how m uch you and your party sp e n t in the U pper P eninsula and in the
Munising a re a o n this trip from 12:01 a.m. until midnight on th e day you received this questionnaire.
R e m e m b er to include cash, check, and credit card purchases for the following item s. If you did not
buy a n item during th e 24 hour period, p le a se record a zero in th e blank. For p u rp o se s of this study, the
Munising a re a is within 15 m iles of Munising a n d includes the tow ns of Munising, Christm as, Au train,
S hingleton, a n d C h a th a m . Do not Include M unising a r e a p u r c h a s e s In t h e a m o u n t s r e c o r d e d
fo r th e U p p e r P e n in s u la .
U pper
P e n in s u la

M u n isin g
A rea

G rocery food & b e v e ra g e s

$

$

R e stau ra n t food a n d b e v e ra g e s

$

$

Vehicle related item s (g a s, etc.)

$

$

L odging

$

$

Ski equipm ent

$

$

Clothing

$

$

All o th er item s

$

$

27.

How m any p eople, including yourseH, a re you reporting expenditures for? __________

20.

How m any people, including yourself, a re traveling with y o u ? _________

29. Are the people in your party : Q F am ily Q F r ie n d s

O R e la tiv e s

□ club or organization

30. How m any days have you b e e n aw ay from hom e on your trip ? _______
31. Approximately how m any m iles is Munising from w here you liv e ? ________________Miles
32. Check the m ost im portant so u rc e from which you learn ed about skiing opportunities in the Munising area.
Q F risnds or relatives

QTravel brochure

QM agazine/newspaper article

Q ch am b er of Commerce

Q F orest Service employee

Q F orest Service publication

QHotel/Motel employee

QDNR employee

QD NR publication

Q Previous Trip

Q

Q lJpper Peninsula Travel and
Recreation Association

QMichigan Travel Bureau

QMichigan Welcome Center

Q L ocal ski shop

QOut-of-town ski shop

Qother______________

QNational Park Service
employee

QNational Park Service
publication

Q P a rk Service/Forest Service
Visitor Center

aaa

5

146
33.

W hat is th e highest level of education you have com pleted?
12 3 4 5 6 7 8
elem en tary

34.

{Circle y e ars completed]

13 14 15 16
college

9 10 11 12
high school

17 18 19 20
graduate school

W here is your p erm anent resid en c e?
City:.

County:.

S tate:.

Zip code:.

35. M edian household incom e in M ichigan is $31,645. In 1990, w a s your total household incom e
taxes: (C heck One)
□

M o re

3 6 . How old a re you?

□

L ess

□

b e fo re

Approximately th e sa m e

Your g e n d er?

□ Fem ale

□ Male

T h a n k y o u fo r y o u r a s s i s t a n c e . P le a s e drop off the com pleted questionnaire at the Pictured R ocks
National L akeshore office at S a n d Point, Munising, or mail the com pleted questionnaire in the p ostage-paid
e nvelope provided. If the envelope is m issing, p le a se s e n d the com pleted questionnaire to:
Dr. P e te r Forsberg
Travel, Tourism , a n d R ecreation R esource C enter
172 N atural R e so u rc es Building
M ichigan S ta te University
E ast Lansing, Ml 48824-1222
P le a se w nte in the sp a c e below any com m ents you would like to m ake regarding cross-country skiing in the
Munising a rea .

6

147

Cross-Country Skiing Mail Q uestionnaire 1992
Travel, Tourism , a n d Recreation Resource C enter
172 N a tu ra l R esources B uilding
M ichigan State U n iv ersity
East Lansing, M ichigan 48824-1222
Please answer the following questions in order and as fully as possible. Please answer all the questions on the
front before reading the back. If your answer is none or don't know, write that answer or check the appropriate
box. Most people will probably leave som e of the spaces blank for some of the questions.
1.

2.

Of all the available cross-country ski destinations in the G re a t L akes R e g io n that were available to you
during the winter of 1991-92, w h ic h d e s tin a tio n s c o m e to m in d ? Pleas list the nearest town.
1.

4.

2.

5.

3.

6.

□

None

W hat o n e c o m m u n ity in the G re a t L ak e s R e g io n would you have p re fe rre d to visit for cross-country
skiing during the winter of 1991 -92? Check none if applicable.
________________________________________________

□ None

B.
What considerations
come to mind when deciding on
a X -C destination

List each of these considerations
in order of importance

1.

1

2.

2

3.

3

4.

4

4.

Please list all the communities in Michigan’s Upper Peninsula which you visited for cross-country skiing
between Novem ber 1, 1991 and March 30, 1992.
□ None

5.

Did you fill out a questionnaire distributed at one of the Munising area cross-country ski trails?
□ Yes
□ No
If you responded “yes" to the above question, please go to part 13.

6.

Which newspaper do you read most frequently?

7.

Please circle each winter recreation activity in which you participate.

snowmobiling

ice fishing

ice climbing

camping

hunting

sledding

downhill skiing

snowshoeing

cross-country skiing

ice skating
Other_________

148
8.

H ow Im p o rta n t are the a v a ila b ility ot the following features to you fflfle n ChOOSlUtt a cross-country skiing
destination community? Place a check mark in the blank which reflects how you teel.

Low price motel
Mid-price m o te l
Ski lodge or ski resort

Not important_______________________________ _____ Extremely important
Not important_________________________
............................. Not important__________________________

c ^ m a k ^ m n o rta n t
re
V P 0

Basic kitchen (sink and microwave) . . Hot important________________________________
Extremely important
Child care . . .................................. . . . Not important_____________________________________ Extreme* rmportarrt
Swimming pool
No. important____________________________________
Sauna
.T .
Not important____________________________________ Extreme y importan
Ventilated waxing room ........................... Not important--------------------------------------------------------Extreme y impo a
Laundry ...................................................... Not important
Extremely important
Fast food .................................................... Not important
Extremely important
E x t r e m e l y important
Family r e s ta u r a n t...................................... Not important
Gourmet restaurant ..................................Not important--------------------------------------------------------- E x t r e m e l y important
Niqht C lu b .................................................... Not important_____________________________________ Extremely important
Family oriented entertainment ................Not important----------------------------------------------------------Extremely important
Cultural a ttra c tio n ...................................... Not important-------------------------------------------------------- Extreme y
Extremely important
g ar ............................................................. Not important__________
Gift shoos
Not important______________________________________Extremely important
Clothing shops ...........................................Not important______________________________________Extremely .mportant
Ski shops .................................................... Not important_____________________________________Extreme y important
Local art and craft shops ......................... Not important______________________________________Extremely important
Quality of cross-country ski trails
Not important_____________________________________ Extremely important
Overall community .................................... Not important______________________________________Extremely important

1 0 . W hat is th e h ig h est level of e d u c a tio n you h ave co m p le te d ? [Circle y e a rs com pleted]
1 2 3 4 5
6
7
8
9
10
11
12
13
14
15 16
17
18
19
20
elementary
high school
college
graduate school
1 1 . How old a re y o u ? ________

G e n d e r:

□ fem ale

□ m ale

1 2 . M edian h o u se h o ld incom e in M ichigan is $ 3 1 ,6 4 5 . In 1990, w a s your total h o u se h o ld incom e b efo re ta x e s :
(C h e ck O ne)
O Mor e

□ L ess

3 A pproxim ately th e s a m e

1 3. T hank you for y our a s s is ta n c e . P le a s e mail th e co m p le te d q u estio n n aire in th e p o sta g e -p a id envelope
provided. If th e e n v e lo p e is m issing, p le a se s e n d the c o m p leted q u e stio n n a ire to :
Dr Peter Forsberg
Travel. Tourism, and Recreabon Resource Center
172 Natural Resources Building
Michigan S tate University
East Lansmg. Ml 48824-1222

P le a s e u s e th e s p a c e below or a n additional s h e e t for any c o m m e n ts you w ould like to m ak e regarding c r o s s ­
country skiing.