INTRODUCTION
Providing excellent customer interaction has always been a fundamental aspect of the hospitality industry. It is a field which combines tangible products with intangible services to create lasting experiences. Due to the interactive dynamics of tourism, with the close connections between service providers and customers during product production and service creation, experiences are co-created and require ongoing participation from both parties (Kusluvan et al. 2010). According the Zeithaml, Berry and Parasuraman:
“Quality in services is not engineered at the manufacturing plant, then delivered intact to the consumer... In most services, quality occurs during service delivery, usually in an interaction between the customer and contact personnel of the service firm. For this reason, service quality is highly dependent on the performance of employees, an organizational resource that cannot be controlled to the degree that components of tangible goods can be engineered.” (1988, 35).
It is not surprising to find research emphasizing how the service orientation of hospitality employees is critical for success within service operations (Hennig-Thurau 2004;Bitner, Booms and Tetreault 1990). Service orientation has been defined as, “a set of attitudes and behaviours that affects the quality of the interaction between the staff of any organization and customers” that demonstrate a “disposition to be helpful, thoughtful, considerate, and cooperative” (Hogan et al. 1984, 167). From an operational management perspective, positive employee service behaviours have been found to improve the quality of service perceived by customers (Kim 2011), as well as increasing overall customer satisfaction (Pizam, Shapoval and Ellis 2016). Guest satisfaction has also been linked to value perceptions crafted by customers (Worsfold et al. 2016) and an antecedent to a guest’s intention to return as a future customer (Susskind, Kacmar and Borchgrevink 2018). When viewed from the perspective of managing talent, positive customer orientation traits have been shown to favourably moderate the negative impacts of stress on employee engagement and lessen turnover intentions (Babakus, Yavas and Karatepe 2017;Dusek et al. 2014). Positive employee service orientation has also been indirectly connected to increased retention, through increases in both job satisfaction and organizational commitment (Kim et al. 2005).
It would then appear to be quite valuable for there to be a clear and replicable measure for which to gauge an employee’s service orientation. Yet, service orientation continues to be unpacked with a variety of tools, varying in degrees of depth, range, and quality. Dimensions within the larger construct of service orientation fluctuate widely, ranging from 10 dimensions in work conducted within the retail service field (Lytle, Hom and Mokwa 1998) down to 3 dimensions found in various foodservice studies (Groves 1992;Dienhart et al. 1991).Lytle’s work (1998) takes on an operational focus to explore how organizations may be able to positively influence the service orientations held by employees through the application of policies and p rocedures.
This was the intention around the scale designed byGroves (1992) – to enrich our understanding of the service orientations held by workers in the hospitality industry, as measured through their attitudes. Yet, revisitation of Groves’ work within the foodservice industry, which was the original context for his study, has challenged the both the found dimensions and measurement items included (Kim et al. 2003). Furthermore, it appears that prior research shows conflicting impacts around the influence of demographical characteristics on customer service orientation. WhereasGroves (1992) showed differences by gender, results found byKim et al. (2003) were not able to duplicate these gender differences. Additionally, calls have been made for the need to explore the dimensionality of service orientation characteristics beyond the foodservice field (Kim 2011). Scale development fromDienhart et al. (1991;1992) throughKim et al. (2005) have all made attempts to validate a measure of service orientation through the lens of restaurant employees, while the hospitality industry encompasses a variety of operations outside of foodservice.
This was the intention around the scale designed byGroves (1992) – to enrich our understanding of the service orientations held by workers in the hospitality industry, as measured through their attitudes. Yet, revisitation of Groves’ work within the foodservice industry, which was the original context for his study, has challenged the both the found dimensions and measurement items included (Kim et al. 2003). Additionally, calls have been made for the need to explore the dimensionality of service orientation characteristics beyond the foodservice field (Kim 2011).
The main focus of this research study was to examine service orientation of hotel employees within the Canadian tourism sector, with the goal of validating the service orientation model originally design byGroves (1992), subsequently altered byKim(2003). From this analysis, a new multi-dimensional service orientation instrument is proposed which is both balanced and streamlined in design. The effects of employee demographical characteristics on an employee’s service orientation was also investigated, with a small number of significant differences surfacing by gender and job position.
1. LITERATURE REVIEW
This literature review explores research surrounding the reliable measurement of the service orientation construct over the last 40 years. As a validated construct, service orientation remains in the early stages of understanding. Prior to 1980, ABI/Inform database only shows 20 published journal articles mentioning the term ’service orientation’, with none proposing a reliable measurement scale or instrument. Very little formal work had been undertaken to formally unpack the characteristics and dimensions of employee service orientation. Reviewing hospitality sector-specific literature within the Hospitality & Tourism Complete database, some of the first attempts to capture employee service orientation was completed in the mid-1990s. This early work focused on validating the small amount of a priori research conducted in other sectors (Cran 1994). In his exploration, Cran stated that “given the contemporary business focus on customer service and the recognition of the critical role of front-line employees [it was quite surprising that] more effort has not gone into the identification and assessment of service orientation prior to selection” (1994, 43). Nearly fifteen years later, there remained a paucity of work responding to Cran’s call for additional research, with Teng and Barrows stating that “service orientation (has) received little attention by hospitality researchers, especially in the recent decade” (2009, 1416).
Although several researchers have surfaced service orientation within their studies, it has often been achieved by applying a few questions as a surrogate for the full construct measure.Susskind, Kacmar and Borchgrevink (2003) moved past validated scales, instead creating eight items to study service orientation, with their data only supporting five of those questions.Smith et al. (2012) selected just five items from the service orientation scale developed twenty years earlier (Dienhart et al. 1992). In one study,Lee and Ok (2015) sought to explore employee attitudes towards service with six items, only one of which was drawn from extant literature (Butcher 1994). Just a few years later,Lee et al. (2018) opted to use a 12-item dispositional service orientation scale proposed byBrown et al. (2002) but found only nine items to sufficiently load in a unidimensional outcome.Donavan, Brown and Mowen (2004) did attempt to explore service orientation multi-dimensionally; however, their 13-item measure spread across four dimensions failed to gain traction as a replicable model, although recently some researchers exploring the service worker and customer interactions (Fellesson and Salomonson 2020) borrowed five of the 13 items initiated in Donavan’s work. When utilized in a study to examine the effects of personality traits on service orientation with Turkish undergraduate students, all 13 items failed to align to Donavan’s four dimensions, instead collapsing unidimensionally (Köşker, Unur and Gursoy 2019).
To view service orientation through a clear lens, there is value returning to the origins of the research. Early work conducted to understand the personal attitudes that employees have about service began in the healthcare field approximate 40 years ago. After completing an exhaustive search of existing work, Hogan, Hogan and Busch expressed that, “we could find no existing measures that adequately assess service orientation” (1984, 168). They sought to fill this gap in stages (Hogan 1983), by developing a larger measurement tool that might shed psychological insight into self-reported personality traits of employees, from which came the Hogan Personality Inventory (HPI). The HPI was an extremely rich and detailed tool, originally containing 310 items that capture six scales (intelligence, adjustment, prudence, ambition, sociability, and likeability), within which were 45 distinct homogeneous item composites (Hogan et al. 1984). This was followed by a more practical effort to improve employee hiring practices through increased understanding of an individual’s personality and attitudes towards service, leading to the development of a more tightly focused index.
When developing their index,Hogan et al. (1984) included employee characteristics such as courteousness, consideration of others, high awareness of the needs of others, and strong interpersonal communication skills. In total, their initial design to surface service attitudes held by workers contained 92-items along 14 composite indices derived from the HPI; they called this the Service Orientation Index (SOI). Subsequent studies uncovered associations in some underlying characteristics in service orientation that included the ability to cope with normal life challenges (adjustment), an employee’s likeability, and the lack of hostile feelings when under stress (prudence) (Cran 1994;Dale and Wooler 1991). Some researchers have challenged the limited nature of looking solely at personality traits, noting that external environmental factors, such as the organizational policies, culture and role restrictions, could have strong contextual influences on how employees construct and reconstruct their orientations (Cran 1994;Solomon et al. 1985).
Drawing together the initial personality traits presented byHogan et al. (1984) while taking into account the contextual pressures surfaced bySolomon et al. (1985),Dienhart, Gregoire and Downey (1991) explored the construct of service orientation from the perspective of employees working in the high-context field of the hospitality industry. The ideation phase of their work was grounded in the concepts surfaced by previous work in financial institutions (Schneider et al. 1980) and the healthcare field (Hogan et al. 1984). However, they chose to veer away from the extremely large number of items contained within the SOI (Hogan et al. 1984). Instead of building on past indices, they started their own methodological exploration of service characteristics by grounding their work in direct consultation with foodservice industry executives as well as their collective personal experiences in the hospitality field. This resulted in a more concentrated focus around nine key service-focused areas. Questions in each of these areas were then developed and presented to 852 workers and 309 managers in the restaurant field. Results found that the initial nine areas condensed into three distinct factors: organizational support, customer focus, and service under pressure.
Groves (1992) was encouraged by their exploration into the characteristics of service orientation within the hospitality field and proceeded to build upon their work in his doctoral dissertation. Focusing on the restaurant industry, Groves identified 34 items of service orientation that also divided into a similar set of three factors (Yen, Yeh and Lin 2016): organizational support, customer focus, and providing service under pressure. These dimensions brought together the idea of service orientation as a combination of individual traits (Hogan et al. 1984), as captured in the lens of customer focus, with the influences of situational and environmental variables, found in the positive aspects of organizational support and negative impacts of service under pressure (Solomon et al. 1985). The integrative approach to understanding service orientation as an interaction of the individual within the context of their environment has ongoing support in the extant literature (Donavan, Brown and Mowen 2004;Dienhart et al. 1992).
However, replication and validation of Groves’ three factor service orientation scale has proved problematic in item loading and dimensionality attempts at replication (Kim et al. 2003). While some studies (Kim, et al. 2005) have been able to retain item loading onto the original three dimensions established by Groves, this was only accomplished by removing approximately 74% of the original measurement items, or 25 of 34 questions. This deep cut into the individual measures deeply reduced the richness, leaving, for example, only 2 items to represent customer focus and just a single item to capture service under pressure.
Some others have had better success teasing out a four factor model, with organizational support and service under pressure holding consistent while the interpersonal traits of employees divided into two areas: customer focus and prior customer relationships (Kim et al. 2003). In this case, 50% of the original items were removed; however, at least three items fell into each factor.Kim (2011) applied this shortened 17-item scale in a subsequent project and, although results confirmed earlier work, the exclusion of Groves’ original 34-items om Kim’s study made it impossible to verify whether Groves’ fuller model held consistent over time and across studies.
Clearly, there remain inconsistencies in research to date and challenges in replication. Although a multi-dimensional approach to understanding an employee’s attitudes towards service has shown rich results, the aspects of these dimensions and which measurement items best capture them within various scales has varied across studies. The purpose of this current study is to push deeper into our knowledge of service orientation dimensionality from an employee perspective by revisiting the various attempts to validate Groves scale (Kim et al. 2003,2005;Groves 1992) while attempting to retain a sufficiently rich measurement tool that provides practical value for industry application. This review of the extant literature clearly shows inconsistencies in prior measurement scales which serves that the primary intention for this study, as presented in the following sections. The current study’s focus is on presenting an improved, validated multi-dimension scale which confidently captures the service orientation held by employees.
2. RESEARCH METHODOLOGY
2.1 Sample
In this study, 348 individuals working in the Canadian lodging industry participated. Data was collected in 2016 directly from lodging properties and through panel collection with Qualtrics. A series of independent t-tests comparing means between respondents gathered directly from industry and through Qualtrics panel services showed no significant differences between means, validating the blending of data from two different methods. All data was collected through the Qualtrics online survey tool. Senior managers at 12 Canadian hotel properties distributed an invitation to participate to all their employees through their internal communication system, adopting a census approach. Panel data collection targeted applied inclusion criteria to only reach respondents who were currently working in the Canadian lodging industry at the time they completed the survey. To confidently complete factor analysis,Comrey and Lee (1992) noted that a sample size exceeding 300 is considered good.Meyers, Gamst and Guarino (2006) held that a strong sample target ratio should be 10 participants per variable, with a lower ratio could be allowed as the number of variables rise. As all 34 variable items from Groves’ original index were included, the resulting sample of 348 participants exceeds all requirements for an examination through factor analysis.
2.2 Instrumentation
The service orientation instrument utilized was the 34-item scale designed byGroves (1992). This scale contained nine items exploring organizational support (OS) which were all written in the positive direction, 12 items looking at service under pressure (SUP) that were all composed in the negative direction, and 13 items for customer focus (CF) which had seven questions worded in the positive direction and six in the negative direction. To analyse the scale, individual measurement items were recoded as necessary so that directionality was consistent, and interpretation was clearer.
2.3 Analysis
The analysis of this scale seeks to answer two questions in relation to past work (Kyriazos and Stalikas 2018). First, can the factors be explained within distinct variables and, second, are the factors correlated investigated against results found first byGroves (1992) and subsequently refined byKim et al. (2005). To complete this, confirmatory factor analysis (CFA) was applied to validate results against each of the two prior studies respectively. With outcomes failing to provide reliable, replicable results, exploratory factor analysis (EFA) was then used to establish both where and how the measures within the original 34-items fall into various latent variables. The intercorrelations of items was also be tested to ensure discovered variables were reliable between each other within the construct of service orientation.
3. RESULTS
3.1 Demographic Details
Table 1 shows the breakdown across demographics. A total of 134 of respondents identify as males (38.5%), 212 as female (60.9%), with 2 individuals identifying as other (0.6%). Subjects range in age, with 46.2% of respondents representing people 30 years old or younger, 40.0% between 31 and 50 years old, with those over 50 years of age representing the remaining 13.8%. Nearly half of respondents (48.3%) work as line employees, many directly in contact with customers. Just over one third of subjects (38.8%) earn less than $30,000 each year, with another 43.1% earning between $30,000 and $60,000. The majority of subjects (83.3%) are employed full time, working over 30 hours each week. While 20.4% of respondents have only worked for their current organization for less than a year, 42.6% have worked between 1 and 5 years with their company, with 37.1% of all participants having stayed with their organization for over 5 years.
| Dimension | Items | Groves (1992) | Kim et al. (2003) | Current Study |
| Organizational Support | 9 | 0.77 | 0.69 | 0.81 |
| Customer Focus | 13 | 0.77 | 0.67 | 0.74 |
| Service Under Pressure | 12 | 0.80 | 0.76 | 0.86 |
3.2.1. Reliability
The questions regarding organizational support and service under pressure initially remained written in their respective positive and negative directions, as these were internally consistent. Reliability coefficients of the original 34 items produced a Cronbach’s alpha of 0.57. However, when the directionality of all items was transformed to a positive direction, the alpha score increased to 0.89, satisfying the 0.70 standard for internal reliability(Taber 2018).
3.2.2. CFA of Prior Models
Prior studies have focused on the examination of reliability at the dimensional level established byGroves (1992).Table 2 compares the internal consistencies found by dimension of this study compared against past work (Groves 1992;Kim et al. 2003). Alpha levels for organizational support, customer focus, and service under pressure of 0.81, 0.74, and 0.86 respectively demonstrate dimension reliability levels which are slightly stronger in most areas than previous studies. The three dimensions also demonstrated strongly significant Pearson correlations between each other, indicating good dimension convergent validity within the scale. Reviewing the individual inter-item relationships, no pairs within the 34 items were found to have correlations exceeding 0.80, indicating no issues involving multicollinearity (Field 2013). Inter-item reliability within the dimensions of organizational support and service under pressure were all significant. However, 8 of 13 items within the customer focus dimension demonstrated insignificant inter-item correlations between 1 and 3 other items, indicating discriminate validity pressures and high potential for dimensionality challenges.
Data was evaluated into a 3-dimensional service orientation scale, as established byGroves (1992), using AMOS 26.0. Standardized estimate loadings are shown in Table 3. Goodness of fit indexes indicated that the original model was less than suitable for the new observed data. While the RMSEA value of 0.075 fell just within the range of acceptable fit (Brown and Cudeck 1993), GFI and CFI values were 0.780 and 0.737 respectively, thereby failing to meet the accepted inclusion level of 0.90 (Joreskog and Sorbom 1984;Bentler 1990) for good fit. Construct validity was assessed with factor loadings above 0.45 (Comrey and Lee 1992), with 13 of the original 34 measurement items within Groves’ model lacking sufficient factorial weight.
Kim et al. (2003) also found challenges with the model fit proposed by Groves’ and, through their study in Korean foodservice, proposed a 4-dimension model with only 17 of the original items. They found four items held to the dimension of organizational support and six held to the dimension of service under pressure. However, the original dimension of customer focus proposed by Groves was divided in two distinct factors byKim et al. (2003) that of customer focus (CF) and prior customer relationships (PCR). When the current data was evaluated against this 4-dimension model, problems surfaced within the standardized estimate loadings of the dimensions (shown inTable 4), and goodness
| Items (34) | Factor Loadings | ||
| 1 | 2 | 3 | |
| Service under pressure (SUP)- 12 items | |||
| X23 We often get too busy to pay proper attention to the customer | 0.82 | ||
| X22 At peak times, we often find ourselves forgetting service | 0.79 | ||
| X11 We are often too busy to cater to the customer | 0.76 | ||
| X8 At peak hours, it is so busy that we cannot provide excellent customer service | 0.71 | ||
| X3 We have too many customers, making it difficult to provide good service | 0.68 | ||
| X15 A busy time can spoil good customer service | 0.65 | ||
| X9 Customers become real "nags" when we are busy | 0.62 | ||
| X25 Sometimes my smile slips when we are busy | 0.56 | ||
| X26 Often customers try to tell me what my job should be | 0.44 | ||
| X13 I dislike this job when the customer is unhappy | 0.37 | ||
| X31My manager expects us to follow routine procedures (e.g. cleaning) even if it means giving less than excellent customer service | 0.34 | ||
| X24 The manager pushes us to go-go-go when we get too busy | 0.25 | ||
| Organizational support (OS) | |||
| X17 Our service procedures make it easy for me to give excellent customer service | 0.75 | ||
| X21 The manager sets a good example for service | 0.69 | ||
| X2 People come to our business because of its good service | 0.64 | ||
| X16 Our manager pushes us to provide excellent customer service | 0.62 | ||
| X19 In our company, service is given the same importance as most other procedures | 0.59 | ||
| X1 The employees in our company provide excellent service | 0.58 | ||
| X14 When we are too busy to provide good service, the manager brings in help | 0.52 | ||
| X4 Customers treat me with respect | 0.39 | ||
| X30 Customers show understanding and patience when we are busy | 0.36 | ||
| Customer focus (CF) | |||
| X12 I try to please the customers | 0.68 | ||
| X10 I like it when the customers have a good time | 0.68 | ||
| X32 I provide good service even when a customer is in a bad mood | 0.64 | ||
| X28 I enjoy working when we are very busy with customers | 0.52 | ||
| X5 Customers I have previously served talk to me in a personal way when they come in | 0.49 | ||
| X34 We do not take breaks while customers wait | 0.47 | ||
| X29 I will go out of my way to provide good service to customers | 0.43 | ||
| X33 I was trained to give excellent customer service | 0.40 | ||
| X6 People I have previously served ask for me | 0.37 | ||
| X18 I have my special customers | 0.36 | ||
| X20 My manager will not give employees a good rating if they provide poor customer service | 0.33 | ||
| X27 We are not allowed to take breaks if the customers must wait | 0.26 | ||
| X7 I often have customers express satisfaction with me | 0.25 | ||
| Goodness of Fit Indexes: RMSEA = 0.075; GFI = 0.780; CFI = 0.767 | |||
| Factor loading values > 0.45 highlighted as fair or better to the model (Comrey and Lee 1992) | |||
| Items (17) | Factor Loadings | ||||
| 1 | 2 | 3 | 4 | ||
| Service under pressure (SUP) - 6 items | |||||
| X23 We often get too busy to pay proper attention to the customer | 0.74 | ||||
| X22 At peak times, we often find ourselves forgetting service | 0.68 | ||||
| X11 We are often too busy to cater to the customer | 0.54 | ||||
| X8 At peak hours, it is so busy that we cannot provide excellent customer service | 0.49 | ||||
| X15 A busy time can spoil good customer service | 0.38 | ||||
| X25 Sometimes my smile slips when we are busy | 0.28 | ||||
| Organizational support (OS) - 4 items | |||||
| X17 Our service procedures make it easy for me to give excellent customer service | 0.60 | ||||
| X21 The manager sets a good example for service | 0.53 | ||||
| X16 Our manager pushes us to provide excellent customer service | 0.35 | ||||
| X14 When we are too busy to provide good service, the manager brings in more help | 0.33 | ||||
| Customer focus (CF) - 4 items | |||||
| X34 We do not take breaks while customers wait | 0.64 | ||||
| X27 We are not allowed to take breaks if the customers must wait | 0.30 | ||||
| X29 I will go out of my way to provide good service to customers | 0.12 | ||||
| X33 I was trained to give excellent customer service | 0.10 | ||||
| Prior customer relationship (PCR) - 3 items | |||||
| X6 People I have previously served ask for me | 0.53 | ||||
| X5 Customers I have previously served talk to me in a personal way when they come in | 0.52 | ||||
| X18 I have my special customers | 0.37 | ||||
| Goodness of Fit Indexes: RMSEA = 0.110; GFI = 0.566; CFI = 0.421 | |||||
| Factor loading > 0.45 highlighted as fair or better to the model (Comrey & Lee 1992) | |||||
3.3. EFA for New Model
As neither model was confirmed, exploratory factor analysis (EFA) was initiated to surface model dimensionality. Principal component analysis (PCA) was applied, initially applying direct oblimin oblique rotation. The correlation matrix for the extracted factors failed to demonstration inter-factor correlations exceeding 0.32 (Tabachnick and Fiddell 2007), thereby necessitating the use of orthogonal rotation. PCA was reapplied with a Varimax rotation, with 9 components extracted using the Kaisen-Guttman criterion for eigenvalues that exceed 1, explaining 61.57% of variance. Correlations demonstrated sufficient sampling adequacy for factor analysis with a KMO measure of 0.869 (Kaiser 1974) and a significant Bartlett’s test of sphericity (p > 0.001). Review of the scree plot showed a sharp drop in eigenvalues from the 4th factor (1.89) to the 5th factor (1.30), after which levels evened off for the remaining extracted dimensions. Therefore, EFA solutions for both 4- and 5-dimensions were conducted.
| Items (22) |
Eigen value | % of variance | Factor loading | |||||||
| Organizational support (OS) - 7 items | 5.75 | 26.15 | ||||||||
| X17 Our service procedures make it easy for me to give excellent customer service | 0.73 | |||||||||
| X16 Our manager pushes us to provide excellent customer service | 0.72 | |||||||||
| X2 People come to our business because of its good service | 0.72 | |||||||||
| X1 The employees in our company provide excellent service | 0.69 | |||||||||
| X21 The manager sets a good example for service | 0.68 | |||||||||
| X19 In our company, service is given the same importance as most other procedures | 0.63 | |||||||||
| X14 When we are too busy to provide good service, the manager brings in more help | 0.56 | |||||||||
| Service under pressure (SUP) - 6 items | 2.19 | 9.93 | ||||||||
| X23 We often get too busy to pay proper attention to the customer | 0.76 | |||||||||
| X22 At peak times, we often find ourselves forgetting service | 0.74 | |||||||||
| X25 Sometimes my smile slips when we are busy | 0.68 | |||||||||
| X9 Customers become real "nags" when we are busy | 0.66 | |||||||||
| X26 Often, customers try to tell me what my job should be | 0.58 | |||||||||
| X13 I dislike this job when the customer is unhappy | 0.53 | |||||||||
| Customer relationships (CR) - 6 items | 1.99 | 9.03 | ||||||||
| X5 Customers I have previously served talk to me in a personal way when they come in | 0.73 | |||||||||
| X6 People I have previously served ask for me | 0.71 | |||||||||
| X18 I have my special customers | 0.70 | |||||||||
| X4 Customers treat me with respect | 0.58 | |||||||||
| X30 Customers show understanding and patience when we are busy | 0.48 | |||||||||
| X28 I enjoy working when we are very busy with customers | 0.47 | |||||||||
| Customer focus (CF) - 3 items | 1.51 | 6.85 | ||||||||
| X34* We do not take breaks while customers wait | 0.66 | |||||||||
| X12 I try to please the customers | 0.65 | |||||||||
| X10 I like it when the customers have a good time | 0.63 | |||||||||
| 51.97 | ||||||||||
| Exploratory. Factor Analysis: PCA w/ Varimax; loading > 0.45 (Comrey & Lee 1992) | ||||||||||
| * Indicates reverse coding / question wording for positive directionality | ||||||||||
Although the five-factor model explained 48.71% of variance, only one item loaded on the fifth dimension above 0.45, a level that was classified byComrey and Lee (1992) only as ‘fair’. As this additional factor did not add to the overall model structure, further work continued with a focus on a four-factor analysis. With all 34-items included, the EFA on four factors explained 44.88%, with at least four individual measurement items loading in each factor. However, seven items failed to load adequately on any of the four factors, suggesting further removal and continued analysis. Using the remaining 27 items, EFA explained 50.67% of total variance and produced distinct dimensions in service orientation surrounding service under pressure, organizational support, and customer focus, along with a new dimension titled “customer relationships”. This new factor blends items whichGroves (1992) had classified in either organizational support or customer focus.
| Dimension (4) | Items (22) | Alpha |
| Organizational Support | 7 | 0.815 |
| Service Under Pressure | 6 | 0.773 |
| Customer Relationships | 6 | 0.722 |
| Customer Focus | 3 | 0.653 |
An iterative review of the remaining measurement items surfaced some face-value redundancies which were further examined. The concept of time management that exists between providing customer service and allowing for employee rest appeared in both X34 (“We do not take breaks while customers wait”) and X27 (“We are not allowed to take breaks if it means the customers must wait”), with the only difference being the idea of ‘taking’ a break versus being ‘allowed’ to take a break. The nuance here was very slight, so it was concluded that X27, which had the lower loading factor, would be removed. Within service under pressure, 6 questions (X23, X22, X8, X11, X15, & X3[1]) all centred around the concept of providing quality service when busy, including: X8 (“At peak hours, we can still provide excellent customer service when it is busy”), X11 (“We are rarely too busy to cater to the customer”), X15 (“A busy time cannot spoil good customer service”), and X3 (“Even when we have too many customers, we still can provide good service”). In total, 4 of these 6 questions were removed to achieve a more parsimonious measurement item list; items X23 and X22 remaining as they possessed strongest loading factors. In both cases of item reduction, the total variance explained increased incrementally as items were removed. In total, 22 items remain within the service orientation scale along 4 factors, explaining 51.97% of total variance (seeTable 5).
When testing against the service orientation index presented byGroves (1992), the original three constructs of organizational support, service under pressure, and customer focus were all supported, similar to the support found byKim et al. (2003) in their examination of employees in the foodservice industry. This was expected as hotels and restaurants share many service-based characteristics common to the hospitality industry, including high-touch customer interactions. However, much likeKim et al.(2003) the analysis of data suggested that Groves’ original unidimensional factor of customer focus was better divided into two dimensions: customer focus and customer relationships. The internal reliability of the new factors was strong (Table 6) while each dimension contained multiple measurement items which loaded quite well (Table 5). For clarity moving forward, this refined 22-item index will be referred to as the SOS-22.
| Factors | Meana | S.D. | OS | SUP | CR | ||||
| OS | 4.010 | 0.652 | |||||||
| SUP | 3.461 | 0.744 | -0.357 | ** | |||||
| CR | 3.797 | 0.591 | 0.390 | ** | -0.292 | ** | |||
| CF | 4.513 | 0.533 | 0.301 | ** | -0.364 | ** | 0.368 | ** | |
| a Scale ranges 1 - 5; **p<0.001 | |||||||||
3.4. Correlation of Factors
Pearson correlations were calculated to reveal the Intercorrelations between the four factors within the service orientation scale in the new proposed structure (seeTable 7). All dimensions were significantly correlated with each other. Organizational support’s positive relations with both customer relationships and customer focus suggest that positive levels of support at a unit level should also be positively reflected in front line service focus on customers. Conversely, increased organizational support appears to have a negative relationship with pressures experiences during service. Service under pressure was negatively related to all other factors around service, inferring that increased experienced pressures could decrease focus on service processes and relationships. As these four factors are only dimensions within a large service orientation scale, their interrelations only reflect correlated results and not causal pathways.
3.5. MANOVA Results
To determine significant differences between demographic groups within the sample population, MANOVA analysis was employed. When faced with multiple dependent variables, MANOVA analysis considers intercorrelations that would be overlooked through univariate tests (Meyers, Gamst and Gaurino 2006), enables the examination of relationships between dependent variables, allows for group differences to be discovered due to the increased power of analysis, as well as providing “researchers with statistical guidance to reduce a large set of dependent measures to a smaller assemblage”. (Meyers, Gamst and Gaurino 2006, 368). Evaluations of normality and equality of variance-covariance determined that the data met the statistical assumptions to support these analyses. Through the use of Wilks’s criterion, the composite dependent variable of service orientation was significantly affected by both gender (Wilks’s λ, F[2, 346]=3.03, p<0.018)[1] and job position (Wilks’s λ, F[4, 348]=3.29, p<0.000). Categorical analysis by gender revealed that only service under pressure demonstrated significant difference (seeTable 8), with female respondents expressing greater experienced service pressures (M=3.53) that male respondents (M=3.36). When the various categories for job position were unpacked, only two of the four factors demonstrated significant variations. Relationships with customers appear to grow in perceived strength as job level rises, as do the perceived experienced pressures around service. No other demographical categories showed significant differences between groups for the four factors within service orientation.
[2] Two respondent surveys were omitted list-wise as outliers, leaving N=346.
4. DISCUSSION
The dimensions of customer relationships and customer focus within the service orientation perceptions of hospitality employees appear to be quite distinct. The orientation around customer focus centres around the intensity in which hotel employees paid attention to each customer interaction. Guests stay at hotels for significantly periods of time ranging from a single overnight to days or weeks; however, each interaction is a singular instant of impact, whatJan Carlzon (1987) coined as ‘moments of truth’. Each of these moments are opportunities for in which service providers can invest directed attention to influence guest happiness, enjoyment, and positive experiences. Customer focus brings heightened awareness to the enjoyment of customers and emphasizes that employees can make a distinct impact on service experiences. Customer relationships appear to relate more to interpersonal associations felt between service providers and their customers. It represents the personal connections and affiliations that can commonly develop when you see customers again and again, learning and understanding their particular set of needs and preferences. On the other side of the service relationship, customers may well find that developing deeper connections with certain hospitality employees can decrease cognitive pressures while increasing the differentiation between various service providers. When you have confidence that the Assistant Front Desk Manager at a certain hotel knows who you are and makes a personal investment in ensuring your particular needs are consistently met, the decision making process of deciding which hotel you will stay at disappears.
When applying our new data from the Canadian lodging industry into Groves’ original three factor model, 13 of the original individual measures proved unreliable and failed to load sufficiently (Table 3). EFA results showed that four of seven items focused on service pressures during busy times failed to load well, indicating that too many items were concentrated in this area. Seven of the original items within organizational support continued to load well. The remaining two items from OS (customers show understand and patience; customers treat me with respect) pivoted into the new customer relationships dimension; these did not appear to naturally align well within the realm of organizational support, but fit much more intuitively with relationships with customers. The dimension of customer relationships was rounded out with four other items pulled from the original customer focus area that all had a better match with interpersonal associations. Only three items continued within the dimension of customer focus, removing a few repetitive measures around making customers wait and extending extra effort to pleasure customers.
Although the analysis in this study did align with the four-factor model proposed byKim et al.(2003) two important distinctions emerged. First, while the measurement instrument presented byKim et al. (2003) attempted to be lean and efficient, only incorporating 17 measurement items across four dimensions, results from the current data was only able to support keeping nine of these items; the remaining eight failed to load sufficiently for inclusion (Table 4). With nearly half of their items omitted, it was necessary to return to the complete set of 34 items originated by Groves, revisit the dimensionality of the service orientation construct, and recalculate factor loads for each item.
The EFA in this study found six items loaded into the service under pressure dimension. Although this was the same number of items as found byKim et al. (2003) half of the items that fell into this area differed. The three items dropped fromKim et al. (2003) concentrated on service perceptions during busy times to the point of repetition. In the new analysis, they were replaced by pressures exerted directly by customers (customers feel like nags when we are busy; customers try to tell me what my job should be) and a single statement of job satisfaction connected to service (I like this job when the customer is happy).
Three additional items were found to load well into the dimension of organizational support, expanding this dimension to seven from the four proposed byKim et al. (2003). The new items centring on the organization’s reputation (people come to our business because of its good service; the employees in our company provide excellent service) as well as organizational culture (In our company, service is given the same importance as most other procedures).
As was found with the prior comparison with Groves, three items held in the dimension of customer focus.Kim et al. (2003) included four items in their customer focus dimension; three of these failed to replicate sufficient loading weight to remain. Although customer focus does hold the smallest number of measurement items at three, there is little repetition between items, each looking at a unique aspect of focus, from attention when busy to the enjoyment received when providing positive service experiences.
The fourth dimension, customer relationships, was the only dimension to keep all the items proposed by Kim et al. However, the new results doubled the category item count from three to six. A new item captures the pleasure of interacting with customers even during busy times (I enjoy working when we are very busy with customer), while two items surrounding the respect, understanding and patience experience with customer interactions that Groves had originally included with organizational support found a better fit within the customer relationship grouping.
The primary focus of this research was to assess and validate the service orientation models presented by Groves and later adapted byKim et al. (2003). The current study proposes an updated multi-dimensional measurement scale that exists somewhere between the two earlier pieces of work, presenting a streamlined design of only 22 items, while increasing both richness and depth within each respective dimension.
4.2. Demographical Impacts on Service Orientation
This study challengesKim et al. (2003) who reported that none of the demographical characteristics they collected in their study of the Korean foodservice industry influenced employee service orientation. Their discovery contradictedGroves (1992) findings that multiple personal variables impacted service perspectives. Groves found that females expressed lower stress under pressure, greater organizational support, and held stronger focus on service. Both duration of employment and organizational positions positively impacted relationships with customers, and college educated participants showed higher scores in the wider customer focus realm, although Groves was limited in his conclusions around education due to power and sample size (Kim et al., 2003).
| (N=346) | Line Employee | Supervisory | Managerial | Executive | Total |
| Male | 33.1% | 39.7% | 43.8% | 53.1% | 38.7% |
| Female | 66.9% | 60.3% | 56.3% | 46.9% | 61.3% |
Falling between these two prior studies, this current exploration found two areas in which demographical variables appeared to impact service orientation areas. First, female employees expressed greater stress under pressure during service than male colleagues, a finding that is in opposition to Groves. This could well be attributed to the fact that positional roles in the lodging industry continue to be skewed by gender. Female respondents in this study represented 61.3% of the total sample, aligning with Canadian demographical data showing that females constitute 60.3% of the accommodations workforce (Tourism HR Canada 2016). As can be seen inTable 9, two thirds of all line employees in this study identified as female, with relative representation by gender decreasing with each major job position level. It is rational to attribute higher levels of stress under pressure expressed by female workers to the reality that they make up a significantly larger proportion of direct customer-contact employees who also possess lower levels of institutional power to affect service changes and implement solutions.
Job position also demonstrated two dimensional impacts. First, customer relationships appear to grow with job level. This could be attributed to the reality that employees who have been in their jobs for longer periods of time have both moved up within their organization and have had a longer period of time to nurture deeper relationships with regular customers. Additionally, more senior managers spend less of the daily workday interacting with customers, and those customers they do interact with may get far more time and attention than is available for line employees who have far more customers to manage in a given period of time. However, expressed stress during service under pressure also showed a positive relationship with job position. When under pressure, senior managers are likely dealing with customers representing increased operational value, such as meeting planners, conference organizers, or larger corporate clients. Additionally, issues that require senior management intervention at the service level could well involve more serious product or service-related problems, as well as higher levels of emotional intensity.
5. CONCLUSION
5.1. Implications for Industry Practitioners
Increased understanding of an employee’s service orientation, the intangible temperament of workers to be “helpful, thoughtful, considerate, and cooperative” with customers (Hogan et al. 1984, 167), has numerous benefits for hospitality operators. From the side of customers, these benefits include increased customer quality (Kim 2011) and value (Worsfold et al. 2016) perception, improved intentions for guests to become repeat customers (Susskind, Kacmar and Borchgrevink 2018), and overall customer satisfaction (Pizam, Shapoval and Ellis 2016). Higher service attitudes have also been shown to improve organizational commitment and job satisfaction (Kim et al. 2005).Kim et al. (2005), while positively impacting retention with decreased turnover intentions (Babakus, Yavas and Karatepe 2017;Dusek et al. 2014). The new model proposed in this study provides practitioners with a streamlined 22-question assessment of service orientation, well-balanced across four core dimensions.
It is an instrument that could be used by operators at various opportunities to provide insight into employee attitudes. As a hiring tool, it can surface the value that potential employees place in nurturing deeper relationships with their customers as they draw reflectively on their past experiences when answering. It can also highlight the type of focus that an employee brings to their interpersonal service work and the importance they place on customer engagement. When used with current talent, this scale provides a richer interpretation about how employees view customers and service during situations of pressure and stress. Questions such as ‘can they provide good service when busy?’ and ‘do customers become bothersome during high stress moment?’ provide insight into the minds of employees that operators can use to evolve supports when business levels predictably increase pressure. In addition, assessing how employees perceive available organizational supports to service can provide valuable opinions about whether existing supports are achieving intended outcomes, as well as offer some insight about an operation’s reputation. The multi-dimensional design of this scale can be utilized as a whole, or in part, based on the goals of use.
Furthermore, it could also be utilized by operators with employees to gauge service attitudes on a longitudinal basis. Due to the straightforward design of the instrument and speed in which respondents can answer the questions, operators could easily ask employees to revisit the survey tool on a recurring basis without being overly intrusive on their time or developing survey fatigue. There is good fit with this tool to be incorporated into a simple pre and post application. Operators could quickly measure employee service attitudes prior to engaging in a change initiative, such as starting a new service training course or enacting different organizational supports, then follow up with remeasuring service attitudes once a reasonable period has passed to allow for the changes to have impacted employees and the operation.
Finally, the use of this scale allows practitioners to see clear evidence of differing pressures experienced by employees according to gender. The heightened awareness and desire for fact-based knowledge around gender, diversity, and inclusion issues in the workplace make insight into employee perspectives extremely valuable. Assumptions that talent experience service interactions in the same way or interpret organizational supports equally, can result in invisible stressors felt by employees which, in turn, have detrimental operational outcomes.
5.2. Limitations and Future Steps
Scale development is an iterative process as measurement tools evolve once they are applied in new contexts, both at various points in time and within different industries. This study responded to calls for additional exploration of service orientation (Kim 2011). While significant success was achieved in validating a balanced but approachable measurement tool of four dimensions across 22 questions, there needs to be additional supporting work before this tool can gain more generalized adoption. There is a justifiable rationale that the characteristics found working within foodservice and lodging would be similar, as these two sectors constitute the pillars of the hospitality industry. However, this study is the first to focus exclusively on the accommodations sector, while earlier studies byGroves (1992) andKim et al. (2003) looked at foodservice. Additionally, the concentration of study within the Canadian lodging industry limited the range of cultural and social contexts explored. Much as this study builds upon the rich work of others, there is need for the SOS-22 to be tested further, both within the high-service contexts of hospitality and beyond.