A Systematic Approach to Scale Development in Tourist Shopping Satisfaction

Tourism Shopping

Most researchers agree in their findings that shopping is a major travel motive and an enjoyable leisure activity during a trip (Swanson and Horridge 2004; Timothy 2005; Hsieh and Chang 2006). Some scholars believe that for most tourists a trip is not complete without shopping (Keown 1989b; Heung and Cheng 2000). A survey conducted by the Travel Industry Association of America showed that 63 percent of tourists went shopping (Gentry 2001). Shopping is also found as the main objective of tourists in popular shopping destinations such as Hong Kong and Paris (Mak et al. 1999; Lehto et al. 2004).Tourists often spend more money on shopping than on other activities such as dining, lodging, or entertainment (Kent, Shock, and Snow 1983; Turner and Reisinger 2001). The Travel Industry Association of America (2000) reported that tourist shoppers often spend three to four times more than regular shoppers (cf. Rosenbaum and Spears 2005). Hence, a well-managed tourist shopping experience is essential to travel destinations not only because of the economic impacts of tourism shopping on host communities (Wong and Law 2003; Hsieh and Chang 2006), but also because it helps build a more favorable tourism image (Tosun et al. 2007).

The literature refers to tourism shopping as a different concept than regular shopping activity (LeHew and Wesley 2007). Tourists generally like to spend more money on purchasing high-quality items from well-known manufacturers and with renowned brand names because these attributes help enhance their confidence in making the right purchase decision in a foreign country (Littrell and Baizerman 1994; LeHew and Wesley 2007). Tourist shoppers also look for unique products that are not available in their hometown or are unique to a destination (Reisinger and Turner 2002). Many tourists do not have a specific shopping list; instead, shopping often occurs as an unplanned travel experience (Thomas and LeTourneur 2001). Therefore, tourists are often motivated to shop by a number of tangible, intangible, and impulsive factors such as the salespeople whom they interact with, the uniqueness of products, and the sites they visit in a travel destination.

Shopping Satisfaction

Contemporary research on consumer satisfaction is largely based on the expectancy disconfirmation paradigm (Oliver 1980). This theory suggests that consumers’ satisfaction is a comparative judgment between expectation and disconfirmation. Other definitions of satisfaction commonly refer to a comparison between a consumer’s expectation and his perceptions of the performance of a service or product (Fornell et al. 1996; Seiders et al. 2005). Consumers are dissatisfied if the performance falls short, while they are satisfied if the performance exceeds their expectations.

While travel satisfaction has been widely acknowledged as a forward indicator of destination success and a crucial predictor of tourist behaviors (Kozak 2001; Rittichainuwat, Qu, and Leong 2003), the literature is just beginning to understand the role of tourist satisfaction in shopping (Hsieh and Chang 2006). Wong and Law (2003) noted that “the concern for measuring customer satisfaction in the tourism industry has been precipitated by the need to position destinations competitively in the worldwide marketplace” (p. 402). Tourist satisfaction has been found positively associated with tourists’ behaviors such as return visit intention, loyalty, and positive word-of-mouth (Dube and Renaghan 1994; Chadee and Mattsson 1995). Tourist satisfaction, therefore, serves as an indicator of whether a destination meets tourists’ needs. A multidimensional measure of satisfaction could help destinations to better understand their specific areas of strength and weakness in this increasingly competitive global tourism market (Wong and Law 2003).

Shopping satisfaction particularly depends on the shopping enjoyment that tourists experience (Murphy et al. 2011). It is often a result of the difference between tourists’ expectations and their perceptions of the performance of the retail stores and of the product they purchased (Wong and Law 2003). The existing literature has identified multiple attributes to measure satisfaction. Early studies tend to link product value with customer satisfaction, as the perceived value of a product is likely to fulfill customer needs (Zeithaml 1988; Fornell et al. 1996). More recent research, however, reveals that shopping satisfaction could be achieved beyond product attributes, for instance by the service offered by retailers (Buttle 1992; Christiansen and Snepenger 2002). In fact, service quality, including the service product, service environment, and service delivery, has been widely acknowledged to positively influence customer satisfaction (Rust and Oliver 1994; Brady and Cronin 2001).

The work of Lindquist (1975) further suggests that shopping satisfaction can broadly be categorized into nine facets: merchandise quality, service, clientele, physical facilities, convenience, promotion, store atmosphere, institutional factors, and posttransaction satisfaction. Merchandise quality refers to the quality, price, and selection of goods; service and clientele represent the intangible service provided by salespeople; physical facilities and store atmosphere indicate satisfaction with tangible service attributes such as the lighting, air conditioning, cleanliness, store layout, and appearance of a store; convenience refers to the location and parking access to a store; promotion corresponds to the advertising and promotional offerings; institutional factors include store reputation and specific service products that project a store’s image; and posttransaction refers to the return policy and other after-sales services (cf. Lin and Lin 2006).

In regard to specific tourist shopping satisfaction attributes, Keown (1989a) found that wide section of merchandise, faster and more efficient service, and good value for money are the three most important shopping attributes that tourists in Hong Kong seek. Heung and Cheng (2000) investigated international tourists in regard to their shopping experience in Hong Kong and identified four factors with 14 attributes that could influence tourists’ shopping satisfaction. These factors include tangible quality, staff service quality, product value, and product reliability (cf. Murphy et al. 2011). Turner and Reisinger (2001) surveyed the shopping satisfaction of Japanese tourists visiting Hawaii and the Gold Coast. They found 17 elements of shopping satisfaction, which cover tourists’ assessment of product value, staff service, price, and tangible quality of retail shops such as salespeople’s appearance, shop cleanliness, and convenience (cf. Reisinger and Turner 2002). Lin and Lin (2006) studied mainland Chinese tourists in Taiwan and found 20 attributes that contribute to shopping satisfaction, including knowledge of the sales staff, product price, product commemoration, hours the shops were open, and product packaging and size. This wide range of shopping attributes is primarily germane to retail store and merchandise characteristics and provides the theoretical foundation for the proposed shopping satisfaction scale.

Conceptualization of Tourist Shopping Satisfaction

On the basis of the aforementioned literature, we define tourist shopping satisfaction as a tourist’s subjective evaluation of his or her shopping experience with retail stores and merchandise purchased during his or her stay in a travel destination. Shopping satisfaction may be based on a tourist’s evaluation of the products he or she purchased, the staff service quality he or she received, the tangible store environment and service product he or she encountered, or special services and offerings that he or she found intriguing. We propose tourist shopping satisfaction (TSS) as a multidimensional construct that differs from other conceptualizations of shopping satisfaction in that it explicitly considers tourists rather than regular consumers. Our conceptualization is also different from the work of Heung and Cheng (2000) and Lin and Lin (2006) in that we also consider stores’ ability to provide special services, an important factor in enhancing satisfaction. In addition, prior scale-development studies on tourist shopping satisfaction were not carried out systematically as the literature recommends (Churchill 1979; Thompson 2004; Hair et al. 2006); hence scale validity and reliability are in question.

Destination Tourism Facility

The tourism literature is beginning to understand how destination-level attributes impact tourist behaviors. Yet it still remains an understudied area as “many regional factors affecting tourists’ perceived satisfaction with shopping have been ignored” (Tosun et al. 2007). Prior studies have revealed that tourists are motivated to travel to a destination by a number of push and pull factors such as safety and security, leisure entertainment options, accessibility and transportation systems, and hygiene and cleanliness of a destination and its facilities (Goeldner and Ritchie 2003). For example, most researchers agree that the safety of a destination is a primary motive and concern for most tourists (Reisinger and Mavondo 2006). Kim, Guo, and Agrusa (2005) note that safety is the most important destination attribute for Chinese tourists. Geographic accessibility and the local transportation system have also been documented as motives that pull tourists to specific destinations (Fenich 2001). The cleanliness and hygiene of a destination are further factors that motivate travel activities (Sangpikul 2008). It follows that if the aforementioned needs are not fulfilled, tourists are unlikely to find their trips satisfactory. In addition, the safety and cleanliness of a destination’s tourism facilities are found to be attributes that enhance tourists’ travel preference (Timothy 2005; Hsieh and Chang 2006; Sangpikul 2008). However, it is still unclear how these factors might influence tourists’ shopping satisfaction.

In regard to shopping facilities, LeHew and Wesley (2007) note that the location, convenience, and size of shopping centers are some of the attributes that shoppers use to assess their shopping satisfaction. Hsieh and Chang (2006) contend that tourists are concerned about the safety, traffic, accessibility, and cleanliness of a destination’s tourism facilities. Murphy et al. (2011) also find that these attributes play a role in attracting tourists to a shopping village. In addition, tourists prefer a large-scale shopping area where they can find a large selection of items and ample leisure and dining options.

The literature suggests that shopping areas or centers themselves are tourist attractions. Many are embedded within a larger tourist facility, such as the shopping area in an amusement park (e.g., Disney World or Fisherman’s Wharf), in a hotel resort or an integrated casino complex (e.g., the Venetians), along a beach (e.g., Waikiki Beach), in a block of streets or a public open area (e.g., Ladies’ Market in Hong Kong), or at a heritage site (e.g., the Korean Folk Village or the Senardo Square in Macau), or in other tourist facilities (Johnson 1990; LeHew and Wesley 2007). Accordingly, attributes of destination facilities should influence tourists’ shopping satisfaction. Given that shopping activities of the current study take place in stores or shopping centers within major tourist attractions, we believed that the safety, transportation system, accessibility, cleanliness, and scale of destination’s tourist facilities would have positive impacts on tourist shopping satisfaction. This assumption leads to the following hypotheses.

Shopping Experience

The extant literature has acknowledged the central role of customer experience in nurturing loyalty behaviors (Pine and Gilmore 1998; Baron and Harris 2008). Hsieh and Chang (2006) suggest that tourists ultimately seek a complete travel experience. Given that shopping is often a major travel activity, satisfied tourist shoppers are likely to enhance their overall travel experience including their shopping experience. This is confirmed by Wong and Law (2003), who found that satisfaction with shopping attributes leads to overall satisfaction with the shopping experience.

Other studies present similar findings showing that tourists develop positive attitudes toward their shopping experience based on a number of shopping attributes relating to the serving staff, the shopping environment and product, the merchandise value, and other service characteristics (Keown 1989a; Heung and Qu 2000; Tosun et al. 2007). For example, Heung and Qu (2000) find that tourist satisfaction of staff service quality, product value, and product reliability lead to a positive shopping experience. Tosun et al. (2007) also contend that “the tourist-shopping experience is the sum of tourist satisfaction or dissatisfaction gained from individual attributes of products and services purchased” (p. 88). Hence, we propose the following hypothesis.

Scale Generation and Initial Purification

We began the scale-development process by generating a pool of items from prior literature (e.g., Berry 1969; Heung and Cheng 2000; Reisinger and Turner 2002; Lin and Lin 2006; Tosun et al. 2007). In addition, we conducted in-depth interviews with 54 tourists who had shopping experience in Macau, China. We used open coding and carefully examined individual words, phrases, and sentences to develop the properties and categories pertaining to shopping satisfaction from the data. Codes and categories were compared and contrasted until saturation, which enhanced the fit between data and the results (Strauss and Corbin 1998).

Based on the literature and the findings from the interviews, we generated an initial set of 31 items. We then asked a panel of three scholars, who have domain knowledge in tourism shopping and satisfaction, to judge whether the items were appropriate and important as measures of shopping satisfaction. In particular, we asked the panel members to assess the importance of the 31 items using a 5-point semantic differential scale. The interrater reliability is high (Rwg = .94; cf. James, Demaree, and Wolf 1993). Eight items with a mean value below 3.0 were removed from the process, and 23 items remained.

Scale Refinement

To refine the scale, we collected data in six popular shopping areas including large shopping malls and open shopping areas in the world gambling capital: Macau, China. The research context is appropriate as it is a major travel destination in Asia and is emerging as a major tourism shopping locale. The population of interest are mainland Chinese tourists. Hence, the questionnaire was available in Chinese and was administered by Chinese native speakers. Items on the questionnaire were translated by two independent bilingual individuals who are fluent in both Chinese and English. The translated version of the questionnaire was compared for any inconsistencies, mistranslations, and lost meanings or words by the authors. Finally, the questionnaire was pilot tested, and respondents did not appear to have difficulty understanding the items. Each of the 23 items on the shopping satisfaction scale is measured with a 5-point option ranging from 1 (dissatisfied) to 5 (satisfied).

Data were collected in two waves, and a combined 1,213 surveys were distributed by means of face-to-face mall-intercept interviews during several weekends and public holidays. A total of 554 Chinese tourists agreed to participate in the survey, which corresponds to a 46 percent response rate. However, 55 of the responses were incomplete; hence, a total of 499 complete responses remained for data analysis.

Of the respondents, 59 percent were female, 36 percent were between 26 and 35 years of age while 29 percent were between 36 and 45, 64 percent earned a monthly income between USD 160 and 1,270, and 28 percent were white-collar workers while 14 percent were self-employed. In regard to respondents’ travel characteristics, 23 percent of the respondents had visited Macau more than five times while 25 percent of them were first-timers. A good portion of the respondents (43 percent) reported leisure as the primary travel motive, and 63 percent of them were spending less than two nights on their trip in Macau. The average expected spending on shopping was USD 720 and the actual shopping expense was 1,005.

We followed the recommendation from Hair et al. (2006) and randomly split the data set into two subsamples. The first subsample was subjected to exploratory factor analysis (EFA) using principal component extraction, while the second subsample was subjected to confirmatory factor analysis. Results of the EFA (with Promax oblique rotation) indicated that three items should be removed because of low factor loadings or cross-loadings, and the remaining 20 items were able to load to four factors of satisfaction: service product and environment, merchandise value, staff service quality, and service differentiation (see Table 1). The four-factor solution was able to account for 66.16 percent of variation in tourist shopping satisfaction. In addition, the Kaiser–Meyer–Olkin test (KMO = .92) revealed that the sample size was adequate. Cronbach’s alpha indicates that the four subscales are reliable with α >.70.

OPEN IN VIEWER

Table 1.

Item Descriptions and Measurement Model Results for Tourism Shopping Satisfaction

Scale Item Descriptions	Pattern Coefficient	Variance Explained	Cronbach’s α	Composite Reliability	AVE a
Service product and environment	γ b = 9.29	46.46%	.91	.90	.56
Payment methods the stores accept	.82
Decoration of stores	.80
Display and windowing of goods	.81
Variety of product categories the stores carry	.76
Reputation of the stores	.78
Variety of brands the stores carry	.77
Accessibility of the stores	.69
Location of the stores	.67
Ambient condition of the stores	.60
CFI = .99, RMSEA = .08, SRMR = .03
Staff service quality	γ = 1.57	7.85%	.88	.87	.64
The stores’ employee attitude	.86
The stores’ employee communication ability	.90
The stores’ employee promptness	.84
The stores’ employee product knowledge	.68
CFI = .99, RMSEA = .08, SRMR = .02
Merchandise value	γ = 1.21	6.04%	.86	.83	.55
Quality of goods the stores carry	.74
Authenticity of goods the stores carry	.74
Price of goods the stores carry	.85
Style of goods the stores carry	.66
CFI = 1.00, RMSEA = .02, SRMR = .01
Service differentiation	γ = 1.16	5.82%	.72	.75	.51
Stores’ ability to provide home delivery service	.82
Stores’ ability to offer unique or specialized merchandise	.78
Stores’ ability to offer special deals	.63
CFI = 1.00, RMSEA = .00, SRMR = .00

Note: Items with primary loadings greater than or equal to .5 and secondary loadings less than or equal to .30 are retained. All other pattern coefficients of .30 or lower are not reported. The fit statistics for the four-dimensional confirmatory factor model are χ²/df = 1.78; comparative fit index (CFI) = .99; root mean square error of approximation (RMSEA) = .06; standardized root mean square residual (SRMR) = .05.

a.

AVE: average variance extracted.

b.

γ = eigenvalues.

Service product and environment has the highest eigenvalue (9.29), which suggests that this facet has the largest proportion of variance explained in our satisfaction measure. The subscale includes nine items that represent properties related to the tangible aspects of store service environment, accessibility, window display, reputation, payment methods, and merchandise selection. We borrow the term from the service literature (Rust and Oliver 1994; Brady and Cronin 2001) and use it to refer to a conglomerated set of items related to the service product and service environment. Staff service quality includes four items that represent the level of gratification tourists have with stores’ customer-service performance during the service delivery. Merchandise value contains four items that embody tourists’ perceptions of the benefits and costs associated with the merchandise. Finally, the three-item service differentiation subscale embodies value-added services such as retailers’ ability to provide special items, deals, and delivery services.

The four-factor solution extracted from PCA depends on Kaiser’s K1 rule in that a factor is extracted with an eigenvalue equal to or greater than 1.0 (Kaiser 1958). However, the literature suggests that the Kaiser criterion may overestimate the number of extracted factors (O’Connor 2000). Horn (1965) suggests a more advanced strategy called parallel analysis, which appears to be the most accurate method for deciding how many factors to extract or retain (Thompson and Daniel 1996; Hayton, Allen, and Scarpello 2004). We performed parallel analysis based on the procedure presented by O’Connor (2000) and found four factors with raw data eigenvalues greater than their random-order counterparts; this result supported our decision to retain four factors.

Scale Validation

We used confirmatory factor analysis in LISREL 8.8 to validate the four-dimensional structure of the proposed satisfaction scale. The fit of the four subscales and the overall scale are satisfactory, with comparative fit index (CFI) ≥ .99, root mean square error of approximation (RMSEA) ≤ .08, and standardized root mean square residual (SRMR) ≤ .03. We examined convergent validity by means of average variance extracted (AVE). As Table 1 reveals, the AVE for the four subscales is above the .50 threshold (Fornell and Larcker 1981).

Next, we followed the literature ((Brakus, Schmitt, and Zarantonello 2009) and ran three competing models to identify the measurement model that fits the data best, as Table 2 shows. The first model—the baseline model—assumes that all items of the shopping satisfaction scale are loaded on a single construct (i.e., one-factor model); the second defines shopping satisfaction as a four-factor model; and the third operationalizes the scale as a second-order construct with four first-order factors as subdimensions. The results reveal that the third model with a second-order construct with four first-factor factors provides the best fit statistics: CFI = .99, RMSEA = .06, SRMR = .05, and Akaike’s information criterion (AIC) = 383.31. The findings indicate that shopping satisfaction is best defined as a second-order construct composed of four first-order dimensions rather than a one-factor or a four-factor first-order construct.

OPEN IN VIEWER

Table 2.

Confirmatory Factor Analysis Model Fit Comparisons

Model	Description	χ2	df	χ2/df	p	CFI	RMSEA	SRMR	AIC
1	One factor	715.66	166	4.31	< .001	.95	.12	.07	803.66
2	Four factor	286.02	160	1.78	< .001	.99	.06	.05	386.02
3	One second-order factor with four first-order factor	287.31	162	1.77	< .001	.99	.06	.05	383.31

Note: CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; AIC = Akaike information criterion.

To cross-validate the scale, we assessed factorial invariance in two invariance tests: metric invariance and scalar invariance (Hair et al. 2006). Metric invariance is accomplished by comparing whether the factor loadings of two (sub)samples are equivalent. We used the aforementioned two subsamples to examine metric invariance. As Table 3 shows, the results indicate that the measurement models for both subsamples are invariant, Δχ²(16) = 9.22, p = .90 (see Model 1 and 2), in support of full metric invariance.

OPEN IN VIEWER

Table 3.

Test of Measurement Model Invariance

Model	Description	χ2	df	Δχ2 (Δdf)	p	CFI	RMSEA	SRMR
1	Baseline model for Subsample 1 and 2	590.73	350	–	–	.99	.05	.05
2	Factor pattern invariance for Subsample 1 and 2	599.95	366	9.22 (16)	.90	.99	.05	.06
3	Baseline model for male and female	618.33	350	–	–	.98	.07	.11
4	Factor pattern invariance for male and female	629.30	366	10.97 (Δ16)	.81	.98	.07	.11
5	Scalar invariance for male and female	647.16	382	17.86 (Δ16)	.33	.98	.07	.11

Note: CFI = comparative fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.

Scalar invariance is achieved by comparing the means between two theoretically distinct groups (e.g., male shoppers vs. female shoppers). To assess scalar invariance, we first divided the data set into two groups by gender, as males and females are theoretically different in their shopping preferences and behaviors (Lehto et al. 2004; Oh et al. 2004; Kuruvilla, Joshi, and Shah 2009). Before we test scalar invariance, we must establish a baseline model with metric invariance for the two groups as Hair et al. (2006) suggest. Next we tested the metric invariance of the two gender groups. The findings indicate that the two groups are invariant, Δχ²(16) = 10.97, p = .81 (see Model 3 and 4 in Table 3). We then tested scalar invariance by constraining the latent factor means (or factor intercepts) of the structural equation of the observed variables on the latent factors to be equivalent across the two groups; this can be achieved through latent mean analysis (Schmitt and Kuljanin 2008). A path diagram of the latent mean analysis is depicted in Figure 1.

OPEN IN VIEWER

Figure 1.

Path diagram for latent mean analysis

We used the female group as a reference group and fixed the factor intercepts (β) of the four shopping satisfaction constructs to zero, while freely estimating the intercepts of the four constructs for the male group. In addition, the parameters of the paths leading from the constant term (represented by a triangle) and then from the constructs to the observed variables were freely estimated albeit constrained equally across both groups. The results indicate that the two groups are invariant across the intercepts, Δχ²(16) = 17.86, p = .33 (see Model 5 in Table 3), in support of full scalar invariance. In sum, results of the invariance tests warrant a strong factorial invariance (i.e., both full metric and scalar invariance requirements are fulfilled; Cheung and Rensvold 2002; Hair et al. 2006) of the postulated four-dimensional tourism shopping satisfaction scale.

Measures of Destination Tourism Facility and Shopping Experience

The aforementioned survey also includes a number of destination-level tourism variables. These variables include tourists’ satisfaction with the safety of, transportation connected to, location of, cleanliness of, and scale of the tourist facilities of a destination as well as their overall satisfaction with their shopping experience in the destination. Measures of these variables were adopted from the extant literature (e.g., Wong and Law 2003; Hsieh and Chang 2006; Reisinger and Mavondo 2006; LeHew and Wesley 2007; Sangpikul 2008). Each scale is measured with a 5-point option ranging from 1 (dissatisfied) to 5 (satisfied). We diagnosed common method variance based on the Harman one-factor analysis (Podsakoff et al. 2003). The results indicate that common method bias is not a concern in the study, χ²(299) = 1,278.89, p < .001, χ²/df > 2.0. We also tested the normality of the variables of interest. Results presented in Table 4 suggest that the distributions of all the variables are approximately symmetric, with skewness and kurtosis values below 1.00 (Bulmer 1979).

OPEN IN VIEWER

Table 4.

Construct Correlations, Means, and Standard Deviations

	Skew-ness	Kurtosis	M	SD	1	2	3	4	5	6	7	8	9	10
1. Service product and environment	−.33	−.15	4.01	.76	1.00
2. Merchandise value	−.93	.62	4.05	.72	.73	1.00
3. Employee service quality	−.44	.18	3.84	.83	.71	.64	1.00
4. Service differentiation	−.36	.01	3.51	.81	.56	.56	.64	1.00
5. Safety of facilities	−.81	.54	4.21	.79	.74	.68	.58	.46	1.00
6. Transportation system	−.66	−.15	3.80	1.03	.64	.54	.60	.52	.62	1.00
7. Location of facilities	−.54	−.02	3.82	.94	.64	.55	.58	.52	.63	.77	1.00
8. Cleanliness of facilities	−.61	−.26	4.09	.88	.71	.61	.57	.42	.72	.62	.61	1.00
9. Size of facilities	−.05	−.84	3.44	1.00	.53	.41	.44	.44	.49	.58	.58	.56	1.00
10. Overall shopping experience	−.48	−.23	3.83	.74	.79	.67	.68	.59	.79	.84	.85	.81	.78	1.00

Note: Correlations are all significant at .001. All measures are assessed in a 5-point scale.

Results

We first examined the hypothesized relationships through Pearson’s correlation analyses. As Table 4 shows, all the zero-order correlation coefficients are significant (p < .001). We then tested the effects of the proposed model using structural equation modeling in LISREL 8.8, as Figure 2 depicts. We also controlled for tourists’ travel and demographic characteristics such as income and number of visits to the host destination. Furthermore, we diagnosed whether multicollinearity was an issue in the proposed model. The results indicate that it is not a concern (variance inflation factor [VIF] < 3.00).

Results reported in Table 5 reveal full support for hypotheses 1a–d in that the safety of tourist facilities positively relates to all four types of shopping satisfaction. The transportation system connecting to the tourist facilities positively influences satisfaction with service product and environment, staff service quality, and service differentiation, in support for hypotheses 2a, c, and d. Similarly, the location of the tourist facilities also positively affects tourist satisfaction with staff service and service differentiation, hence supporting hypotheses 3c and d. We also find that the cleanliness of the tourist facilities has a positive impact on tourist shopping satisfaction except for service differentiation; thus hypotheses 4a–c are supported. However, the scale of the tourist facilities has a positive influence only on service differentiation satisfaction, in support of hypothesis 5d.

OPEN IN VIEWER

Table 5.

Results of Testing Antecedents and Consequence of Shopping Satisfaction

Variable	Service Product and Environment Satisfaction	Merchandise Value Satisfaction	Staff Service Quality Satisfaction	Service Differentiation Satisfaction	Overall Shopping Experience
Safety of facilities	.37***	.38***	.34***	.15†	–
Transportation system	.12*	.05	.12†	.17*	–
Location of facilities	.09	.09	.15*	.16*	–
Cleanliness of facilities	.23***	.18**	.21**	.02	–
Size of facilities	.05	.03	.02	.10†	–
Service product and environment	–	–	–	–	.54***
Merchandise value	–	–	–	–	.25*
Staff service quality	–	–	–	–	.20*
Service differentiation					.36**
R 2	.70	.57	.48	.47	.37

Note: χ²/df = 2.42; comparative fit index = .98, root mean square error of approximation = .08, standardized root mean square residual = .08. Parameter estimates are unstandardized.

†

p < .10, *p < .05, **p < .01, ***p < .001.

Results of the structural model reveal that all the satisfaction factors are positive predictors of tourists’ overall shopping experience, as expected; hence, hypotheses 6a–d are supported. In sum, results from the hypothesized model reveal that safety is the most salient driver of service product and environment satisfaction, as well as of merchandise value satisfaction, while transportation is the most important driver for service differentiation satisfaction. The overall cleanliness of the tourist facilities also plays an important role in shopping satisfaction except for service differentiation. In turn, satisfied tourist shoppers are likely to enjoy their shopping experience in a destination. The findings indicate that satisfaction with the service product and environment is the most important predictor of a positive shopping experience. In turn, shopping satisfaction plays a mediating role between tourists’ satisfaction with destination tourism facilities and overall shopping experience. The proposed model also reveals reasonable fit, with R² ranging from .33 to .70.

Theoretical Implications

Despite its acknowledged importance, tourist shopping satisfaction has received little attention in the tourism and hospitality literature. Satisfied shoppers are likely to stay longer and spend more on their trip, while dissatisfied shoppers are, on the contrary, likely to constrain their shopping budget and curtail their shopping sojourn. However, an incomplete conceptualization of tourist shopping satisfaction has limited a fuller understanding of the dimensionality of the construct and its effect on tourist behaviors. This research bridges the research gap by proposing a 20-item scale that conceptualizes the multidimensionality of shopping satisfaction for tourists.

This research extends prior studies on tourist shopping satisfaction. Based on a rigorous and systematic scale-development process, we define tourist shopping satisfaction as a multifaceted construct that includes four dimensions of satisfaction, regarding service product and environment, merchandise value, staff service quality, and service differentiation. Although these results may mimic similar findings from prior research on tourist shopping satisfaction (Heung and Cheng 2000; Lin and Lin 2006), our study suggests that service differentiation also plays a role in shopping satisfaction. This finding is particularly germane in tourism and hospitality contexts because differentiation has become a strategic imperative to firms (MacMillan and McGrath 1997) and perhaps to tourism destinations themselves, as we show in the current study.

More importantly, we explore the antecedents and consequences of the four shopping satisfaction dimensions. The results provide scholars with new insights on the role of destination tourism facilities in tourists’ shopping satisfaction. Given that contentment with the four facets of shopping directly leads to tourists’ overall destination shopping satisfaction, we build a linkage between tourist facilities and individual attributes. That is, a safe and clean destination with large retail/leisure entertainment facilities that are easily accessible is likely to lead to a pleasant shopping experience of the tourists during their sojourn, if they are gratified with the service product and environment, the value of the merchandise, the staff service quality, and/or the service differentiation of the retailers. Hence, this article provides a better understanding of how the destination-level offerings could influence industry- and individual-level evaluation.

The literature has acknowledged a need to investigate the relationship between regional factors and shopping satisfaction attributes (Tosun et al. 2007); this study therefore offers a timely update to the literature by delineating the role of various destination tourism facilities on tourist satisfaction during shopping excursions. Furthermore, our findings contribute to the literature by presenting a mediating role of shopping satisfaction between destination attributes and tourists’ overall shopping experience. This research is also the first study that links destination tourism facilities and shopping experience through the mediating role of service differentiation. These findings may help to explain why some tourists prefer to shop at popular destinations, such as Hawaii, Hong Kong, and Paris, over the less popular counterparts.

This study provides an update to research on consumer behavior, retailing, and service. The tourism industry has emerged as a global market. While it continues to expand and, conversely, many domestic markets continue to shrink, marketing scholars should reevaluate how well the existing marketing frameworks apply to tourists in order to better understand their behaviors as well as the impact of these behaviors on domestic markets. For example, the evidence collated indicates that safety of a destination has by far the most important effect on shopping satisfaction. Hence, the service and hospitality literature should take into account the destination-level contextual factors in order to better explain tourists’ perception of and satisfaction with specific services.

Methodological Implications

This study adopts a systematic process in developing the TSS scale based on the literature. First, a pool of scale items is generated through a combination of in-depth interviews and prior literature. Next, measurement validity is established based on the results from exploratory and confirmatory factor analyses. Both Cronbach’s alpha and composite reliability are used to assess scale reliability. Parallel analysis is adopted to validate the number of factors that best represent the underlying TSS scale. Three different measurement models—a one- factor model, a four-factor model, and a second-order factor with four first-order factors model—are compared to derive the best-fit measurement model. Scale validation is performed through both metric and scalar invariance tests. While metric invariance is achieved through a factor pattern invariance test, scalar invariance is assessed through latent mean analysis by comparing the latent mean values of the four shopping satisfaction dimensions between male and female tourists. To our knowledge, this is the first study in tourism research that uses latent mean analysis to test the mean differences of latent constructs and examine scalar invariance. Finally, nomological validation is assessed by examining the interfactor correlations as well as the antecedents and consequences of TSS. Although these steps are recommended by the literature (Nunnally 1978; Thompson 2004; Hair et al. 2006), few if any studies in the field of tourism and hospitality have systematically applied them in scale development. This study advances the literature by presenting a means for systematically applying this rigorous scale development process to develop a reliable and valid scale in tourism studies.

Managerial Implications

The results have managerial implications for tourism marketers, retailers, and policy makers. The TSS scale not only provides practitioners with a comprehensive and standardized measure of tourist satisfaction with their shopping, it also provides a means to benchmark tourism destinations’ ability to satisfy outbound shoppers. Hence the TSS instrument could serve as a diagnostic tool to find out areas tourists are satisfied with and areas they are not happy with. Because service product and environment is the most important dimension in shopping satisfaction, policy makers, merchants, and marketers should strive to work together to develop state-of-the-art shopping facilities, provide ample shopping options, and allow easy access to these shopping venues. Although these suggestions may seem trivial, as they may seem to portray an image of Honolulu, Hong Kong, Paris, or Singapore, our findings reveal that places that are able to offer tourists excellent merchandise value, staff service quality, and differentiated service may also be able to compete with the aforementioned popular shopping locales. This finding is imperative to developing tourist destinations such as Macau, Manila, and Bali, as well as other Asian Pacific and Eastern European cities.

Our findings indicate that transportation is a key factor influencing tourists’ shopping satisfaction. A good transportation network that connects major tourist attractions with shopping areas is a necessary and important condition to foster shopping satisfaction. The results of this study also reveal that tourist satisfaction with shopping is affected by the location and the size of tourist facilities where shopping takes place (e.g., an integrated casino complex or a theme park). This finding is consistent with Hsieh and Chang (2006) in that tourists generally prefer to shop in easily accessible areas with large-scale conglomerated shopping offerings. Large-scale shopping facilities, such as the Grand Canal shopping mall inside the Venetian casino, the Waikiki Beach shopping area, and the Causeway Bay shopping centers in Hong Kong, have emerged as major tourist attractions.

In fact, the findings reveal that the size of a tourist facility has a positive impact on tourists’ satisfaction on service differentiation (e.g., stores’ ability to offer unique or specialized merchandise and their ability to offer special deals). This evidence should help justify why destinations should develop large integrated tourist facilities that combine multiple leisure entertainment, shopping, and dining offerings into a single complex or location (e.g., the Las Vegas Strip and Disney World). Travel planners and destination authorities should also develop infrastructure that specifically targets busy tourist shoppers and allows them to better utilize their travel time in combining shopping with other types of travel activity.

Our research findings further suggest that safe and clean destination tourism facilities have positive effects on tourists’ shopping satisfaction and travel experience; hence destination stakeholders must pay close attention to the safety and sanitation of destination tourism facilities. Major tourist facilities and streets need regular surveillance and patrolling in order to reduce the crime rate and to maintain cleanliness. The measures are of particular importance for developing destinations, such as China, Macau, Bali, and Bangkok, which are known for high crime rates and poor sanitation. Shopping areas may also need to provide adequate outdoor lighting, surveillance cameras, and security guards to project the image of a safe and pleasant environment.

Maintaining safety and cleanliness in developing tourist destinations could be a challenging mission, as there is evidence suggesting it could be difficult to maintain high service quality for every touch point (Testa and Sipe 2006). Although traditional methods such as employee training, listing hot spots (i.e., areas that are likely to be problematic), and service blueprinting are useful (Testa and Sipe 2006; Zeithaml, Bitner, and Gremler 2006), the effectiveness of these techniques is unclear at the macro level. We believe that an alternative approach is to educate residents and link tourism’s contribution to economic progress. Policymakers should invest in education and hospitality programs in order to enhance residents’ civic-mindedness, motivating them to help maintain good hygiene and fight against crime. In addition, policymakers should strive to link tourist satisfaction with economic growth. The intertwining relationships among tourist satisfaction, tourism development, and economic development should be highly publicized in order to gain support from communities that intend to develop into world-class tourist destinations. Furthermore, given that satisfaction is a forward indicator, international retailers would be better able to justify their foreign investment by linking a place’s tourism infrastructure with consumer/tourist behaviors and financial outcomes.

Limitations and Future Research Directions

This article is not free of limitations. First, although this research utilizes a systemic scale-development process to validate the proposed scale, cross-validation from a different sample is needed. Second, this study uses single-item measures to assess destination tourism facilities and shopping experience. Although these scales are primarily adopted from prior literature, we believe that an avenue of future research is to explore the dimensionality of these constructs. Third, the findings of the current research may only be generalized to travel destinations with similar tourist and shopping facilities. Also, because our sample is Chinese tourists, the findings should be interpreted with caution if they are to be interpreted for other nationalities.

As we seek to understand shopping satisfaction from a tourism perspective, there is a great need for additional studies. We encourage future efforts to extend this paper by replicating the study in other tourist destinations and with other types of tourist. Although the research context is based on tourists, the shopping satisfaction scale may be applicable to other types of shoppers. Hence, it is worthwhile to explore the potential moderating effect of shopper type in future studies. We also believe it is worthwhile to investigate the antecedents of shopping satisfaction in regard to individual characteristics and perceptions (e.g., store branding and media exposure). International business research commonly suggests intra- and international heterogeneity (Burgess and Steenkamp 2006); hence we encourage future studies to investigate the direct and moderating social, cultural, and economic effects on shopping satisfaction and experience at the regional level.