contact:
Dr. Robert C. Burns
Recreation, Parks, and Tourism Resources Program
Division of Forestry
PO Box 6125
West Virginia University
Morgantown, WV 26506-6125.
robert.burns@mail.wvu.edu

The first question, focusing on comparing reported satisfaction levels across various user segments, found many significant differences between subgroups of recreationists.  For example, campers were generally more satisfied than members of the other activity groups with all aspects of their outings.  Likewise, visitors traveling with children and older visitors tended to report higher satisfaction levels.  Those traveling greater distances from home were also more satisfied with the facilities, services, and information they encountered than local visitors.

Regarding the second research question, the comparison of regression models across various user characteristics showed only minor differences.  Thus, although satisfaction levels varied across segments of visitors, the variables influencing overall satisfaction were remarkably stable.  Overall satisfaction was correlated to all four customer service domains, and varying combinations of the domains accounted for 13 to 19% of the variance in overall satisfaction.  The strongest predictive model was found for day users, in which 19% of the variance in overall satisfaction was accounted for by satisfaction with facilities, services, and information.

Key words: customer satisfaction, market segmentation, service quality, water-based recreation

Over the past 15-20 years, customer satisfaction has become one of the primary goals of managers of service or product-related organizations.  Service agencies in particular tend to focus on customer satisfaction as a way to differentiate themselves from their competitors through delivery of high quality service (Parasuraman, Zeithaml, & Berry, 1985; 1988).  Many federal agencies in the US, including those that provide recreation opportunities, have begun focusing on understanding the needs of customers.  One of the methods that these agencies have utilized to understand their customers has been customer satisfaction surveys.  This article seeks to differentiate between the satisfaction levels of different user groups, or market segments, that recreate in water-based outdoor recreation areas.

Service Satisfaction Literature
For various reasons, service agencies often have a greater challenge than product-related agencies in demonstrating their superiority over their competition (Parasuraman, Zeithaml, & Berry, 1985; 1988).  Service agencies differ from product-related agencies in three ways.  The first of these, according to Parasuraman et al. (1985, 1988), is the intangibility of services, the second is the heterogeneity of services, and the third is the inseparability of production and consumption.  Unlike products that can be purchased, stored, and consumed at a specific moment, services are consumed as they are produced.  This simultaneous production and consumption puts more emphasis on the delivery of services than is seen in product delivery.  Many authors have discussed the inseparability of production and consumption that causes services to stand out from products (Carmen & Langeard, 1980; Gronroos, 1982; Upah, 1980; Parasuraman et al., 1985).  According to Zeithaml (1981), it is difficult to understand how consumers perceive services and how they evaluate customer satisfaction.

Once the peculiarities of service-related customer satisfaction are understood, the definition of satisfaction itself must be addressed.  Satisfaction has been defined as a post-consumption evaluation of perceived quality related to expected quality (Rust & Oliver, 1994; Parasuraman et al., 1985).  A second important definition is that of service quality, which has been described as “the postchoice evaluative judgment of a specific purchase occasion” (Rust & Oliver, 1994, p. 245).  From a service manager’s perspective, service quality entails the tasks that involve interaction with customers to provide efficiency and result in satisfied customers (Bearden & Teel, 1983; Churchill & Suprenant, 1982; Parasuraman et al., 1985, 1988).  Accordingly, customer satisfaction can be defined as an individual's overall evaluation of a series of service quality events.  These events may include a trip to the restroom or an interaction with a service provider or any other item that is measured on the survey instrument.

Segmentation
Market segmentation has been suggested by recreation researchers as a method of managing different recreation users (Andereck & Caldwell, 1994; Donnelly, Vaske, DeRuiter, & King, 1996).  In his study of freshwater trout anglers, Bryan (1977) showed that management concerns differ according to an angler’s specialization level.  By examining anglers on a continuum from occasional fishermen to technique-setting specialists, Bryan suggested that specific settings should be managed differently for those with very specific levels of specialization to meet variability in recreation needs.  In a related effort, Graefe (1981) identified different subgroups among saltwater anglers, based on socio-economic differences, reasons for fishing, and participation levels.

Tinsley and Kass (1978) conducted research focusing on the differences in leisure activity needs of males and females, finding that many recreation activities differ in their need satisfying properties.  Absher and Lee (1981) noted that visitor characteristics and prior experience had an effect on perceptions of crowding in a National Park.  Westover (1984) found gender and age to be valid predictors of perceived safety in urban parks and forests.  Bullaro and Edginton (1986) indicated that segmentation is useful in recognizing that people are different and that managers must understand their differences in order to meet the needs of recreationists.

Kuss, Graefe, and Vaske (1990) examined the different needs of diverse user groups in outdoor recreation settings, based on the notion that a single management strategy cannot satisfy all visitors.  Their research attempted to develop visitor typologies, based on participation rates, preferences, demographics, and geographical locations.  Backman (1994) focused on a similar concept, namely a person-situation approach to market segmentation.  This approach focuses on segmentation as a function of both the recreationist and the situation in which the individual recreates.  In this study, significant differences were noted in benefits sought across seasons and geographic location.

Andereck and Caldwell (1994) used cluster analysis to examine motivations, trip characteristics, and demographics of zoo visitors.  These authors noted that "understanding the diversity of participant needs and desires allows organizations to manage resources in the most efficient manner” (p. 19).  Absher, Howat, Crilley, and Milne (1996) demonstrated that customer characteristics such as gender, age, and disability status impacted overall satisfaction levels of users to sporting events and leisure centers in Australia and New Zealand.  This study also revealed visit characteristics that defined specific market segments of visitors with significantly different levels of satisfaction.

Donnelly et al. (1996) pursued the concept of person-occasion segmentation, which focuses not only on the different user groups visiting a recreation area, but the different natural resource attributes of the area that they were visiting.  Distinct subgroups of visitors were found at state parks offering varying levels of facilities and activities.  For the park type offering the highest degree of amenities, user-based segmentation revealed three clusters of park users.  The three groups included people who placed a great deal of importance on facility attributes, those who placed a greater level of importance on activity attributes, and thirdly, respondents who perceived both facility and activity attributes were similarly important.  This group showed lower levels of importance with both attributes.

More recently, Park, Yang, Lee, Jang, and Stokowski (2002) examined Colorado gamblers’ characteristics through the use of involvement profiles, socio-economic characteristics, and behavioral patterns.  These authors determined that understanding a recreationist’s involvement in their chosen activity resulted in a better understanding of the recreationists.  According to the authors, this is particularly helpful in understanding the differences in groups that are very homogenous in their socio-economic characteristics.

Jang, Morrison, and O’Leary (2002) used benefit segmentation to understand the characteristics of Japanese pleasure travelers visiting the USA and Canada.  This segmentation focused on identifying a specific target market with relation to profitability.  The benefits sought were used to classify the visitors into three clusters, including novelty/nature seekers, escape/relaxation seekers, and family/outdoor activity seekers.  The results showed that marketers in small destination locations should pursue pleasure travelers who sought novelty/nature benefits, while larger destination sites might pursue family/outdoor seekers in their quest for profit from pleasure travelers.  Galloway (2002) segmented park users by their degree of motivation for sensation seeking, as well as traditional socio-economic characteristics.  He suggests that understanding a park visitor’s motivation for sensation seeking contributes to an overall understanding of park visitors.

Borrie et al. (2002) proposed examining respondents’ trust in the agency, commitment, and attitudes of social responsibility.  These authors lamented the use of a marketing view of recreation services, and the current focus on “…on-site satisfaction and repeat visitation as an indicator of success…” (Borrie, et al. p 54, 2002).  The findings of this study showed significant differences across the independent variables of trust, commitment and social responsibility, and across the typical segmentation variables of socio-demographics and trip characteristics.  This research effort showed the importance of understanding different segments of users so that land managers may provide specific informational messages that meet the needs of as many constituents as possible.

In a study conducted in Tanzania’s National Parks, Wade and Eagles (2003) segmented their respondents into four market segments; lodge safari visitors, special campers, camping safari visitors, and overland safari visitors.  Using the Importance-Performance Analysis (IPA) method introduced by Martilla and James (1977), Wade and Eagles found significantly different market segments regarding on-site satisfaction levels, as well as levels of crowding.  The results of this study showed that information regarding visitor perceptions goes undetected without the use of market segmentation.  Additionally, this research effort exposed the issue of possible skewing of the data, either positively or negatively, when specific segments are not identified and separated from the larger database.

The role of segmentation as an influence on satisfaction has been the focus of much research in the outdoor recreation field.  This study seeks to pull together the various socio-demographic segmentation variables upon which previous researchers have focused, and explain the role of each.

Project Background
Data for this study were collected as part of a larger study of customer satisfaction levels funded by the US Army Corps of Engineers Recreation Research Program.  The nationwide study of customer satisfaction levels at Corps of Engineers lakes ran from 1995 to 1998.  Preliminary research regarding lake users’ satisfaction levels was conducted at Lake Sakakawea, North Dakota in 1995 to create a baseline level of customer perceptions regarding satisfaction-related issues (Burns & Titre, 1996).  That study used open-ended questions to identify satisfaction attributes that visitors felt were important, as well as performance levels for these same attributes.  The study was then replicated and extended in 1996 at Lake Francis Case and Gavins Point, both located in South Dakota (Burns, Graefe, Thapa, & Titre, 1997).  The study from which these data were drawn built on the 1995 and 1996 studies and was conducted at various Corps of Engineers lakes throughout the United States.

Table 1.  Profile of Survey Locations

Reservoir Name	Location	Shoreline  Miles	Suface Area
Allatoona Lake	Georgia	270	12,000
Bonneville Lock and Dam	Oregon/ Washington	106	9,000
Dale Hollow Lake	Tennessee	600	30,000
Mark Twain Lake	Missouri	285	56,000
Lake Mendocino	California	11	1,822
Norfolk Lake	Arkansas	400	22,000
Lake Oahe	South Dakota	2,250	55,000
Raystown Lake	Pennsylvania	78	8,300
Saylorville Lake	Iowa	60	6,000
Somerville Lake	Texas	85	1,160

A stratified sampling plan was devised to ensure that weekdays, weekends/holidays as well as various times of the day were appropriately represented.  Only persons eighteen years old or older were included in the sample, and no users were interviewed more than once.  Interviewers were provided with a sampling plan and were in regular telephone communication with the investigators to ensure that data were collected in a systematic manner.  The data were mailed to the US Army Corps of Engineers Waterways Experiment Station in Vicksburg, MS on a weekly basis, and inputted into a database and provided to the authors for analysis.

Interviewers at each of the ten lakes were instructed to collect data from approximately 300 lake users.  Approximately 100 boaters, 100 day users, and 100 campers were surveyed at each lake.  Respondents were approached by the interviewers while they were in a recreation setting, such as a campground, boat ramp area, or day use area (beach, picnic area, etc.).  The data were collected in entirety through the use of on-site, face-to-face interviews.  Refusals were very limited (29 returned refusal sheets) because of the on-site methodology of the study.  At lakes where only one or two of the user groups existed, a total of 300 surveys were collected at the various recreation sites that were available.  In total, minus refusals, 2,933 recreationists at the ten Corps of Engineers lakes participated in this study.

The visitors were asked what recreational activities they were pursuing and then asked to rank their primary, secondary, and tertiary activities by level of importance.  The respondents were categorized accordingly, falling into one of three primary user segments (ramp use, camping, or day use).  The users were fairly evenly distributed across the three categories as indicated in Table 2.  Of the 2933 respondents interviewed, 35.2% reported that their primary activity was day use, another 34.5% indicated that camping was their primary activity, and 30.3% reported ramp-use as their primary activity.

Table 2.  Socio-demographic Profile of Respondents.

Frequencies Table	Number	Percent
Primary Activity
Ramp use	720	30.3
Camping	820	34.5
Day use	837	35.2
Gender
Male	1816	63.1
Female	1062	36.9
Children In Group
Yes	1597	55.2
No	1297	44.8
Age
40 years or younger	1486	53.0
41 years or older	1319	47.0
Number Of Miles From Primary Residence
1-50 miles	1519	53.5
51 or more miles	1320	46.5
Previous Experience at Recreation Area
First-time Visitor	514	17.6
Repeat Visitor	2413	82.4

Respondents were asked several questions regarding their socio-demographic characteristics and visitation patterns (Table 2).  The sample was approximately two-thirds male (63.1%) and one-third female (36.9%).  The mean age of respondents was 42.2, and the vast majority was Caucasian (92.3%).  A small proportion of respondents indicated that they were Hispanic (3.9%), or African American (2.4%).  The mean group size for respondents in this sample was 4.6, with over one-half of the groups (58.5%) including 3-4 people, and almost one-quarter (22.7%) consisting of 5-10 people.  Over one-half of the respondents (55.2%) indicated that their group included no children.

The average distance traveled from a respondent's primary residence to the recreation area was 128 miles.  Over half of the respondents (53.5%) reported traveling 50 miles or less.  Regarding previous experience, most of the visitors (82.4%) had been to the lake before, while about one-fifth of them (17.6%) were first-time visitors.

The socio-demographic makeup of the recreationists sampled in this study was quite similar across the ten lakes.  Although we do not have specific information on Corps of Engineer recreationists nationwide, the fact that there were no significant differences across ten lakes in ten states suggests that the respondents in this sample are similar to those nationwide.

Respondents were asked to rate their satisfaction with the four satisfaction domains (facilities, services, information, and recreation experience) using a five point Likert scale ranging from “not at all satisfied” to “extremely satisfied.  The final satisfaction measure was an overall measure of satisfaction, designed to query visitors as to their satisfaction with their overall experience on that day or trip.  A 10-point scale, ranging from "1" to "10" (where 1 is worst and 10 is best) was used to measure overall satisfaction.

Limitations
Respondents were asked only about the primary activity in which they were participating, regardless of the level of overlap between two or more recreation activities.  Thus, their perceptions of satisfaction/dissatisfaction about one particular activity may have been influenced by another activity that they had just experienced, or other variables, such as setting and crowding/conflict issues.  The multiple regression analysis strategy focused only on examining the extent to which overall satisfaction was a function of satisfaction with the various customer service domains, and did not include other variables that might contribute to the explanation of visitor satisfaction.

Table 3.  Comparison of Satisfaction with Facilities, Services, Information, and Recreation Experience Domains by Primary User Group.

Satisfaction Measure	Ramp users	Campers	Day users	F value
Satisfaction with facilities	4.18a	4.30b	4.27ab	4.84**
Satisfaction with services	4.06a	4.28b	4.09a	18.78***
Satisfaction with information	3.98a	4.14b	3.97a	11.30***
Satisfaction with recreation experience	4.32a	4.42b	4.35a	4.60**
Overall satisfaction	8.28	8.61	8.49	8.53***

***= Significant at .001  ** =Significant at .01
^a   Means with different superscripts differ significantly at the .05 level

The first part of this research question examined the differences in overall satisfaction and satisfaction within each domain between three primary lake user groups (ramp users, campers, and day users).  One-way analysis of variance revealed significant differences for all of the satisfaction measures (Table 3).  In each case, the campers showed the highest mean scores among the three user groups.  The greatest differences were noted for satisfaction with services.  Campers showed significantly higher scores for this domain (mean = 4.28) than day users (4.09) and ramp users (4.06).  Satisfaction with information showed the second greatest degree of difference across the three user groups, with campers (4.14) again indicating the highest mean score, and no significant difference noted between ramp users (3.98) and day users (3.97).  Similarly, campers showed the strongest levels of satisfaction with facilities (4.30), followed closely by day users (4.27).  The ramp users (4.18) were significantly less satisfied with facilities than the campers (4.30).  The smallest differences were noted for satisfaction with the recreation experience, where campers showed a higher mean satisfaction score (4.42) than day users (4.35) and ramp users (4.32).

Among the customer service domains, satisfaction was highest for the recreation experience domain for all three activity groups, while satisfaction with information was consistently the lowest domain for all three groups.  Overall satisfaction was relatively high for all three activity segments.  Campers reported the highest overall satisfaction levels (8.61 out of a 10-point scale) followed by day users (8.49) and ramp users (8.28).

Table 4.  Comparison of Satisfaction with Facilities, Services, Information, and Recreation Experience Domains, and Overall Satisfaction, Across Various Market Segments.

Satisfaction Measure                                                   Segmentation Variable                  F value

	Males	Females	F value
Satisfaction with facilities  Satisfaction with services  Satisfaction with information  Satisfaction with recreation experience  Overall satisfaction	4.22 4.15 4.04 4.35 8.32	4.21 4.14 4.04 4.40 8.58	.06     .12     .00   3.19 18.84***
	Without children	With children	F value
Satisfaction with facilities  Satisfaction with services  Satisfaction with information  Satisfaction with recreation experience  Overall satisfaction	4.18 4.12 4.04 4.33 8.34	4.26 4.18 4.04 4.41 8.52	6.11*   3.09     .00   7.50**   9.68**
	40 or younger	41 or older	F value
Satisfaction with facilities  Satisfaction with services  Satisfaction with information  Satisfaction with recreation experience  Overall satisfaction	4.14 4.09 3.97 4.36 8.40	4.30 4.21 4.13 4.37 8.43	29.06*** 13.62*** 23.50***     .36     .17
	Travel 50 miles or less	Travel 51 miles or more	F value
Satisfaction with facilities  Satisfaction with services  Satisfaction with information  Satisfaction with recreation experience  Overall satisfaction	4.19 4.10 3.99 4.33 8.54	4.26 4.22 4.11 4.42 8.48	4.51* 15.38*** 15.62*** 13.01***   4.03*
	Previous visitor	First time visitor	F value
Satisfaction with facilities  Satisfaction with services  Satisfaction with information  Satisfaction with recreation experience  Overall satisfaction	4.32 4.21 4.05 4.42 8.66	4.20 4.13 4.04 4.36 8.37	9.83**   4.26*     .09   3.84* 15.53***

***= Significant at .001  ** =Significant at .01  * =Significant at .05

Additional tests pertinent to the first research question focused on other potential visitor segmentation variables including gender, children in group, age, distance traveled, and previous experience (Table 4).  The analysis of satisfaction scores between males and females showed no significant differences for the four customer service domains, but a significant difference between males and females regarding overall satisfaction.  Females showed a significantly higher level of overall satisfaction (8.58) than was noted for males (8.32).

Differences in satisfaction levels between respondents with children in their group versus those with no children in their group were noted for satisfaction with facilities and satisfaction with the recreation experience.  In both cases, the visitors whose groups included children showed significantly higher satisfaction scores than respondents with no children in their group.  Additionally, visitors with children reported higher overall satisfaction than visitors whose groups did not include children.

Comparisons based on age showed significant differences for three of the four satisfaction domains, while no significant difference was noted for overall satisfaction.  Older visitors tended to show higher mean scores for the satisfaction domains.  The largest differences were noted for the facilities domain (older mean = 4.30; younger mean = 4.14) and for satisfaction with information (older mean = 4.13; younger mean = 3.97).

Satisfaction levels were also compared between those who traveled less than 51 miles to the recreation area and those who traveled more than 50 miles.  Significant differences were noted for all four domains and for overall satisfaction.  Respondents who had traveled more than 50 miles consistently showed higher satisfaction levels.  The greatest differences were noted for satisfaction with information (far mean = 4.11; near mean = 3.99) and satisfaction with services (far mean = 4.22, near mean = 4.10).  Interestingly, however, overall satisfaction was slightly higher for visitors who had traveled 50 miles or less (mean = 8.54) than among respondents who had traveled 51 miles or more (mean = 8.48).  Although this difference was in the opposite direction compared to the customer service domains, the difference was small and probably not significant for management purposes.

Finally, with respect to previous experience at the recreation area, significant differences were found for three of the four customer service domains and for overall satisfaction.  In each case, the visitors who had previous experience at the recreation area showed higher mean satisfaction scores.  The greatest differences were found for overall satisfaction, with previous visitors reporting a mean score of 8.66, compared to 8.37 for first time visitors.  Among the four satisfaction domains, the facilities domain showed the greatest difference (previous visitors mean = 4.32; first time visitor mean = 4.20), followed by the services domain (previous visitors mean = 4.21; first time visitor mean = 4.13), and the recreation experience domain (previous visitors mean = 4.42; first time visitor mean = 4.36.).

Table 5.  Results of Multiple Regression Between Facilities, Services, Information, and Recreation Experience Domains and Overall Satisfaction for Primary User Groups.
                                                                               Ramp users          Campers         Day users

Independent variables	R	Beta	R	Beta	R	Beta
Satisfaction with facilities	.28***	.08	.32***	.14***	.38***	.18***
Satisfaction with services	.31***	.18***	.31***	.10*	.39***	.15**
Satisfaction with information	.22***	-.01	.29***	.07	.35***	.12**
Satisfaction with recreation experience	.31***	.18***	.33***	.17***	.32***	.06

                                                                                R² = .13               R² =  .15           R²  = .19
                                                                         F = 24.758***        F = 34.075***    F = 44.341***

Dependent variable: Overall satisfaction
***= Significant at .001  ** =Significant at .01  * =Significant at .05

To understand the extent to which each of the domains was related to overall satisfaction, the four domain satisfaction scores were regressed against overall satisfaction for each of the visitor segments.  Regarding the three primary activity groups (ramp user, camper, day user), there were at least two significant predictors of overall satisfaction for each group, but the significant predictors varied for the different groups (Table 5).

An examination of the ramp users showed that the recreation experience domain (Beta = .18) and the services domain (Beta = .18) were significant predictors of overall satisfaction.  These variables accounted for about 13% of the variance associated with overall satisfaction.  The campers showed significant effects for the recreation experience domain (Beta = .17), the facilities domain (Beta = .14), and the services domain (Beta = .10), accounting for about 15% of the variance in overall satisfaction.  Day users’ results also revealed three significant predictors of overall satisfaction.  The facilities domain (Beta = .18), services domain (Beta = .15), and information domain (Beta = .12) together accounted for 19% of the variance associated with overall satisfaction.

One pattern that emerged from these regression models was that all three user groups showed a significant effect from the services domain.  The facilities domain showed a significant influence for both the campers and the day users, while the ramp users and the campers showed a significant influence from the recreation experience domain.  The information domain was significant for only one user group (day users).

A similar analysis was conducted for the other segmentation variables (Table 6).  The regression analysis for males showed that there was a significant influence on satisfaction from the services domain (Beta = .16), recreation domain (Beta = .15), and facilities domain (Beta = .14).  This model accounted for about 17% of the variance associated with overall satisfaction.  The female respondents showed a similar pattern, with the facilities, services, and recreation experience domains showing nearly equivalent influence (Beta = .13 for each domain) on overall satisfaction.  The regression model for the females accounted for about 13% of the overall satisfaction variance.  The information domain was the single non-significant independent variable for both males and females.

Table 6.  Results of multiple regression between facilities, services, information, and recreation experience domains and overall satisfaction for various market segments.

Independent variables      Satisfaction with facilities       Satisfaction with services       Satisfaction with information       Satisfaction with recreation experience	Males     R          Beta .34***   .14*** .36***   .16*** .30***   .04 .34***   .15***       R2  .17 F = 85.225***	Females      R        Beta .29***   .13*** .31***   .13*** .26***   .05 .28***   .13*** R2   .13 F = 36.433***
Independent variables      Satisfaction with facilities       Satisfaction with services       Satisfaction with information       Satisfaction with recreation experience	Without children R Beta .34***   .16*** .35***   .14*** .30***   .06 .33***   .13*** R2  .16 F = 72.038***	With children R Beta .28***   .10** .31***   .15*** .24***   .02 .31***   .19*** R2   .13 F = 47.648***
Independent variables      Satisfaction with facilities       Satisfaction with services       Satisfaction with information       Satisfaction with recreation experience	40 or younger R Beta .34***   .13*** .35***   .14*** .31***   .06 .36***   .19*** R2  .17 F = 74.089***	41 or older R Beta .30***   .13*** .32***   .16** .24***   .01 .30***   .13*** R2   .13 F = 46.743***
Independent variables      Satisfaction with facilities       Satisfaction with services       Satisfaction with information       Satisfaction with recreation experience	50 miles or less R Beta .31*** .09** .35*** .19*** .29*** .04 .32*** .15*** R2  .15 F = 63.206***	51 miles or more R Beta .32*** .13*** .31*** .13*** .25*** .05 .33*** .13*** R2   .15 F = 55.469***
Independent variables      Satisfaction with facilities       Satisfaction with services       Satisfaction with information       Satisfaction with recreation experience	Previous visitor R Beta .31***   .08 .32***   .08 .32***   .12 .36***   .21*** R2  .15 F = 22.739***	First time visitor R Beta .29***   .14*** .31***   .16*** .26***   .03 .28***   .15*** R2   .15 F = 99.557***

Dependent variable: Overall satisfaction
***= Significant at .001  ** =Significant at .01  * =Significant at .05

When examining respondents without children at the recreation area, the multiple regression model (Table 6) indicated that there was a significant relationship for the facilities domain (Beta = .16), the services domain (Beta = .14), and the recreation experience domain (Beta = .13).  This model accounted for about 16% of the variance associated with overall satisfaction.  The users who were accompanied by children at the recreation area showed a similar pattern, with the recreation experience domain being the strongest predictor (Beta = .19), followed by the services domain (Beta = .15), and facilities domain (Beta = .10).  These independent variables accounted for about 13% of the variance in overall satisfaction.  As seen previously, the information domain was the single non-significant independent variable in this regression analysis.

Considering the visitor segments based on age, for respondents who were 40 or younger, the strongest predictor of overall satisfaction was the recreation experience domain (Beta = .19), followed by the services domain (Beta = .14) and the facilities domain (Beta = .13).  The information domain was not a significant predictor of overall satisfaction.  The pattern of results was very similar for the older respondents, with the same three significant predictors and a slightly lower R² value.

For those respondents who traveled 50 miles or less, three of the four satisfaction domains were significant predictors of overall satisfaction.  These were the services domain (Beta = .19), the recreation experience domain (Beta = .15), and the facilities domain (Beta = .09).  This regression model accounted for 15% of the variance in overall satisfaction.  Visitors who had traveled a distance of more than 50 miles showed a similar pattern, with the services (Beta = .13), facilities (Beta = .13), and recreation experience (Beta = .13) domains contributing equally to overall satisfaction.  The independent variables in this model accounted for 15% of the variance in overall satisfaction.  Once again, information was the only domain that was not a significant predictor of overall satisfaction for either group.

The last independent variable examined under this research question was previous experience (first-time visitors versus repeat users).  This multiple regression analysis broke from the pattern of the previous regression models tested in this analysis.  The single significant predictor for first-time visitors was the recreation experience domain (Beta = .21), and this model accounted for about 15% of the variance in overall satisfaction.  For repeat visitors, the pattern was much like that seen for the other independent variables tested.  The strongest predictor was the services domain (Beta = .16), followed by the recreation experience domain    (Beta = .15), and the facilities domain (Beta = .14).  This model also accounted for about 15% of the variance in overall satisfaction.

It should be noted that the proportion of variance accounted for in this study is relatively low; similar to the findings in many previous recreation research efforts.  For example, Pizam and Milman (1993) accounted for only 15% of the variance associated with satisfaction in a pleasure trip to Spain.  The use of only four domains of satisfaction as independent variables, coupled with a single-item indicator of the respondents' overall satisfaction levels, should also be addressed as possible factors causing the low levels of variance accounted for in this study.  The use of a multiple-item scale as the dependent variable, instead of single-item measures, has been shown to account for a greater proportion of the variance in multiple regression analyses (Graefe & Fedler, 1986)

This study examined several potential segmentation variables to better understand differences in satisfaction levels across various user groups.  First, respondents were segmented based on their self-described primary recreation activity (ramp use, camping, or day use), focusing on those activities that typically occur at Corps of Engineers recreation areas.  Visitors were then segmented based on other demographic variables in order to identify other meaningful differences between various groups of lake users.

Very different results were found for the two research questions addressed in this study.  The first question, focusing on comparing reported satisfaction levels across various user segments, found many significant differences between subgroups of recreationists.  For example, campers were generally more satisfied than the other activity groups with all aspects of their outings.  Perhaps the level of development and services in Corps of Engineers campgrounds is greater than that offered at the boat ramps and day use areas, resulting in more satisfied users.

Likewise, visitors traveling with children and older visitors tended to report higher satisfaction levels.  Those traveling greater distances from home were also more satisfied with the facilities, services, and information they encountered.  The finding that visitors from greater distances reported higher satisfaction levels may be a result of the degree of investment that they have in their trip.  It may be that they had traveled quite a distance to recreate, and they were going to have a good time regardless of any minor annoyances encountered on the trip.  Women reported higher overall satisfaction than men, while the two genders did not differ in satisfaction with any of the individual domains.

The higher satisfaction scores reported by repeat visitors, compared to first-time visitors, depart from previous literature on satisfaction and experience use history.  Earlier studies suggest that experienced visitors may be harder to please based on their exposure to presumably less crowded conditions during earlier visits (Vaske, Donnelly, & Heberlein, 1980), or as a result of more specific or refined expectations for their experience (Schreyer, Lime & Williams, 1984).  The higher satisfaction scores reported by repeat visitors in this study may reflect improvements that have been made at the recreation areas during recent years.

Regarding the second research question, the comparison of regression models across various user characteristics showed only minor differences.  Thus, although satisfaction levels varied across segments of visitors, the variables influencing overall satisfaction were remarkably stable.  Overall satisfaction was correlated to all four customer service domains, and varying combinations of the domains accounted for 13 to 19% of the variance in overall satisfaction.  The strongest predictive model was found for day users, in which 19% of the variance in overall satisfaction was accounted for by satisfaction with facilities, services, and information.

It is noteworthy that the recreation experience domain was one of the stronger correlates of overall satisfaction, and was a significant predictor of overall satisfaction for all user segments except day users.  Perhaps day users rely more heavily than the other user segments on the more tangible aspects of customer service (i.e. facilities, services and information).

The information domain, while significantly correlated with overall satisfaction, rarely contributed to the explanation of overall satisfaction after the other customer service domains were already accounted for.  Again, day users stood out as the only visitor segment for which information was a significant predictor of overall satisfaction.  While information is important to all types of users, managers of day-use areas should pay special attention to providing up-to-date information for their visitors.

The same management process should be used in focusing on the people who use the recreation areas.  Although the sample in this study appears to be quite homogenous, the results of the study show that significant differences emerge when the sample is closely examined.  Examining the differences in the various segments of the population who recreate at the lakes should result in a better understanding of how to provide improved facilities, services, information and, ultimately, the recreation experience.

Further research is necessary that would focus upon including a broader range of domains, or specific items, that might influence outdoor recreation satisfaction.  In addition, this study should be replicated in different types of settings, including commercial recreation areas, that might allow greater generalizability to various types of outdoor settings.

Absher, J.D., Howat, G., Crilley, G., & Milne, I.  (1996).  Toward customer service: Market segment differences for sports and leisure centres.  Australian Leisure,
        7(1): 25-28.

Absher, J.D., & Lee, R.G.  (1981).  Density as an incomplete cause of crowding in backcountry settings. Leisure Sciences, 4, 231-247.

Andereck, K.L., & Caldwell, LL.  (1994).  Motive-based segmentation of a public zoological park market.  Journal of Park and Recreation Administration, 12(2),
        19-31.

Backman, S.  (1994).  Using a person-situation approach to market segmentation.  Journal of Park and Recreation Administration, 12(1), 1-16.

Bearden, E.W., & Teel, J.E.  (1983).  Selected determinants of consumer satisfaction and complaint reports.  Journal of Marketing Research, 20, 21-28.

Borrie, W. T., N. Christensen, A. E. Watson, T. A. Miller, & McCollum, D.W.  (2002). Public purpose recreation marketing: A focus on the relationships between
        the public and public lands. Journal of Park and Recreation Administration, 20(2), 49–68.

Bryan, H.  (1977).  Leisure value systems and recreation specialization: The case of trout fishermen.  Journal of Leisure Research, 9(3), 174-187.

Bullaro, J.J. & Edginton, C.R. (1986).  Commercial leisure services.  New York: Macmillan Publishing Company.

Burns, R. C., & Titre, J.  (1996).  Customer satisfaction analysis at Lake Sakakawea, North Dakota, Technical management report for US Army Corps of
        Engineers, Omaha District, Omaha, NE.

Burns, R. C., Graefe, A.R., Thapa, B. & Titre, J.  (1997).  Customer satisfaction analysis at Fort Randall Project, South Dakota: Technical management report for
        US Army Corps of Engineers, Omaha District, Omaha, NE.

Carman, J., & Langeard, E.  (1980).  Growth strategies of service firms.  Strategic Management Journal, 1, 7-22.

Churchill, G. A., & Suprenant, C.  (1982).  An investigation into the determinants of customer satisfaction.  Journal of Marketing Research, 16, 491-504.

Donnelly, M.P., Vaske, J.J., DeRuiter, D.S., & King, T.B.  (1996).  Person-occasion segmentation of state park visitors. Journal of Park and Recreation
        Administration, 14(2), 95-106.

Graefe, A.R. (1981).  Understanding diverse fishing groups:  The case of drum fishermen.  In Marine recreational fisheries VI  (pp. 69-79).  Washington, DC:  The
    Sport Fishing Institute.

Graefe, A.R., & Fedler, A.J.  (1986).  Situational and subjective determinants of satisfaction in marine recreational fishing. Leisure Sciences, 8, 275-295.

Galloway, G. (2002).  Psychographic segmentation of park visitor markets:  Evidence for the utility of sensation seeking.  Tourism Management 23, (6), 581-596.

Gronroos, C.  (1982).  Strategic management and marketing in the service sector.  Helsingfors, Sweden:  Swedish School of Economics and Business
        Administration.

Howat, G., Absher, J.D., Crilley, G., & Milne, I.  (1996).  Measuring customer service quality in sports and leisure centres. Managing Leisure, 1, 77-89.

Jang, S. C., Morrison, A.M., & O’Leary, J.T.  (2002).  Benefit segmentation of Japanese pleasure travelers to the USA and Canada: Selecting target markets
        based on the profitability and risk of individual market segments.  Tourism Management, 23 (4), 367-378.

Kuss, F.R., Graefe, A.R., & Vaske, J.J.  (1990).  Visitor impact management: A review of research, Volume 1.  Washington, D.C.:  National Parks and
        Conservation Association.

MacKay, K.J., & Crompton, J.L.  (1990). Measuring the quality of recreation services.  Journal of Park and Recreation Administration, 8 (3), 47-56.

Parasuraman, A., Zeithaml, V.A., & Berry, L.L.  (1988).  SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality.  Journal of
        Retailing, 64, 12-40.

Parasuraman, A., Zeithaml, V.A., & Berry, L.L.  (1985).  A conceptual model of service quality and its implications for further research.  Journal of Marketing, 49,
        41-50.

Park, M., Yang, X., Lee, B., Jang, H., & Stokowski, P.  (2002).  Segmenting casino gamblers by involvement profiles: A Colorado example. Tourism
        Management, 23 (1), 55-65.

Pizam, A., & Milman, A.  (1993).  Caution in the use of difference scores in consumer research.  Journal of Consumer Research, 19 (March), 655-662.

Rust, R.T. & Oliver, R.L. (1994).  Service quality: Insights and managerial implications from the frontier.  Service Quality: New Directions in Theory and Practice.
        Sage Publishing, Thousand Oaks, CA.  pp.1-19.

Schreyer, R., Lime, D., & Williams, D.  (1984).  Characterizing the influence of past experience on recreation behavior.  Journal of Leisure Research 16(1), 34- 50.

Tinsley, H.E., & Kass, R.  (1978).  Leisure activities and need satisfaction.  Journal of Leisure Research, 9, 109-120.

Upah, G.D.  (1980).  Mass marketing in service retailing:  A review and synthesis of major methods.  Journal of Retailing, 56, 59-76.

Vaske, J., Donnelly, M. & Heberlein, T.  (1980).  Perceptions of crowding and resource quality by early and more recent visitors. Leisure Sciences, 3(4),
        367-381.

Wade, D.J. and Eagles, P.J.  (2003).  The use of importance-performance analysis and market segmentation in parks and protected areas:  An application to
        Tanzania’s National Parks. Journal of Ecotourism, 2(3), 196-212.

Westover, T. N.  (1984).  Perceptions of crime and conflict in urban parks and forests.  Unpublished doctoral dissertation, Michigan State University.

Zeithaml, V.A. (1981).  How consumer evaluation processes differ between goods and services.  In J. Donnelly and W. George (Eds.), Marketing of Services
        (pp.186-190). Chicago: American Marketing Association.