Construction and Validation of the Customer Participation Scale

Abstract

Although the importance of a mandatory customer participation construct in service delivery has been much discussed in the literature, little research has been devoted to conceptualizing and measuring one. To fill this void, this study followed a seven-step process for creating and analyzing scales in order to develop a customer participation scale and evaluate its generalizability, reliability, and validity. In theoretical terms, this scale extends the service quality literature, which has heavily emphasized the service provider’s responsibility for service quality, and will facilitate further studies in customer participation. In practical terms, the scale provides practitioners with useful mechanisms that could enhance their interactions with customers through facilitating the latter’s mandatory role in service delivery.

Keywords

mandatory customer participation customer roles service quality scale development customer loyalty

Introduction

Customer participation is not a new concept in the service quality literature. In fact, the number of studies addressing its importance has increased in the past decade (e.g., Bowen & Ford, 2004; Dong & Sui, 2013; Ford & Heaton, 2001; Namsivayam, 2003). These studies have advocated that service managers treat customers as active participants or service coproducers rather than as passive recipients or buyers. They have also offered specific managing techniques, such as treating customers like quasi-employees (Bowen & Ford, 2004) and letting them have a sense of control or fairness in service delivery (Namsivayam, 2003). These thoughts are in line with the service quality literature on recognizing the difference between a service and a goods product. As the service marketing literature has revealed, a key characteristic of a service product is service inseparability, defined as the need for both the service provider and the customer to participate in service production so that the service can be successfully created, purchased, and consumed; in other words, service quality is influenced by inputs from both parties.

Thus, the consensus in the service marketing literature is that managing service production is a core component in managing service quality. This is further explained by a classification framework of properties of offerings (Zeithaml, Bitner, & Gremler, 2006), which posits that service offerings such as vacations, hotel rooms, and restaurant meals are high in experience qualities (e.g., comfort, excitement, taste, etc.) because their attributes cannot be fully known or evaluated by consumers until they have been purchased and are being experienced and consumed. Consequently, the management of service quality during service production becomes a difficult task that distinguishes successful service companies (e.g., Ritz-Carlton, Disney) from those that fail.

To date, despite this considerable importance of service production in service quality management and an increasing interest in the concept of customer participation in the literature, empirical research has focused largely on the role of the service provider. Meanwhile, research on critical issues such as defining, conceptualizing, measuring, and determining the customer’s mandatory participatory role in service production remains scarce, and a reliable and valid customer participation scale has yet to be developed. The purpose of this study, therefore, is first to explore the nature of customer participation and second to develop and test a customer participation scale by collecting empirical data. In theoretical terms, this scale fills a void in the service quality literature and will facilitate further studies in customer participation. In practical terms, the scale provides practitioners with systematic and useful mechanisms that could enhance their interactions with customers so as to effectively create high-quality service.

In the following text, the researchers first examine definitions of customer participation, explores its dimensionality and measurement as a construct, and proposes a model of participation. The researchers then report a series of steps in developing a measure of customer participation and assessing the new measure’s reliability and validity.

Literature Review

The service marketing literature reveals that customers can participate in and influence a service firm’s business through two types of behavior: voluntary and nonvoluntary. As Zeithaml et al. (2006) contended, service inseparability implies that customers have mandatory production roles that are necessary for creating the service successfully, such as showing up at the service production site or offering basic personal information. This nonvoluntary behavior is also termed customer in-role behavior by some researchers (e.g., Yi & Gong, 2013). At the same time, as Gruen (1995) commented, many activities in which a firm engages customers for the firm’s own benefit are voluntary, such as taking a customer satisfaction survey or referring new customers to the firm. This type of voluntary behavior has also been termed customer citizenship or extra-role behavior (Yi & Gong, 2013). It is not too bold to argue that these two types of customer behavior are distinct and so should be treated and examined differently when discussing their influence over service quality. Consequently, this study focuses on the mandatory customer role in service production, in other words, customer participation.

Definition

The broad definition of customer participation as nonvoluntary behavior is likely insufficient for understanding the nature and composition of the construct, and so a survey of the thought-provoking discussion in the relevant literature is of value. One perspective focuses on how customers involuntarily participate in service production. Silpakit and Fisk (1985) argued that customer participation refers to a customer’s active role, which includes supplying activities and inputs rather than simply being present or having contact with service employees during the service encounter. Mills and Morris (1986); Kelley, Donnelly, and Skinner (1990); Kelley, Skinner, and Donnelly (1992); and Risch-Rodie and Kleine (2000) similarly contended that the customer role includes customer activities such as providing information, which constitutes the “raw material” of production, and exhibiting particular behaviors required for the conversion process to produce a satisfactory outcome.

Another perspective includes both activities and attitude. Kelley et al. (1990) suggested the use of “customer technical and functional quality” to explain how customers interact with service providers during service delivery: What the customer provides to the service encounter (e.g., providing personal information, carrying a tray), and how he or she behaves when services are provided (e.g., being cooperative, friendly, and respectful). Uzkurt (2010) concluded that customer participation is an expression of a customer’s informational, physical, behavioral, and emotional contributions to the stages of the service process (i.e., provision, production, presentation, and evaluation) and his or her willingness and ability to contribute to increasing customer satisfaction and service quality as well as create value. Yi and Gong (2013) in turn summarized that customer participation behavior involves not only information seeking, information sharing, and responsible behavior but also personal interaction between customer and service provider in a certain way.

Another group of researchers, Bitner, Faranda, Hubbert, and Zeithalm (1997), proposed that understanding customer participation requires knowing that not all services require equal involvement. They suggested that customer participation be classified into three levels: low, in which only a customer’s physical presence is involved; moderate, which requires not only a physical presence but also some input from the customer, such as information, time, effort, or physical possessions; and high, in which the customer is involved in cocreating the service, such as in a self-service buffet.

Although none of the above perspectives appears to define customer participation completely, together they reveal three critical components associated with this construct: what customers do to participate (e.g., providing personal information or locating the business), how they participate (e.g., being friendly, indifferent, rude, or passive), and how much they participate (measured as low, moderate, or high) in coproducing the service. This observation is in line with the concept of relationship marketing, where relationship building between two parties involves what they do, how they do it, and how much of it they do. It also sets a foundation for identifying the dimensionality of customer participation.

Dimensionality

Given that a universally accepted definition for customer participation has not been established, debates over its dimensionality should not be surprising. The service marketing literature shows at least four schools of views on this issue. First, some researchers have considered customer participation as a single-dimension concept and so have measured it with a single item (e.g., Cermak, File, & Prince, 2011; Dean, 1996); this methodology, however, has been criticized as potentially misleading in measure reliability (Andreason, 1983; Silpakit & Fisk, 1985). The second view regards this concept as a two-dimension construct. Researchers such as Bateson (1985), Dabholkar (1996), and Lee (1996) have examined customer participation in technological self-services using a dichotomous approach, that is, either full participation (self-service) or little-to-no participation (full service), but this method focuses on the magnitude of customer participation and has also provoked much criticism. As Chua and Sweeney (2003) noted, when customers have joint production roles with contact employees, the extent of their participation is less black and white and more appropriately measured along a continuum made up of different behaviors.

Studies describing customer participation as a three-dimensional construct can also be found. For instance, Ennew and Binks (1999) believed that customer participation should include information sharing, responsible behavior, and personal interaction with the service provider. Claycomb, Lengnick-Hall, and Inks (2001) held a slightly different view, contending that the construct should consist of the three dimensions of customer attendance, information provision, and coproduction behavior (e.g., helping out, being onsite).

Additionally, some researchers have used a four-dimensional approach to examine customer participation. For instance, while investigating the interactions of bank clients with bank managers, Kellogg, Youngdahl, and Bowen (1997) identified four forms of client participation: (a) preparation, such as an information search; (b) information exchange with the service provider to clarify service requirements and ensure that the customer understands his or her role in service delivery; (c) relationship building with the service provider in delivering the service; and (d) intervention if the customer believes the service provider is unlikely to produce a satisfactory outcome. Similarly, Auh, Bell, McLeod, and Shih (2007) identified four forms of clients’ investment activities when using financial services, namely, communication (timely information sharing), client expertise (customer knowledge about the product or service), affective commitment (customer’s attachment to, identification with, or involvement in the organization based on a sense of belonging to or being “part of the family”), and interactional justice (perceived fairness when a customer receives a service). Table 1 provides an overview of these studies.

Table 1

Examples of Participation Measures in the Literature

Author	Measures and Dimensions of Customer Participation
Groth (2005)	One dimension (five items, α = .94)
	1. I performed all the tasks that are required
	2. I helped the organization with those things that are required
	3. I adequately completed all expected behaviors
	4. I met formal performance requirements
	5. I fulfilled responsibilities to the organization
Auh, Bell, McLeod, and Shih (2007)	One dimension (three items, α = .80)
	1. I try to work cooperatively with my advisor
	2. I do things to make my advisor’s job easier
	3. I prepare my queries before contacting my advisor
Yi, Nataraajan, and Gong (2011)	One dimension (two items)
	1. I perform all the tasks that are required of me
	2. I fulfill customer responsibilities to employees of the firm
Kelley, Donnelly, and Skinner (1990)	Two dimensions
	1. Technique quality (i.e., what customers do)
	2. Functional quality (i.e., how customers do what they do)
Ennew and Binks (1999)	Three dimensions
	1. Information sharing
	2. Responsible behavior
	3. Personal interaction
Ennew and Binks (1999)	Three dimensions
	1. Information sharing (five items, α = .82)
	2. Absence of fear (two items, α = .66)
	3. Monitoring (two items, α = .32)
Claycomb, Lengnick-Hall, and Inks (2001)	Three dimensions
	1. Attendance (one item)
	2. Information provision (five items)
	3. Coproduction (three items)
Kellogg, Youngdahl, and Bowen (1997)	Four dimensions
	1. Preparation
	2. Relationship building
	3. Information exchange
	4. Intervention
Uzkurt (2010)	Four dimensions
	1. Information exchange
	2. Behavioral participation
	3. Emotional/interactive participation
	4. Willingness or ability to participate

These studies have undoubtedly enhanced our understanding of customers’ participatory roles in service production from different angles, yet not without criticism. In the words of Chua and Sweeney (2003), many of the studies examining customer participation dimensions have focused mainly on customer behaviors that occur during the service encounter while ignoring those at other purchase stages, including some mandatory behaviors that occur outside the service encounter. For example, in the financial services sector, it is critical that customers collate and bring along tax information for the accountant before entering the service (Zeithaml & Bitner, 1996), whereas in health care they may count calories in preparing food for a weight loss course (Hubbert, 1995). In the hotel industry, hotel customers must decide which hotel they want to stay at before checking in, and in the restaurant industry, customers must review their bill before paying it.

In this study, it was deemed essential to explore the boundaries of customer participation as they apply to hospitality and tourism settings before developing and testing a customer participation scale. To do so, the researchers first devised a framework derived from the well-established consumer behavior discipline: the model of the consumer decision-making process.

Proposed Model: Service Customer Decision Making

Understanding how consumers make decisions in purchasing and consuming products has proved vital in understanding what they do, how they do it, and why they do it. Customer participation, as an element of consumer behavior in service settings, involves all these questions. Therefore, examining the widely recognized model of consumer decision making is appropriate if we wish to better understand customer participation.

Originating in the consumer goods market, the consumer decision-making model consists of critical steps such as need recognition, information search, information evaluation, purchase decision, purchase, consumption, and postpurchase evaluation (Erasmus, Boshoff, & Rousseau, 2001). Applying this model to the context of service suggests that customer participation starts with recognizing a need (e.g., a traveler needs a hotel for an overnight stay), searching for information (e.g., the traveler checks what accommodation options are available at the destination), evaluating the information (e.g., the traveler compares different options based on certain criteria—location, price, amenities, etc.), deciding to purchase (e.g., the traveler selects one hotel), coproducing and consuming services (e.g., the traveler checks into the hotel and stays there for one or more nights), and making a postpurchase evaluation (e.g., the traveler writes to friends about the good time he had staying in the hotel).

The distinction between goods and services models is apparent: The purchase and consumption stages for a typical service product in a service setting are no longer two independent events; instead, they happen simultaneously, with the service consumption portion completed when the customer leaves the coproduction process. In other words, they form a cohesive process consisting of numerous “moments of truth,” each of which is an encounter that occurs every time the customer interacts with the service provider (Normann, 2001). In this process, the customer, consciously or unconsciously, carries out his or her own side of the role as a service coproducer as well as consumer in a personal manner. The customer may or may not be actively building a relationship, exchanging information, or solving problems with the service provider, and may or may not be cooperative, friendly, kind, or respectful. But the customer must be present in the service setting to consume the service product.

It is easy to see from Figure 1 that a service customer’s role starts before the service coproduction stage. One might argue that a customer is not necessarily required to engage in searching for and evaluating information. But a customer must have certain level of information about the service provider before making a purchase decision. For instance, a customer walking down the street in a completely foreign location begins to feel hungry. He sees a restaurant nearby and, approaching the server, asks what type of food the restaurant serves. The information engagement between the customer and restaurant service is minimal, but might be sufficient for the customer to decide whether to eat in this restaurant. But although there is no doubt that customers play a mandatory role in service coproduction, do they still play such a mandatory role after the service is produced and consumed? Normally, postpurchase evaluation and other activities conducted by the customer are considered voluntary behavior: customers can choose to evaluate or not evaluate a service experience, they can choose to complain or not complain about a bad experience, and they can choose to share or not share their positive experience with others.

Figure 1

Service Consumer Decision-Making Process

Therefore, it may be argued that a customer’s mandatory participation in service delivery occurs between Steps 2 and 5 in Figure 1 and is likely to include information participation, behavioral participation, and attitudinal participation. Unlike customer citizenship behavior, which can continue as long as the customer’s passion toward the service firm remains, a customer’s mandatory role as a service coproducer appears to cease once a service coproduction and consumption process has been completed and all the bills are paid. Management of customer mandatory participation therefore needs to be kept at a micro-level, with one service product life cycle beginning with an information search and ending with service coproduction and consumption (see Figure 1).

Method

This study followed the seven-step process for developing and analyzing scales as presented by Hinkin, Tracey, and Enz (1997). The procedure consists of generating items, assessing content adequacy, administering questionnaires, performing factor analysis, evaluating internal consistency, assessing construct validity, and replicating results.

Generation of Items

Guided by the model of customer participation (see Figure 1), the researchers first generated scale items from the construct definition as well as existing studies. Table 1 lists major articles with developed measures. From them the researchers derived 19 participation items covering the three core components of mandatory customer participation: information participation, attitude, and behavior, as identified in the literature review. The researchers then undertook three steps to assess the content validity of these items. First, the researchers asked industry experts to offer their insights on mandatory customer roles in service delivery. Such use of experts as judges of a scale’s domain is common in marketing (Sweeney & Soutar, 2001; Zaichkowsky, 1985). These hospitality executives included one CEO of a large resort, one vice president of a medium-size hospitality group, a restaurant owner, and two upscale boutique hotel general managers. All of their properties were located in the Northwest of the United States. Every executive, four male and one female, with an average age of 40, had worked in the hospitality industry more than 10 years and had earned high respect from their counterparts through building successful businesses.

These executives were sent an open-ended question via e-mail: “From your experience, what must a guest do to ensure a positive experience in your property?” They all used words such as cooperative, considerate, friendly, respectful, supportive, and trusting, which explained the attitudinal dimension of customer participation. Four executives said that guests should actively interact with their staff, have high expectations, demand outstanding service in a respectful way, and let staff know whether they were doing a good job and what they could do to make them happy. These comments reflected the necessary behavioral and information participation from the customer’s side. Three executives commented that guests should have some knowledge of the property or the property’s service delivery system. This also reflected the necessary customer information participation. All these responses fell within the content scope of the 19 items derived from the literature and provided evidence that 13 of these items could well represent a reasonable scope of customer participation measures.

Assessment of Content Adequacy

To further examine the content validity of these items, the researchers then asked 18 students to map their service experience in a casual dining restaurant they had visited at least once in the previous 6 months by applying the service mapping model (Shostack, 1984), with an emphasis on what they needed to do as a customer in order to have a good dining experience at their chosen restaurant. These students were seniors enrolled in a business school at a major university located in the Northwest. Among them, 60% were female, and their average age was 28. Through content analysis, the researchers grouped their answers into three blocks: prepurchase, during purchase, and after purchase participation. The prepurchase block showed that some students made reservations, checked online menus, or called to check rush hours, whereas others simply walked in because they were frequent customers and already knew the restaurant and its services well. The during-purchase block showed that all the students were very active in the restaurant: They asked for special offers or about the ingredients in dishes, and even advice on deciding what to order; gave instructions on how their food should be cooked, and so on. More than 90% of the students described themselves as being friendly, kind, and respectful of the employees and other customers. The third block showed patterns such as paying the bill and leaving tips, and being courteous when leaving the restaurant.

All these students’ responses confirmed that 13 out of these 19 items should be sufficient to explain the content scope of mandatory customer participation. The researchers thus set up these 13 items to be scored on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree) to describe the extent of mandatory customer participation. Three senior services marketing professors from two major universities in the Northwest were then asked to fill out the survey as customers to check the readability and logical flow of the questionnaire’s structure as well as the face validity of the participation measures. Some typos were deleted, while some sentences were restructured in response to these professors’ suggestions.

Questionnaire Administration

The researchers next delivered an online survey questionnaire to the target population—casual dining customers—through Qualtrics Lab, a survey research company. The questionnaire included three sections: the screening questions, the 13 scale items, and two related questions, one probing the respondents’ overall dining experience and emotional attachment to their favorite casual dining restaurant, and the other gathering demographic information. The respondents were instructed that to be eligible to participate, they had to be at least 18 years old and have eaten at a casual dining restaurant in the previous 6 months. A total of 402 responses were received, 24 of which were eliminated owing to missing responses or outliers, leaving a sample of 378 valid responses.

Sample Characteristics

The sample consisted of 52.2% female respondents. More than half the respondents were married. Approximately 55.6% were age 45 and older, which is the baby boomer group currently dominating the consumption market (Hudson, 2010). Another 20% were between ages 35 and 44, and 24% were between 18 and 34. Most respondents indicated they worked either full- or part-time. On average, annual household income before taxes was low. More than 50% of the sample indicated an income less than $50,000, with 19% at less than $25,000. All respondents had visited a casual dining restaurant at least once in the past 6 months, 69% had done so at least three times, and 31% had visited six or more times. Most respondents (56.5%) spent an average of $11 to $20 per person in the restaurant at each visit. In terms of education, 40.5% had some college or an associate degree, followed by 29.5% with trade or high school or less, 16% with a bachelor’s degree, and 13.7% with graduate degrees.

Assessment of Construct Validity and Reliability

The next step was to evaluate how the items performed to determine whether they adequately constituted the scale. Since it was important to repeat testing of the new scale and the initial sample (378) was large enough, the researchers randomly split the sample in about half and conducted parallel analyses for scale development in two separate studies.

Exploratory Factor Analysis (EFA)

For the first study, the researchers began with principle component factor analysis on the first subsample using SPSS 20.0 to reduce the set of observed variables to a smaller, more parsimonious set. The objective of the factor analysis was to identify those items that most clearly represented the content domain of the underlying construct. The initial sample size for the first half of the full sample was 188. Missing case analysis and outlier examination showed that three cases had an absolute value of a z score higher than 3.0, and thus were deleted from the datasheet. This resulted in a valid sample size of 185 for this analysis. The rotation method used was varimax with Kaiser normalization. The researchers then used the following criteria in extracting the factors: (a) all factors must have an eigenvalue greater than 1, and each must explain at least 4% of the total variance among the reason items; and (b) a generally suggested iterative process should eliminate items with a factor loading below 0.50, high cross-loadings above 0.40, and low commonalities below 0.30 (Hair, Anderson, Tatham, & Black, 2009).

The latent root criterion for the number of factors to derive indicated that three components should be extracted for these variables; in other words, the cumulative proportion of variance criteria could be met with three components to satisfy the criterion of explaining about 73% of total variance. This process led to deleting four variables, resulting in a clear factor structure matrix with nine items (see Table 2). The alpha values for all factors ranged from .71 to .87, exceeding the.70 cutoff value recommended by Nunnally (1994). The Kaiser–Meyer–Olkin (KMO) value of 0.80 and a significant chi-square value for Bartlett’s test of sphericity (χ² = 665.76, p < .000) indicated that EFA was appropriate for the data. Table 2 shows the final list of nine items retained for the next step of confirmatory factor analysis (CFA). In general, the results were very interpretable and showed good structure.

Table 2

Study 1: Principle Component Factor Analysis

Item^a	Factor
Item^a	1	2	3
Factor 1: Attitudinal Participation (abb. “Attitudinal”)
I try to be cooperative with the restaurant staff	.891
I am friendly to the restaurant staff	.880
I respect the restaurant staff	.805
Factor 2: Information Participation (abb. “Information”)
I read reviews of other customers about the restaurant		.809
I spend time searching for information about the restaurant		.750
I ask people I know for their opinions about the restaurant		.725
Factor 3: Actionable Participation (abb. “Actionable”)
I intervene when I feel something is not right when dining in the restaurant			.804
I openly discuss questions and concerns with the restaurant staff			.801
I ask questions if I don’t know how to get a service in the restaurant			.531
No. of items	3	3	3
Alpha value	.87	.81	.71
Variance explained (total = 73%)	26	25	22

Note: Extraction method: principal component analysis; rotation method: varimax with Kaiser normalization with KMO value = 0.79, p < .000.

Items were scored on a 7-point Likert-type scale anchored by strongly disagree (1) and strongly agree (7). The valid sample size = 185. Items deleted due to low communities or factor loadings: (1) I have provided the restaurant with my personal information, (2) I trust that the restaurant management puts my interest ahead of its own, (3) I try to make things easy for the restaurant staff, and (4) I try to build a good relationship with the restaurant.

A closer look at the results showed that the items loading on the first factor described how respondents interacted with restaurant staff, and so this factor was labeled Attitudinal Participation (abbreviated as Attitudinal). Since items loading highly on the second factor indicated that respondents took time to seek out information about a restaurant, this factor was labeled Information Participation (abbreviated as Information). The third factor included the “actionable” participation of respondents in service delivery, and thus was labeled Actionable.

Confirmatory Factor Analysis

The researchers next performed CFA to assess the quality of the factor structure by statistically testing the significance of the overall model (e.g., distinction between scales) as well as the relationships between items and scales. For the CFA, the researchers used AMOS version 22 and first evaluated the fit of the three-factor model using the sample variance–covariance matrix as input and a maximum likelihood solution; specifically, the researchers used seven commonly used fit indexes to measure model fit, namely, chi square (χ²/df), the goodness-of-fit index (GFI), the adjusted goodness-of-fit index (AGFI), the normed fit index (NFI), the comparative fit index (CFI), the root mean square error of approximation (RMSEA), and PCLOSE, the p value testing the null that RMSEA is not greater than .05.

Results from the CFA provided further strong support for a three-factor model. Using the sample variance–covariance matrix as input and a maximum likelihood solution, the researchers found that a slightly modified model with deletion of one item in the third factor of Action achieved a good model fit (see Figure 2). In summary, the overall chi-square was statistically nonsignificant, χ² = 15.97, df = 17, p = .0.526; GFI = .98, AGFI = .96, NFI = .98, CFI = 1.0, RMSEA = .000, PCLOSE = .87. These values were all above acceptance levels, indicating a good fit between the model and the observed data. Figure 2 provides standardized parameter estimates.

Figure 2

Study 1: Confirmatory Factor Analysis

Assessment of Construct Validity

To assess the extent to which both the set of measured items derived from the literature review and the qualitative data collection in fact reflected the theoretical latent construct they were designed to measure, the researchers used the CFA results to examine the convergent validity of the measures, discriminant validity of the constructs, and nomological validity (i.e., whether the correlations between the constructs in the measurement theory made sense).

Convergent validity

First the researchers looked at the factor loadings and average variance extracted (AVE) in the model for evidence of convergent validity. As Bollen (1989) has stated, the larger the factor loading or standardized coefficients, the stronger is the evidence that the observed variables represent the underlying constructs. All factor loadings in this study were equal to or greater than 0.70, indicating that the observed variables were good measures of their latent construct. Since the AMOS software does not provide values for AVE, the researchers calculated the measure based on factor loadings using Formula 1. As Table 3 shows, the AVE for each construct was 0.5 or higher, indicating adequate convergent validity.

Table 3

Calculation for Convergent Validity and Discriminant Validity

Convergent Validity
Operation	Construct		Factor Loading	Sq. Factor Loading		Commonalities (1 − Sq. Loading)		CR
	Information		0.871	0.758		0.241
	Information		0.698	0.487		0.512
	Information		0.741	0.549		0.450
Sum			2.310	1.794		1.205
Sq.Sum			5.336
								0.709
	AVE			0.598
	Attitudinal		0.768	0.589		0.411
	Attitudinal		0.876	0.767		0.333
	Attitudinal		0.859	0.737		0.362
Sum			2.503	2.095		1.106
Sq.Sum			6.265
								0.849
	AVE			0.698
	Actionable		0.746	0.556		0.435
	Actionable		0.734	0.538		0.461
Sum			1.480	1.095		0.896
Sq.Sum			2.19
								0.708
	AVE			0.547
Discriminant Validity
			IC	SIC		AVE	SIC
Information	↔	Attitudinal	−0.247	0.061	Information	0.598	−0.247	0.458
Information	↔	Actionable	0.458	0.209	Attitudinal	0.698	0.402	−0.247
Attitudinal	↔	Actionable	0.402	0.161604	Actionable	0.547	0.458	0.402

Note: CR = construct reliability; AVE = average variance extracted; IC = interconstruct correlations; SIC = squared interconstruct correlations.

A V E = \frac{\sum_{i = 1}^{n} λ_{i}^{2}}{n},

where $λ_{i}^{2}$ is the sum of the squared factor loadings and n is the number of measures.

Discriminant validity

The rule of thumb states that all construct AVE estimates should be larger than the corresponding squared interconstruct correlation (SIC) estimates. If such is the case, this indicates that the measured variables have more in common with the construct they are associated with than with other constructs. The calculation based on interconstruct correlations (IC), as shown in Table 3, indicates that all AVE estimates were indeed larger than the corresponding SIC. Therefore, the customer participation three-construct CFA model demonstrated discriminant validity.

Assessment of Construct Reliability

Afterward the researchers assessed the scale’s reliability using construct reliability (CR), which was computed from the sum of factor loadings (λ_i) squared for each construct and the sum of the error variance terms for a construct (δ_i) using Formula 2. The rule of thumb for a CR estimate is that a value of .70 or higher suggests good reliability. In this study, all CR values were greater than .70 (as shown in Table 3), indicating that there was internal consistency and that the measures all consistently represented something.

C R = \frac{{(\sum_{i = 1}^{n} λ_{i})}^{2}}{{(\sum_{i = 1}^{n} λ_{i})}^{2} + (\sum_{i = 1}^{n} δ_{i})},

where δ_i denotes the error variance for each measure, calculated by subtracting the squared factor loading from 1.

Nomological validity

Cronbach and Meehl (1955) argued that a researcher must develop a nomological network for the measure of validity. This network includes the theoretical framework for what will be measured, an empirical framework for how it will be measured, and specification of the linkages within and between these two frameworks. Here the researchers established the nomological network of customer participation by examining its antecedents, consequent variables, and its relationship with some of these variables; the detailed work is reported by Chen, Raab, and Tanford (in press) in another article using the same data set for this article. Briefly, prior research on a customer’s influence on service quality has indicated that customer role clarity (e.g., Webb, 2000), self-efficacy (e.g., Hibbert Winklhofer, & Temerak, 2012), importance of the purchase (e.g., Cermak et al., 2011), servicescape (e.g., Booms & Bitner, 1982), and other factors can significantly influence customer participation in a service life cycle. Meanwhile, customer participation has been found to be an important predictor of customer loyalty (e.g., Hyun, 2010; Namsivayam, 2003; Oliver, 1999; Ryu, Han, & Jang, 2010). Chen et al. (in press) used path analysis to test relationships between and among these variables. Their results (seen in Figure 3) show that the model-fit indices indicated good model fit with a statistically nonsignificant chi-square value of 14.8 (df = 10, p = .139) and good model-fit indicators, specifically CMIN (chi-square value)/df = 1.4, GFI = .99, AGFI = .96, CFI = .99, RMSEA = .035, with a PCLOSE of .71. Also, the exterior environment, purchase importance, role clarity, self-efficacy, and interior environment predicted the constructs of mandatory customer participation, which in turn predicted customer loyalty. This provides adequate evidence for criterion validity as well as generalizability of the scale.

Figure 3

Nomological Network of Mandatory Customer Participation: Results of Path Analysis With Significant Coefficients (Chen et al., 2015)

Validation

The researchers next sought to validate the above results using the remaining half of the data set for Study 2. Validation would enhance confidence that these results could be generalized to the referent population while confirming the validity of the constructs. The valid sample size for this study was 194. The data analysis replicated all the steps in Study 1, including (a) conducting an EFA to identify the most representative measurement items and a CFA to confirm the dimensionality of customer participation and (b) assessing construct validity and reliability.

The results of this replication process were fairly consistent with Study 1. The researchers therefore decided to test the measurement invariance between two randomly split subsamples in order to determine whether the same CFA model would be applicable across groups. In other words, the observed scores should depend only on latent construct scores, and not on group membership or occasion; the observed differences between groups should reflect true differences in the amount or variability of the construct.

The general testing procedure involved testing measurement invariance between the unconstrained model for all groups combined, and then for a model with constrained parameters (parameters such as factor loadings are constrained to be equal between groups). If the chi-square difference statistic is not significant between the original and constrained models, then we may conclude that the model has measurement invariance across groups. Therefore, the researchers merged the two randomly split subsamples, coded as Study 1 and Study 2, and set the test parameters in both models as the variance, covariance, and regression weights. The model associated with Study 1 was set as “Study 1 model” and the model with Study 2 as “Study 2 model.”

As Table 4 shows, the model summaries for both models had close scores: Both chi-square values were nonsignificant, and the model fit index values were very close. The standardized regression weights were all above 0.5, and the squared multiple correlations were all equal to or above 0.49, thus falling between the acceptable range of R² values from .213 to .780 (Doll, Raghunathan, Lim, & Gupta, 1995). Finally, the nested model comparison showed a nonsignificant chi-square difference between the two models. As expressed by CMIN, the nonsignificant chi-square difference (4.78, df = 6, p = .57) suggested that imposing the additional restrictions of six equal factor loadings across two groups did not result in a statistically significant worsening of overall model fit. AMOS assumed that the baseline model (Study 1 model) was true. The second model (Study 2 model), which specified a group invariant factor pattern, was supported by the sample data.

Table 4

Comparison of CFA Results Between Two Randomly Split Subsamples (Study 1 Model Versus Study 2 Model)

			Study 1 Model		Study 2 Model
Model summary
Number of distinct sample moments			72		72
Number of distinct parameters to be estimated (df)			25 (47)		19 (53)
Chi-square (df, p)			55.16 (47, .19)		59.93 (53, .24)
Fit index for model fit
GFI			.97		.96
AGFI			.95		.95
NFI			.96		.95
CFI			.99		.99
RMSEA (PCLOSE)			.02 (.99)		.01 (.99)
Standardized regression weights
Info1	←	Information	.890		.887
Info2	←	Information	.698		.719
Info3	←	Information	.777		.797
Att1	←	Attitudinal	.765		.766
Att2	←	Attitudinal	.836		.837
Att3	←	Attitudinal	.831		.838
Act1	←	Actionable	.729		.727
Act2	←	Actionable	.746		.722
Squared multiple correlations
Info1			.557		.522
Info2			.532		.529
Info3			.691		.702
Att1			.699		.701
Att2			.586		.586
Att3			.604		.636
Act1			.487		.517
Act2			.793		.786
Nested model comparisons: Assuming Study 1 model to be correct
Model	df	CMIN	p	NFI	IFI	RFI	TLI
Study 2 Model	6	4.776	.573	.004	.004	−.002	−.002

Note: CFA = confirmatory factor analysis; GFI = goodness-of-fit index; AGFI = adjusted goodness-of-fit index; NFI = normed fit index; CFI = comparative fit index; RMSEA root mean square error of approximation; IFI = incremental fit index; TLI = Tucker–Lewis index; RFI = relative fit index.

Discussion

This study makes a number of theoretical contributions. First, it appears to be one of the first empirical studies to examine systematically a scale of mandatory customer participation. As mentioned earlier, many previous studies have been conceptual or have not distinguished the differences between customer roles when discussing customer influence on service outcomes. In contrast, this study first considered the definition of mandatory customer participation, which is quite different from the other type of participation, namely, voluntary. It has argued that although both types of customer participation can influence service outcomes, the mandatory type deserves individual and focused attention since its influence on service outcomes is more direct and immediate as well as being measurable. Second, using both qualitative and empirical methods, the study developed and validated a mandatory customer participation scale, which fit a three-dimension construct model. These dimensions—information participation, attitudinal participation, and actionable participation—were significantly explained by eight measures. A series of studies suggested that the scale exhibits construct validity, reliability, and a certain level of generalizability. Overall, the scale appears to be conceptually sound and psychometrically valid.

The findings of this study also suggest a number of important managerial implications. First, it is important for managers to realize that customers, as service purchasers, do have a level of self-awareness of their own obligation in service production. As shown in Namsivayam’s (2003) empirical study, customers want a sense of control when purchasing hospitality services, and if they do not have control, they want at least some level of fairness from the service provider. That is why some customers blame themselves if something goes wrong in their service experience. Managers can make good use of this customer self-awareness and engage customers in service delivery by giving them a sense of control of their own service purchase experience. One good example to look at is how the LEGO Group engages its customers. As Kalcher (2012), the Global Leader of Consumer Experiences of LEGO Group, states, the LEGO culture seeks to engage customers: The more customers engage with the company’s products, the more fun and rewards they receive. So, hospitality managers might come up with creative ideas that will lead customers to participate happily in service coproduction out of a sense of responsibility and ownership instead of goodwill.

Second, the multidimensional nature of mandatory customer participation should remind managers that interacting with customers before service coproduction and consumption might be better timing than a postpurchase approach. It is less intrusive and can establish a good first step for the customer to have a successful service experience. It has been good to see some top hospitality companies such as Hilton, Marriott, Starwood, and Disney leading this practice by contacting their immediate customers in advance to open the communication channel. In addition, managers could directly use the specific measures associated with the three dimensions of mandatory customer participation in this study as a way to diagnose both how and to what degree their customers should take care of their own responsibilities in service delivery, and so take necessary actions to “train” new customers or “educate” repeated customers about their “new” products.

Third, it is time for both researchers and managers to rethink how to measure service outcomes. The traditional SERVQUAL approach with five dimensions does not include customer roles in service delivery. But the empirical evidence in this study confirms that mandatory customer participation does in fact influence service outcomes such as customer loyalty. Therefore, it makes sense to include mandatory participation measures when measuring service quality, customer loyalty, and other service outcomes. Again, the LEGO Group could serve as a good example for service managers to think about linking customer participation with their overall product experience, in turn leading to profitability. Finally, managers could consider segmenting customers by their levels of mandatory participation. In this respect, service managers could consider the insights provided by health care researchers such as Shaffer and Sherrell (1995), who found that assertive and active patients tended to engage actively in information seeking and exchange with their physicians and thus were able to achieve higher levels of satisfaction in treatment, but that passive patients tended to place lots of trust in their physicians but were usually less satisfied with the service outcomes.

This study has several limitations, which suggest areas for future research. The sample population of the online survey included only customers in the full-service casual dining sector. But the consensus is that different types of services demand different levels of participation from customers. If we used a continuum anchored from left to right from lots of participation to little participation as a measure, the required customer participation in a casual dining restaurant should fall at a middle point. Therefore, the findings in this study should be applied mainly to explain relevant issues in the casual dining food service sector or a sector with similar services. But this also opens opportunities for carrying the framework and scale in this study to other service sectors that require different levels of customer participation. Also, the survey was conducted online. As Wright (2005) argued, online surveys have both advantages and disadvantages, and this study undoubtedly shares such disadvantages as sampling and access issues. Finally, the second study conducted to cross-validate the customer participation model was based on a subsample split from one large sample. As experts suggest, using new data improves the validation of a model. Thus, future research should be carried out in other service sectors (e.g., fine-dining sector, hotels, resorts, etc.) to further validate the scale tested in this study.

Footnotes

Authors’ Note:

The authors would like to thank the Caesars’ Foundation for partially funding this research.

Sandy C. Chen, PhD (e-mail: chens5@ohio.edu), is an associate professor of hospitality management, Ohio University Athens, OH 45701 Office: Grover Center E172. Carola Raab, PhD (email: carola.raab@unlv.edu), is an associate professor in the hotel college at University of Nevada-Las Vegas, Las Vegas, Nevada.

References

Andreason

A. R.

(1983). Consumer research in the service sector. In Berry

L. L.

Shostack

Upah

G. D.

(Eds.), Emerging perspectives in services marketing (pp. 63-64). Chicago, IL: American Marketing Association.

Auh

Bell

S. J.

McLeod

C. S.

Shih

(2007). Co-production and customer loyalty in financial services. Journal of Retailing, 83, 359-370.

Bateson

J. E. G.

(1985). The self-service customer: An exploratory study. Journal of Retailing, 61(3), 49-77.

Bitner

M. J.

Faranda

W. T.

Hubbert

A. R.

Zeithalm

(1997). Customer contributions and roles in service delivery. International Journal of Service Industry Management, 8, 193-205.

Bollen

K. A.

(1989). Structural equations with latent variables. New York, NY: John Wiley.

Booms

B. H.

Bitner

M. J.

(1982). Marketing services by managing the environment. Cornell Hotel and Restaurant Administration Quarterly, 23(1), 35-39.

Bowen

T. J.

Ford

R. C.

(2004). What experts say about managing hospitality service delivery systems. International Journal of Contemporary Hospitality Management, 16, 394-401.

Cermak

D. S. P.

File

K. M.

Prince

R. A.

(2011). Customer participation in service specification and delivery. Journal of Applied Business Research, 10(2), 90-98.

Chen

Raab

Tanford

(2015). Antecedents of Mandatory Customer Participation in Service Encounters: An Empirical Study. International Journal of Hospitality Management, 46(2015), 65-75.

10.

Chua

Sweeney

J. C.

(2003). Customer participation in service production: Development of a multidimensional scale. Paper presented at the Anzmac 2003 Conference, Adelaide, South Australia, Australia.

11.

Claycomb

Lengick-Hall

C. A.

Inks

L. W.

(2001). The customer as a productive resource: A pilot study and strategic implications. Journal of Business Strategies, 18(1), 47-69.

12.

Cronbach

L. J.

Meehl

P. D.

(1955, July). Construct validity in psychological tests. Psychological Bulletin, 52, 281-302.

13.

Dabholkar

P. A.

(1996). Consumer evaluations of new technology-based self-service options: An investigation of alternative models of service quality. International Journal of Research in Marketing, 13, 29-51.

14.

Dean

A. M.

(1996, December). The impact of consumer participation on perceived service quality. Paper presented at the Australian and New Zealand Academy of Management, Wollongong, New South Wales, Australia.

15.

Doll

W. J.

Raghunathan

T. S.

Lim

J. S.

Gupta

Y. P.

(1995). A confirmatory factor analysis of the user information satisfaction instrument. Information Systems Research, 6, 177-188.

16.

Dong

Siu

N. Y.

(2013). Servicescape elements, customer predispositions and service experience: The case of theme park visitors. Tourism Management, 36, 541-551.

17.

Ennew

C. T.

Binks

M. R.

(1999). Impact of participative service relationships on quality, satisfaction, and retention: An exploratory study. Journal of Business Research, 46, 121-132.

18.

Erasmus

A. C.

Boshoff

Rousseau

G. G.

(2001). Consumer decision-making models within the discipline of consumer science: A critical approach. Journal of Family Ecology and Consumer Sciences, 29, 82-90.

19.

Ford

R. C.

Heaton

C. P.

(2001). Managing your guest as a quasi-employee. Cornell Hotel and Restaurant Administration Quarterly, 29(8/9), 46-54.

20.

Groth

(2005). Customers as good solders: Examining citizenship behaviors in Internet service deliveries. Journal of Management, 31, 7-27.

21.

Gruen

T. W.

(1995). The outcome set of relationship marketing in consumer markets. International Business Research, 4, 447-469.

22.

Hair

J. F.

Anderson

R. E.

Tatham

R. L.

Black

W. C.

(2009). Multivariate data analysis: A global perspective (7th ed.). Upper Saddle River, NJ: Prentice Hall.

23.

Hibbert

Winklhofer

Temerak

M. S.

(2012). Customers as resource integrators: Toward a model of customer learning. Journal of Service Research, 15, 247-261.

24.

Hinkin

T. R.

Tracey

J. B.

Enz

C. A.

(1997). Scale construction: Developing reliable and valid measurement instruments. Journal of Hospitality and Tourism Research, 21, 100-120.

25.

Hubbert

A. R.

(1995). Customer co-creation of service outcomes: Effects of locus of causality attributions (Unpublished doctoral dissertation). Arizona State University, Phoenix.

26.

Hudson

(2010). Wooing zoomers: Marketing to the mature traveler. Marketing Intelligence and Planning, 28, 444-461.

27.

Hyun

(2010). Predictors of relationship quality and loyalty in the chain restaurant industry. Cornell Hospitality Quarterly, 51, 251-267.

28.

Kalcher

(2012, April). Engaging consumers: Building the Lego brand and culture one brick at a time. Cornell Hospitality Industry Perspectives, 2(2). Retrieved from http://www.hotelschool.cornell.edu/research/chr/pubs/perspective/perspective-16119.html

29.

Kelley

S. W.

Donnelly

J. H.

Jr. Skinner

S. J.

(1990). Customer participation in service production and delivery. Journal of Retailing, 66, 315-335.

30.

Kelley

S. W.

Skinner

S. J.

Donnelly

J. H.

Jr. (1992, November 25). Organizational socialization of service customers. Journal of Business Research, 197-214.

31.

Kellogg

D. L.

Youngdahl

W. E.

Bowen

D. E.

(1997). On the relationship between customer participation and satisfaction: Two frameworks. International Journal of Service Industry Management, 8, 206-219.

32.

Lee

(1996). Predictability, controllability, and outcome desirability as moderators of the perceived risk, perceived value, and adoption intention of a technology-based service innovation: A personal control perspective (Unpublished doctoral dissertation). University of Alabama, Tuscaloosa.

33.

Mills

P. K.

Morris

J. H.

(1986). Clients as “partial” employees: Role development in client participation. Academy of Management Review, 11, 726-735.

34.

Namsivayam

(2003). The customer’s role in the service encounter: The effects of control and fairness. Cornell Hotel and Restaurant Administration Quarterly, 44(3), 26-36.

35.

Normann

(2001). Service management: Strategy and leadership in service business (3rd ed.). New York, NY: John Wiley.

36.

Nunnally

J. C.

(1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.

37.

Oliver

R. L.

(1999). Whence consumer loyalty? Journal of Marketing, 63(1), 33-44.

38.

Risch-Rodie

Kleine

S. S.

(2000). Customer participation in services production and delivery. In Swartz

T. A.

Iacobucci

(Eds.), The handbook of services marketing and management (pp. 111-126). Thousand Oaks, CA: Sage.

39.

Ryu

Han

Jang

(2010). Relationships among hedonic and utilitarian values, satisfaction and behavioral intentions in the fast-casual restaurant industry. International Journal of Contemporary Hospitality Management, 22, 416-432.

40.

Shaffer

T. R.

Sherrell

D. L.

(1995). Exploring patient role behaviors for health care services: The assertive, activated and passive patient. Health Marketing Quarterly, 13(1), 19-35.

41.

Shostack

G. L.

(1984, January-February). Designing services that deliver. Harvard Business Review, 133-139.

42.

Silpakit

Fisk

R. P.

(1985). “Participatizing” the service encounter: A theoretical framework. In Bloch

Upah

G. D.

Zeithaml

V. A.

(Eds.), Services marketing in a changing environment (pp. 117-121). Chicago, IL: American Marketing Association.

43.

Sweeney

J. C.

Soutar

G. N.

(2001). Consumer perceived value: The development of a multiple item scale. Journal of Retailing, 77, 203-220.

44.

Uzkurt

(2010). Customer participation in service processes: A model and research propositions. International Journal of Services and Operations Management, 6(1), 17-37.

45.

Webb

S. S.

(2000, August). Project managers as client relationship managers: Implications for client loyalty. Paper presented at the National Academy of Managerial Meetings, Toronto, Ontario, Canada.

46.

Wright

K. B.

(2005). Researching internet-based populations: Advantages and disadvantages of online survey research, online questionnaire authoring software packages, and web survey services. Journal of Computer-Mediated Communication, 10(3). doi:10.1111/j.1083-6101.2005.tb00259.x

47.

Gong

(2013). Customer value co-creation behavior: Scale development and validation. Journal of Business Research, 66, 1279-1284.

48.

Nataraajan

Gong

(2011). Customer participation and citizenship behavioral influences on employee performance, satisfaction, commitment, and turnover intention. Journal of Business Research, 64(1), 87-95.

49.

Zaichkowsky

J. L.

(1985, December). Measuring the involvement construct. Journal of Consumer Research, 12, 341-352.

50.

Zeithaml

V. A.

Bitner

M. J.

(1996). Services marketing. Singapore: McGraw-Hill.

51.

Zeithaml

V. A.

Bitner

M. J.

Gremler

D. D.

(2006). Services marketing: Integrating customer focus across the firm. Boston, MA: McGraw-Hill.