University of Amsterdam
Amsterdam Business School
The influence of physical environment, personal factors and wait expectations
on customers’ willingness to wait in restaurants
Mater thesis: final version Student Name: Ximin Juan Student Number: 11086394 Date: 24th June, 2016
Program: MSc Business Administration-Marketing Track-2015/2016 Faculty: Faculty of Economics and Business, University of Amsterdam Supervisor: Adriana Krawczyk
2
Statement of originality
This document is written by Student Ximin Juan who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Acknowledgement
I would like to thank my supervisor, Adriana Krawczyk for her valuable support and encouragement. Our communication was efficient and effective, which really helped me a lot in the past few months. I was so lucky to write my thesis under her guidance. Additionally, I would like to thank my second examiner who will take time to read my thesis and all our customers who help me to fill in the questionnaires.
4
Abstract:
On account of the nature of service industry, waiting sometimes is inevitable. Therefore service companies always try to find their own way to deal with the possible negative influence brought by waits or implement some strategies to reduce customers’ waiting time. However, the anticipations brought by waits can make customers willing to wait. By introducing a new concept willingness to wait, this study examined the possible influence of several factors on customers’ willingness to wait.
In general, this study aims to discuss the factors that could be linked to customers’ willingness to wait and to examine the relationship between these factors and customers’ willingness to wait. Firstly, this study identified the possible relation between main variables including physical environment, personal factors and wait expectations based on literature reviews. Then, main scales that are suitable for the restaurant setting were used. Finally, hypotheses were examined.
Both managers and academics can get benefits from the findings got in this study. Although some of the findings are found mixed, this study provides some valuable insights for future research and can be used for marketers to either enhance customers’ service experience or improve customers’ purchase intentions.
5
Contents
1. Introduction………6
2. Literature review………...7
3. Research gap and research objectives………...14
4. Research questions………....15
5. Theoretical framework….……….16
6. Research design……….23
7. Measures………....26
8. Statistical results and analysis………29
9. Discussion………...39
10. Research implications………..45
11. Limitations and future research………49
Reference……….51
Appendix……….59
List of tables Table 1 Hypotheses………. 22
Table 2 Reliability of measurement items………... 28
Table 3 Profile of respondents……… 29
Table 4 One-way ANOVA……….. ...32
Table 5 One-way ANOVA………. 33
Table 6 Correlation analysis………...34
Table 7 Simple regression model………....36
6
1.
IntroductionResearch on waiting is essential, because the knowledge of how to control the timing of the delivery of services’ elements can be an important competitive advantage to firms especially when services cannot be offered instantaneously (Bitran, Ferrer and Rocha e Oliveira, 2008). However, waiting is always regarded as an unpleasant thing, since waiting is associated with psychological and economic costs. Waiting can produce negative emotions (Antonides, Verhoef and van Aalst, 2002), for example, most of customers feel unhappy and bored when waiting. Therefore, numerous researchers focus on the negative impacts brought by waiting and try to deal with these negative impacts by distracting customers, increasing efficiency or making waiting time under control. For example, Dragicevic and Rakidzija (2012) recommend managers to adopt music in store to reduce customers’ negative mood while waiting. However, since waiting can be categorized into active waiting and un-active waiting (Winter Nie, 2000), there are some occasions that customers are willing to wait, waiting can lead to hope and anticipation as well (van Riel et al., 2012).
In some cases, waiting is a necessity, for example, waiting at check out in a supermarket cannot be avoided, and waiting in this case is always unpleasant for customers (van Riel et al., 2012). But sometimes people are willing to wait even high opportunity costs exist. For example when customers are queuing in front of the recreational facilities in an amusement park, facilities that need waiting always signal their popularity, therefore waiting for these facilities can improve customers’ expectations on following recreational activities (Koo and Fischbach, 2010). In this situation, customers would like to wait rather than enjoy other facilities that do not require waiting in no time. The same phenomenon can be seen in some restaurants and boutiques. Some restaurants attract lots of customers to queue in front of it even with smiles although these restaurants are crowded enough inside. From this perspective, waiting does not cause negative emotions on customers but makes people expect the services.
7
Then, the key questions are why they are willing to wait instead of getting services from a service provider that can serve them at any time. Combining physical environment, personal factors and wait expectations, this study aims to use physical environment, personal factors and wait expectations to evaluate customers’ willingness to wait as well as giving possible reasons why customers are willing to wait.
2. Literature review 2.1 Waiting
A wait is defined as the duration from the moment the customer is ready for the service encounter to the moment the encounter actually starts (Taylor, 1994). Since operational capability is limited and optimal volume is hard to be controlled, waits usually occur when demand exceeds supply (Wu, Lu and Ge, 2013) or when there is low operational efficiency. On account of the limited capacity of service and varied demands of service, waits sometimes are inevitable for customers.
Basically, waiting time can be divided into two categories: actual waiting time and perceived waiting time (subjective time) (Wu, Lu and Ge, 2013). Perceived waiting time is defined as how individuals perceive and feel about the length of time durations (Baker and Cameron, 1996). Owing to individual and situational factors, perceived waiting time can predict customer satisfaction better than actual waiting time (Nie, 2000). As a result, most academics concentrate on customers’ subjective perception of waiting time.
Because of the growing significance of waiting to both customers and managers, this subject has attracted lots of attentions in the past few years, especially in the field of production/operation management. Considering the growing number of service activities and the nature of service, researchers begin to pay attention to the influence of waiting time on service industry by incorporating psychological elements regarding waiting (Seawright and
8
Sampson, 2007). Ad Pruyn and Ale Smidts (1998) find that waiting has a strong impact on customers’ overall satisfaction, but it is the waiting condition instead of the objective waiting time that influences customers a lot. As waiting is regarded as a waste of time and an economic and psychological cost, it can cause negative emotions on consumers, thereby lowering overall satisfaction with service (van Riel et al., 2012). Besides the effect of waiting on customers’ mood, waiting also influences the prosperity of customers’ repurchase and word of mouth in the future (Nie, 2000).
Waiting can occur before, during and after purchase (Durrande ‐M oreau, 1999). Customers’ waiting time during and after purchase has attracted a majority of attention. When customers wait after purchase, service delays occur. Customers hate waiting and service delays, especially when a firm makes a commitment that it will deliver service in a guaranteed waiting time (Kumar, Kalwani and Dada, 1997) .If waiting time far exceeds customers’ degree of tolerance, managers will have to deal with the influence of service failure on customers through using service recovery strategies. Consequently, the optimization of waiting experience, the relief of customers’ negative emotions and the reduction of customers’ perceived waiting time are main focuses in academic field. For example, Jim Browne (1999) thinks that the key to managing service operations is balancing customer waiting time and staff idle time. Because of the varied demand of services, managers have to determine how many staffs they need to meet these demands at various times. Butcher and Kayani (2007) find that managers can use service intervention strategies such as providing length of waiting or reasons of a delay to relieve customers’ negative emotions on certain circumstances.
Although plenty of research is conducted on waiting during and after purchase, research on waiting before purchase is rare. However, the influence of wait before purchase on
9
customers’ satisfaction outweighs that of wait during purchase (Davis, 1990). This study aims to fill the blank in the subject of waiting before purchase.
Besides waiting before, during and after purchase, Durrande‐M oreau (1999) indicates other factors associated with waiting can be classified on basis of two main criteria, namely, individual and situational factors, duration and other factors, this author also puts forward that distraction, expectation, environment and uncertainty are four major elements that influence customers’ perceived waiting time.
2.2 Wait expectation
Although, waiting, as an increasingly important subject, is brought to service field, wait expectation and its determinations remain relatively unexplored and should attract more attentions (Durrande‐M oreau, 1999) . It is widely accepted that the key to improving service quality is to narrow down the gap between expected service and perceived service (Sarkar, Mukhopadhyay and Ghosh, 2011). Like expectations’ significance in the whole service industry, wait expectations also play a key role in customers’ wait decisions. As Nie (2000) mentions, by measuring expected waiting time, customers make decisions on whether they will wait for products (services) or not. Kumar, Kalwani and Dada (1997) also find that prior wait expectation has a significant impact on both in-process and end-of- process satisfaction and overall experience. Therefore, understanding wait expectation is necessary.
Durrande‐M oreau (1999) defines wait expectation as “the probable waiting time in the mind of the customer” and illustrates that waiting time is a subjective estimate of time before customers’ purchase. Based on different levels of customer expectations, Nie (2000) classifies wait expectation into wait expectation of desired level and tolerable level (“the minimum tolerable expectation of waiting time that customer can accept” [Nie, 2000, p621]), and he also suggests that the region between these two levels could be influenced by personal
10
and situational factors. Customers’ wait expectations also have a significant impact on store patronage, thereby being a reason of customers’ waiting before purchase. Because when customers expect short waiting time, they are more likely to enter the queue (Grewal et al., 2003). Thus, it is reasonable to infer that positive wait expectation contributes to customers’ waiting behaviors. According to the findings of Zeithaml, Berry and Parasuraman (1993), expectations are usually divided into several standards including expectations-as-predictions standards, expectations-as-ideal standard and other standards. Expected standard is defined as “an objective calculation of probability of performance” (Zeithaml, Berry and Parasuraman, 1993). In this study, expectations are seen as customers’ predictions about what may happen in the purchase process, thus predictive expectations are used in this study.
2.3 Physical environment
Physical environment is always important in service settings in that customers can actually experience the service facilities and join in the co-production process (Bitner, 1992). It is also proven to be a strong determinant of service satisfaction and an influential factor in customers’ reaction to waiting. Because of the nature of service, physical environment serves as a functional cue in service business. Physical facilities and other forms of tangible communication are general elements of physical evidence for service businesses (Puspita, 2015). And the key to causing desirable consumer behaviors is achieving the synergy of all elements of physical environment and making service places recognizable and special for customers (Dragicevic and Rakidzija, 2012).
The influence of specific elements of physical environment on perceived waiting time and customer satisfaction has been explored a lot. Choi and Sheel (2012) suggest that sitting services such as providing a comfortable space and enough chairs for customers can significantly enhance customer satisfaction with a restaurant. Webster and Sundaram (2009)
11
measure the relationship between customers’ evaluation on service and service provider’s communication style and indicate that there is a high correlation index between these two elements (customers rely on service provider’s communication style to evaluate the received service). Storage image can play a mediating role in the impact of customers’ negative emotional response to wait on customers’ overall satisfaction (van Riel et al., 2012). As the most uncontrollable part of physical environment, service providers’ communication style can influence customers’ waiting experience: the more satisfied they feel during communication processes, the more positive they will evaluate the service (Webster and Sundaram, 2009). However, the influence of physical environment on waiting may be subtle on account of situational factors and personal factors. As a normally used element in physical environment to improve customers’ waiting experience, music is indicated not always as useful and positive as service providers expect (Cameron, Baker and Peterson, 2013). In the study of Ad Pruyn and Ale Smidt (1998), when they use the presence of TV as a variable that can distract customers from waiting and assume the presence of TV can shorten customers’ perceived waiting time, the results go opposite to their hypothesis, which suggests the function of physical environment can be influenced by personal factors. For example when customers are not anxious, they are less likely to be influenced by the attractiveness of physical environment. Basically, physical environment can influence customers’ subjective perception of the length of waiting duration.
Although physical environment is regarded as an important element that influences customers’ perceived waiting time and overall satisfaction, research on physical environment as a reason of customers’ waiting is rare. Based on attribution theory, when customers lack previous experience or knowledge of the services or products that they want to buy, they tend to use peripheral cues to make decisions. When customers have no idea about their choices among numerous restaurants, they are more likely to choose a restaurant that is physically
12
attractive. For example, the presence of music is proved to cause positive emotions on consumers and to improve their evaluation on store atmosphere (Grewal et al., 2003), thereby making customers willing to wait.
In most studies, researchers think that the more customers wait in line, the longer waiting time customers will expect, which means high perceived customer density is always associated with a small likelihood of customers’ waiting behaviors. Therefore, in most studies, researchers think customer density can lower customers’ purchase intention, and crowding influences customers’ decision to enter the service environment by diminishing desired privacy (Hwang, Yoon and Bendle, 2012). However, crowding can only cause negative emotions such as anger, frustration on a part of customers (Mattila and Hanks, 2012). According to attribution theory and cognitive dissonance theory, crowding or queuing sometimes can signal products or services of high values, even it requires lots of efforts (Koo and Fishbach, 2010). Therefore, customer density may also contribute to customers’ waiting. Given that physical environment can serve as an important indicator of customers’ perception and future decision (Hightower, Brady and Baker, 2002), it can be inferred that physical environment possibly influences customers’ willingness to wait.
2.4 Perceived customer density
Customer density can be examined in either physical aspect or psychological aspect (Noone and Mattila, 2009), the latter one is always associated with customers’ perceptions, which is more valuable in service encounters. Another word similar to customer density is crowding which signals a high perceived customer density situation, in other words, perceived customer density is thought to reflect an individual’ conceptualization of crowding. This means individuals who enter in a crowded service setting should perceive the crowding in their own ways (Baum and Greenberg, 1975). Early research on crowding puts too much
13
emphasis on retail environment (Baum and Greenberg, 1975), instead, this study will focus on restaurants where high perceived customer density also strongly influences customers’ behavioral intentions.
Normally, crowding is assumed aversive and believed to produce negative emotions on people. A large group of researchers find that potential customers could avoid the crowdedness in service settings by canceling their purchase plans (Michael and John 1990). Customers’ attitudes towards crowding could be attributed to their past experience and tolerance of crowding (Baum and Greenberg, 1975), when individuals enter in the crowded setting, their behaviors consistent with their expectations on crowding arouse. In some studies, academics find that crowdedness may hinder the goal-oriented customers (Eroglu & Harrell, 1986; Eroglu & Machleit, 1990, Eroglu, et al., 2005), which means that the negative effects of crowding may be limited to some circumstances that overcrowding interfere with the achievement of customers’ hedonic and utilitarian goals. Noone and Mattila (2009) indicate that when customers are seeking a utilitarian dining experience, a crowded service area always leads to a perception of lower service quality, yet when they are seeking a hedonic dining experience, a crowded restaurant will contribute to a perception of higher service quality. Generally, crowding has a direct impact on customers’ approach-avoidance response, in other words, it appears to influence customers’ decisions to enter the restaurant (Hwang, Yoon and Bendle, 2012). In this study, customers’ perception of density will be linked to a variable relevant to behavioral intention, willingness to wait. Given that the crowdedness in restaurant always can attract more customers, the inference that perceived customer density could result in customers’ willingness to wait can be made.
14
2.5 Personal factors
The value of time is not constant but depends on contexts (Lin, Xia and Bei, 2014). Customers who decide to wait for the service are always not a homogenous group as well. As the subjective estimate of waiting time is more influential than real waiting duration, it is also important to understand the impacts brought by personal factors. Some subjective waiting variables such as customers’ personal characteristics possibly have impacts on the appraisal of waiting time (Ad Pruyn and Ale Smidts, 1998). Although the significance of personal factors such as moods and gender seems not clear to managers especially before wait,
customer categories are still meaningful to managers (Durrande ‐M oreau, 1999), and research on interaction of crowding and personal factors is needed (Mattila and Hanks, 2012).
For example, male and female or the old and the young should be treated differently. New or infrequent customers tend to feel a longer waiting than frequent visitors (Jones and Peppiatt, 1996), customers without pre-existing knowledge may attribute waiting to the fact that longer waiting time may lead to products or services of high quality. Meanwhile, although customers with queuing information tend to overestimate the perceived waiting duration, when customers have affective responses to service providers, they are likely to have positive attitudes toward the acceptable wait (Hui and Tse, 1996). Besides, according to social identity theory, female and male do experience different levels of satisfaction, and female customers tend to be more empathy while they are asking for services from female owners (Jones and Peppiatt, 1996), therefore female customers are more likely to wait for a longer time while there is a female service provider.
3. Research gap and research objectives
Although academic research on waiting undoubtedly grows in these decades, this topic is still rarely investigated. Overall, pervious research mainly focuses on customers’ subjective
15
perception of waiting time during waiting, customers’ evaluations on service after purchase and strategies that can reduce customers’ subjective perceived waiting time. Research on customers’ willingness to wait before purchase is rarely conducted. Although the influence of physical environment and personal factors on customers’ subjective waiting time has been researched a lot, this study will use waiting expectation as another indicator to explore customers’ willingness to wait besides those two factors.
Managers especially in service industry will strongly benefit from this study. The exploration of customers’ willingness to wait can increase their understanding of customers’ waiting behaviors and help them to enhance a specific element in service scape or implement an effective strategy to improve customers’ willingness to wait during peak time. Since this research focuses on waiting before purchase, managers can increase customers’ purchase intention or waiting motivations by optimizing the physical environment and categorizing customers into different groups, thereby increasing sale volume and improving profits.
4. Research questions
Unlike previous research that focuses on either customers’ waiting experience during or after purchase or strategies of reducing the influence of waiting time in service encounters, this study will use three constructs---physical environment, personal factors and wait expectations to evaluate customers’ willingness to wait in restaurant contexts.
As mentioned, over the past decades, research mainly focuses on reliving the impacts of waiting duration on customers and exploring waiting experience on customers’ evaluation in service encounters, while research on customers’ wait expectations and willingness to wait is rather limited. Additionally, compared with the previous literature that emphasizes the negative side of crowding, this study will creatively address the positive influence that crowding can implement on waits. Overall, the research question can be illustrated as below:
16
How do personal factors, wait expectations and physical environment influence customers’ willingness to wait?
5. Theoretical framework
In this section, based on previous literature review, the conceptual framework in which all relationships among variables are built is showed and hypotheses that are going to be tested further on are presented.
5.1 Physical environment and customers’ willingness to wait
Physical environment can cause different behaviors of customers: approach and avoidance, these two types of behaviors are both influenced by customers’ perception of environment (Mehrabian and Russell, 1974). Here, the term “approach” refers to customers’ patronage intention, while “avoidance” represents customers’ getting away from the service. On account of the function of physical environment as cues, customers may evaluate the
17
attractiveness of physical environment from different perspectives before making decisions on approach or avoidance.
In addition, physical environment can also perform as an element that is beneficial to customers’ “duration neglect”. The term “duration neglect” originates from a paper written by Bitran, Ferrer and Rocha e Oliveira (2008). In the research, they emphasize the role of duration in evaluation of service quality and indicate that “duration neglect” is one of the two main ways in which waiting/duration impacts customers’ experience. The term “duration neglect” shows that the influence of duration on customers’ actual experience depends on the weight of duration in customers’ overall service encounter, in other words, if the service encounter is salient, customers are more likely to neglect the waiting time they spend on service encounter, yet if the experience is unpleasant or the duration is salient compared with other attributes of the experience, customers will care more about the duration and will have to trade off between waiting and other attributes. Consequently, to trigger “approach” behaviors, the duration in service encounter should not be salient. In this case, an attractive physical environment can be a tool that distracts customers from waiting and leads to “duration neglect”.
To measure the attractiveness of physical environment, it is said that the key to causing desirable consumer behaviors is achieving the synergy of all elements of physical environment (Dragicevic and Rakidzija, 2012). Therefore in this study we will evaluate the overall attractiveness of physical environment before specific elements. Bitner (1992) uses three dimensions to evaluate physical environment in service encounters, namely, ambient conditions; spatial layout and functionality; signs, symbols and artifacts, which are commonly used to categorize different items of physical environment. Here, ambient conditions represent “background characteristics of the environment such as temperature,
18
lighting, noise, music and scent” (Bitner, 1992). Spatial layout refers to the arrangement of furnishings, equipment and machinery, and signs, symbols, artifacts serve as cues of locations for customers. Mehrabian and Russell (1974) identify that physical environment can cause customers’ emotion variation, and they divide these emotions into three groups: pleasure, arousal, and dominance.They also indicate that these three factors can be used to categorize the elements of physical environment. Additionally, Baker (1986) also uses ambient elements and design elements as two main constructs for measuring physical environment. The biggest difference Baker (1986) uses to study waiting time is that he replaces symbols and signs with social elements as an important part of categories of physical environment on the basis of the experience that customers go through in each service encounter. Based on diverse literatures above, this study will use ambient elements and design elements which are both mentioned in two researchers’ models. Given that when customers make decisions on whether join the queue or not, they are already close to restaurants and rarely have chance to interact with service providers, the third dimension---signs and symbols or social elements will be excluded.
Willingness to wait (WTW), as a relatively novel concept, aims to “assess the willingness of customers to wait before a certain service” (Riganti and Nijkamp, 2008, p8). By referring to stated preference methods to indicate customers’ attitudes or preferences, this method can show some information on customers’ satisfaction with service or their behavioral intentions (Riganti and Nijkamp, 2008). In most literatures, WTW is measured by the maximum estimated waiting time customers can accept (Riganti and Nijkamp, 2008; Poelmans and Rousseau, 2014). In this study, willingness to wait mainly serves as a dependent variable that reflects customers’ attitudes towards waiting behaviors.
19
H1: the attractiveness of design elements of physical environment is positively related to customers’ willingness to wait.
H2: the attractiveness of ambient elements of physical environment is positively related to customers’ willingness to wait.
Additionally, considering the features of the restaurant that we will conduct a research on, we explicitly manipulate one specific factor---perceived customer density, in that this restaurant is run as an open kitchen. In this restaurant, the cooking process and food delivery are openly showed to customers. Usually this restaurant can attract more people when numerous customers are waiting for the service in front of it than when few people are waiting there. Consequently, compared with traditional literature which assumes that crowding increases perceived customer density and lowers purchase intention (Grewal et al., 2003), we propose that perceived customer density is positively related to willingness to wait, namely, the more seeable customers wait in front of the restaurant, the more willing to wait potential customers will be.
H3: perceived customer density is positively related to customers’ willingness to wait.
5.2 Personal factors and customers’ willingness to wait
As suggested above, customers’ waiting is always associated with personal factors such as mood, personality traits and past experience. For instance, Giebelhausen, Robinson and Cronin (2010) find that customers who lack past experience use waiting time as a proof of high quality services, and as a result, their satisfaction with service increases while they are experiencing a long duration.
Personal factors refer to the sub-categories that customers fall into (Zeithaml, Berry and Parasuraman 1993), as thestudy points out, customers with high social status may have high
20
requirement and expectation on a certain kind of service. In this study, customers will be categorized by gender, age and past experience given that different categories may implement multiple impacts.
As mentioned before, gender, age and past experience can cause a difference on waiting behaviors. Socialization theory indicates that based on the social and work experiences, men have more time pressure than women, and men tend to be more time conscious than women (Sharma, Chen and Luk, 2012). For example, men care more about the outcomes they get from shopping while women enjoy the experience, thus women tend to make shorter
estimations of waits and more willing to wait than men (Durrande ‐M oreau, 1999. Moreover, females tend to be more empathetic than males from either social or psychological
points of view (Lin, 2009), which means females possibly can better understand service providers while they see service providers’ efforts during peak time. Their empathy originating from what they see may lead to their stronger willingness to wait than men’s. From this perspective, in terms of willingness to wait, we can assume that:
H4: women are more willing to wait than men.
The perception of time can be different for people at different ages on account of the accumulated life experiences and the aging process (Moschis, 1994). With aging, older people have more purchase experiences based on which they can compare values than younger people do (Ligas and Chaudhuri, 2012). John and Cole (1986) also point out that although older people are more tolerant of and less sensitive to negative information, they are more conscious about their time and efforts they input in the transactions. Therefore, similar to women in gender category, younger people tend to be more willing to wait considering their characteristics. Then the following hypothesis can be developed:
21
Consumers with experience of a product or service may have different expectations from those without the experience (Rodríguez Molina, Frías-Jamilena and Castañeda-García, 2013). Here, past experience refers to “the level of familiarity a shopper has with a particular store (Ligas and Chaudhuri, 2012, p4)” or a store providing similar types of services. It is a “belief updating process” (Ligas and Chaudhuri, 2012, p4), based on which intentions are developed. According to schema congruity, similar information is less likely to be dealt with carefully than unfamiliar information (Ligas and Chaudhuri, 2012). For customers who drop by the restaurant frequently or have similar purchase experiences, they may compare experiences that are similar on most dimensions (Kim and Moon, 2009), customers who experience the service or similar service before are more likely to evaluate the service based on their past service experiences. Here, based on uncertainty reduction theory, past experience performs as additional information for reducing uncertainty during transactions, therefore customers with past experience may feel more confident in their choices (Ligas and Chaudhuri, 2012), leading to greater willingness to wait. In this case, another hypothesis can be developed as below:
H6: customers with past experience are more willing to wait than customers without past experience.
5.3 Wait expectations and customers’ willingness to wait
Expectation represents “the prediction of what customers think they will receive from service providers (Devlin, Gwynne and Ennew, 2002, p4)”. Therefore, expectation can serve as an important indicator of behavioral intentions such as patronage or purchase intention. For example, customers under high time pressure may refuse to patronize a store if they think they will wait for a long time or have negative wait expectations on the services or products (Grewal et al., 2003). From this perspective, customers with negative wait expectations probably assume that the cost of waiting is much higher than the value they can get from the
22
products, resulting in the fact that their willingness to wait will be lower than customers with positive wait expectations. The hypothesis based on this inference can be developed as below:
H7: customers with positive wait expectations are more willing to wait than customers with negative wait expectations.
Table 1 Hypotheses
All hypotheses are advanced:
H1: the attractiveness of design elements of physical environment is positively related to customers’ willingness to wait.
H2: the attractiveness of ambient elements of physical environment is positively related to customers’ willingness to wait.
H3: perceived customer density is positively related to customers’ willingness to wait.
H4:women are more willing to wait than men.
H5:young people are more willing to wait than older people.
H6: customers with past experience are more willing to wait than customers without past experience.
H7: customers with positive wait expectations are more willing to wait than customers with negative wait expectations.
23
6. Research design
Normally, experiments are widely used in the studies regarding physical environment and customers’ waits on account of the chief advantages of experiments such as the control of variables and great flexibility (Cameron, Baker and Peterson, 2013; Hul, Dube and Chebat, 1997; Giebelhausen, Robinson and Cronin, 2010). However, it can be difficult to generalize the results got from highly controlled experiment to real life situations (Alwin, 1991). For this reason, surveys, or rather, questionnaires that can maximize external validity were designed so that the sample can be representative of the whole population (Alwin, 1991). The research setting is a small restaurant located in north-eastern part of China. Since the restaurant is operated in the mode of an “open kitchen” (customers can see what is going on in the restaurant from outside), it requires customers to wait outside the restaurant before purchase so that there is a high possibility that more customers could be attracted by chiefs’ skillful cooking process or the pleasant odor. This type of restaurant is especially attractive to customers with strong curiosity.
The whole questionnaire was divided into two parts. In the first part, the influence of personal factors and the elements of physical environment were tested based on the real restaurant and customers’ personal information. Additionally, customers’ evaluations on the physical environment in the restaurant were required to input. Uhrich and Luck (2012) state that varying customer density conditions should be set for the reason that the comparison can shed light on influence of different levels of crowdedness on customers’ evaluation apprehension. Thus, they suggest that future researchers should endeavor to explore customers’ behavioral consequences of either low customer density or high customer density (Uhrich and Luck, 2012). As a consequence, in the second part, customers were given two photos that describe both uncrowded and crowded situations of the restaurant. To simulate the physical
24
environment, I used photos which are applied widely and proven an effective method of testing customers’ perception of customer density (Rompay et al., 2008; Pan and Siemens, 2011). In total, 15 photographs of either crowded situation or uncrowded situation were taken. Two photos that can show different levels of crowdedness were chosen finally. In the first photo, two employees are preparing for cooking with six customers waiting in front of the restaurant. Considering the size of the restaurant, the six customers’ situation can be seen as “uncrowded situation”. In the second photo, twelve customers are waiting in front of the restaurant in the same situation, with some of them sitting or standing, even the tables on the platform are fully occupied, and employees are busy with cooking and charging. After either of the two photos was shown randomly, following questions with respect to customers’ perception of crowdedness based on the photo and their willingness to wait in this situation were asked. Considering that peak hours suitable for the setting of crowding last from 6:00 pm to 9:00 pm, all the photos were taken in the evening to keep other variables the same in the environment. In general, two versions of questionnaires including either of two photos were designed with other parts totally the same. The two chosen pictures are shown below:
25
Photo 2 Crowded situation:
In order to ensure the measurement of wait before purchase, all questionnaires were distributed to potential customers who stop and look through the menu in front of the restaurant before actually paying for their products. Accordingly, all respondents who fill in the questionnaires were guaranteed to be potential customers of this restaurant. Using potential customers can lead to a significant reduction of biased results resulting from sampling population. In view of the restaurant setting, the questionnaire was administered in a paper-pencil version. All questionnaires were distributed by the owner of the restaurant who monitored the daily management of the restaurant at the same time. With regard to ethical issues, all respondents were informed about the purpose of this study and the confidentiality of their personal information before they fill in questionnaires. Moreover, in order to improve survey response, questionnaires were suggested to be designed “respondent-friendly” in terms of specific graphical aspects (Chisnall, 2007). Hence, all questionnaires in this study were color-printed.
Since all questionnaires were planned to be delivered in a restaurant located in China, all questionnaires were initially written in English, and then were translated into Chinese. Before
26
being formally distributed, questionnaire was checked by 10 Chinese students who study in graduate program in University of Amsterdam in order to correct all typos and possible translation mistakes. Then a prior survey was done among a small group of customers who came to the restaurant on the first day to improve the quality of results and ensure that all questions are understandable (Saunders and Lewis, 2012). After being refined and printed, all questionnaires were formally delivered on the second day. Given that the second part of the questionnaire is different, two versions of questionnaires were delivered randomly. Meanwhile the numbers of delivered questionnaires in each version were kept almost equal. The distribution of questionnaires lasted for more than three weeks. In total, 139 questionnaires were collected with the first version accounting for 70 and the second version representing 69. 135 questionnaires are effective. Then all collected data were input into computer and saved in digital formats.
7. Measures
Most items in multiple scales were borrowed from the previous literature, and some of them were modified slightly so as to be applied to the restaurant context. Besides, since there are no items available for a few variables, a small group of items were developed for this study. Ambient elements in physical environment are normally intangible and related to background conditions such as smell, music and lightning (Yalch and Spangenberg, 1990; Milliman, 1982). Considering the specific restaurant context, in this study, smell and lighting can be the most attractive elements. However, by virtue of the mode of an “open kitchen”, background music will make the waiting environment annoying. Therefore, normally background music is not applied in this restaurant, and in this study, it will not be tested. As for design elements, usually, this variable includes tangible and visual stuff such as displays, color, cleanliness and layout. The items of the attractiveness of ambient elements (Cronbach’s α=0.81) and design
27
elements (Cronbach’s α=0.81) were both borrowed from the study of Baker, Grewal and Parasuraman (1994). In ambient elements, a two-item, 5-point likert scale (strongly agree/strongly disagree) was used in questionnaires: the odor makes this restaurant pleasant; the lightning is appropriate. To measure design elements, a four-item, 5-point likert scale (strongly agree/strongly disagree) was used: the color theme is pleasing; the colors used in the restaurant appear to be currently fashionable; the physical facilities are attractive; the facilities in the restaurant appear organized.
To measure perceived customer density, a multi-scale was borrowed from Machleit, Kellaris and Eroglu (1994)’s study and modified to fit the context: there are a lot of customers in front of the restaurant; this restaurant seems very crowded to me; this restaurant is a little too busy (anchored by 1=strongly disagree and 5=strongly agree; Cronbach’s α=0.81)
In terms of wait expectations, a semantic differential scale (Cronbach’s α=0.80) used by Grewal et al. (2003) was adopted and modified to fit the situation in this study. In this semantic differential scale, two questions were asked: “how short/long would this amount of time feel? Would this amount of time be reasonable to wait for service in this restaurant?”
In the final part of the questionnaire, instead of the normal method of letting participants estimate the maximum time they are willing to wait for (Riganti and Nijkamp, 2008; Poelmans and Rousseau, 2014), a semantic differential scale was developed for this study so that respondents can indicate their willingness to wait: how willing would you be to wait in front of this restaurant? (weak/ strong).
To measure the consistency and reliability of all scales, reliability analysis was performed. Reliability refers to “the ability of a measure to produce consistent results when the same entities are measured under different conditions (Field, 2014, p1502)”. Since variables developed for other purposes will be accepted only when they can provide stable results over
28
a repeated administration of the test (Santos, 1999), reliability analysis is necessary in this study. In Table 2, Cronach’s Alphas of all scales are above 0.7, which suggests that all scales are acceptable (Nunnally, 1967).The results of reliability analysis are shown in Table 2.
Table 2 Reliability of measurement items
Variables Measurement items Cronbach’s Alpha
Wait expectation
1. How short/long would this amount of time feel?
2. Would this amount of time be reasonable to wait for service in this restaurant?
0.80
Design elements
1. The color theme is pleasing.
2. The colors used in the restaurant appear to be currently fashionable.
3. The physical facilities are attractive.
4. The facilities in the restaurant appear organized.
0.81
Ambient elements
1. The odor makes this restaurant pleasant.
2. The lightning is appropriate.
0.81
Perceived customer
density
1. There are a lot of customers in front of the restaurant;
2. This restaurant seems very crowded to me;
3. This restaurant is a little too busy.
29
8. Statistical results and analysis
8.1 Sample characteristics
Before proceeding to testing all models, a missing data check was done, which produced acceptable results. Particularly, the variable wait expectation was recoded. All descriptive statistics with regard to all items can be seen in Table 3. The number of effective questionnaires in this study was 135 in total, with female respondents (n=80) and male respondents (n=55) representing 59.3 percent and 40.7 percent respectively. The age category of respondents ranged mainly from 18 to 52, and in these respondents, respondents at 24 years old accounted for the biggest portion (26.2 percent), followed by respondents at 23 years old (12.3 percent) and 25 years old (12.3 percent). In this investigation, 70.9 percent of respondents said that they had past experience, while a small portion of respondents (29.1 percent) were new customers.
Table 3 Profile of respondents
Characteristic Frequency Percentage
Gender Female Male 80 55 59.3 40.7 Age ≦20 21-30 31-40 8 101 17 6.2 77.6 13.1
30 41-50 ≧51 3 1 2.3 .8 Past experience Yes No 95 34 70.9 29.1 8.2 Hypotheses testing
Usually, statistical analysis is considered to be one of the most useful methods of examining hypotheses about the relationship among variables (Zvizdojevic and Vukotic, 2015). In this section, to investigate all the hypotheses developed in other parts, correlation analysis was run to pre-test the model, then on account of the different categories of independent variables, simple linear regression analysis was performed to examine the relationship between each independent variable and willingness to wait, and one-way ANOVA was performed to examine the difference between two groups. Finally, multiple regression analysis was done to test the entire model. During this process, the main reported results are standardized regression coefficients.
To test the difference between groups, one-way ANOVA was performed, due to the fact that both gender and past experience are measured as categorical variables (dummy coding). Dummy codes, as suggested, are usually used to code categorical data before a further analysis is performed, which can be seen as the dichotomization of data (Irwin and McClelland, 2003). According to Field (2014), the basic theory behind ANOVA (acronym for analysis of variance) is using a linear model to compare group means. As illustrated
31
before, H4 predicts that compared with men, women are more patient to their services and more willing to wait, while H6 predicts that past similar experience will make people more willing to wait since they know what they can expect. In this part, by comparing the means of two subgroups in each category, we can find whether there is a significant difference between genders and customers with/without past experience. Past experience, serving as a categorical variable, was measured and coded (“yes”=1, “no”=2). After performing one-way ANOVA, I found a statistically significant effect of past experience F (1, 130) = 20.938, p < .01 between two subgroups, which means there is a significant difference between customers with past experience and without past experience. While comparing willingness to wait between female and male, the effect brought by gender is not statistically significant (F (1,131) =.064, p=.800), thus H4 is not supported, suggesting that female customers do not differ from male customers in this case.
Specially, in this study, a specific element, perceived customer density in physical environment was manipulated, and two photos describing different levels of crowdedness were shown to respondents in questionnaires. To sum up, 69 respondents were shown a picture of an uncrowded situation (6 customers are waiting), and 70 respondents were shown a picture of a crowded situation (12 customers are waiting), among these questionnaires, 135 are effective. To examine whether customers’ perception of density is different in these two situations, one-way ANOVA was performed. In Table 4 and Table 5, respondents’ perceptions of density in two groups are significantly different as illustrated by the statistics (F (1,130) =144.354, p<.01).
32
Table 4 One-way ANOVA
Variables Mean SD N Past experience Yes 3.245 1.206 94 No 2.211 1.094 38 Gender Female 2.963 1.206 80 Male 2.906 1.348 53 Crowd Crowded 3.199 .962 67 Not crowded 2.497 .816 65
Table 5 One-way ANOVA
Variables SS DF MS F Sig.
Past experience
Between groups 28.941 1 28.941 20.938 .000
Within groups 179.688 130 1.382
33 Gender Between groups .103 1 .103 .064 .800 Within groups 209.416 131 1.599 Total 209.519 132 Crowd Between groups 16.239 1 16.239 20.334 .000 Within groups 103.818 130 .799 Total 120.057 131
As mentioned, correlation analysis was performed to pre-test whether there is a relationship among variables. Correlation analysis is widely applied to show the strength of statistical relationships between statistical variables (Zvizdojevic and Vukotic, 2015). Before regression analysis is run, correlation analysis is necessary in that correlation analysis can build the linear relationship between pairs of variables, which is usually the starting point of regression analysis (Zvizdojevic and Vukotic, 2015). The lower the correlation coefficient is, the more discrepancies exist. In Table 6, the Pearson coefficient between age and willingness to wait is .207 (p<.05), which means the relation is relatively weak and positive (Zvizdojevic and Vukotic, 2015). For physical environment and willingness to wait, the Pearson coefficients are .244 (p<.01) for ambient elements and .211 (p<.01) for design elements, which also indicate a positive relation between physical environment and willingness to wait. In terms of wait expectation, it is also positively related to willingness to wait with a Pearson coefficient .220 (p<.01). In addition, the relationship between gender and willingness to wait
34
is still not statistically significant, while past experience is negatively related to customers’ willingness to wait with a Pearson coefficient -.372 (p<.001). The results of correlation analysis also show a significant positive relationship between perceived customer density and willingness to wait with a high Pearson coefficient (.376, p<.01) (Zvizdojevic and Vukotic, 2015).
Table 6 Correlation analysis
Variables Mean SD 1. 2. 3. 4. 5. 6. 7. 8. 1. Age 26.439 6.006 1 2.Wait expectation 3.769 .816 .116 1 3.Ambient elements 3.959 1.042 .157 .459** 1 4.Design elements 3.856 1.059 .154 .413** .818** 1 5.Perceived customer density 2.854 .957 -.152 .096 .115 .140 1 6.Willingness to wait 2.940 1.260 .207* .220** .244** .221* .376** 1 7.Gender 1.593 .493 -.123 -.139 -.032 -.059 .022 .022 1 8.Past experience 1.291 .456 .037 -.304** -.204* -.168 .001 -.372** -.100 1
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
35
Then, to examine the predictive validity of each variable and test the hypotheses, the relationship between each dependent variable (age, the attractiveness of ambient elements, the attractiveness of design elements, past experience, wait expectation and perceived customer density) and willingness to wait was examined separately in this section, and gender was excluded since the relationship between gender and willingness to wait was found not statistically significant in correlation analysis. Different from correlation analysis, regression analysis was used for examining a dependency relationship of parameters (Farrar and Glauber, 1967), when there is one predictor variable, it is called simple regression, but when there are several predictors, it is called multiple regression (Field, 2014). In this study, I used simple regression before proceeding to using multiple regression analysis.
H1 and H2 predict that attractive ambient elements and design elements are associated with strong willingness to wait. The relationship between the attractiveness of ambient elements and willingness to wait is statistically significant (F (1, 131) = 8.26; p < .01) and explains 5.9 % of variance in customers’ willingness to wait. As for the attractiveness of design elements, it explains 4.9% variance in willingness to wait (F (1, 130) = 6.69; p < .05). Therefore, both H1 and H2 are supported, with person-organization fit for ambient elements and design elements recording Beta value .24 (p < .01) and .22 (p < .05) respectively. Then, younger people are predicted to be more willing to wait than old people, and customers with positive wait expectations would be more willing to wait. As the statistics show, age and wait expectation both have statistically significant relationships with willingness to wait separately (age: F (1,126) =5.635, p<.05; wait expectation: F (1,116) =5.905, p<.05), which means that age can explain 4.3 percent of variance in customers’ willingness to wait while wait expectation can explain 4.8 percent. However, the statistical result of age goes opposite to H5. The statistics show that older people tend to be more willing to wait than young people. Consequently, as can be seen in Table 7, the statistics indicate that H7 is supported, while H5
36
is not supported. In terms of past experience, it shows a negative statistically significant relation with willingness to wait, which is contrary to the prediction. As can be seen in Table 7, the variable past experience explains 37.2 percent (F (1,130) =20.938, p<.01) of the variance in customers’ willingness to wait, yet the negative relationship means that customers without past experience tend to be more willing to wait in front of the restaurant in this situation. Therefore, the results suggest that H6 is not supported. Perceived customer density can also be a good predictor of willingness to wait, which explains a substantial percentage of the variation (37.6 percent, F (1,130) =21.430, p<.01). As a result, H3 is supported.
Table 7 Simple regression model
Variables R R2 R2 Change B SE β t (Constant) 1.773 .420 4.224** Ambient elements .244 .059 .059 .295 .103 .244 2.875** (Constant) 1.916 .407 4.704** Design elements .221 .049 .049 .263 .102 .221 2.586* (Constant) 1.794 .498 3.605** Age .207 .043 .043 .044 .018 .207 2.374* (Constant) 1.662 .539 3.086** Wait expectation .220 .048 .048 .339 .140 .220 2.430* (Constant) 2.849 .375 7.597** Gender .022 .000 .000 .057 .224 .022 .254
37 (Constant) 4.279 .309 13.868** Past experience .372 .139 .139 -1.034 .226 -.372 -4.576** (Constant) 1.549 .320 4.841** Perceived customer density .924 .854 .854 1.090 .040 .924 27.583**
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
After the separate testing of each hypothesis, the multiple regression analysis was performed since multiple regression analysis can examine the influence of several predictors simultaneously (Field, 2014). In this full regression analysis, the attractiveness of ambient elements, the attractiveness of design elements, age, gender, past experience, perceived customer density and wait expectation serve as independent variables with willingness to wait a dependent variable. Table 8 shows information on the results of multiple regression analysis. In this multiple regression analysis, all independent variables explain 62.1 percent of the variance of customers’ willingness to wait (R2= .621). As the table shows, the impact of age on willingness to wait is still significant. This result is in line with that in simple linear regression analysis, yet contrary to the hypothesis. Therefore, H5 is still not supported. Past experience is negatively related to customers’ willingness to wait, as predicted in hypothesis, suggesting that H6 is not supported in either simple linear regression analysis or multiple regression analysis. Similarly, the multiple regression analysis shows a positive relationship between perceived customer density and willingness to wait, thus H3 is supported. However, some of the results differ from those results got in simple linear regression analyses, namely, the results with regard to H1, H2 and H7 are all mixed, since the statistics got in simple linear
38
regression analysis are significant, while in multiple regression analysis, these results are not significant. For example, in simple linear regression analysis, the variable ambient elements explains 5.9 percent of the variance in willingness to wait, while the attractiveness of design elements explains only 4.9 percent (ambient elements: F (1, 131) = 8.26; p < .01; design elements: F (1, 130) = 6.69; p < .05). Although even the combination of them represents a really small percentage of the variance in customers’ willingness to wait, H1 and H2 are still supported. In this multiple regression analysis, the relationship between either physical environment (ambient elements and design elements) or wait expectation and willingness to wait confirmed in simple linear regression analysis is not supported in multiple regression analysis (ambient elements: p=.703; design elements: p=.965; wait expectation: .908). Consistent with previous results, the relationship between gender and willingness to wait is still not supported (p=.923), so H4 is fully rejected.
Table 8 Multiple regression analysis
R R2 R2 Change B SE β t Model .621 .386 .345 .901 (Constant) .872 .968 Ambient elements .063 .164 .047 .382 Design elements -.007 .152 -.005 -.045 Age .071 .018 .319 4.014**
39 Wait expectation -.016 .141 -.010 -.116 Gender -.020 .206 -.008 -.097 Past experience -1.139 .235 -.406 -4.851** Perceived customer density .530 .109 .386 4.870**
*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).
9. Discussion
The scope of this research is fourfold: firstly, although early studies mostly focus on the influence of specific elements in physical environment such as music, light on customers’ attitude towards stores (Grewal et al., 2003) or customers’ complicated moods and feelings caused by the unpredictable waiting time (Grewal et al., 2003), this study put more emphasis on the overall influence of physical environment, which is said to be more influential than specific elements in physical environment (Grewal et al., 2003). In this study, the overall attractiveness of physical environment was divided into two parts: ambient elements and design elements, both of which are suggested to implement a big influence on customers’ behavioral intentions. Compared with early literature which mainly measures customers’ subjective estimate of waiting time, this study used a relatively new variable, willingness to wait, to evaluate customers’ attitudes toward possible waiting time. Additionally, this study empirically brought wait expectation into the research on customers’ waiting behaviors, given
40
that in service encounters, expectations established by past experience or WOM can lead to a big difference. Thirdly, since service-providing process combines production and delivery simultaneously, personal differences were taken into account while customers’ willingness to wait was measured. Finally, in terms of the specific situation of the sample restaurant (open-kitchen mode), a specific element (perceived customer density) was manipulated in this study.
9.1 The attractiveness of physical environment and willingness to wait
As suggested, physical elements may have more immediate effects on customers’ decision making process than any other marketing efforts such as promotion activities or advertising (Baker, Grewal and Parasuraman, 1994). This study also proved the functions of physical evidence in intangible service encounters. In this study, specifically, physical environment was divided in two variables, ambient elements including colors and lighting and design elements including the organization of facilities. Then I examined the relationship between these two variables and willingness to wait separately and found that the relationship between physical environment and willingness to wait is mixed. In simple linear regression analysis, as previous literature proved, physical environment can serve as cues in customers’ decision making processes: the results regarding either ambient elements or design elements indicate significant relationships. From this perspective, a gorgeous physical environment such as fashionable color theme, proper background music and organized structure does make people more willing to wait for the services. The results in the simple linear regression analysis are in line with the findings of Hightower, Brady and Baker (2002), they find that service scape has a strong impact on customers’ behavioral intentions including perception of service quality or value perception, but this impact can be mediated by a number of constructs. For example, Baker and Cameron (1996) indicate that affect may mediate the relationship between the physical environment and customers’ time perception, and physical environment
41
is always composed of a majority of elements that can be impacted by customers’ feelings and time perceptions. Bitner (1992) also finds that the relationship between physical environment and customers’ behavioral intention is mediated by either cognitive or affective factors. However, in multiple regression analysis, the relationship between either ambient elements or design elements and willingness to wait is not significant, leading to a mixed relationship between physical environment and willingness to wait. This insignificant relationship possibly can be linked to the interaction effect of other variables, namely past experience, age and perceived customer density, as Hightower, Brady and Baker (2002) point out, the relationship between service-scape and other variables tend to be complex to some degree. Antonides, Verhoef and van Aalst (2002) examine the effect of music in their study and find that no evidence is found that music can reduce the negative effect of waiting and create duration neglect. They conclude that attractive physical surroundings can improve customers’ evaluation on waiting, while customers are still not willing to wait when they realize that they will still pay for waiting. This opinion can partly explain the insignificant relationship between physical environment and customers’ willingness to wait. While customers’ personal factors such as age, past experience and other significant elements in physical environment are taken into account, the influence of overall physical environment will decline. In other words, the influence of physical environment depends on whether there are other factors that are weighed more in customers’ minds.
9.2 Personal factors and willingness to wait
Bennett (1998) indicates that customers with some characteristics are more likely to be averse to waiting, for example, he mentions that affluent customers tend to be less tolerant of unanticipated waiting than poor people. Therefore, personal factors can be an effective tool to
42
categorize customers into various groups. In this study, mainly three personal factors, gender, age and past experience were examined.
Normally, gender is regarded as an important segmentation variable in marketing issues, due to the fact that it is identifiable and accessible (Yelkur and Chakrabarty, 2006). There are indeed some gender differences in consumers’ ratings of service performance (Snipes, Thomson and Oswald, 2006). In this study,the results in either simple regression analysis or multiple regression analysis are consistent, resulting in an insignificant relationship between gender and willingness to wait. In this study, the difference between female group and male group was not confirmed, which is contrary to extant researches. When Grewal et al. (2003) examinethe difference brought by gender on evaluations on store atmosphere, they find that since men are achievement-oriented, they always have lower evaluations on store atmosphere than women. Sharma, Chen and Luk (2012) also point out that time, money or other values that customers care about weigh more for men than for women, thus men appear to be more value-conscious than women. However, this view cannot be confirmed by this study, in which the willingness to wait of either male customers or female customers is not significantly different. This insignificant difference between two groups probably can be attributed to the utilitarian categories. Sharma, Chen and Luk (2012) shed light on the influence that utilitarian category or hedonic category can implement. They find that in the utilitarian category, female customers’ behavioral intentions are no stronger than male customers’. From this perspective, both female and male customers may regard this restaurant as a place where they can seek good food instead of a signal of social identity.
With regard to age, its relationship with willingness to wait was found significant in either simple regression analysis or multiple regression analysis, while this relationship still differs from the hypothesis. The results show that with age, customers become more patient and
43
willing to wait for the services, although the sample size is not equal (77.6 percent of the respondents are 21-30 years old). Generally, extant literature holds the opinion that older customers are less likely to wait since they have numerous experience on which they can rely to make right decisions (Sharma, Chen and Luk, 2012). In contrast, young people can be hastier and less value-conscious. For example, considering the value of time to the young and the old, the elderly people’s perception of waiting time is generally poorer than the young people (Wu, Lu and Ge, 2013), this could be the reason why the older are more willing to wait than the young.
In terms of past experience, contrary to the prediction, customers without past experience tend to be more willing to wait than customers with past experience. This result is inconsistent with the opinion of Jones and Peppiatt (1996) who indicate that new or infrequent customers tend to feel a longer wait than frequent visitors. This is because as customers get more experienced and more familiar with this service encounter, their wait expectations can reflect reality more accurately (Rodríguez Molina, Frías-Jamilena and Castañeda-García, 2013). However, the inference theory may explain this negative relationship between past experience and willingness to wait. Customers without past experience are more likely to choose the restaurant that requires waits although they have less available information to make an inference on the quality of services or products than frequent customers, in that keeping consistent with the crowd can help them minimize the risks (Grewal et al., 2003) and givethem a sense of security. For first-time users or customers without past experiences, they may have no access to additional information about the quality of services or products, or rather, they may regard the wait as a cue of good quality (Giebelhausen, Robinson and Cronin, 2010), while for frequent customers who can use past experience as a reference, the presence of wait cannot be a signal of good quality. Similar opinions are held by other academics. Becherer and Richard (1978), Cacioppo, Petty, Feng
44
Kao (1984) suggest that customers who can get other information will be less influenced by social cues while evaluating product quality, therefore, product familiarity may need to be researched more (Giebelhausen, Robinson and Cronin, 2010).
9.3 Wait expectation and willingness to wait
Prior expectations are usually used as a reference by customers while they make some evaluations on services (Choi and Mattila, 2008). Expectations exist as standards with which customers’ experiences can be compared (Zeithaml, Berry and Parasuraman, 1993) .In previous literature, wait expectations are proven to be good predictors of customers’ behavioral intentions. For instance, Grewal et al. (2003) suggest that when customers find nothing positive about waiting, they may cancel their decided patronage plan. In this study, similar to previous studies, wait expectation is also predicted positively related to customers’ willingness to wait, indicating that when customers have positive wait expectations before purchasing, they are more willing to wait than customers with negative wait expectations. However, this prediction is just confirmed in simple regression analysis, although it can only explain 4.8 percent of the variance in customers’ willingness to wait, in multiple regression analysis, that confirmation is not supported, which makes the relationship between wait expectation and customers’ willingness to wait complex. The insignificant relationship between wait expectations and customers’ willingness to wait is possibly due to the interaction effect brought by the influence of other factors. That means, while both physical environment and personal factors are taken into account, wait expectations do not matter a lot for customers. Devlin, Gwynne and Ennew (2002) also suggest that the influence of expectations can be influenced by context specific factors, especially when customers face the difference levels of crowdedness, wait expectations may not weigh as much as when all
