Tourism Management 23 (2002) 521–530

A structural equation model of residents’ attitudes for tourism

development
Dong-Wan Koa,*, William P. Stewartb
aDivision of Tourism Science, Kyonggi University, San 94-6, Yiui-Dong, Paldal-Gu, Suwon-Si, Kyonggi-Do 442-760, South Korea bDepartment of

Leisure Studies, University of Illinois at Urbana-Champaign, 104 Huff Hall, 1206 S. Fourth Street, Champaign, IL 61820, USA
Received 3 July 2001; accepted 4 January 2002
Abstract
This study tests the structural equation model between residents’ perceived tourism impacts and attitudes toward host community.
The model consisted of five latent constructs and nine path hypotheses and is based upon 732 mailback questionnaires returned by
residents of Cheju Island, Korea, a major domestic tourism destination. It was found that residents’ ‘community satisfaction’ was
closely related to ‘perceived positive’ and ‘perceived negative’ tourism impacts. These constructs were directly causing ‘attitudes
toward additional tourism development’. But the hypothesized path relationships between ‘personal benefits from tourism
development’ and the constructs of ‘perceived negative tourism impacts’ and ‘overall community satisfaction’ were rejected. In
conclusion, community satisfaction was influenced by perception of tourism impacts, and may be useful in planning for additional
tourism development. r 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Tourism impacts; Overall community satisfaction; Residents’ perception and attitudes
1. Introduction
Tourism development is widely viewed as an im- portant set of economic activities to enhance local economies.
Many studies have suggested that develop- ment and promotion of tourism is a source of new employment, revenues,
additional tax receipts, foreign exchange benefits, and enhances community infrastruc- ture that will, in turn, attract
other industries (Lankford & Howard, 1994). Until recently, the development and promotion of the tourism industry
has been widely accepted as a positive economic step, especially in less developed countries (Cooke, 1982).
The term ‘tourism impact’ has been gaining increasing attention in the tourism literature. A number of studies in
recent years have examined host residents’ perception of the impact of tourism development on their commu- nity,
and it continues to be an important issue. A major reason for rising interest has been the increasing evidence that
tourism development leads not only to
positive, but also has the potential for negative, outcomes at the local level (Lankford & Howard, 1994). Liu and Var
(1986) noted that tourism develop- ment is usually justified on the basis of economic benefits and challenged on the
grounds of social, cultural, or environmental destruction. Furthermore, the economic benefits traditionally associated
with tourism development are now being measured against its potential for social disruption (Cooke, 1982). Huang
and Stewart (1996) indicated that tourism development may change residents’ relationships to one another and to
their community. It is generally felt that the perception and attitudes of residents toward the impacts of tourism are
likely to be an important planning and policy consideration for successful development, mar- keting, and operation
of existing and future tourism programs (Ap, 1992).
Although many studies have been performed to identify residents’ perception of tourism impacts and attitudes
toward tourism, just a few have assessed relationships between tourism development and com-
*Corresponding author. Tel.: +82-31-249-9509; fax: +82-31-249- 9503.E-mail addresses: [email protected] (D.-W. Ko), wstewart
munity satisfaction. To date, little research has exam- ined the relationships between residents’ perceived impacts of
tourism on their community and attitudes @uiuc.edu (W.P. Stewart).
toward their own community. Residents must perceive
0261-5177/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 1 - 5 1 7 7 ( 0 2 ) 0 0 0 0 6 - 7
tourism in overall positive terms to sustain development of esearch.
tourism programs. In this context, relationships between A significant portion of the social impact of tourism
residents’ perception of tourism impacts and community terature suggests that stakeholder involvement and
satisfaction are an important, yet not well understood, area of ommunity-based planning should be a part of the early stages
of tourism development (Jamal & Getz, 1995). It has been appear to be confined to certain dimensions of community life
argued that when residents are involved in the planning process, related to public services and opportunities for civic
then tourism development will be socially responsible and involvement. Un- fortunately, their study did not discuss
social impacts will be perceived as appropriate by the host relationships between residents’ perception of tourism impacts
community (Robson & Robson, 1996). However most studies and community life satisfaction. Also, Allen, Haffer, Long, and
directed at residents’ perceptions of tourism and Perdue (1993) found that residents agreed that their community
community-based planning have been conducted in the west: should attract more tourists because this would lead to a higher
Canada (e.g., Ritchie, 1993), the US (e.g., Ap & Crompton, quality of life.
1993), the UK (e.g., Robson & Robson, 1996), or Australia Although very little research has directly examined the
(e.g., Brown & Giles, 1994). Whether or not community influence of personal benefits from tourism on perception of
involvement effects social impacts within Asian coun- tries is impacts, numerous authors have inter- preted the observed
still an open point that has yet to be fully examined. relationship between resident characteristics and perceptions of
Currently, there is limited understanding of the impact as supporting a positive relationship between personal
relationships between residents’ perception of tourism impacts benefits from tourism and favorable perceptions of tourism
and community satisfaction. The lack of such research limits impacts (Perdue, Long, & Allen, 1990). However, most of these
the current literature on understanding residents’ behavior studies do not test the assertion that residents’ percep- tion of
toward the impacts of tourism (Ap, 1992). For a tourism-related tourism impacts influences attitudes toward community
economy to sustain itself, residents must be willing partners in satisfaction.
the process. Because of the frequency of interaction between As a notable exception, Perdue et al. (1990) developed a
residents and tourists, their willingness to serve as graciousmodel that examined relationships between resident’s
hosts is critical to the success of tourism. Therefore, residents
perception of tourism impacts and their support for it. They
must be involved in the planning and their attitudes toward tested a model that hypothesized relationships among rural
tourism and perceptions of its impact on community life must resident perceptions of tourism impacts, support for additional
be continually assessed (Allen, Long, Perdue, & Kieselbach, tourism development, restrictions on tourism development, and
1988). The purpose of this study is to demonstrate a structural
support for special tourism taxes. They found that when
model that explains the relationships between the residents’ controlling for personal benefits from tourism development,
percep- tion of tourism impacts and attitudes toward host perception of impacts were unrelated to socio-demographic
community. char- acteristics and that support for additional development
was positively related to perceived positive impacts of tourism.
Support for additional tourism development was negatively
2. Conceptual model integrating community satisfaction related to the perceived positive future of the community. Their
conclusions were based on a multivariate regression analysis,
Although most of the studies of host community however goodness-of- fit was not fully addressed in their
tourism attitudes and perceptions have focused on differences reported process of model development.
in the perceived impacts of tourism among different types of The hypothetical model (Fig.1) is adapted from Perdue
local residents, a few have discussed relationships between et al. (1990) whose model consisted of five latent constructs
residents’ perception of tourism impacts and attitudes toward about tourism development and community satisfaction. Also,
their own community. it has nine path hypotheses, which are the relationships among
McCool and Martin (1994) found that Montana five latent constructs: personal benefits from tourism
residents were concerned that increasing levels of tourism development, positive perceived tourism impacts, negative
would crowd them out of local fishing, hunting, and other perceived tourism impacts, overall community satisfaction, and
recreation areas. Allen et al. (1988) revealed that relationships attitudes for addi- tional tourism development. Each path
between tourism development and satisfaction of various represents an hypothesized relationship with the direction of
dimensions of community life effect identified as either positive (+) or negative (À).
were generally nonlinear with citizen involvement, public Primary research questions are directed at the influence
services, and the environment being most sensitive to tourismof perception of tourism impacts on overall community
development. They argued that negative attitudes about tourismsatisfaction, and the extent to which community satisfaction
 effects attitudes for additional
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 522
tourism development. The model also hypothesizes relationships among five latent constructs. This research tests the
goodness-of-fit of the model and hypotheses with structural equation modeling.
3. Method
3.1. Study site
The study was conducted in Cheju Island in Korea during November and early December 1997. Cheju Island is the
largest and southernmost island of the Korean peninsula, is 1846 km2, oval-shaped, and in 1996 held a population of
524,000 people. The highest point is Halla Mountain at 1950m in elevation; most of the residents live on the coastal
plains at elevation of 300 m or less.
Cheju Island is one of the most popular tourist destinations in Korea. The Korean government initiated tourism
development on Cheju Island in the 1960s, and the local governments of the island made it a top priority in the
1970s. The main attractions of Cheju Island are its outstanding natural scenery of mountains rising from beautiful
coastlines, the cultural heritage of the island people, the historic ambience of the rural villages, and opportunities for
playing golf. According to the Cheju Statistical Yearbook (Cheju Do, 1997), there are 41 hotels with a collective
capacity of 5168 rooms. In 1996, 4,140,000 visitors (209,000 were international) visited the island and spent the
equivalent of US$ 1205 million that accounted for 28% of the gross income of Cheju Island. Tourism is the primary
business sector of the Cheju economy, with the tangerine- growing industry as second at US$ 720million. In
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 523
Perceived Positive Tourism Impacts
Overall
H5 +
Attitudes for Additional Tourism Development H1 + H4 + Personal Benefit from Tourism Development
H3 + Community
8-
Satisfaction
H2 - H6 -
Perceived Negative Tourism Impacts
H9 +
H8 - H7 -
Fig. 1. Relationships between residents’ perceived tourism and attitudes toward host community (Adapted from Perdue et al., 1990).
1997, various local governments initiated the develop- ment of 23 additional tourism sites for the island.
3.2. Procedures
The 1041 mailback questionnaires were delivered to a proportional stratified random sample of adults in each
community of Cheju Island. A 70% response rate resulted from 732 usable questionnaires returned. Respondents
consisted of 448 males (61%) and 284 females (39%). Their age ranged from 20 to 63 years with 61% being 40–49
years and 23% being 30–39 years old. Native residents of the island comprised 83% of respondents.
The items of this study were originally derived from a comprehensive review of existing literature. The items for
residents’ perception of tourism impacts were taken from seven existing empirical studies (e.g., Pizam, 1978; Belisle
& Hoy, 1980; Liu & Var, 1986; Milman & Pizam, 1988; Perdue, Long, & Allen, 1987; Lankford & Howard, 1994;
Haralambopoulos & Pizam, 1996). Although these combined studies reported 39 items about residents’ perception of
tourism impacts, this study used 24 items due to exclusion of uncommon items (e.g., increases morality, honesty,
politeness and manners, mutual confidence, and attitude toward work, increases exploitation of local natives) and
redundant items.
The items of community satisfaction were derived from studies related to Allen, Long, Perdue, and their colleagues.
Allen and Beattie (1984) and Allen et al. (1988) developed 33 items that grouped into seven dimensions of
community satisfaction: public services, economic, environment, medical services, citizen invol- vement, formal
education, and recreation services and
construct. opportunities. This study employed all but two elements
The criteria used in deciding whether to (adult education, dentists) that were not relevant in the
delete an item were its corrected item-to-total correla- case of Cheju Island.
tion and whether the elimination improved the corre- A 5-point Likert-type scale was utilized for most items
sponding alpha values (Parasuraman, Zeithaml, & in this study (5=strongly agreed or very satisfied;
Berry, 1988). In general, items with corrected item-to- 1=strongly disagreed or very dissatisfied). Maddox
total correlations below 0.30 were eliminated. (1985) recommended the use of a Likert-type scale in
Reliability analysis was performed in two stages. The tourism impact research due to its superior validity
first stage of reliability analysis was directed at scales (convergent and discriminant). An exception was the use
related to perceived tourism impact and overall com- of a binary scale (2=yes; 1=no) for ‘attitude for
munity satisfaction. The variable of these latent additional tourism development’.
constructs used summated rating scales comprised of Data analysis was performed in two stages. In the first
many items. The corrected items-total correlation and stage, reliability analysis was conducted using SPSS (ver.
Cronbach Alpha Coefficients for perceived tourism 7.5) to evaluate the stability and consistency for
impacts are shown in Table 1 (positive tourism impact) measured items. In the second stage, the evaluation of
and Table 2 (negative tourism impact). goodness-of-fit indices for the proposed structural
The items of positive perceived economic impacts equation model and testifying hypotheses were per-
(four items), positive perceived social and cultural formed by using Analysis of MOment Structure
impacts (six items), negative perceived social and (AMOS, ver. 3.6), and were estimated using asympto-
cultural impacts (five items) and negative perceived tically distribution free (ADF) method because the items
environmental impacts (three items) in Table 1 had of attitudes for additional tourism development were
Cronbach Alpha Coefficients of over 0.74 with no measured via binary scale. Standardized scores were
increase resulting if any of the items were deleted. This used in the second stage of the analysis.
Cronbach Coefficient Alpha exceeds Nunnally and Bernstein’s (1994) recommendation of 0.70, and sup- ports the
use of these items in each scale. 4. Results
The perceived positive environmental impacts scale (three items, Table 1) and perceived negative economic 4.1.
Reliability analysis
impacts scale (three items, Table 2) had Cronbach Alpha Coefficients under 0.70. Even though the three Reliability
analysis was used to evaluate the stability
items for perceived positive environmental impact had a and consistency for measured items of each latent
Cronbach Alpha Coefficient of 0.54, it would increase to
Table 1 Reliability of perceived positive tourism impact scales
Variable name and description Mean (SD) Items-total correlation Alpha if items
deleted
Economic impacts 0.80 Improves investment, development, and infrastructure in
the economy
3.3(0.9) 0.61 0.74
Increases employment opportunities 3.0(0.9) 0.66 0.72 Contributes to income and standard of living 3.1(0.9) 0.64 0.73 Improves town’s overall
tax revenue 3.3(0.8) 0.53 0.78
Social and cultural impacts 0.74 Improves quality of life 3.2(0.9) 0.51 0.70 Increases availability of recreational facilities and entertainment
2.8(0.9) 0.50 0.70 Improves understanding and image of different communities/cultures 3.0(0.8) 0.55 0.69 Increases demand for historical and
cultural exhibits 3.0(0.9) 0.48 0.71 Encourages variety of cultural activities 3.0(0.9) 0.49 0.71 Improves quality of police and fire protection
3.1(0.9) 0.37 0.74
Environmental impacts 0.54 Preserves environment and improves the appearance (and images) of
an areas
3.0(0.9) 0.23 0.64
Improves living utilities infrastructure (supply of water, electric, and
telephone, etc.)
3.4(0.8) 0.41 0.34
Improves public facilities (pavement, traffic network, and civic center) 3.6(0.8) 0.43 0.33
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 524
 atisfaction, this study used the scale as part of community
atisfaction after the item ‘cost of living’ was deleted. The
eak reliability of
able 2 Reliability of perceived negative tourism impact scales

0.64 if the item of ‘preserves environment and improves the

appearance (and images) of an areas’ was deleted. This item is
qualitatively distinct from the other two items that deal directly
the economic satisfaction sub-scale remains as a limitation of
with community infrastructure, which explains the high alpha
the empirical portion of this study.
(0.64) if this item is deleted. The items of negative perceived
economic impact had a Cronbach Alpha Coefficient of 0.65
with no increase resulting if any of the items were deleted. 4.2. Evaluation of proposed model
However, these two scales had items-total correlation of over
0.30, which is a general criterion for acceptable reliability. The proposed model hypothesized that there were
Therefore, this study used the two scales for perceived positive significant causal relationships among five latent con- structs of
environmental impact (after the item of ‘preserves environment ‘personal benefits of tourism development’, ‘perceived tourism
and improves the appearance (and images) of an areas’ was impacts (positive and negative)’, ‘overall community
deleted) and the scale for perceived negative economic impact satisfaction’ and ‘attitudes for additional tourism development’.
without any items deleted. The causal relation- ships represented the nine hypotheses in
the path model. Empirical evaluation of such hypotheses is
The corrected items-total correlation and Cronbach
complicated by the fact that latent constructs are not directly
Alpha Coefficient for community satisfaction are shown in
observable. Evaluation is based on sets of observed or
Table 3. The latent construct of community satisfac- tion
measured variables that serve as indicators of latent variables,
consisted of seven scales. Three of the scales had Cronbach
with the relationship between the observed and latent variables
Alpha Coefficient above 0.70, and three additional scales were
being estimated using factor analysis. The relationship among
above 0.60. Each of the scales whose coefficient was above
latent variables or among latent variables and surrogate latent
0.60 had items-total correlations greater than 0.30, and thus
variables (observed variables serving as single-item proxies for
were retained for analysis. The economic satisfaction scale
latent variables) is typically estimated using regression analysis.
(five items) had a Cronbach Alpha Coefficient of 0.42, which
would increase to 0.52 if the item ‘cost of living’ were deleted. Structural equation modeling (SEM) is a technique for
Two items within this scale had items-total correlation lower simultaneously estimating the relationships between observed
than the general criterion of 0.30. The results of reliability and latent variables (the measurement model), and the
analysis showed that the items of economic satisfaction have relationships among latent variables (the structural model).
weak reliability. However, since the dimension of economic SEM is a method that has gained popularity because it
satisfaction is a very important factor in discussing community combines confirmatory factor analysis and regression analysis
 to model a variety of psychological, sociological, and othe second stage of reliability analysis are
relationships (Lindberg & Johnson, 1997). ariable name and description Mean (SD) Items-total correlation
Alpha if items
The descriptions, corrected items-total correlation, and
deleted
Cronbach Alpha Coefficients for five observed variables of the

Economic impacts 0.65 Unfairly increased real estate cost and property taxes 3.3(0.9) 0.39 0.64 Increases cost of living 3.6(0.8) 0.52 0.47
Increases price of goods and services 3.6(0.9) 0.47 0.53

Social and cultural impacts 0.81 Increases traffic accidents 3.6(0.9) 0.44 0.82 Increases crime/robberies/vandalism 3.5(0.9) 0.64 0.77 Increases
alcoholism, prostitution, and sexual permissiveness 3.5(1.0) 0.70 0.75 Increases gambling/illegal games 3.4(1.0) 0.65 0.76 Increases exploitation
of local natives 3.3(1.0) 0.59 0.78

Environmental impacts 0.79 Damage natural environment and landscape 3.4(1.0) 0.70 0.70 Destroy local ecosystem 3.6(0.9) 0.66 0.66 Increases
environmental pollution (litter, water, air, and noise) 3.7(0.9) 0.78 0.78
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 525
 em, that if deleted would increase the scale’s Alpha
oefficient. If the item ‘positive environ-
able 3 Reliability of community satisfaction scales

presented in Table 4. The observed variables of ‘personal

benefits of tourism development’ and ‘atti- tudes for additional
tourism development’ were items measured directly. The mental tourism impacts’, or ‘PPTI’ was deleted, the Cronbach
observed variables of ‘positive perceived tourism impacts’, Alpha Coefficient would increase to 0.75. But, the marginal
‘negative perceived tourism impacts’, and ‘overall community improvement in Alpha was not deemed significant enough
satisfaction’ were the summated rating scales of the first stage compared to the items’ value in this study. Therefore, it was not
of the reliability analysis (Tables 1–3). deleted. Finally, the results of reliability for the latent constructs
in this study support the use of these observed variables.
In Table 4, Cronbach Alpha Coefficient exceeds
Fig. 2 shows the standardized model as estimated by
Nunnally and Bernstein’s (1994) recommendation of 0.70, and
supports the use of these observed variables. There was one AMOS. Each of the observed variables is displayed in a
 ariable name and description Mean (SD) Items-total correlation
rectangle, and each of the latent constructs is displayed in an
Alpha if items
oval. The evaluation of goodness-of-fit indices supported the
deleted
model. The χ2 test provides that the model generated χ2 1⁄4
376; df 1⁄4 110; po0:01; which

Public service satisfaction 0.73 Fire protection 3.5(0.9) 0.40 0.71 Welfare and social services (public assistant) 3.1(0.9) 0.45 0.70 Public
transportation to and from other community 2.5(1.1) 0.39 0.71 Police protection 2.9(0.9) 0.51 0.69 Local government 3.1(0.9) 0.45 0.70 Road
and highway 2.6(1.0) 0.44 0.70 Public health services 2.8(1.0) 0.46 0.70

Formal education satisfaction 0.63 Public schools (K through 12 programs) 3.2(0.9) 0.39 0.59 College university courses (for credit) 2.8(0.8)
0.57 0.32 Technical and/or vocation training for career 2.6(0.9) 0.35 0.63

Environment satisfaction 0.69 Physical geography or terrain 3.5(1.1) 0.38 0.67 Environmental cleanliness (air, water, soil) 3.4(0.9) 0.50 0.61
Climate and weather 3.8(0.8) 0.38 0.66 General appearance of your living town 3.7(0.8) 0.58 0.58 General appearance of your region (Cheju
Island) 4.0(0.8) 0.41 0.65

Recreation opportunities satisfaction 0.80 Private/commercial recreation (health clubs, movies, etc) 2.7(1.0) 0.61 0.77 Publicly funded recreation
(social, cultural, sports/fitness) 2.5(0.9) 0.74 0.64 Park and open space 2.6(0.9) 0.61 0.77

Economics satisfaction 0.42 Shopping facilities 2.7(1.1) 0.30 0.28 Cost of living 2.2(0.8) À0.30 0.52 Housing (cost and availability) 2.7(0.9) 0.28
0.31 Utilities (water, gas, electricity, sewage) 3.8(0.8) 0.20 0.37 Job opportunities 2.3(0.9) 0.32 0.28

Citizen involvement and social opportunities 0.66 Opportunities to be with friends and relatives 3.2(0.9) 0.47 0.58 Citizen input into community
decisions 2.8(0.9) 0.46 0.59 Religious organizations (church/temple) 2.9(0.9) 0.31 0.65 Opportunities in civic and fraternal organizations 3.0(1.1)
0.41 0.61 Opportunities to become familiar with other residents 2.7(0.9) 0.42 0.60

Medical services satisfaction 0.79 Hospital and medical facilities 2.6(1.0) 0.57 0.78 Medical doctors 2.6(0.9) 0.67 0.67 Emergency services
2.5(0.9) 0.65 0.69
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 526
indicates a marginal fit (normed χ2 1⁄4 3:42). Because this χ2 test is sensitive to sample size (n 1⁄4 732 in this
study), supplementary measures have been developed. The
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 527
Table 4 Reliability analysis of observed variables
Latent constructs and description of observed variables Corrected items-total correlation Alpha if items deleted
PBTD Personal benefits from Tourism Development 0.73 pnjobrl Relationship with tourism of personal job 0.57 — fmjobrl Relationship with
tourism of family job 0.57 —
PPTI Positive perceived tourism impacts 0.73 pecoimp Positive economic tourism impacts 0.58 0.60 pscimp Positive social and cultural tourism
impacts 0.63 0.57 penimp Positive environmental tourism impacts 0.47 0.75
NPTI Negative Perceived Tourism Impacts 0.72 necoimp Negative economic tourism impacts 0.49 0.68 nscimp Negative social and cultural
tourism impacts 0.62 0.52 nenimp Negative environmental tourism impacts 0.51 0.67
OCS Overall community satisfaction 0.77 satps Public service satisfaction 0.57 0.73 sated Formal education satisfaction 0.55 0.74 satenv
Environmental satisfaction 0.32 0.77 satrec Recreation satisfaction 0.48 0.75 sateco Economic satisfaction 0.58 0.73 satci Citizen involvement
satisfaction 0.50 0.75 satms Medical service satisfaction 0.52 0.74
AATD Attitudes for additional Tourism Development 0.80 atdcheju Support level in Cheju regional contexts 0.67 — atdcomn Support level in
living community contexts 0.67 —
.69 .81 .52
a
.277
a
.360
a
.634
.080
.66 .63 .41 .80
.74
.78
pecoimp pscimp nenimp
PPTI
 PBTD
necoimp nscimp
nenimp satps sated
pnjobrl satenv

-.121 .96
atdcheju
atdcomn
satci
Fig. 2. Standardized estimated hypothetical model (a) and (b) indicate significance at the 0.01 and 0.05 levels, respectively. Dashed lines indicate
paths that are not significant at 0.05 or better.
OCS
AATD fmjobrl
b a
-.101 .66 .63 satrec .106

-.244 a sateco -.013 .62

.68
satms
PNTI
.60 .81 .69
other goodness-of-fit indices indicated a good fit within accepted exhortation levels. The goodness-of-fit index (GFI)
is acceptable at 0.926, the root mean square error
of approximation (RMSEA) is acceptable at 0.058, and also the ourism development’ are positively related to ‘overall
root mean square residual (RMR) is acceptable at 0.044. ommunity satisfaction’) was rejected at a significant level of
Path hypothesis 1 (i.e., ‘personal benefits from 5% (two tailed test, t > 1:96), but supported with a marginal
tourism development’ are positively related to ‘positive vel at t 1⁄4 1:625 (po0:05) and β of 0.08. If the path
perceived tourism impacts’) was supported with an optimal elationship between ‘personal benefits from tourism
level at t 1⁄4 6:168 (po0:001) and β 1⁄4 0:36: But path evelopment’ and ‘perceived negative tourism impacts’ was
hypothesis 2 (i.e., ‘personal benefits from tourism eleted (due to rejection of hypothesis), the t-value decreased to
development’ are negatively related to ‘perceived nega- tive 1⁄4 1:587 (p > 0:10). These results suggest that ‘personal
tourism impacts’) was rejected at t 1⁄4 À0:026 and β 1⁄4 enefits from tourism develop- ment’ does not contribute to
À0:013: ttitude toward ‘overall community satisfaction’. Tourism
Many studies have supported a causal relationship evelopment is widely perceived as an important community
between ‘personal benefits from tourism development’ and evelopment strategy. It converges with intuition that ‘overall
‘perception of tourism impacts’ (e.g., Lindberg & Johnson, om- munity satisfaction’ would be positively related to
1997; Madrigal, 1993; Perdue et al., 1990). Support for this personal benefits from tourism development’. However, this
relationship aligns with common sense that residents (or their elationship was not significant at po0:05:
relatives, friend, and neighbors) who depend upon Path hypothesis 4 (i.e., ‘perceived positive tourism
tourism-based employment would be more favorable toward mpacts’ are positively related to ‘overall community
tourism (e.g., Liu & Var, 1986; Milman & Pizam, 1988; atisfaction’) was supported with an optimal level at t 1⁄4
Murphy, 1983; Pizam, 1978). Although the related hypothesis 0:10 (po0:001) and β 1⁄4 0:634: Also, path hypothesis 5
between ‘personal benefits from tourism development’ and .e., ‘perceived positive tourism impacts’ are positively related
‘perceived negative tourism impacts’ was supported at t 1⁄4 o ‘attitude for additional tourism development’) was supported
J7:9 (po0:001) and β 1⁄4 À0:157 by Perdue et al. (1990), this ith an optimal level at t 1⁄4 4:142
study did not find a significant relationship. It could be that the (po0:001) and β 1⁄4 0:291: Path hypothesis 6 (i.e., ‘perceived
relationship between ‘personal benefits from tourism negative tourism impacts’ are negatively related to ‘overall
development’ and ‘perceived negative tourism impacts’ was community satisfaction’) was sup- ported with an optimal level
strongly dependent upon the level of tourism development or at t 1⁄4 À2:319 (po0:05) and β 1⁄4 À0:101: Path hypothesis
residents’ education level. An important finding of these results 7 (i.e., ‘perceived negative tourism impacts’ are negatively
is the lack of a significant relationship between benefits from related to ‘attitude for additional tourism development’) was
tourism and perceived negative impacts. supported with an optimal level at t 1⁄4 À5:923 (po0:001) and
Path hypothesis 3 (i.e., ‘personal benefits from β 1⁄4 À0:244: These results generally converge with those of
 previous research. Perdue et al. (1990) have reported that thi tourism tend to favor tourist activity (e.g., Pizam, 1978;
relationship (hypothesis 7) was supported at β 1⁄4 À27Brougham & Butler, 1981). Perdue et al. (1990) also found a
(po0:001). Finally, path hypothesis 8 (i.e., ‘overall communitystrong positive relationship between personal benefits from
satisfaction’ is negatively related to ‘attitude for additiona tourism and attitudes for additional tourism development.
tourism development’), although nega- tive, was no
statistically significant at po0:05 (the t 1⁄4 À1:895 (po0:10
and β 1⁄4 À0:121). 5. Conclusions
Many researchers have suggested that residents
attitudes toward tourism may be related directly to the degree The purpose of this research was to test a model of
and/or stage of development within the host community (e.g. residents’ perceptions and attitudes of tourism, examin- ing not
Doxey, 1975; Williams, 1979; Butler, 1980; Cooke, 1982; Getz only perceptions of tourism impacts, attitudes toward
1983; Haywood, 1986). These studies suggest that communitie community satisfaction, and additional tourism development,
have a certain capacity to absorb tourists (Allen et al., 1988) but also the path relationships between perceptions of benefits
Therefore, the lack of a significant negative relationshipfrom tourism development, positive and negative tourism
between ‘community satisfaction’ and ‘attitude toward addi impacts, community satisfaction, and attitudes for additional
tional tourism development’ may be due to the long history o tourism devel- opment. The results support six hypotheses from
tourism development at Cheju Island. previous studies, but three of the hypotheses were not supported
Path hypothesis 9 (i.e., ‘personal benefits from tourismat po0:05 level.
development’ are positively related to ‘attitude for additiona This study found that ‘residents’ community satisfac-
tourism development’) was supported with an optimal level at tion’ was closely related to ‘perceived positive tourism impacts’
1⁄4 2:635 (po0:01) and β 1⁄4 0:106: Many studies haveand ‘perceived negative tourism impacts’. Both perceived
found that residents who are economically dependent onpositive and negative impacts were directly
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 528
causing ‘attitudes toward additional tourism develop- ment’. Although this study supported some important
Perdue et al. (1990) found that resident support for additional elationships between residents’ perceptions of tourism impacts
tourism was negatively associated with a positive future for the nd community satisfaction, the limitations include a weak
community. Tourism develop- ment is viewed not as a goal but eliability on community economic satisfaction sub-scale along
as a tool or means of community development. ith a sample that may not be representative of the population
Community satisfaction may be a useful concept for e.g., 84% of respondents were between 30 and 49 years of
evaluation of residents’ perception of tourism impacts and ge). Many investigations have suggested that residents’
attitudes for additional tourism development. Therefore, further ttitudes toward tourism may be directly related to the degree or
research in this field is needed to discuss integrating age of development within the host community, which may
community satisfaction with tourism development. But, the artially explain the findings. Further research needs to address
hypothesis that ‘personal benefits from tourism development’ ommunity satisfaction, level of tourism development, and
would be negatively related to ‘perceptions of negative tourism ttitude toward additional tourism development.
impacts’ was rejected, and contrasts to findings of past research
(e.g., Perdue et al., 1990). The implications suggest that
personal benefits from tourism are relevant to under- standing
perceptions of positive impacts but, in this case, irrelevant toReferences
understand perceptions of negative impacts. This may be due to
llen, L. R., & Beattie, R. (1984). The role of leisure as an indicator of overall
the level of development on Cheju Island, which has a tisfaction with community life. Journal of Leisure Research, 16(2), 99–109.
longstanding history of tourism as an integral part of its llen, L. R., Haffer, H. R., Long, P. T., & Perdue, R. R. (1993). Rural residents’
economy. In other words, the length of time and history of titudes toward recreation and tourism development. Journal of Travel
tourism development within a host community may be relevant esearch, 31(4), 27–33. Allen, L. R., Long, P. T., Perdue, R. R, & Kieselbach,
to understanding residents’ perceptions of tourism (see also (1988). The impacts of tourism development on residents’ perceptions of
ommunity life. Journal of Travel Research, 26(1), 16–21. Ap, J. (1992).
Brown & Giles, 1994; Ryan, Scotland, & Montgomery, 1998), esidents’ perceptions on tourism impacts. Annals of
and should be addressed as part of future research efforts. Tourism Research, 19(4), 665–690.
 Ap, J., & Crompton, J. (1993). Respondents’ strategies for responding (1995). Collaboration theory and community tourism planning. Annals of
to tourism impacts. Journal of Travel Research, 32(1), 47–50. Belisle Tourism Research, 22(1), 186–204. Lankford, S. V., & Howard, D. R. (1994).
F. J., & Hoy, D. R. (1980). The perceived impact of tourism by residents: a caseDeveloping a tourism impacts attitude scale. Annals of Tourism Research,
study in Marta Colombia. Annals of Tourism Research, 7(1), 83–101. 21(1), 121–139. Lindberg, K., & Johnson, R. L. (1997). Modeling resident
Brougham, J. E., & Butler, R. W. (1981). A segmentation analysis of resident attitudes
attitudes to social impact of tourism. Annals of Tourism Research, 7(4), toward tourism. Annals of Tourism Research, 24(2), 402–427. Liu, J. C., & Var,
569–590. Brown, G., & Giles, R. (1994). Resident responses to the social T. (1986). Resident attitudes toward tourism impacts
impact of tourism. In A. Seaton, et al. (Ed.), Tourism: A state of the art (pp. in Hawaii. Annals of Tourism Research, 13(2), 193–214. Maddox, R.
755–764). Chichester: Wiley. Butler, R. W. (1980). The concept of a tourism N. (1985). Measuring satisfaction with tourism. Journal of
areas cycle of evaluation: implications for management of resources. Canadian
Travel Research, 23(3), 2–5. Madrigal, R. (1993). A tale of tourism in
Geographer, 24(1), 5–12. Cheju, DO. (1997). Cheju Statistical Yearbook,
two cities. Annals of Tourism
Cheju: Cheju Province
Research, 20(2), 336–353. McCool, S. F., & Martin, S. R. (1994).
Government. Cooke, K. (1982). Guidelines for socially appropriate
Community attachment and attitudes toward tourism development. Journal of
tourism development in British Columbia. Journal of Travel Research, 21(1),
Travel Research, 32(3), 29–34. Milman, A., & Pizam, A. (1988). Social impacts
22–28. Doxey, G. V. (1975). A causation theory of visitor-resident irritants’
of tourism on Central
methodology and research inferences. Proceedings of the Sixth Annual
Florida. Annals of Tourism Research, 15(2), 191–204. Murphy, P. E.
Conference of the Travel Research Association (pp. 195–198), San Diego CA:
(1983). Perceptions and attitudes of decision making groups in tourist centers.
Travel and Tourism Research Association. Getz, D. (1983). Capacity to absorb
Journal of Travel Research, 21(3), 8–12. Nunnally, J. C., & Bernstein, I. H.
tourism: concepts and implica- tions for strategic planning. Annals of Tourism
(1994). Psychometric theory (3rd
Research, 10(2), 239–263. Haralambopoulos, N., & Pizam, A. (1996).
Perceived impacts of tourism: the case of Samos. Annals of Tourism Research, ed). New York: McGraw-Hill. Parasuraman, A., Zeithaml, V., &
23(3), 503–526. Haywood, K. M. (1986). Can the tourist area life cycle be Berry, L. L. (1988). Servqual: a multiple-item scale for measuring consumer
made perceptions of services quality. Journal of Retailing, 64, 12–40. Perdue, R. R.,
Long, P. T., & Allen, L. R. (1990). Resident support for tourism development.
operational? Tourism Management, 7(3), 154–167. Huang, Y., &
Annals of Tourism Research, 17(4), 586–599.
Stewart, W. P. (1996). Rural tourism development: shifting basis of community
solidarity. Journal of Travel Research, 36(4), 26–31. Jamal, T., & Getz, D.
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 529
Perdue, R. R., Long, P. T., & Allen, L. R. (1987). Rural resident of tourism issues for tourism marketers. Tourism Management, 17(7), 533–540. Ryan, C.,
perceptions and attitudes. Annals of Tourism Research, 14(3), 420–429. Pizam, Scotland, A., & Montgomery, D. (1998). Resident attitudes to tourism
A. (1978). Tourism impacts: the social costs to the destination community as development: a comparative study between the Rangitikei, New Zealand and
perceived by its residents. Journal of Travel Research, 16(4), 8–12. Ritchie, J. Bakewell, United Kingdom. Progress in Tourism and Hospitality Research,
(1993). Crafting a destination vision: putting the concept of resident-responsive 4(2), 115–130. Williams, T. A. (1979). Impact of domestic tourism on host
tourism into practice. Tourism Management, 14(5), 379–389. population the evolution of a model. Tourism Recreation Research, 14(1),
Robson, J., & Robson, I. (1996). From shareholders to stakeholders: critical 15–21.
D.-W. Ko, W.P. Stewart / Tourism Management 23 (2002) 521–530 530