Journal of Administrative and Business Studies JABS

2017, 3(5): 221-233

PRIMARY RESEARCH

Investigating consumer preferences in selecting buffet

restaurants in Davao region, Philippines
Dunhill Z. Bilog *

Accountancy & Business Administration Program

University of the Immaculate Conception, Bonifacio Street, Davao, Philippines

Keywords Abstract. The buffet restaurant industry players in Davao region of the Philippines have contributed to eco-
nomic growth. However, with the in lux of national buffet restaurants, the need to revisit the focus of the home-
Consumer preferences
grown buffet restaurant owners and managers has become apparent. Along this is the need to look into the pref-
Buffet restaurant
erences of the buffet restaurant consumers and the segments that patronize the industry. This study primarily
Conjoint analysis
sought to identify the attributes consumers in Davao region would look for in a buffet restaurant. It employed an
Two-step cluster analysis
experimental research design where conjoint analysis as a technique was used. Buffet restaurant pro iles were
Davao region
generated based on the attributes of types of food, quality of food, price, and cleanliness. These were rated by
the respondents. For uncovering the segments, a two-step cluster analysis was utilized. The results of the study
indicate that buffet restaurant consumers in the region placed the highest utility on cleanliness. Both junior and
Received: 29 September 2017
senior professionals have the same order of priority as regards the four buffet restaurant attributes. However,
Accepted: 2 October 2017
the utility values differ. Moreover, three consumer groups, namely single junior professional oriental food lovers,
Published: 9 October 2017
married junior professional oriental food lovers, and married senior professional oriental food lovers were uncov-
ered. These results could prove helpful to buffet restaurant owners and managers in coming up with sustainable
service designs and appropriate positioning strategies.

©2017 TAF Publishing. All rights reserved.

INTRODUCTION from 2013 (Riehle, Grindy & Stensson, 2014). Hence, given
One of the most highly-competitive industries in the market these seemingly increasing consumer demand and an in-
today is food service. Being so, it is of utmost importance creasing dif iculty in satisfying restaurant consumers (Han,
that food service operators take the time to understand 2009; Intan, 2016), a thorough analysis to look into ways
their market consumers better considering that it is also that would expand consumer base and consequently tap
the global market’s fastest growing industry (Gu & Kim, into new opportunities in the market becomes essential
2002). Notably, its players have described their experience (International Markeys Bureau, 2010).
as being in the eye of a perfect storm (Haas, 2008). Equally According to Bartlett & Han (2007), the expectations
important to note is that the food service industry is con- and evaluations of a consumer will have an in luence on
tinuously impacted by the swift-evolving preferences of the the restaurant dining experience. This could well be the
consumer (Waldfogel, 2008). In the US where the restau- reason why today’s dining consumers are more impatient
rant industry is considered a powerhouse, sales volume (Saputra & Dewi, 2016; Soderlund & Ohman, 2005). Due
was projected at 683.4 billion USD, a 3.6 percent increase to high expectations coupled with many choices in all these

* Corresponding author: Dunhill Z. Bilog

† Email: dunhilljigz@gmail.com

Electronic copy available at: https://ssrn.com/abstract=3744862

2017 D. Z. Bilog – Investigating consumer preferences in . . . . 222

restaurant segments, a consumer leaving for another es- Theoretical Framework

tablishment is a common sight (Kang, Nobuyuki & Herbert, This study was anchored primarily on consumer behavior
2004). theory which has two schools of thought: the cognitive and
According to HRI Food Service Sector (2015), the Philip- behavioralist paradigms. Using the cognitive view, the The-
pines has experienced a spike in the growth of its food ser- ory of Reasoned Action is the primary basis of this study.
vice industry. This was driven by the increasing number of It espoused that personal beliefs towards the attributes
hotels and shopping malls as well as the arrival of restau- of a good or service become the basis for consumer atti-
rants that are of foreign origin. With these developments in tude (Ajzen & Fishbein, 1980). The behavioralist view of
the country and as food service players continue to mush- the consumer behavior theory could be clearly indicated
room, the playing ield will become much more competitive by the Behavioral Perspective Modeling which emphasized
(Akbar & Alaudeen, 2012). that purchasing behavior is in luenced by reinforcing and
Olsen (2002) asserted that there are other signi icant punishing consequences. These consequences of human
factors that traditional restaurants should focus apart from behavior are driven by physical and social environment
food quality and taste. These are location, atmosphere, (Foxall, 2010). Another interestingly essential theory cen-
serving, tangibles, and interior decorations. Akinyele tral to the direction of this study was the theory of choice.
(2010) supported this with a claim that the restaurant expe- This highlights the normative and descriptive aspects that
rience would include more than what the customer wants are taken into consideration when an individual makes a
to eat. Hence, for food service industry players, it is not decision. The consumer’s beliefs and the individuals’ col-
enough anymore to only offer good food and service to at- lective description of beliefs and values would come into the
tract and retain consumers (Soriano, 2002, Kuo & Chen, picture when making a sequential decision-making (Slovic,
2015). Fischhoff, & Lichtenstein, 1977).
In Davao region, a consistently growing economic pic- Lastly, the market segmentation theory was adapted.
ture through the years has become apparent. The services According to Dilworth (2000), a necessary part of achieving
sector where food service belongs to recorded a 4.95% av- quality is understanding who the customers are and what
erage annual growth rate. Along with this, the sector con- their needs and expectations are. Further, Kotler (2013)
tributed immensely to the region’s economy followed by mentioned that preferences could be classi ied as homoge-
the industry sector (Regional Development Council Region neous, diffused, or clustered.
XI, 2011). Considerably, the food service industry players
in Davao City, particularly home-grown buffet restaurants, Conceptual Framework
brought much help. The result of the Focus Group Discussion provided the fol-
With this growth in the region comes the in lux of na- lowing: the independent variables, which are the attributes
tional buffet restaurant brands. With these operating in the of buffet restaurants, and the dependent variable, which is
city, the local buffet restaurant consumers may have altered the customer preferences towards selecting a buffet restau-
their view on the homegrown buffet restaurant. Hence, the rant. Hence, the different attributes of the buffet restaurant,
researcher believes that there could exist a gap between which constitute the independent variables in the study,
the local buffet restaurant attributes that are being con- were type of food, quality of food, price, and cleanliness.
centrated on by the owners and the consumer preferences These are thought to affect the dependent variable of the
for selecting buffet restaurants. The local buffet industry study which is the preferences of diners indicated by the
owners might be using their perception in offering buffet total utility they give for a buffet restaurant. The types of
restaurant service vis-à -vis the national buffet restaurants. food were sub-classi ied into Filipino, Korean, Japanese,
In this light, the researcher proposes this study to narrow Chinese, European, and American. The quality of food was
the said gap and to provide a signi icant contribution to the divided into three levels: excellent, average, and poor. The
body of knowledge regarding buffet restaurants and their price levels were above P500, P300-P500, and below P300.
management. Lastly, cleanliness was classi ied into excellent, average,
and poor.

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
223 J. Admin. Bus. Stud. 2017

FIGURE 1 . Conceptual framework

Statement of the Problem This is helpful in predicting consumer acceptance towards

This study aimed to identify the attributes that would com- combinations of attributes. Cochran’s (1963) sampling for-
prise the preference structure of consumers for a local buf- mula was used in generating the minimum sample size.
fet restaurant in Davao City. With these results, the study covered 500 buffet restaurant
Speci ically, it sought to answer these questions: patrons in Davao City. This fell within the median range
1. What attributes of buffet restaurants form the preference of 300-550 according to Cattin & Wittink (1982), which
structure of consumers classi ied as junior professionals were adapted in the conjoint studies of (Cruz, 1997, 2004;
and senior professionals? Patayon, 2008). Orme (2010) de ined that conjoint-related
2. What are the utility values that junior professionals studies should generally have sample sizes from 150 to
attach to the different attributes forming the preference 1,200. Further, a probabilistic sampling method employing
structure for a buffet restaurant? the concept of Poisson distribution for the total sample size
3. What are the utility values that senior professionals of 500 buffet restaurant patrons was used.
attach to the different attributes forming the preference A survey questionnaire was adopted as the method of
structure for a buffet restaurant? data collection which was administered in two stages. The
4. What is the relative importance of each attribute forming irst stage involved the Focus Group Discussion wherein
the preferred buffet restaurant of junior professionals? the survey questionnaire was given to participants to ex-
5. What is the relative importance of each attribute forming tract the relevant buffet restaurant attributes consumers
the preferred buffet restaurant of senior professionals? would consider important. Subsequently, once important
6. Are there other segments of buffet restaurant consumers attributes were known, another survey questionnaire was
with signi icantly different preferences? What are the dis- given geared towards the evaluation of 34 buffet restaurant
tinguishing characteristics of the local buffet restaurants if pro iles through rating, seven of which were hold-out stim-
they signi icantly exist? uli.
The data gathered for this study were summarized, in-
METHODOLOGY terpreted, and analyzed using frequency counts and per-
This study utilized an experimental research design since centages which described the buffet restaurant consumers’
it employed conjoint experiments (orthogonal design). Ac- response to the survey items, conjoint analysis to deter-
cording to Hair, Black, Babin, & Anderson (2014), an exper- mine the utility of attribute which represents the value con-
imental design used in analyzing consumer preferences has sumers put on an attribute level and the importance of at-
the objective of determining the predictor variables’ contri- tribute, which is simply the difference between the lowest
bution and forming a valid consumer judgments model. and highest utilities across the levels of attributes

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
2017 D. Z. Bilog – Investigating consumer preferences in . . . . 224

and two-step cluster analysis used to group objects based signi icance at .01 level which means that this model rep-
on the characteristics they possess so that each object is resents the buffet restaurant attributes identi ied by the
very similar to others in the cluster with respect to some respondents. The correlation for holdouts indicates the
predetermined selection criterion, in this case, segments. model is it, at .05 level of signi icance.

RESULTS AND DISCUSSION TABLE 2 . Overall conjoint correlationsresults

between observed and estimated
Pro ile of Buffet Restaurant Consumers in Davao Region
preferences of buffet restaurant
As presented in Table 1, around 80.4 percent of the 500 re-
consumers based on location
spondents came from Davao del Sur and 2.0 percent were
Correlations Value Sig.
from Compostela Valley. 31.2 percent were 18-25 years
Pearson’s R .984 .000**
old while 2.4 percent were above 60 years old. The College
Kendall’s tau .954 .000**
degree holders comprised 78.2 percent whereas Doctor-
Kendall’s tau for Holdouts .905 .002*
ate degree holders registered 1.0 percent. Further, 50.6
Correlations between observed and estimated
percent were male and 40.4 percent were female. More- preferences
over, 51.8 percent were single while only .6 percent were * Signi icant at 0.05 level of signi icance.
divorced. Lastly, of the 500 respondents, 73.6 percent were **.Signi icant at 0.01 level of signi icance.

Junior professionals.
As shown in Table 3, Cleanliness earned the high-
TABLE 1 . Distribution of buffet restaurant consumers according est utility value (35.765). This is in consonance with the
to demographics studies done by Fatimah, Boo, Sambasivan, & Salleh (2011),
Category Frequency Percentage wherein they asserted the importance of food service hy-
Davao del Sur 402 80.4 giene or cleanliness.
Address Davao del Norte 68 13.6
Davao Occidental 2 .4 Further, this aspect remains far more essential since the
Davao Oriental 18 3.6 associated risk could be really substantial. The lowest util-
Compostela Valley 10 2.0
Age 18-25 156 31.2
ity value was registered by Price (8.427) attribute. This
26-30 103 20.6 means that buffet restaurant consumers no longer consider
31-40 120 24.0 price as the primary attribute to look into in selecting a buf-
41-50 79 15.8
51-60 30 6.0 fet restaurant. Instead, they place the highest value on the
Above 60 12 2.4 cleanliness of a buffet restaurant. Chung & Kim (2011) clar-
Education High School 58 11.6
College 391 78.2
i ied that price remains one of the major factors that is used
Masteral 36 7.2 in consumer decision-making as well as its subsequent be-
Doctorate 5 1.0 haviors. Kafel & Sikora (2013) deepened the insight by as-
Others 10 2.0
Sex Male 253 50.6 serting that consumers utilize price as a gauge for the qual-
Female 247 40.4 ity of the restaurant. Hence, if a restaurant serves quality
Civil Status Single 259 51.8
Married 228 45.6
food, then it must be expensive.
Widowed 10 2.0
Divorced 3 .6 TABLE 3 . Respondents’ overall average
Current Position Junior Professional (25-45) 292 73.6 importance values on the
Senior Professional (above 45) 105 26.4
Does not belong 103
buffet restaurant attributes
Total 500 100.0 based on location
Buffet Restaurant Attributes Value
Utility Values of Buffet Restaurant Attributes by Buffet Type of Food 25.541
Restaurant Consumers from the Selected Provinces in Quality of Food 30.267
Davao Region Price 8.427
As seen in Table 2, all utility values are highly signi icant. Cleanliness 35.765
Correlations between observed and estimated
Hence, all the buffet restaurant attributes are looked upon
preferences
as the basis for the selection of a buffet restaurant. The * Signi icant at 0.05 level of signi icance.
Pearson’s R and Kendall’s tau correlations both showed **.Signi icant at 0.01 level of signi icance.

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
225 J. Admin. Bus. Stud. 2017

Partial utility estimates (or part-worths) can be used to its taste. However, tremendous appreciation is seen gradu-
compute the overall utility for each of the four buffet restau- ally given the noticeable proliferation of Korean culture in
rant attributes. This is done as part-worth utility estimates the Philippines. Kimchi and Ramyeon are popular Korean
are naturally converted into a common scale. Generally, dishes that have become household names which are all the
these estimated utilities are scaled in such a way that they more reinforced by Koreans coming to the country to study
total to zero. or do business.
Consequently, a negative number will not mean that a Also, the Chinese food earning a negative utility estimate
particular level has “negative utility”. It just means that the could have come from the indigenization and transforma-
factor level is on average less preferred than a factor level tion of the Chinese food in the country. Fernandez (2002)
having positive estimated utility (Chapman, 2013). As pre- stressed that while there are a lot of foods that are of Chi-
sented in Table 4, the utility estimates showed a greater nese origin, a Filipino consumer would not be able to ind
preference for Japanese (.042), American (.195), and Fil- a Chinese dish done exactly the way they are in China, or
ipino (.236) foods. Greater preference is evident as well tasting the same. This indigenizing process eventually Fil-
for average (.139) to excellent (.406) quality of food. As ipinized Chinese foods.
regards the price, preference is less than P300. All of the Lastly, the European food, according to Roman & Russell
buffet restaurant attributes are even enhanced if the buffet (2009), has always had their in luence in Southeast Asia.
restaurant sports average (.168) and excellent (.469) clean- The love for food such as pan de sal, empanada, kilawin,
liness. As presented, the Korean (-.221), Chinese (-.040), paella, and a variety of other seasoned meat dishes cannot
and European (-.212) types of food earned negative util- be denied. However, these, as time went on, experienced
ity estimates. Temblor (2016) emphasized that the hesi- indigenization. Hence, most of us perceived them as typical
tation about Korean food stemmed from the uniqueness of Filipino foods.

TABLE 4 . Respondents’ overall utility estimates of factors based on buffet restaurant

attributes
Attribute Factor Utility-Estimate (Part-worths) Std. Error
Type of Food Filipino .236 .060
Korean -.221 .060
Japanese .042 .060
Chinese -.040 .078
European -.212 .078
American .195 .078
Quality of Food Excellent .406 .042
Average .139 .042
Poor -.545 .042
Price Above P500 .112 .036
P300-P500 .224 .072
Less than P300 .337 .108
Cleanliness Excellent .469 .042
Average .168 .042
Poor -.637 .042
Constant 2.561 .078

As seen in Table 5, all utility values are highly signi icant. presents the buffet restaurant attributes identi ied by the
Hence, all the buffet restaurant attributes are looked upon respondents. The correlation for holdouts indicates the
as the basis for the selection of a buffet restaurant. The model is it, at .05 level of signi icance.
Pearson’s R and Kendall’s tau correlations both showed sig-
ni icance at .01 level which means that this model re-

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
2017 D. Z. Bilog – Investigating consumer preferences in . . . . 226

TABLE 5 . Overall conjoint correlations fet restaurants. In particular, cleanliness was revealed to
results between observed and generate the largest utility value, which is indicative of the
estimated preferences of buffet fact that most existing buffet restaurants in Davao Region
restaurant consumers based on
look into the cleanliness as a primary attribute factor.
current position
Correlations Value Sig. TABLE 7 . Preference score on
Pearson’s R .984 .000** the overall utility
Kendall’s tau .954 .000** estimates of factors
Kendall’s tau for Holdouts .905 .002* based on the buffet
a. Correlations between observed and estimated restaurant attributes
preferences
Attribute Factor Utility
*.Signi icant at 0.05 level of signi icance.
Type of Food: Filipino .236
**.Signi icant at 0.01 level of signi icance.
Quality of Food: Excellent .406
Price: Less than P300 .337
As shown in Table 6, cleanliness earned the highest util- Cleanliness .469
ity value (32.442). This is in consonance with the studies Constant 2.561
Preference Score 4.009
done by Fatimah et al., (2011) wherein they asserted the im-
portance of food service hygiene or cleanliness.
Further, this aspect remains far more essential since the On Job Position. Table 8 presents that there is a distinc-
associated risk could be substantial. The lowest utility value tion for both the Junior and Senior professionals when it
was registered by Price (11.386) attribute. This means that comes to buffet restaurant preferences. This is evidenced
buffet restaurant consumers no longer consider price as the by the Kendall’s tau for holdouts which is measured at a .05
primary attribute to look into in selecting a buffet restau- level of signi icance. The Pearson and Kendall’s correlation
rant. Instead, they place the highest value on the cleanli- outcomes both indicate signi icance at a .01 level, which is
ness of a buffet restaurant. Chung & Kim (2011) clari ied very representative of the buffet restaurant attributes iden-
that price remains one of the major factors that is used in ti ied for both junior and senior professionals.
consumer decision-making as well as its subsequent behav-
TABLE 8 . Conjoint correlations results between observed
iors. Kafel & Sikora (2013) deepened the insight by assert-
and estimated preferences of buffet restaurant
ing that consumers utilize price as a gauge for the quality
consumers when compared between junior and
of the restaurant. Hence, if a restaurant serves quality food, senior professionals
then it must be expensive. Junior Professional Senior Professional
Value Sig. Value Sig.
TABLE 6 . Respondents’ overall average Pearson’s R .982 .000** .973 .000**
importance values on the Kendall’s tau .940 .000** .920 .000**
Kendall’s tau for holdouts .810 .005* .905 .002*
buffet restaurant attributes
a. Correlations between observed and estimated preferences
based on current position *.Signi icant at 0.05 level of signi icance.
Buffet Restaurant Attributes Value **.Signi icant at 0.01 level of signi icance.
Type of Food 29.502
Quality of Food 28.496 In Table 9, a clear distinction between Junior and Senior
Price 8.640
Professionals can be seen through the average importance
Cleanliness 33.362
values on the buffet restaurant attributes. A senior profes-
sional places high preference on the type of food. As re-
As exhibited in Table 7, the overall preference gards the quality of food, a junior professional puts more re-
(4.015) is derived from the sum of utility part-worths for gard compared to a senior professional. On the other hand,
each attribute plus the regression constant term. The high a senior professional is less price-conscious than a junior
preference score is an indication of greater desired level of professional. When it comes to cleanliness, a senior profes-
attributes which then adds a lot to the overall utility of buf- sional places more importance than a junior professional.

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
227 J. Admin. Bus. Stud. 2017

TABLE 9 . Conjoint correlations results between observed and estimated

preferences of buffet restaurant consumers when compared
between junior and senior professionals
Junior Professional Senior Professional
Type of Food 16.587 21.328
Quality of Food 35.172 29.062
Price 8.127 6.998
Cleanliness 40.115 42.612

The responses of the junior and senior professionals (β = .125) with excellent cleanliness (β = .492). On the other
towards the attribute factors can be seen in Table 10. The hand, the senior professionals show a preference for a buf-
junior professionals place the highest preference on a buffet fet restaurant that offers Filipino type of food (β = .291) with
restaurant that would offer Filipino type of food (β = .229) excellent quality (β = .300), priced at less than P300 (β =
with excellent quality (β = .427) and is priced less than P300 .079) with excellent cleanliness (β = .418).

TABLE 10 . Respondents’ utility estimates of factors based on the buffet restaurant attributes when compared
according to junior and senior professionals
Attribute Factor Junior Professional Senior Professional
Utility-Estimate Std. Error Utility-Estimate Std. Error
Type of Food Filipino .229 .065 .291 .069
Korean -.245 .065 -.199 .069
Japanese .066 .065 -.048 .069
Chinese -.048 .086 .008 .090
European -.207 .086 -.220 .090
American .205 .086 .168 .090
Quality of Food Excellent .427 .045 .300 .048
Average .151 .045 .097 .048
Poor -.578 .045 -.397 .048
Price Above P500 .116 .039 .084 .041
P300-P500 .232 .079 .168 .083
Less than P300 .348 .118 .252 .124
Cleanliness Excellent .492 .045 .418 .048
Average .161 .045 .184 .048
Poor -.654 .045 -.603 .048
Constant 2.533 .086 2.651 .090

In Table 11, the total utility for junior (4.029) and se- restaurant that offers Filipino foods, then these acquired
nior (3.912) professionals suggests that the predetermined tastes and dining pattern will be carried forward as con-
buffet restaurant attributes for junior professionals have a sumer ages (Reynolds & Hwang, 2006).
much higher overall preference score over the senior pro- On the other hand, the junior professionals tend to pre-
fessionals. Moreover, the senior professionals tend to favor fer more excellent quality of food and cleanliness than the
more the type of food than the junior professionals. This is senior professionals. This is consistent with the indings of
in conjunction with the study of Reynolds & Hwang (2006) Soriano (2002) where in the 31-40 year old respondents
wherein they stressed that older consumers tend to veer af irmed more importance in the quality of food than any
away from exotic foods and newer food trends. Hence, if other restaurant attribute. Interestingly, the junior profes-
these consumers have been so entrenched in selecting a sionals tend to be more price-conscious than the senior

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
2017 D. Z. Bilog – Investigating consumer preferences in . . . . 228

professionals. Pucciarelli & Thomas (2011) underscored ported by Al-Khatib & Al-Mitwalli (2009), who underlined
that as consumers expand careers and combine incomes, that safe handling of food in restaurants is a fundamental
purchasing power for food becomes greater, thus increas- component in the reduction of foodborne diseases, which
ing foods bought from all food categories. also was emphasized by one of the Focused Group Discus-
As regards cleanliness, the junior professionals place sion participants as far as it being important in securing
higher preference than the senior professionals. This is sup- and retaining a job.

TABLE 11 . Preference score on the utility estimates of factors based on the buffet
restaurant attributes when grouped according to position
Attribute Factors Utility Estimates
Junior Professional Senior Professional
Type of Food: Filipino .229 .291
Quality of Food: Excellent .427 .300
Price: Less than P300 .348 .252
Cleanliness: Excellent .492 .418
Constant 2.533 2.651
Preference Score 4.029 3.912

Conjoint analysis as a statistical tool can be used for seg- settings. However, in spite of several tries, the result still
mentation. However, for it to create more meaning, it has shows poor it.
to be complemented with cluster analysis. This exploratory
tool procedure is intended to expose natural groupings or
clusters within a dataset that would otherwise not be evi-
dent. This procedure employs an algorithm that possesses
several desirable features that distinguish it from tradi-
tional clustering techniques. Further, with this method, it
gives the user the ability to determine the appropriate num-
ber of clusters, and then proceed with classifying through
the use of a nonhierarchical routine (Facca & Allen, 2011).
Clusters, therefore, can be created based on both cate-
gorical and continuous variables using this tool. Also, the
number of clusters are automatically selected as well as it
can analyze large data iles ef iciently. FIGURE 2 . Model summary view
Figure 2 is a model summary view which presents a
snapshot of the cluster model, including a Silhouette mea-
The two-step cluster analysis is superior to the tradi-
sure of cluster cohesion and separation which can be seen
tional clustering techniques since its algorithm can deal
shaded to indicate results which are poor, fair or good. Ac-
with both categorical and continuous variables, automat-
cording to Kaufman & Rousseeuw (1990), a cluster struc-
ically determines the optimal number of clusters, and al-
ture that has a reasonable or strong evidence will manifest
lows the analysis of large data iles. With its auto-clustering
a good result while fair result shows weak evidence and the
function, the largest ratio of distances generated from the
poor result signi ies no signi icant evidence. As presented, a
dataset signi ies the optimal number of clusters. This auto-
poor silhouette measure of cohesion and separation was de-
mated cluster selection can be done in two ways: Bayesian
rived through the Two-Step clustering algorithm and forty
Information Criterion (BIC) or Akaike’s Information Crite-
(40) predictors. Nonetheless, it resulted into three cluster
rion (AIC).
solutions. To produce a better result, one can always go
According to Jones (2011), the BIC should be seriously
back to the modeling node to amend the cluster model
considered for model selection wherein there are large sam-

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
229 J. Admin. Bus. Stud. 2017

ple sizes. Further, he af irmed that small changes that are Of the 500 total cases, 103 were excluded from the analysis
too trivial to be of practical importance might be significant. due to missing values on one or more of the variables. Of
Hence, the BIC holds a higher penalty for over itting com- the 397 cases assigned to clusters, 131 were assigned to the
pared with AIC. However, both clustering algorithms were irst cluster, 144 to the second, and 122 to the third.
presented. Table 12 presents the frequency of each cluster.

TABLE 12 . Cluster distribution

Cluster N % of Combined % of Total
1 131 33.0% 26.2%
2 144 36.3% 28.8%
3 122 30.7% 24.4%
Combined 397 100.0% 79.4%
Excluded cases 103 20.6%
Total 500 100%

Figure 3 shows the cluster predictor importance view. of cluster names, pro iles, sizes, and other pertinent vari
With this, the relative importance of each predictor in es- ables for each cluster. Hence, from the dataset, three clus
timating the model is shown. The most important predic- ters were labeled and described based on the signi icant
tor has to generate 1.0. In this study, current position regis- categorical and continuous variables as seen in Figure 4.
tered 1.0. This is closely followed by age (0.96). The buffet
restaurant pro ile 25 (Chinese type of food, excellent qual-
ity of food, P300-P500, excellent cleanliness) came in next
which showed 0.61. This was followed by civil status (0.56).
Buffet restaurant pro iles 30 (Chinese type of food, excel-
lent quality of food, less than P300, excellent cleanliness),
7 (Japanese type of food, excellent quality of food, less than
P300, average cleanliness), and 28 (Korean type of food, ex-
cellent quality of food, above P500, Excellent cleanliness)
trailed the predictor importance.

FIGURE 4 . Cluster

Analysis and Interpretation of Results

Based on the outcomes presented, it is hereby posited that
the present buffet restaurant consumers in Davao Region
would look up to cleanliness (“Excellent”; β = .469) as the
primary attribute factor in selecting a buffet restaurant.
This is followed by quality of food (“Excellent”; β = .406),
price (“Less than P300”; β = .337), and type of food (“Fil-
ipino”; β = .236). With the total utility estimate (Уk = 4.009),
the researcher came up with a general conjoint model be-
low:
Total Utility = Utility (Excellent cleanliness) + Utility (Ex-
FIGURE 3 . Predictor importance cellent quality of food) + Utility (Less than P300) + Utility
(Filipino Food) + Constant
The clusters view is another output of the cluster tech- = .469 + .406 + .337 + .236 + 2.561
nique which contains a cluster-by-features grid composed = 4.009

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
2017 D. Z. Bilog – Investigating consumer preferences in . . . . 230

This means that the people in Davao Region ind a ity of food) + Utility (Korean type of food) + Utility (Above
buffet restaurant more attractive if it possesses the bun- P500) + Constant
dled attributes (excellent cleanliness, excellent quality of =-.673 + -.545 + -.221 + .112 + 2.561
food, less than P300, Filipino food). With these bundled at- =1.234
tributes. these people will be more inclined to avail more
frequently a buffet restaurant service. Moreover, Table 13 TABLE 13 . Preference scores of
con irms this. The simulation card number 4 has a pref- simulations
erence score of 4.009 and a maximum utility of 59.4%. As
Card Number ID Score
can be seen, simulation 4 earns the highest BTL, 19.0% and
1 1 1.269
Logit, 29.7%. 133 out of 144 subjects were used in the
2 2 2.639
BTL and Logit methods because the subjects have all non-
3 3 3.320
negative scores. On one hand, the least preferred bundle of
4 4 4.009
attributes is that of poor cleanliness, poor quality of food,
5 5 1.962
Korean type of food, and above P500 price. This combina-
6 6 3.287
tion would generate total utility of:
7 7 3.741
Total Utility = Utility (Poor cleanliness) + Utility (Poor qual-

TABLE 14 . Preference probabilities of simulations

Card Number ID Maximum Utility Bradley-Terry-Luce Logit
1 1 0.7% 7.3% 4.0%
2 2 5.8% 13.3% 9.4%
3 3 6.8% 15.9% 15.0%
4 4 57.3% 19.5% 28.7%
5 5 2.8% 9.8% 5.8%
6 6 11.7% 15.8% 14.8%
7 7 14.9% 18.3% 22.3%

TABLE 15 . Simulation cases

Simulation Type of Food Quality of Food Price Cleanliness
1 Korean Poor Above P500 Poor
2 Japanese Poor Above P500 Excellent
3 Chinese Excellent P300-P500 Average
4 Filipino Excellent Less than P500 Excellent
5 European Average Above P500 Poor
6 American Average P300-P500 Average
7 Filipino Average Less than P300 Excellent

The simulations were conducted to understand how the Total Utility (Junior) = Utility (Excellent cleanliness) + Util-
set of respondents would choose among speci ied set of ity (Excellent quality of Food) + Utility (Less than P300) +
pro iles shown in Table 15. For junior professionals, their Utility (Filipino Food) +Constant
most preferred attribute factor is cleanliness (“Excellent”; = .492 + .427 + .348 + .229 + 2.533
β = .492). This is followed by quality of food (“Excellent”; β = 4.029
= .427), price (“less than P300”; β = .348), and type of food For senior professionals, their most preferred attribute
(“Filipino”; β = .229). With the total utility estimate (Уk = is cleanliness (“Excellent”; β = .418). This is followed by
4.029), the researcher came up with a conjoint model for quality of food (“Excellent”; β = .300), type of food (“Fil-
junior professionals as shown below: ipino”; β = .291), and price (“Less than P300”; β = .252).
ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
231 J. Admin. Bus. Stud. 2017

With the total utility estimate (Уk = 3.912), the researcher 21.328, quality of food – 29.062, price – 6.998, cleanliness
came up with a conjoint model for senior professionals as – 42.612. Three clusters of buffet restaurant consumers
shown below: in Davao Region were revealed and these clusters share
Total Utility (Senior) = Utility (Excellent cleanliness) + Util- the same preference for Oriental foods (Chinese, Japanese,
ity (Excellent quality of Food) + Utility (Filipino Food) + and Korean). The irst cluster (33.0%) is comprised of ju-
Utility (Less than P300) + Constant nior professionals who are single and are in the age bracket
= .418 + .300 + .291 + .252 + 2.651 of 31-40 years old. The second cluster (36.3%) is com-
= 3.912 posed of married, 31-40-year-old junior professionals. The
In sum, having a greater preference score over the se- third cluster (30.7%), on the other hand, is primarily made
nior professionals means that the junior professionals are up of senior professionals who are married and are in the
more inclined to favor the four identi ied buffet restaurant age bracket of 41-50. Based on these conclusions, the re-
attributes. Cluster Analysis Based on the outcomes, the searcher recommends that buffet restaurant owners and
cluster analysis yields three uniquely pro iled clusters. The managers should develop a design that its the preferences
irst cluster (Single Junior Professional Oriental food lovers) of the junior professionals and senior professionals. The
is comprised of junior professionals, 31-40 years old, sin- buffet restaurant owners and managers, when targeting ju-
gle. The second cluster (Married Junior Professional Orien- nior professionals, should also put considerable efforts to
tal food lovers) is composed of junior professionals, 31-40 establish cleanliness standards in the premises of the buf-
years old, married. The third cluster (Married Senior Pro- fet restaurant. Further, creating and nurturing an array of
fessional Oriental food lovers), on the other hand, is made excellent quality foods should be a priority; when target-
up of senior professionals, 41-50 years old, married. All ing senior professionals, owners and managers of a buffet
clusters share the same preference for a buffet restaurant restaurant should give more focus on cleanliness. Along
that offers Oriental (Chinese, Japanese, Korean) foods with with this, the quality of food, as well as the type of food,
excellent quality and excellent cleanliness regardless of the should be prioritized; the utility value of the type of food
price. came close to the quality of food. Therefore, for senior pro-
fessionals, the spirit of adventurism in trying out foreign
CONCLUSION AND RECOMMENDATIONS dishes is minimal. Thus, Filipino type of food should be
Given the results of the study, the researcher concludes that prioritized. Moreover, the buffet restaurant owners and
the order of priority as regards the four attributes (Clean- managers targeting junior professionals should place much
liness, Quality of Food, Type of Food, and Price) of both importance on cleanliness since it is their utmost consid-
junior and senior professionals is the same. However, the eration in buffet restaurant selection. The owners should
utility values differ. For junior professionals, the utility invest in getting sanitation certi ications as these would
values of each attribute are as follows: Cleanliness: excel- streamline hygienic processes and procedures in the orga-
lent - .492, Quality of Food: excellent - .427, Type of Food: nization. The owners and managers of buffet restaurant
Filipino - .229, and Price: less than P300 -.348 while for targeting senior professionals should consider cleanliness
senior professionals, the utility values of each attribute are as its priority. This means that the buffet restaurants should
as follows: Cleanliness: excellent - .418, Quality of food: invest in the establishment of processes that would elevate
excellent - .300, Type of Food: Filipino - .291, and price: its sanitation and hygiene standards. Lastly, the owners
less than P300 -.252. For junior professionals, the relative and managers of buffet restaurants should study the needs
importance of each attribute is as follows: Type of food and unique characteristics of the identi ied clusters so that
– 16.587, quality of food – 35.172, price – 8.127, cleanli- right positioning strategies could be crafted.
ness – 40.115 while for senior professionals, the relative
importance of each attribute is as follows: Type of food –

REFERENCES
Ajzen, I., & Fishbein, M. 1980. Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-
Hall.
Akbar, Y., & Alaudeen, M. 2012. Determinant of factors that in luence consumer in chooising normal full-rervice restaurant:
Case in seri iskandar, perak. South East Asian Journal of Contemporary Business, Economics and Law, 1(4): 137-145.

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
2017 D. Z. Bilog – Investigating consumer preferences in . . . . 232

Akinyele, S. 2010. Customer satisfaction and service quality: Customer’s re-patronage perspectives. Global Journal of Man-
agement & Business Research, 10(6): 83-90.
Al-Khatib, I., & Al-Mitwalli, S. 2009. Food sanitation practices in restaurants of Ramallah and Al-Bireh district of Palestine.
Eastern Mediterranean Health Journal, 15(4): 951-958.
Bartlett, J., & Han, F. 2007. Analysis of service quality in restaurants in China: an eastern perspective. Paper presented
at the ABR (Business) Conference & the 2007 TLC (Teaching) Conference, Oahu, HI.
Cattin, P., & Wittink, D. 1982. Commercial use of conjoint analysis: A survey. Journal of Marketing, 46(3): 44-53.
DOI: 10.2307/1251701
Chapman, C. 2013. 9 things clients get wrong about conjoint analysis. Paper presented at the conference of Sawtooth
Software, Dana Point, CA.
Chung, J., & Kim, S. 2011. Restaurant selection criteria: Understanding the roles of restaurant type and customers' socio-
demographic characteristics. Global Journal of Management & Business Research, 5(7): 70-86.
Cochran, W.G. 1963. Sampling techniques. New York, NY: Wiley Online.
Cruz, F.D. 1997. An application of conjoint analysis in the design of a coconut drink. Unpublished doctoral dissertation,
University of the Philippines Diliman, Quezon, PH.
Cruz, M. 2004. Consumer preferences and market segments for 1.3-liter cars in Davao city, Philippines: An application
of conjoint analysis as a predictive technique. Unpublished doctoral dissertation, De La Salle University, Manila, PH.
Dilworth, J. 2000. Operations management: Providing value in goods and services. Orlando, FL: Harcourt College Pub-
lisher.
Facca, T.M., & Allen, S.J. 2011. Using cluster analysis to segment students based on self-reported emotionally intelligent
leadership behaviors. Journal of Leadership Education, 10(2): 72-96. DOI: 10.12806/v10/i2/rf4
Fatimah, U.U., Boo, H., Sambasivan, M., & Salleh, R. 2011. Foodservice hygine factor: The consumer perspective. Interna-
tional Journal of Hospitality Management, 30: 38–45.
Fernandez, D.G. 2002. Chinese food in the Philippines. In D.Y. Wu & S.C. Cheung (Eds.), The globalization of Chinese food,
Honolulu, HI: University of Hawai'i Press.
Foxall, G. 2010. Interpreting consumer choice. New York, NY: Routledge.
Gu, C., & Kim, Y.J. 2002. Penalized likelihood regression: General formulation and ef icient approximation. Canadian Jour-
nal of Statistics, 30(4): 619-628. DOI: 10.2307/3316100
Haas, T.J. 2008. Today's foodservice industry demands strong leadership. New York, NY: Business Publications.
Hair, J., Black, W., Babin, B., & Anderson, R. 2014. Multivariate data analyis. Harlow, ESX: Peason Education Limited:
Pearson New International Edition.
HRI Food Service Sector. 2015. Global agricultural information network report. Manila, PH: Philippines Food Service
Hotel Restaurant Institutional.
Hu, M. 2009. Developing a core competency model of innovative culinary development. International Journal of Hospital-
ity Management, 6(4): 63-87. DOI: 10.1016/j.ijhm.2009.10.024
International Markets Bureau. 2010. Market analysis report. URL: goo.gl/Wy5zXb. Last accessed on 19 June 2017.
Intan, W.S. 2016. The analysis factors of experential marketing, product quality, and customer satisfaction of motor bike
as a main transportation mode in bandung-indonesia. International Journal of Business and Administrative Studies,
2(1): 6-8. DOI: 10.20469/ijbas.2.10002-1
Jones, R.H. 2011. Bayesian information criterion for longitudinal and clustered data. Statistics in Medicine, 30(25): 3050-
3056. DOI: 10.1002/sim.4323
Kuo, M.P., & Chen, Y.M. 2015. A Study on the relationships among body sensory experience, customer satisfaction and cus-
tomer loyalty-beauty SPA center as an example. International Journal of Business & Administrative Studies, 1(2):
61-67. DOI: 10.20469/ijbas.10003-2
Orme, B. 2010. Getting started with conjoint analysis: Strategies for product design and pricing research. California,
CA: Research Publishers.
Kafel, P., & Sikora, T. 2013. The usage of quality management methods and tools in food. Food Science Technology Quality,
1(86): 204-216.
ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862
233 J. Admin. Bus. Stud. 2017

Kang, S., Nobuyuki, O., & Herbert, D. 2004. Service quality and its effects on customer satisfaction and customer behavioural
intention: Hotel and ryokan guests in Japan. Asian Paci ic Journal of Tourism Research, 9(2): 189-203.
Kaufmann, L., & Rousseeuw, P.J. 1990. Finding groups in data. New York, NY: John Wiley & Sons, Inc.
Kotler, P. 2013. Principles of marketing. Harlow, ESX: Prentice Hall.
Olsen, S. 2002. Comparative evaluation and the relationship between quality, satisfaction, and repurchase loyalty. Academy
of Marketing Science, 30(3): 240-249. DOI: 10.1108/mbe.2002.26706dae.004
Patayon, E.C. 2008. Parental preferences concerning preschools. Unpublished doctoral dissertation, Ateneo de Davao Uni-
versity, Davao, PH.
Pucciarelli, D., & Thomas, A. 2011. Determinates that in luence food consumption among older members of a midwest com-
munity. Food and Nutrition Sciences, 2(9): 1018-1026. DOI: 10.4236/fns.2011.29136
Regional Development Council Region XI. 2011. Davao regional development plan 2011-2016. URL: goo.gl/4rR6DJ. Last
accessed on 18 May 2017.
Reynolds, J.S., & Hwang, J. 2006. In luence of age on customer dinning experience factors in US Japanese restaurants. An
International Multidisciplinary Journal of Tourism, 1(2): 29-43.
Riehle, H., Grindy , B., & Stensson, A. 2014. Restaurant industry forecast. Washintong, DC, WA: National Restaurant Asso-
ciation.
Roman, B., & Russell, S. 2009. Southeast asian food and culture. URL: goo.gl/CzYXrj. Last accessed on 20 May 2017.
Slovic, P., Fischhoff, B., & Lichtenstein, S. 1977. Behavioural decision theory. Annual Review of Psychology, 28(1): 1-39.
Soderlund, M., & Ohman, N. 2005. Assessing behaviour before it becomes behaviour: An examination of the role of inten-
tions as a link between satisfaction and repatronising behaviour. International Journal of Service Industry Manage-
ment, 16(2): 169-185. DOI: 10.1108/09564230510592298
Soriano, R. 2002. Customers’ expectations factors in restaurants: The situation in Spain. The International Journal of
Quality & Reliability Management, 19(8/9): 1055-1067.
Saputra, R., & Dewi, C.K. 2016. The impact of brand trust on brand loyalty mediated by customer satisfaction: Case of Tokoba-
gus.com (now OLX.co.id). Journal of Administrative & Business Studies, 1(1): 8-13. DOI: 10.20474/jabs-1.1.2
Temblor, S.B. 2016. News section. URL: goo.gl/u2hBnQ. Last accessed on 18 June 2017.
Waldfogel, J. 2008. The median voter and the median consumer: Local private goods and population composition. Journal
of Urban Economics, 63(2): 567-582. DOI: 10.3386/w11972

— This article does not have any appendix. —

ISSN: 2414-309X
DOI: 10.20474/jabs-3.5.2
Electronic copy available at: https://ssrn.com/abstract=3744862