Research in Hospitality Management

ISSN: 2224-3534 (Print) 2415-5152 (Online) Journal homepage: www.tandfonline.com/journals/rrhm20

Consumer patronage and willingness-to-pay at

different levels of restaurant attributes: A study
from Kenya

David Njite, Joyce Njoroge, Haragopal Parsa, Rahul Parsa & Jean-Pierre van
der Rest

To cite this article: David Njite, Joyce Njoroge, Haragopal Parsa, Rahul Parsa & Jean-Pierre van
der Rest (2015) Consumer patronage and willingness-to-pay at different levels of restaurant
attributes: A study from Kenya, Research in Hospitality Management, 5:2, 171-180, DOI:
10.1080/22243534.2015.11828342

To link to this article: https://doi.org/10.1080/22243534.2015.11828342

Published online: 18 Jan 2017.

Submit your article to this journal

Article views: 918

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rrhm20
Research in Hospitality Management 2015, 5(2): 171–180 Copyright © NISC (Pty) Ltd
Printed in The Netherlands — All rights reserved
RHMISSN 2224-3534
http://dx.doi.org/10.2989/RHM.2015.5.2.8.1281

Consumer patronage and willingness-to-pay at different levels of restaurant

attributes: A study from Kenya

David Njite1, Joyce Njoroge2, Haragopal Parsa3, Rahul Parsa4 and Jean-Pierre van der Rest5*

1
The Lester E. Kabacoff School of Hotel, Restaurant and Tourism Administration, University of New Orleans, New Orleans,
Louisiana, USA
2
School of Accounting, Drake University, Des Moines, Iowa, USA
3
Fritz Knoebel School of Hospitality Management, University of Denver, Denver, Colorado, USA
4
Department of Supply Chain and Information Systems, College of Business, Iowa State University, Ames, Iowa, USA.
5
Department of Business Studies, Institute of Tax Law and Economics, Leiden Law School, Leiden University, Leiden, The Netherlands
*Corresponding author email: [email protected]

This study investigates the effect of three key restaurant attributes – food quality, service quality and ambiance – on consumers’
(i) willingness-to-pay (WTP) and (ii) intentions to patronage (ITP) for limited service and upscale restaurants in Kenya. A full-profile
2 (types of restaurants) × 2 (levels of attributes) × 3 (service attributes) factorial design was developed in which type of restaurant
(upscale dining and limited service) was varied as a between-subjects factor, and the variations in the attributes and levels a
within-subject factor. Findings indicate that food quality is the only attribute to have a positive relationship with both consumer
patronage and willingness-to-pay, in high-end as well as limited service restaurants. Service quality and ambiance both have
an effect on consumer intentions in upscale dining, but not in a limited service restaurant. Moreover, an interaction effect was
found. In high-end restaurants, and not so in limited service restaurants, increase in ambience was found to increase patronage
and willingness-to-pay by a significant amount if service quality was also increased. The study extends previous work on the
nature of restaurant attributes, as it (1) examines consumer preferences in a developing country, and (2) shows how attributes
interact and affect ITP and WTP differently in a limited service and high-end restaurant context.

Keywords: attributes; consumer behavior; willingness to pay; consumer patronage; price

Introduction with the exception of Andersson (1991) and Dutta et al.

(2014), who assess willingness-to-pay (WTP). Understanding
One of the most important issues in consumer research is the WTP is of particular interest in studying consumer prerferences
determination of consumer preferences and wants so that as it is rich in individual information. The ability to measure
marketers can adapt or generate new products accordingly WTP enables managers to estimate demand and maximise
(Griffin & Hauser, 1993; Verma, Thompson & Louviere, 1999; revenue through price optimisation.
Bujisic, Hutchinson & Parsa, 2014). This implies that managers This study therefore investigates the effect of key restaurant
need insight into the profitable attributes that should be attributes – specifically food quality, service quality and
considered in making investment and strategic decisions. For ambiance – on consumers’ (1) willingness-to-pay (WTP) and
example, restaurants may be obliged to invest in attributes such (2) intentions to patronise (ITP). It focuses on restaurants in
as ambiance in order to remain competitive and profitable in the Kenya and differentiates between limited service and high-end
marketplace. If managers elect to invest in a specific attribute, it restaurants, thereby, significantly extending previous work that
is important for them to understand the long-term implication has examined consumer behavioural intentions in other parts
in relation to profitability and return on investment (ROI). of the world (e.g., Bujisic et al., 2014; Dutta et al., 2014).
There is a vast literature on consumers’ preferences for
services, including studies on restaurant attributes (e.g. Literature review
Perutkova and Parsa, 2010; Bujisic et al., 2014; Harrington,
Ottenbacher & Kendall, 2011; Choi & Zhao, 2010; Kim, Raab Utility theories such as multi-attribute theory (MAUT)
& Bergman, 2010; Njite, Dunn & Kim, 2008; Knutson, Beck and subjective expected utility theory (SEUT) suggest that
& Elsworth, 2006; Moschis, Curasi, & Bellenger, 2003), but products (goods and services) comprise of several attributes
with little attention for consumer preferences in non-western that influence consumer purchase and consumer evaluation
countries (e.g. Dutta et al., 2014; Khan & Oyewole, 2014; of the product. As Mittal, Kumar and Tsiros (1999) suggest,
Upadhyay, Singh & Sharma, 2009; Lord, Putrevu & Shi Yi, consumers are likely to evaluate their satisfaction with a
2005; Bhuian, 2000). Moreover, work focusing on the financial product at the attribute level rather than at the product level
implications of consumer preferences is virtually non-existent, as a whole.

Research in Hospitality Management is co-published by NISC (Pty) Ltd and Routledge, Taylor & Francis Group
172 Njite, Njoroge, Parsa, Parsa and van der Rest

Restaurants possess several attributes and extant literature customers not only expect to meet or exceed all service quality
reveals that food quality level, service levels and the ambiance dimensions, but also expect to receive excellent personalised
are the most important attributes of customers’ restaurant service delivery and a “VIP-like” treatment.
dining experience (e.g. Bitner, 1992; Dabholkar, Shepherd & Personalisation of service is of great strategic significance
Thorpe, 2000; Kim, Ng & Kim, 2009; DiPietro, Parsa & Gregory, and businesses invest in personalisation and information
2011; Dutta et al., 2014). Moreover, these attributes were processes and acquisition capabilities in order to develop
reported as significantly correlated with consumer behavioural and manage various customer retention strategies (Winer,
intentions ITP and WTP (DiPietro et al., 2011). 2001). Personalisation of service serves the distinct purpose
of increasing customer loyalty and also serves as a deterrent
Food quality to switching (Alba et al., 1997). Personalisation has exclusively
Namkung and Jang (2007) investigated the significance of food been employed by businesses in the luxury goods and services
quality in relation to customer satisfaction and behavioural sector where it often signals high quality with implicit price
intentions in mid- to upscale restaurants. Their findings premiums for personalised products or services (Mattila, 1999).
revealed that, overall, food quality significantly influenced Other instances where personalisation can be observed are
consumer satisfaction and behavioural intentions. A further when consumer-service provider relationships are strong
analysis of their results revealed that taste and presentation and repetitive, for example at those high-end restaurants
were the two most important contributing aspects of food where customers are greeted by their names and niceties
quality that enhanced consumer satisfaction with the food. are exchanged between the customers and service provider.
Similarly, Min and Min (2011) examined service quality and Therefore, it is reasonable to propose that personalised service
customer satisfaction in fast food restaurant franchises in the is positively related to behavioural intentions of consumers
USA. The results of their study revealed that the taste of the of high-end restaurants. Thus, the following hypotheses are
food was the most important restaurant attribute in fast food formulated:
restaurant customers’ perceptions and eventual satisfaction. • H2: The relationship between service quality and consumers’
These studies indicate that there is a positive and significant behavioural intentions is moderated by the type of restaurant
relationship between food quality and consumers’ behavioural based on the service levels offered.
intentions (DiPietro et al., 2011). That is, the higher the food • H2a: In high-end restaurants, there is a positive relationship
quality, the higher the consumer satisfaction with the product between service quality and intention to patronise (ITP).
and the more likely the consumer is expected to engage in • H2b: In high-end restaurants, there is a positive relationship
a certain positive behavioural intentions. Given the reported between service quality and consumers’ willingness-to-pay
importance of the food attribute for consumers of both limited (WTP).
service and high-end restaurants, it is formulated that: Even though some of the aforementioned studies have
• H1: The relationship between food quality and consumer showed that there is a relationship among service quality and
behavioural intentions is independent of the type of consumer intentions, the extant literature reviewed does not
restaurant. establish nor provide concrete premises to make conclusions
• H1a: In the case of both high-end and limited service about the relationship between these variables in the limited
restaurants, there is a positive relationship between food service industry. It is therefore reasonable to argue that even
quality and consumers’ intention to patronise (ITP) a though there is a relationship, its direction and strength cannot
restaurant. be easily deduced. The following hypotheses are therefore
• H1b: In case of both high-end and limited service restaurants, presented:
there is a positive relationship between food quality and • H2c: In limited service restaurants, there is a positive but
consumers’ willingness-to-pay (WTP). nonsignificant relationship between service quality and
consumers’ willingness-to-pay (WTP).
Service quality • H2d: In limited service restaurants, there is a positive but
Service quality is perhaps one of the most examined constructs nonsignificant relationship between service quality and
in the marketing literature. Kara, Kaynak and Kucukemiroglu consumers’ intention to patronise (ITP).
(1995) investigated customer perception of fast food
restaurants service quality in the United States and Canada Ambiance
using eleven attributes and concluded that service quality is Unlike the consumers of tangible goods, consumers of services
significant to consumer behavioural intention. The work of have a limited number of cues to evaluate their satisfaction.
Stevens, Knutson and Patton (1995), John and Tyas (1996), Qin Therefore, consumers utilise several peripheral cues besides
and Prybutok (2008) and Qin, Prybutok and Zhao (2010) reveal the core product in making their decisions. In many cases, the
that service quality is based on both tangible and intangible physical environment provides a tangible cue for evaluating
aspects of a product. the service product. In this way, physical environment can
Overall, the extant literature seems to indicate that there is a be perceived as an important element in evaluating and
relationship between service quality and customer satisfaction. determining consumer satisfaction with the service provided
Customers of limited service restaurants are mainly concerned (Booms & Bitner, 1982).
with the limited aspects of service quality such as convenience, Indeed, the influence of the physical environment
order accuracy and speed of service (Clark & Wood, 1998; on behaviour and image formation is very apparent in
DiPietro et al., 2011). Service level, however, is one of the most predominantly service oriented businesses such as hotels and
important factors that customers expect from an upscale dining restaurants (Baker, 1987; Booms & Bitner, 1982; Upah &
establishment (Tlapa et al., 2011). In case of upscale dining, Fulton, 1985). Due to the inseparability of services, particularly
Research in Hospitality Management 2015, 5(2): 171–180 173

in restaurants, consumers have to be present in the physical of the largest universities in Kenya located in Nairobi. This
environment to consume and experience the product. Due to group of respondents (students) was identified as a good
this simultaneity, Bitner (1992) suggests that the premises or representative of consumers for limited service restaurants thus
place of consumption cannot be hidden and plays a significant aiding in generalisability. According to the National Restaurant
role in the consumer’s overall perception of satisfaction and Association (2010) in USA, respondents between the ages of
behavioural intentions. 18 and 24 are often considered as the prime users of limited
According to Berry and Clark (1986), consumers commonly service restaurants, thus, providing support for our choice of
seek cues to a business’s capability before, during and after respondents for investigating quick service segment of the
the service encounter, and physical environment provides such restaurant industry. Respondents for limited service restaurants
cues (Rapoport, 1982). Environmental psychologists suggest were full-time students and the average age for the student
that atmosphere, interior design, lighting, noise/music levels respondents was 23 years. A majority of these student
and types, and layout are crucial dimensions of the restaurant respondents indicated that they visit limited service restaurants
ambiance that influence consumer perception, satisfaction and at least twice a week.
behaviour (e.g. Kim et al., 2009; Wall & Berry, 2007). Some
positive outcomes of a customer approach to the environment Attributes
include: affiliation, staying longer, commitment, and carrying Three restaurant attributes were investigated: food quality,
out the purpose of being at the business (Mehrabian & Russel, service quality, and ambiance (DiPietro et al., 2011). With
1974). regard to service quality, personalised service was selected
Since customers of high-end restaurants tend to spend for the purpose of this study. Personalisation refers to the
extended time in a built environment while dining out, the tailoring of products and purchase experience to the tastes of
quality (atmosphere) of the built environment will be important individual consumers based upon their personal and preference
to them. Conversely, since customers of limited service information (Alba et al., 1997). For the high-end restaurants,
restaurants focus on speed of service, order accuracy, menu attribute levels ranged from average food, average service, and
simplification and convenience, the high quality of the built average ambiance (atmosphere) on the low end, to excellent
environment will have lesser importance for their overall food, excellent service, and excellent ambiance (atmosphere)
satisfaction. Thus, the following hypotheses are formulated: on the high end. The scenarios for limited service restaurants
• H3: The type of restaurant moderates the relationship required patrons to make choices ranging from low food
between ambiance and consumer behavioural intentions. quality, non-personalised service, and less appealing place on
• H3a: In high-end restaurants, there is a positive relationship the lower end of criteria, to good food, personalised service,
between ambiance and consumers’ intention to patronise and appealing place on the high end of the criteria.
(ITP).
• H3b: In high-end restaurants, there is a positive relationship Design
between ambiance and consumers’ willingness-to-pay The participants were presented with several scenarios and
(WTP). were asked to consider and make a choice among the eight
• H3c: In limited service restaurants, there is a positive restaurant options with two levels (e.g. excellent and low)
but nonsignificant relationship between ambiance and of three service attributes (food quality, service quality, and
consumers’ willingness-to-pay (WTP). ambiance). It was a full profile factorial design with 2 (types
• H3d: In limited service restaurants, there is a positive of restaurants) × 2 (levels of attributes) × 3 (service attributes).
but nonsignificant relationship between ambiance and Thus, all possible combinations of attributes were presented
consumers’ intention to patronise (ITP). as different restaurant options to the respondents. Each
restaurant option featured either a “high” or “low” service
Methods level, “high” or “low” atmosphere or “high” or “low” food
quality.
Participants The type of restaurant (limited service/high-end) was varied as
The participants analysed in this study consisted of a total a between-subjects factor. Nested in a type, every respondent
of 294 respondents from the city of Nairobi, Kenya, with judged either a limited service or high-end restaurant but not
141 responding to high-end restaurant scenarios and 153 both. The variations in the type of restaurants comprised a
responding to limited service restaurants. The sample consisted within-subject factor. Due to the distinct differences and
of 61% female and 32% males, ranging in age from 19 to characteristics of these two restaurant types, it was felt that it
72. Due to the nature of restaurant types (upscale restaurants would be conceptually easier and more possible for consumers/
and limited service restaurants), two distinctive groups of respondents to solely identify a single restaurant type, and,
participants were mandated. The first consisted of target therefore, easier for the manipulation and communication.
people that had visited and paid for their meals at a high-end
restaurant within the past month. Respondents that did not Independent and dependent variables
meet this eligibility were excluded from the study. For high-end
restaurant scenarios, the respondents were selected randomly The three restaurant attributes, food quality, service quality
by an intercept method at a major retail shopping centre by and ambiance, were manipulated as the independent variables.
trained researchers. By choosing every third retail customer, Criticality of the restaurant encounter was applied to ensure
we were able to ensure randomisation and improve chances that respondents focused and identified the restaurant type
of generalisability to the greater population. The second and the context. This criticality was manipulated as a between-
group consisted of students at a large branch campus of one subject factor. For each restaurant type, the respondents read
174 Njite, Njoroge, Parsa, Parsa and van der Rest

a short scenario describing the conditions under which each Results and findings
restaurant type was being selected. The scenarios described
conditions that were either high or low in terms of criticality, Hypotheses H 1, H 1a and H 1b stated that the relationship
thus, describing greater or lesser outcomes visiting these between food quality and intent to patronise (ITP) as well as
restaurants. willingness-to-pay (WTP) is positive and independent of type
Two dependent variables, indicating the consumer of restaurant. The first observation was that of the overall
behavioural intentions intent to patronise (ITP) and willingness- F-test (overall for ITP and WTP), with a main interest in the
to-pay (WTP), were examined. ITP and WTP were examined Wilks’ lambda (λ) statistic and the F-value associated with it
in both limited service and high-end restaurants. Intention (See Table 1). The one-way MANOVA revealed a significant
to patronise (ITP) was conceptualised as the future action of multivariate main effect for food quality, Wilks’ λ = 0.853, F
the consumer to purchase the product, visit the restaurant or (2, 289) = 24.863, p < 0.001, ηp2 = 0.147 (power = 1.000),
recommend the restaurant to friends, relatives and colleagues. showing a statistically large (ηp2 > 0.14, Cohen, Cohen, West
It was measured on a seven-point Likert scale by asking & Aiken, 2003) main effect of food quality on the dependent
participants the likelihood of selecting a restaurant for their variables ITP and WTP together.
visit. WTP was examined with Kenyan Shillings (KSh.) currency Having obtained a significant multivariate main effect for all
using a price range that was determined appropriate from the factors, a univariate F-test (examining between-subjects
a pretest and multiple discussions held with local restaurant effects) was conducted on each of the dependent variables to
owners. It was defined as the reservation price, the maximum establish if food quality and restaurant type had a significant
price a given consumer willing to pay for a product or service influence on ITP and WTP separately. At this point, since we
(Davenport, 1902; Jennings & Jennings, 2000). The high and were conducting four tests, the experiment-wise alpha rate of
low end of price points for both upscale restaurant and quick 0.05 was divided by four to obtain an acceptable confidence
service restaurants were determined in the pre-test stage with level for each of the four tests, so we set the alpha level to
over 64 students. p < 0.0125. By that criterion, again, all results obtained for
Participants for the high-end restaurants were asked to the univariate tests were significant for the effect of food
choose amounts they were willing to pay while dining out, quality on ITP (F (2, 57) = 24.86, p < 0.0125) and WTP (F (2,
ranging from KSh. 1 500 (US$18.75) to KSh. 2 000 (US$25), 57) = 34.13, p < 0.000).
with an average of about Ksh $1 750 (US$ 22) at the high-end, MANOVA results were analysed for any interaction. The
and KSh. 500 (US$6.25) to KSh.700 (US$8.75), with an average results revealed that the interaction between restaurant type
of about Ksh 600 (US$ 7.50) at the low end (Appendix A). and food quality was non-significant (Wilks’ λ = 0.898, F (2,
Limited service restaurant participants were given a different 289) = 0.366, p = 0.588, ηp2 = 0.102 (power = 1.000)), showing
range of price amounts and they were asked to indicate that the interaction effect of restaurant level and food quality
how much they would be likely to pay for eating at low-end on the behavioural intentions ITP and WTP was non-significant.
restaurants. There were several possible ranges of amounts, Thus, restaurant type did not moderate the influence of food on
with the lowest being KSh.70 (US $0.85) to KSh.100 (US $1.25) consumer behavioural intention (Table 1). Therefore, hypothesis
with an average of about Ksh 85 (US$ 1.05), and the highest H1 was supported as there was no statistically significant
range being KSh.150 (US$1.90) to KSh. 200 (US$. 2.50) with moderating influence of restaurant type (limited service or
an average of about Ksh 175 (US$2.20). The exchange rate at high-end) on the relationship between food quality and the
the time of data collection was about US$1 = KSh. 80. intent to patronise (ITP) and willingness-to-pay (WTP).

Table 1: Multivariate tests for food quality, restaurant types and dependent variables

Effect Value F Hypothesis df Error df Sig. Partial eta squared Observed power
Intercept
Pillai’s trace 0.913 1512.916 2.000 289.000 0.000 0.913 1.000
Wilks’ lambda 0.087 1512.916 2.000 289.000 0.000 0.913 1.000
Hotelling-Lawley trace 10.470 1512.916 2.000 289.000 0.000 0.913 1.000
Roy’s largest root 10.470 1512.916 2.000 289.000 0.000 0.913 1.000
Restaurant level
Pillai’s trace 0.849 814.076 2.000 289.000 0.000 0.849 1.000
Wilks’ lambda 0.151 814.076 2.000 289.000 0.000 0.849 1.000
Hotelling-Lawley trace 5.634 814.076 2.000 289.000 0.000 0.849 1.000
Roy’s largest root 5.634 814.076 2.000 289.000 0.000 0.849 1.000
Food quality
Pillai’s trace 0.147 24.863 2.000 289.000 0.000 0.147 1.000
Wilks’ lambda 0.853 24.863 2.000 289.000 0.000 0.147 1.000
Hotelling-Lawley trace 0.172 24.863 2.000 289.000 0.000 0.147 1.000
Roy’s largest root 0.172 24.863 2.000 289.000 0.000 0.147 1.000
Restaurant level * Food quality
Pillai’s trace 0.102 0.386 2.000 289.000 0.601 0.102 1.000
Wilks’ lambda 0.898 0.366 2.000 289.000 0.588 0.102 1.000
Hotelling-Lawley trace 0.113 0.371 2.000 289.000 0.590 0.102 1.000
Roy’s largest root 0.113 0.386 2.000 289.000 0.651 0.102 1.000
 Research in Hospitality Management 2015, 5(2): 171–180 175

A t-test was conducted to examine the difference in the (limited service and high-end) moderated the influence of
consumer willingness to pay when food was varied from low service quality on consumer behavioural intention (Table 3).
to high. The results of this t-test revealed that in high-end Thus, hypothesis H2 is supported as restaurant type significantly
restaurants when food quality was low, the consumer moderates the relationship between service quality and the
willingness to pay was (x̅ WTP = 1400) and when food quality was intent to patronise (ITP) and willingness-to-pay (WTP).
high, it was (x̅ WTP = 1650), representing a 17.9% change. Also, To further test this relationship (H2a, H2b, H2c and H2d), the
in the case of limited service restaurants, when food quality means of the respondents were examined (Table 4 and Figures
was low, the willingness to pay was ( = 130) and when food 3 and 4). The posited H2a/b states that in high-end restaurants
quality was high (x̅ WTP = 155), a 19.2% change. There was no there is a positive and significant relationship between service
statistical difference (t (139) = 1.23, p > 0.05) in the WTP when quality and consumer intentions. That means, when the
food quality increased in the limited service restaurants. levels of service are raised (e.g. personalisation of service and
Therefore, analysing the means of consumer behavioural attention to detail) in high-end restaurants from low to high,
intention in both restaurants types when food quality was the intent to patronise (ITP) and willingness-to-pay (WTP) also
varied from low to high (Table 2, and Figures 1 and 2), revealed increase. A post priori t-test was conducted to examine the
a positive relationship between food quality and consumer difference in the consumer willingness to pay when service
intentions but it was not significantly moderated by the type was varied from low to high. The results revealed that in
of restaurant. Thus, support was found for hypotheses H1a and the high-end restaurant when service quality was low, the
H1b. consumer willingness to pay was (x̅ WTP = 1 028) and when
Hypothesis H2 stated that the type of restaurant moderates service quality was raised to high then it was (x̅ WTP = 1 515)
the relationship between service quality and intent to patronise with a t-value of t (140) = 4.43, p < 0.001. Thus, there was a
(ITP) as well as willingness-to-pay (WTP). In high-end restaurants statistically significant difference in the consumer WTP when
this relationship was posited as being significantly positive (H2a service was varied from low to high value (47% change).
and H2b), whilst for limited service restaurants it was considered Similarly, a t-test was conducted to examine the influence
positive but nonsignificant (H2c and H2d). One-way MANOVA of varying service on consumer’s intention to patronise the
analysis revealed a significant multivariate main effect for high-end restaurants. The results of the t-test revealed that
service quality, Wilks’ λ = 0.648, F (2, 289) = 78.158, p <. when the level of service personalisation improved the ITP
001, ηp2 = 0.352 (power = 1.000), showing a large main effect also increased (x̅ ITP = 3.44) to (x̅ ITP = 5.87), a 70% change, with
of service quality on ITP as well as WTP (see Table 3). Other a t-value of t (141) = 4.6, p < 0.000. Thus, H2a and H2b are
associated statistics (Table 3) further supported the main effect supported since there is a positive and significant relationship
for quality of service on behavioural intention. between service quality and intent to patronise and willingness-
Further examining the MANOVA table it was found that the to-pay in high-end restaurants.
interaction between restaurant type and service quality was For limited service restaurants (H2c/d) a positive but nonsignificant
significant, Wilks’ λ = 0.767, F (2, 289) = 43.996, p < 0.001, relationship was posited between service quality and consumer ITP
ηp2 = 0.233 (power = 1.000), showing that restaurant type and WTP. In limited service restaurants, the results revealed that

Table 2: Estimated marginal means for willingness-to-pay (WTP) and intention to patronise (ITP) – Food quality

Restaurant type Food quality level Intention to patronise (mean) % change Willingness-to-pay (mean) % change
Limited service Low food quality 4.43 33.4 130 19.2
High food quality 5.91 155
High-end Low food quality 4.03 31.0 1400 17.9
High food quality 5.28 1650
ITP Estimated Marginal Means

6 1400
WTP Estimated Means

1200
5
1000
4
800
3
600 Willingness to pay limited service restaurant
2 Intention to patronage limited service restaurants Willingness to pay high end restaurant
400
Intention to patronage high end restaurants
1 200

Low Food Quality High Food Quality Low Food Quality High Food Quality
Food Quality Food Quality

Figure 1: Estimated Marginal Means for ITP for Food Quality Figure 2: Estimated marginal means for WTP for food quality
 176 Njite, Njoroge, Parsa, Parsa and van der Rest

when service quality was low, the mean value for willingness to with a t-value t (141) = 3.22, p < 0.05. This indicates that
pay was (x̅ WTP = 151) and when service was high quality the WTP improvement in service does increase consumers’ intention to
was (x̅ WTP = 180). The t-test revealed (t (152) = 1.72, p < 0.05) patronise. Thus, the hypothesised nonsignificant relationship,
that there was no statistically significant difference in the WTP if H2d, between ITP and service quality in limited service
service quality was increased from low to high in limited service restaurants was not supported.
restaurants, thus hypothesis H2C was supported. Hypothesis H3 described the relationship between ambiance
A t-test was further conducted to examine the influence and consumer intention, moderated by type of restaurant. The
of varying service on consumer’s intention to patronise (ITP) one-way MANOVA analysis revealed a significant main effect
limited service restaurants. The results of the t-test revealed for ambiance (Table 5) showing a large main effect of ambiance
that when the level of service personalisation increased the on ITP and WTP together. Other statistics observed were Pillai’s
ITP increased from (x̅ ITP = 4.11) to (x̅ ITP = 5.52), a 34% change, trace = 0.374, F(2, 289) = 14.896, p <. 001, ηp2 = 0.154, and

Table 3: Multivariate tests for service quality, restaurant types and dependent variables

Effect Value F Hypothesis df Error df Sig. Partial Eta Squared Observed Power
Intercept
Pillai’s trace 0.939 2208.051 2.000 289.000 0.000 0.939 1.000
Wilks’ lambda 0.061 2208.051 2.000 289.000 0.000 0.939 1.000
Hotelling-Lawley trace 15.281 2208.051 2.000 289.000 0.000 0.939 1.000
Roy’s largest root 15.281 2208.051 2.000 289.000 0.000 0.939 1.000
Restaurant levels
Pillai’s trace 0.867 940.883 2.000 289.000 0.000 0.867 1.000
Wilks’ lambda 0.133 940.883 2.000 289.000 0.000 0.867 1.000
Hotelling-Lawley trace 6.511 940.883 2.000 289.000 0.000 0.867 1.000
Roy’s largest root 6.511 940.883 2.000 289.000 0.000 0.867 1.000
Service levels
Pillai’s trace 0.352 78.518 2.000 289.000 0.000 0.352 1.000
Wilks’ lambda 0.648 78.518 2.000 289.000 0.000 0.352 1.000
Hotelling-Lawley trace 0.543 78.518 2.000 289.000 0.000 0.352 1.000
Roy’s largest root 0.543 78.518 2.000 289.000 0.000 0.352 1.000
Restaurant level * Service level
Pillai’s trace 0.233 43.996 2.000 289.000 0.000 0.233 1.000
Wilks’ lambda 0.767 43.996 2.000 289.000 0.000 0.233 1.000
Hotelling-Lawley trace 0.304 43.996 2.000 289.000 0.000 0.233 1.000
Roy’s largest root 0.304 43.996 2.000 289.000 0.000 0.233 1.000

Table 4: Estimated marginal means for willingness-to-pay (WTP) and intention to patronise (ITP) – Service quality

Restaurant type Service quality level Intention to patronise (mean) % change Willingness-to-pay (mean) % change
Limited service Low service quality 4.11 34.3 151 19.2
High service quality 5.52 180
High-end Low service quality 3.44 70.6 1028 47.4
High service quality 5.87 1515
ITP Estimated Marginal Means

6 1400
WTP Estimated Means

1200
5
1000
4
800
Willingness to pay limited service restaurant
3 Willingness to pay high end restaurant
600
2 Intention to patronage limited service restaurants 400
Intention to patronage high end restaurants
1 200

Low Service Quality High Service Quality Low Service Quality High Service Quality
Service Quality Service Quality

Figure 3: Estimated marginal means for ITP for service quality Figure 4: Estimated marginal means for WTP for service quality
Research in Hospitality Management 2015, 5(2): 171–180 177

Hotelling-Lawley trace = 0.172, F(2, 289) = 14.896, p < 0.001, from (x̅ ITP = 3.97) to (x̅ ITP = 5.36) with a t-value of t (141) = 4.31,
ηp2 = 0.154, providing further support for hypothesis H3. p < 0.001, representing a 35% change (Figure 6). Thus,
The results also revealed that the interaction between hypotheses H3a and H3b were supported.
restaurant type and ambiance was significant (Table 6), Hypothesis H3c/d stated that in limited service restaurants
indicating that the restaurant type moderated the influence of there is a positive but nonsignificant relationship between
ambiance on intent to patronise and willingness-to-pay (Table ambiance and willingness-to-pay (WTP) and intent to patronise
6). These findings provide support for hypothesis H3, indicating (ITP). In limited service restaurants, when ambience quality
that there is a statistically significant moderating influence of was low, the willingness to pay was (x̅ WTP = 138) and when
type of restaurant on the relationship between ambiance and ambience was high the WTP was (x̅ WTP = 145) with a t-value of
consumers’ intention. t (151) = 1.72, p = 0.057. These findings showed no statistically
Hypotheses H3a/b stated that in high-end restaurants a significant difference in the WTP if ambience improved from
positive and significant relationship exists between ambiance low to high in limited service restaurants. Thus, hypothesis H3C
and consumer intentions. Thus, when the ambiance levels of was supported.
the built environment are varied from low to high, ITP and WTP A t-test was further conducted to examine the influence
are expected to increase. The t-test was conducted to examine of varying ambience on consumers’ intention to patronise
the difference in the consumer willingness to pay when (ITP) limited service restaurants. The results of the t-test
ambiance was varied from low to high. The results of this t-test revealed that when the level of ambience increased from low
revealed that in high-end restaurants when ambience was to high the ITP increased from (x̅ ITP = 4.31) to (x̅ ITP = 5.34) with
low, the consumer willingness to pay was (x̅ WTP = 1 180) and a t-value of t (141) = 2.22, p < 0.05, representing a 24%
when ambience was improved to high quality then the WTP change. These results revealed that increase in ambience
was (x̅ WTP = 1 522) with a t-value of t (141) = 6.22, p < 0.000. significantly increases the willingness to patronise, thus the
Thus, a statistically significant difference in consumer WTP hypothesis H3d was not supported. This clearly indicates that
was observed when ambience was improved from low to high even in limited service restaurants consumers care for quality
(29% change). in ambiance and the relationship is positive similar to the
Similarly, a t-test was conducted to examine the influence high-end restaurants. This observation can be supported by
of varying ambience on the consumer’s intention to patronise the fact that most limited service restaurants often tend to
high-end restaurants. The results of the t-test revealed that remodel their restaurants periodically meeting the changing
when the level of ambience increased the ITP has increased needs of the customer.

Table 5: Multivariate tests for ambiance, restaurant types and dependent variables

Effect Value F Hypothesis df Error df Sig. Partial Eta Squared Observed Power
Intercept
Pillai’s trace 0.913 1524.565 2.000 289.000 0.000 0.913 1.000
Wilks’ lambda 0.087 1524.565 2.000 289.000 0.000 0.913 1.000
Hotelling-Lawley trace 10.551 1524.565 2.000 289.000 0.000 0.913 1.000
Roy’s largest root 10.551 1524.565 2.000 289.000 0.000 0.913 1.000
Restaurant type/level
Pillai’s trace 0.848 808.432 2.000 289.000 0.000 0.468 1.000
Wilks’ lambda 0.152 808.432 2.000 289.000 0.000 0.468 1.000
Hotelling-Lawley trace 5.595 808.43 2.000 289.000 0.000 0.468 1.000
Roy’s largest root 5.595 808.432 2.000 289.000 0.000 0.468 1.000
Ambiance level
Pillai’s trace 0.374 14.896 2.000 289.000 0.000 0.154 1.000
Wilks’ lambda 0.651 15.978 2.000 289.000 0.000 0.154 1.000
Hotelling-Lawley trace 0.172 14.896 2.000 289.000 0.000 0.154 1.000
Roy’s largest root 0.168 14.935 2.000 289.000 0.000 0.154 1.000
Restaurant level * Ambiance level
Pillai’s trace 0.078 12.231 2.000 289.000 0.000 0.078 0.996
Wilks’ lambda 0.922 12.231 2.000 289.000 0.000 0.078 0.996
Hotelling-Lawley trace 0.085 12.231 2.000 289.000 0.000 0.078 0.996
Roy’s largest root 0.085 12.231 2.000 289.000 0.000 0.078 0.996

Table 6: Estimated marginal means for willingness-to-pay (WTP) and intention to patronise (ITP) – Ambiance

Restaurant type Ambiance quality level Intention to patronise (mean) % change Willingness-to-pay (mean) % change
Limited service Low ambiance quality 4.31 24 138 5.1
High ambiance quality 5.34 145
High-end Low ambiance quality 3.97 35 1180 29.0
High ambiance quality 5.36 1522
 178 Njite, Njoroge, Parsa, Parsa and van der Rest

WTP Estimated Means (KShs.)

ITP Estimated Marginal Means

6 1400

1200
5
1000
4
800
Willingness to pay limited service restaurant
3
600 Willingness to pay high end restaurant
2 Intention to patronage limited service restaurants 400
Intention to patronage high end restaurants
1 200

Low Ambience Quality High Ambience Quality Low Ambience Quality High Ambience Quality
Ambience Quality Ambience Quality

Figure 5: Estimated marginal means for ITP and ambiance Figure 6: Estimated marginal means for WTP and ambiance (WTP) is
measured in Kenya Shillings (KShs.)

Discussion above discussion can be extended to ambiance and consumer

intention in high-end and limited service restaurants.
The findings suggest, however, that, contrary to common On the other hand, customers of high-end restaurants
belief (DiPietro et al., 2011), the three key restaurant attributes display their preference for personalisation of service by the
are neither equal nor necessarily always significant in their great difference in the means for the low and high service
relationship to consumer intention. The relationship between personalisation. They expect high quality food as well as
food quality and ITP/WTP are positive both in high-end and personalisation of service. Consumers of high-end restaurants
limited service restaurants, but that is not the case with invariably patronise high-end, upscale restaurants for a totally
reference to service quality and ambiance. emotional experience that is beyond the simple physiological
With respect to service quality and ambiance, consumer gratification provided by food alone. In other words, consumers
preferences for high-end and limited service restaurants differ. of upscale restaurants are seeking physiological gratification
In upscale restaurants, the relationship between attributes (food) followed by emotional satisfaction (service) and
quality and ITP as well as WTP is positive and significant. conspicuous consumption (ambiance). In high-end restaurants,
Service quality and ambiance remain unpredictable in limited all three service attributes (food, service and ambiance) are
service restaurants. Unlike food quality, in case of limited important in achieving high consumer patronage.
service restaurants, both service quality (i.e. personalisation)
and ambiance have positive and nonsignificant relationships Conclusions
with WTP but it is significant in case of ITP.
The three key restaurant attributes thus have very unique This study intended to investigate the effect of service attributes
influences on consumers’ behavioural intentions, both in the on consumer patronage and willingness-to-pay. In particular,
high-end restaurants and limited service restaurants. This is it investigated the preferences of consumers in the developing
an interesting finding which deserves further investigation. It economy of Kenya, with respect to three major attributes in the
confirms the questionable universality of consumer preferences restaurant industry: food quality, service quality, and ambiance.
across the different sectors of the restaurant industry The study finds that food quality is the only attribute to have
supporting studies (Mattila, 2001a/b; Min & Min, 2011; a positive and significant relationship with both consumer
Namkung & Jang, 2007). patronage and willingness-to-pay, in high-end as well as limited
The differences in the influences of service quality between service restaurants. Service quality and ambiance both have
upscale and limited service restaurants could be partially positive and significant relationships with consumer intentions
attributed to the unique differences in customer expectations in high-end restaurants. Moreover, in high-end restaurants,
related to these different types of restaurants. Customers of increase in ambience was found to increase patronage and
limited service restaurants usually focus on a limited number willingness-to-pay with a significant amount if service quality
of service dimensions such as convenience, speed of service, was also increased. When ambiance was improved in upscale
order accuracy, menu simplicity, ease of product ordering, high restaurants the WTP improved by 29% compared to the 5%
use of technology in service delivery, drive-through access, and in limited service restaurants. This is definitely a significant
other speed-related attributes (Grönroos, 1984). In addition, difference worthy of further investigation.
consumers at limited service restaurants spend less time with When service was improved, the upscale restaurant WTP
the service staff and, therefore, may not really place emphasis improved by 47% compared to the 19% in limited service
on personalisation of the service process. Most of the limited restaurants. This is definitely a significant difference. Similarly,
service restaurant exchanges are transactional in nature, the change in the ITP for both the upscale and limited service
leaving less scope for service personalisation, and thus it can restaurants was significant with the values for upscale (70.6%)
be expected that ITP and WTP do not significantly change as and limited service (34%) restaurants. This clearly shows
service personalisation varies from low to high. Similarly, the that service has significant impact on restaurant patronage
Research in Hospitality Management 2015, 5(2): 171–180 179

and willingness to pay. Interestingly, in the case of limited Bujisic, M., J. Hutchinson, & H. G. Parsa. (2014). The effect of ambiance,
service restaurants, the relationship between service and food and service quality on restaurant customer behavioral intentions.
consumers’ intentions were mixed. The WTP was positive but International Journal of Contemporary Hospitality Management,
not significant and the ITP was positive and significant. These 26(8), 1270–1291. http://dx.doi.org/10.1108/IJCHM-04-2013-0162
findings also indicate, in the case of limited service restaurants, Clark, M. A., & R. Wood. (1998). Consumer loyalty in the restaurant
industry: A preliminary exploration of the issues. International
that when service improves, even though consumers are
Journal of Contemporary Hospitality Management, 10(4), 139–144.
willing to patronise more, they are not willing to pay more. http://dx.doi.org/10.1108/09596119810222104
In other words, limited service restaurants can expect higher Cohen, J., P. Cohen, S. G. West, & L. S. Aiken. (2003). Applied multiple
revenues through repeated patronage but not necessarily regression/correlation analysis for the behavioral sciences. New York:
higher revenues per customer. Routledge.
Similarly, the relationship between ambiance and WTP and Dabholkar, P. A., C. D. Shepherd, & D. I. Thorpe. (2000). A
ITP are different for high-end restaurants and limited service comprehensive framework for service quality: An investigation of
restaurants. The relationship was positive and significant for critical conceptual and measurement issues through a longitudinal
high-end restaurants and positive but insignificant for limited study. Journal of Retailing, 76(2), 131–139. http://dx.doi.
service restaurants for WTP and significant for ITP. Similar org/10.1016/S0022-4359(00)00029-4
DiPietro, R. B., H. G. Parsa, & A. M. Gregory. (2011). Do restaurant
to service, in the case of limited service restaurants, with an
QSC inspections lead to better financial performance? An
increase in ambiance quality the consumers’ ITP improved empirical investigation. International Journal of Contemporary
(24%) manyfold compared to WTP (5%), but, in the case of Hospitality Management, 23(7), 982–999. http://dx.doi.
high-end restaurants, it remained consistent for ITP (35%) and org/10.1108/09596111111167560
WTP (29%). Dutta, K., H. G. Parsa, A. R. Parsa, & M. Bujisic. (2014). Change in
This study has a few limitations. First of all, it includes only consumer patronage and willingness to pay at different levels of
two levels of service attributes, when in reality attribute levels service attributes in restaurants: A study from India. Journal of
do exist as a continuum versus a dichotomy. It is a scenario- Quality Assurance in Hospitality & Tourism, 15(2), 149–174. http://
based study with stated preferences, thus the findings dx.doi.org/10.1080/1528008X.2014.889533
could be further explored with secondary data from the Griffin, A., & J. Hauser. (1993). The voice of the customer. Marketing
Science, 12(1), 1–27. http://dx.doi.org/10.1287/mksc.12.1.1
industry (i.e. revealed preferences). The data were collected
Grönroos, C. (1984). A service quality model and its marketing
from a developing economy, limiting its generalisability to implications. European Journal of Marketing, 18(4), 36–44. http://
other developed nations and across cultural boundaries. dx.doi.org/10.1108/EUM0000000004784
Furthermore, the three service attributes – food quality, Kara, A., E. Kaynak, & O. Kucukemiroglu. (1995). Marketing strategies
service quality (i.e. personalisation) and ambiance – that were for fast-food restaurants: A customer view. International Journal of
examined, are strongly correlated, and thus any tentative Contemporary Hospitality Management, 7(4), 16–22. http://dx.doi.
conclusion about any one attribute in isolation may not be org/10.1108/09596119510146823
prudent unless the influence of the inter-attribute impact is also Kim, W. G., C. Y. N. Ng, & Y. Kim. (2009). Influence of institutional
better understood. In other words, expectation of high ROI for Dineserv on customer satisfaction, return intention and word-of-
service improvements may not be realistic without proportional mouth. International Journal of Hospitality Management, 28(1),
10–17. http://dx.doi.org/10.1016/j.ijhm.2008.03.005
improvements in food quality and ambiance. Further studies
Mattila, A. S. (2001a). The impact of relationship type on customer
are required to better understand these complex relationships, loyalty in a context of service failure. Journal of Service Research,
including measurement of the slope for change in consumer 4(2), 91–101. http://dx.doi.org/10.1177/109467050142002
patronage when service attributes change from low to high. Mattila, A. S. (2001b). Emotional bonding and restaurant loyalty. The
Suggested studies measuring the slope would demonstrate the Cornell Hotel and Restaurant Administration Quarterly, 42(6), 73–79.
direction as well as the magnitude of the relationships. http://dx.doi.org/10.1016/S0010-8804(01)81012-0
Mehrabian, A., & J. A. Russell. (1974). An approach to environmental
References psychology. Cambridge, MA: MIT Press.
Min, H., & H. Min. (2011). Benchmarking the service quality of
Alba, J., J. Lynch, B. Weitz, C. Janiszewski, R. Lutz, A. Sawyer, fast-food restaurant franchises in the USA. Benchmarking:
& S. Wood. (1997). Interactive home shopping: Consumer, An International Journal, 18(2), 282–300. http://dx.doi.
retailer, and manufacturer incentives to participate in electronic org/10.1108/14635771111121711
marketplaces. Journal of Marketing, 61(3), 38–53. http://dx.doi. Mittal, V., P. Kumar, & M. Tsiros. (1999). Attribute-level performance,
org/10.2307/1251788 satisfaction, and behavioral intentions over time: A consumption-
Baker, J. (1987). The role of the environment in marketing services: system approach. Journal of Marketing, 63(2), 88–101. http://dx.doi.
The consumer perspective. In J. A. Czepiel, C. A. Congram, & J. org/10.2307/1251947
Shanahan (Eds.), The services challenge: Integrating for competitive Namkung, Y., & S. Jang. (2007). Does food quality really
advantage (pp. 79–84). Chicago: American Marketing Association. matter in restaurants? Its impact on customer satisfaction
Berry, L. L., & T. Clark. (1986). Four ways to make services more and behavioral intentions. Journal of Hospitality & Tourism
tangible. Business (Atlanta, GA.), 36(October–December), 53–54. Research (Washington, D.C.), 31(3), 387–409. http://dx.doi.
Bitner, M. J. (1992). Servicescapes: The impact of physical surroundings org/10.1177/1096348007299924
on customers and employees. Journal of Marketing, 56(2), 57–71. National Restaurant Association (2010). http://www.restaurant.org/
http://dx.doi.org/10.2307/1252042 Perutkova, J., & H. G. Parsa. (2010). Consumers’ willingness to pay
Booms, B. H., & M. J. Bitner. (1982). Marketing services by and to patronize according to major restaurant attributes. UCF–
managing the environment. The Cornell Hotel and Restaurant Undergraduate Research Journal, 4(2), 1–11.
Administration Quarterly, 23(1), 35–40. http://dx.doi.
org/10.1177/001088048202300107
180 Njite, Njoroge, Parsa, Parsa and van der Rest

Qin, H., & V. R. Prybutok. (2008). Determinants of customer-perceived Verma, R., G. M. Thompson, & J. J. Louviere. (1999). Configuring
service quality in fast food restaurants (FFRs) and their relationship service operations in accordance with customer needs and
to customer Satisfaction and behavioral intentions. Quality preferences. Journal of Service Research, 1(3), 262–274. http://
Management Journal, 15(2), 35–50. dx.doi.org/10.1177/109467059913007
Qin, H., V. R. Prybutok, & Q. Zhao. (2010). Perceived service quality in Wall, E. A., & L. L. Berry. (2007). The combined effects of the
fast-food restaurants: Empirical evidence from China. International physical environment and employee Behavior on customer
Journal of Quality & Reliability Management, 27(4), 424–437. http:// perception of restaurant service quality. The Cornell Hotel and
dx.doi.org/10.1108/02656711011035129 Restaurant Administration Quarterly, 48(1), 59–69. http://dx.doi.
Rapoport, A. (1982). The meaning of the built environment. Beverly org/10.1177/0010880406297246
Hills, CA: Sage. Winer, R. S. (2001). A framework for customer relationship
Stevens, P., B. Knutson, & M. Patton. (1995). Dineserv: A tool for management. California Management Review, 43(4), 89–105. http://
measuring service quality in restaurants. The Cornell Hotel and dx.doi.org/10.2307/41166102
Restaurant Administration Quarterly, 36(2), 56–60. http://dx.doi. World Bank. (2013). Country and region specific forecasts
org/10.1177/001088049503600226 and data. http://www.worldbank.org/en/publication/
Tlapa, C. A., R. K. Miller, & K. Washington. (2011). The 2011 global-economic-prospects/data?region=SST
restaurant, food & beverage market research handbook. Loganville,
GA: Richard K. Miller & Associates.
Upah, G. D., & J. N. Fulton. (1985). Situation creation in services
marketing. In J. A. Czepiel, M. R. Solomon, & C. F. Surprenant (Eds.),
The service encounter (pp. 255–263). Lexington, MA: Lexington
Books.