Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship
Gadjah Mada International Journal of Business
September-December 2006, Vol. 8, No. 3, pp. 323–342

AN EXAMINATION OF PERCEIVED
QUALITY, SATISFACTION, AND
LOYALTY RELATIONSHIP
Applicability of Comparative and
Noncomparative Evaluation
Licen Indahwati Darsono
C. Marliana Junaedi
This paper evaluates alternative measurement approaches to
investigating the relationship amongst perceived quality, customer
satisfaction, and loyalty. The authors define and measure the constructs within a relative attitudinal framework and compare these
results to a noncomparative or individual evaluation of brands. In
addition, loyalty is measured by and defined as a behavioral and
attitudinal loyalty. The object of this research is teh botol whilst the
subjects are undergraduate students in Surabaya and Yogyakarta.
The proposed model, with satisfaction as a mediator between
quality and loyalty, is found to be an acceptable representation of
data across three brands of teh botol and for both comparative and
noncomparative evaluation. The use of relative attitude, however,
indicates a much stronger relationship amongst perceived quality,
satisfaction, and loyalty than the attitude towards a brand when they
are performed in individual evaluation. With respect to predictive
ability, this study’s findings suggest that comparative evaluation has
higher degree of applicability than does noncomparative evaluation.
Keywords: comparative and noncomparative evaluation; loyalty; perceived quality; satisfaction

323

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

Introduction
The central thrust of the marketing activities of a firm is often viewed
in terms of development, maintenance,
or enhancement of customer loyalty.
Indeed, customer loyalty represents an
important basis for developing sustainable competitive advantage (Dick
and Basu 1994). A loyal customer base
can be a valuable asset for any firm,
which reduces the need to seek new
customers and may act as a barrier to
switching behavior (Ruyter and
Bloemer 1999; Rowley and Dawes
2000). Accordingly, there is substantial ground to believe that there is a
link between loyalty and profitability
(Hallowell 1996; Rowley and Dawes
1999).
Both perceived quality and customer satisfaction are two central issues in promoting customer loyalty.
Many firms allocate substantial resources to measure and monitor quality, satisfaction, and loyalty in order to
retain customers and improve performance. Perceived quality is the antecedent of satisfaction (Szymanski and
Henard 2001) where higher quality
leads to higher customers satisfaction.
Hallowel (1996); Bloemer and Ruyter
(1998); Mittal and Lassar (1998);
Kandampuly and Suhartanto (2000);
Bowen and Chen (2001) report that
customer satisfaction has a positive
relationship with customer loyalty.
Hence, the antecedent and outcome of
satisfaction are perceived quality and
loyalty.

324

Studies on perceived quality and
satisfaction are more focused on how
quality and satisfaction can be improved (Mittal and Lassar 1998), customer satisfaction-purchase intension
and behavioral relationship (Mittal and
Kamakura 2001), customer satisfaction-loyalty relationship (Hallowel
1996; Oliver 1999; Kandampully and
Suhartanto 2000; Bowen and Chen
2001). Despite the extensive discussion of the perceived quality and satisfaction constructs, only a few empirical studies have tested perceived quality-satisfaction-loyalty relationship
simultaneously (Zeithaml 2000; Olsen
2002). It motivates the authors to simultaneously investigate perceived
quality-satisfaction-loyalty relationship.
However, previous studies find
contradictory results. Satisfied customers are not always loyal (Rowley and
Dawes 2000), and the satisfaction-loyalty relationship is not necessarily linear (Soderlund 1998). Parallel with
these arguments, Jones and Sasser
(1995); Mittal and Lassar (1998) document that satisfied customers are still
defective. Meta-analysis towards 50
empirical studies on customer satisfaction finds that the strength of the
relationship between antecedents and
consequences of satisfaction are moderated by measurement and research
methods (Symanski and Henard 2001).
Therefore, the authors are more focused on the measurement issues of
perceived quality, satisfaction, and loyalty.

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

Dick and Basu (1994) propose
that there are advantages to comparing
brands viewed by customers to be relevant in a given consumption context.
They suggest that using a relative attitude terminology in measuring customer loyalty is likely to provide stronger indication of repeat patronage than
the individual evaluation of a product.
Dick and Basu’s (1994) argument is
supported by Laroche et al. (1994).
They report that a potential customer
always conducts comparative evaluation towards a brand when she is in
stage of attitude formation. Hence, relative attitude conceptualization is used
to measure customer loyalty, and both
individual (noncomparative) evaluation and comparative evaluation are
used to measure perceived quality and
satisfaction. The purpose of this research is to compare the predictive
power of comparative evaluation versus noncomparative evaluation in investigating perceived quality-satisfaction-loyalty relationship.
This research differs from previous research (Olsen 2002) with respect to loyalty measurement. Olsen
(2002) uses behavioral measures of
repurchase loyalty, including the proportion of purchase, the purchase sequence, and the probability of purchase. However, this research uses both
attitudinal and behavioral measures.
Lau and Lee (1999) argue that attitudinal measures are much more important and beneficial, provided that the
attitude basically drives behavior.
Whilst behavioral loyalty is partly determined by situational factors, atti-

tudes are more enduring. O’Malley
(1998) states that attitudinal measures
are a poor predictor of behavior. Therefore, it seems that neither attitudinal
nor behavioral measure itself is sufficient to explain or define loyalty. The
use of both attitudinal and behavioral
measures in loyalty definition substantially increases the predictive power of
loyalty (Pritchard and Howard 1997 as
cited by Bowen and Chen 2001). This
research also differs from Olsen’s
(2002) study on research setting. Olsen
(2002) uses people responsible for preparing the family dinner in the household as the subjects and seafood products as the research objects. Meanwhile, this research uses people who
ever drink three brands of teh botol as
the subjects and teh botol per se as the
research object.

Research Model and
Hypothesis
Perceived Quality
Quality can be defined broadly as
superiority or excellence. Specifically,
perceived quality can be defined as
customer’s judgement about product’s
overall excellence or superiority
(Zeithaml 1988). Zeithaml (1988) argues that there are two forms of quality: (1) objective quality and (2) perceived quality. Objective quality refers to the actual technical superiority
or excellence of the products. In this
way, objective quality can be verified
using some predetermined measurable
ideal standard or standards. Concern
325

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

centres on the selection of attributes
and weights to measure objective quality since researchers and experts do
not agree on what the ideal standard or
standards should be. Maynes (1976) as
cited by Zeithaml (1988) claims that
objective quality does not exist, and
that all quality evaluation is subjective. This argument supports the second form of quality, perceived quality.
Zeithaml (1988) defines perceived
quality as consumer’s judgement about
a product’s overall excellence or superiority.
The authors have the same argument as Zeithaml’s (1988). Therefore,
the term of quality in this research
refers to perceived quality. Perceived
quality is defined and measured as a
statement of belief or attribute performance (Churchil and Surprenant 1982
as cited by Olsen 2002).

Customer Satisfaction
The central theme in the satisfaction definition is the expectancydisconfirmation paradigm. According
to this paradigm, customers form expectations, which act as a standard
against which product or service performance will be judged. A comparison of expectations and perceptions
will result in either confirmation or
disconfirmation. Customer’s expectations are confirmed when the product
or service perceptions exactly meet
expectations. Disconfirmation will be
the result of a discrepancy between the
expectations and the perceptions. Two
types of disconfirmation can be identi326

fied: (1) positive disconfirmation occurs when product performance exceeds prior expectations and (2) negative disconfirmation prevails when
expectations exceed performance.
Confirmation and positive disconfirmation will be likely to result in
satisfaction, whereas negative disconfirmation leads to dissatisfaction.
Therefore, Oliver (1996) as cited by
Ruyter and Bloemer (1999) reveals
that satisfaction is perceived to be a
post-consumption evaluation or a pleasurable level of consumption-related
fulfillment.
The authors have the same argument as that of Oliver (1996). Accordingly, satisfaction in this research is
based on Oliver’s (1994) conceptualization, by which satisfaction is defined as pleasurable fulfillment, meaning that consumption fulfils some
needs, desires, goals, or so forth of
customers and this fulfillment is pleasurable. With respect to this definition, satisfaction reflects the impact of
product or service performance on
people’s feeling state (Rosenberg 1960
as cited by Olsen 2002). The customers feel that the level of fulfillment is
pleasant or unpleasant.

Perceived Quality and Customer
Satisfaction
High correlation between perceived quality and customer satisfaction raises contrary arguments amongst
researchers about the order of occurrence between perceived quality and
satisfaction (Parasuraman et al. 1994).

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

However, most marketing researchers
accept a theoretical framework in
which perceived quality leads to satisfaction, especially when perceived
quality is framed as a specific belief
evaluation and satisfaction as a more
general evaluative construct (Gotlieb
et al. 1994; Johnson et al. 2001 as cited
by Olsen 2002). As a result, satisfaction is one facet of a global affective
evaluation that can be predicted from
quality belief as a cognitive component of the evaluation. Johnson (1998)
as cited by Szymanski and Henard
(2001) states that a customer is more
likely satisfied with marketer’s offerings when they have higher capabilities of fulfilling customers’ needs and
wants. Thus, perceived quality is the
antecedent of customer satisfaction and
has a positive impact on customer satisfaction.

Loyalty
The loyal object might be a brand,
product, organization, or service outlet. However, much of the literature on
customer loyalty has looked at brand
loyalty (Rowley and Dawes 1999).
Dharmmesta (1999) points out that
brand is often viewed as a loyal object,
given that it represents a product or
firm identity easily recognized by customers. With respect to this situation,
customer loyalty and brand loyalty
terminologies in this article are of no
difference and can be used interchangeably.

Attempting to define loyalty is
much easier to be said than done. In
general, there are three distinctive approaches to measuring loyalty:
1. Behavioral measurement,
2. Attitudinal measurement, and
3. Composite measurement.
The behavioral measurement considers consistent, repetitious purchase
behavior as an indicator of loyalty.
One problem with the behavioral measures is that repeat purchases are not
necessarily the results of psychological commitment towards the brand
(TePeci 1999 as cited by Bowen and
Chen 2001). Hence, repeat purchases
do not always mean commitment.
Attitudinal measurement uses attitudinal data to reflect the emotional
and psychological attachment inherent in loyalty. The attitudinal measurement is concerned about the sense of
loyalty, engagement, and allegiance
(Bowen and Chen 2001).
The third approach, composite
measurement of loyalty, combines the
first two dimensions and measures loyalty by customer’s product preferences,
propensity to switch, frequency of
purchase, recency of purchase, and
total amount of purchase. The use of
both attitudinal and behavioral measurements in loyalty definition substantially increases the predictive
power of loyalty (Bowen and Chen
2001). The composite approach has
been applied and supported by several
researchers. Mowen and Minor as cited

327

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

by Dharmmesta (1999) define brand
loyalty as consisting of repeat purchases prompted by a strong positive
attitude and commitment towards the
brand. Parallel with Mowen and Minor, Oliver defines brand loyalty as a
deeply held commitment to rebuying
or repatronizing a preferred product/
service consistently in the future,
thereby causing repetitive same-brand
or same brand-set purchasing, despite
situational influences and marketing
efforts having the potential to cause
switching behavior. Dick and Basu
(1994) define loyalty as the strength of
the relationship between individual’s
relative attitude and repeat patronage.
The primary contribution of Dick
and Basu (1994) lies in relative attitude. Not only does the relative attitude focus on attitude towards the
brand, but it also incorporates attitudes
to alternatives. This encapsulates not
only satisfaction measures but also
preference measures. The suggestion
is that the higher the relative attitude
amongst alternatives, the more likely
the attitude will influence behavior.
Dick and Basu (1994) argue that the
nature of relative attitude is likely to
provide a stronger indication of repeat
patronage than of attitude towards a
brand determined in isolation. Dick
and Basu’s (1994) argument for using
relative attitude is supported by
Laroche et al. (1994); Olsen (2002).
They report that a potential customer
always conducts comparative evaluation towards a brand when she is in
stage of attitude formation.
328

Customer Satisfaction and
Loyalty
The satisfaction-loyalty relationship has been much debated in the
literature. Strauss and Neuhaus (1997);
Rowley and Dawes (1999) propose
that the relationship between satisfaction and loyalty is nonlinear, resulting
from the tendency to remain loyal in
spite of the pressure of switching incentives. The role of satisfaction is not
clear as a number of customers who
express satisfaction still defect, and
dissatisfied customers stay loyal.
Hallowell (1996); Soderlund (1998);
Ruyter and Bloemer (1999)
Kandampully and Suhartanto (2000)
state that increasing satisfaction does
not produce an equal increase in loyalty for all customers. Oliver (1999)
offers a possible explanation for the
nonlinear relationship of satisfaction
and loyalty. Satisfaction still has an
important role as the antecedent of
loyalty; however, its role is limited in
affective loyalty. Oliver (1999) proposes that customers first become loyal
in the cognitive sense, then later in
affective sense, and still later in a conative, action, or behavioral manner.
Affective loyalty is based on qualitybased satisfaction or price-based satisfaction. Quality-based and price-based
satisfaction still has the vulnerability
to do brand switching. The vulnerability to do brand switching is caused by
customer dissatisfaction towards focal
brand, other brand’s persuasive efforts,
and the customer’s desire to seek variety and try other brands. Conclusively,

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

the outcome of satisfaction is loyalty,
and satisfaction per se has a positive
impact on loyalty.

Perceived Quality, Satisfaction,
and Loyalty
Regarding the traditional attitude
theory (Fishbein and Ajzen 1975), perceived quality is defined as an evaluation on product performance. This
evaluation hinges on customer’s cognitive process. In other words, perceived quality represents cognitive
component of evaluation. Satisfaction
reflects the customer’s affective evaluation on the impact of product performance on their feeling state. Hence, if
the authors suggest that perceived quality is an evaluation on or appraisal of
attribute performance, and that satisfaction reflects the impact of the performance on customer’s feeling state,
then perceived quality can be utilized
to predict customer’s feelings (satisfaction) or buying behavior.
Previous studies on loyalty (Jones
and Sasser 1995; Hallowel 1996; Mittal
and Lassar 1998; Oliver 1999;
Kandampully and Suhartanto 2000;
Bowen and Chen 2001) document that
the outcome of satisfaction is loyalty.
The authors accordingly propose a
model to be structured in such a way

that satisfaction acts as a mediator
between perceived quality and loyalty
(see Figure 1). This assumption is also
in accordance with the cognition-affect-behavior paradigm (Fishbein and
Ajzen 1995) and the loyalty phase
framework (Oliver 1999).
Dick and Basu (1994) suggest that
relative attitude (comparative evaluation) is likely to provide a stronger
indication of repeat patronage than
attitude towards a brand measured in
isolation. Laroche et al. (1994) find
that a potential customer always does
comparative evaluation towards a
brand when she is in stage of attitude
formation. Van den Putte et al. (1996)
as cited by Olsen (2002) argue that
better prediction of behavior is the
result of framing attitudes and other
constructs as a direct comparison across
relevant alternatives. Therefore, our
hypothesis is:
The strength of relationship
amongst perceived quality and satisfaction and loyalty will be greater
when perceived quality and satisfaction are defined and measured
as comparative evaluation of several brand alternatives functionally substitutable rather than as a
noncomparative or individual
evaluation of brand.

Figure 1. Research Model
Perceived
Quality



Satisfaction



Loyalty

329

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

Method
Objects, Subjects, and Procedures
Three brands of teh botol: Sosro,
Fruitea, and Frestea, are selected as
research objects with the consideration
that preferences and perceived quality
of Sosro, Fruitea, and Frestea seem to
vary. The subjects of this research are
undergraduate students in Surabaya
and Yogyakarta who ever consume
Sosro, Fruit Tea, and Frestea.
The first procedure was to divide
respondents into two groups. One
group received the measuring instrument requiring them to rate all attributes of a brand before moving to
the next brand. This noncomparative
format is so-called the “entity-based
format” by Teas and Wong (1992).
The second group received the instrument that required them to rate all the
three brands on one attribute before
moving to the next attribute. This comparative format is also termed the “attribute-based format” (Teas and Wong
1992).

Questionnaires and Format
Two versions of questionnaire
were developed. The questions and
their order in the two versions are of no
difference. The question order is started
with general, global questions, satisfaction, then followed by more specific questions, perceived quality
evaluation and loyalty (Malhotra
2002). The final section of the questionnaires contains general informa-

330

tion on respondents. The only difference between the two questionnaires
is the comparative and noncomparative format. The noncomparative format has one page with all the satisfaction and perceived quality items for
one brand, followed by the next brand,
and then the third. On the other hand,
the comparative format has one item
(satisfaction items precede perceived
quality attribute performance) for all
the three brands such that all of them
are salient for comparative evaluation.
The order of the items is the same for
all the three brands in the noncomparative format, and the brands are in the
same sequence as they are in the comparative format.

Measurement of the Constructs
Perceived quality. Customer’s
evaluation of perceived quality is defined as and measured by the evaluation of attribute performance (Churchill
and Suprenant 1982 as cited by Olsen
2002), associated with three different
brands in this research. In a pilot study
employing 40 respondents, the authors
investigate positive and negative associations with quality of teh botol. High
quality is associated with nice scent,
sweet taste, and clean packaging.
Meanwhile, low quality is associated
with bad scent, bitter taste, and dirty
packaging. All items are measured with
a seven-point semantic differential
scale, anchored by bad scent/nice scent,
bitter taste/sweet taste, and dirty packaging/clean packaging.

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

Satisfaction. Following Oliver’s
(1999) conceptualization, satisfaction
is defined as and measured by the
global evaluation of feeling state. The
respondents are asked to indicate their
overall feeling towards three different
brands on a seven-point semantic differential scale anchored by dissatisfied/satisfied and unpleasant/pleasant
states.
Loyalty. Predicated on Dick and
Basu’s (1994) conceptualization, loyalty is defined as and measured by
attitudinal and behavioral dimensions.
Attitudinal dimension is measured by
commitment and intention whilst behavioral dimension is measured by
purchase frequency and proportion of
purchase. Three items are adapted from
Lau and Lee (1999), two items are
adapted from Pritchard et al. (1999).
All items are measured using a sevenpoint interval scale.

Results
Respondents Profile
Of the 400 questionnaires distributed, 366 questionnaires were returned.
After a careful examination, including
a check on incomplete responses,
monotonic answers, and outliers, only
173 responses (noncomparative evaluation) were usable and 163 responses
(comparative evaluation) were usable.
Amongst the respondents, 56.4
percent of respondents have pocket
money below IDR500,000, and approximately 34 percent have around
IDR500,000-1,000,000, 9 percent have

more than IDR1,000,000. The youngest respondent was 18 years old, and
the oldest was 25 years old; most of
them were ranging from 20 to 22 years
old.

Validation of Measures:
Reliability and Validity
First, the items measuring perceived quality, satisfaction, and loyalty are subject to a confirmatory factor analysis (CFA) to assess fit statistics (see Table 1). All of the chi-square
values are significant; however, CMIN/
DF value is below the recommended
level of 5 (Hair et al. 1998; Kenny
2002). The GFI and CFI values, however, are between 0.837 and 0.963
across brands and evaluation formats,
far above the suggested fit level of
0.90, and are considered acceptable
(Hair et al. 1998; Kenny 2002). The
other fit indices (AGFI and RMSEA)
represent the marginal fit. Results in
Table 1 imply that a two-factor solution is a reasonable representation of
three brands and of comparative and
noncomparative evaluation.
Additional measurement properties are presented in Table 2. Reliability is assessed utilizing construct reliability and Cronbach’s alpha. Most of
the investigated constructs exhibit construct reliability and alpha values of
greater than 0.60 (see Table 2), suggesting reliability that exceeds the common acceptable level (Hair et al. 1998).
The factor loadings (lambdas) are all
reasonably high (above 0.4) and significant. All measures of variance ex331

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

Table 1 Fit Statistics of Measurement Model
Sosro

Quality < > Satisfaction
Quality < > Loyalty
Satisfaction< > Loyalty
2
Degree of freedom (df)
Pvalue
CMIN/df
GFI
AGFI
CFI
RMSEA

Frestea

Compaparative

Non- Compacompa- rative
rative

Noncomparative

Comparative

Noncomparative

0.864
0.563
0.635
95.024
32
.000
2.97
0.887
0.806
0.942
0.110

0.709
0.564
0.695
65.950
32
.000
2.061
0.930
0.879
0.963
0.079

0.584
0.428
0.572
70.706
32
.000
2.210
0.923
0.868
0.961
0.084

0.574
0.292
0.640
148.908
32
.000
4.653
0.837
0.720
0.891
0.150

0.402
0.223
0.564
146.498
32
.000
4.578
0.843
0.730
0.906
0.144

tracted exceed 0.5, with the exception
for perceived quality. Hence, it can be
concluded that the measurement model
is substantially valid and reliable.

Hypothesis Testing Results
The proposed research model,
with satisfaction as a mediator between perceived quality and loyalty,
provides an acceptable fit across all
brands and for both comparative and
noncomparative evaluation (see Table
3). Three fit indices (GFI, AGFI, and
CFI) are above 0.991, and RMSEA
value is between 0.085 and 1, whilst all
chi-square values are insignificant.
Accordingly, the authors conclude that
the research model is similar and stable
across brands and evaluation formats,
providing us with a good basis for

332

Fruitea

0.859
0.555
0.638
80.700
32
.000
2.522
0.914
0.852
0.947
0.092

comparing both the relationship
amongst the constructs and the model’s
predictive ability or explanatory power.
Table 3 indicates a relatively
strong relationship between perceived
quality and satisfaction, but this relationship varies amongst brands. Therefore, the findings substantiate
Szymanski and Henard’s (2001),
Olsen’s (2002) studies. The perceived
quality-satisfaction relationship is
much stronger for comparative evaluation than it is for noncomparative
evaluation. Standardized path coefficients for comparative evaluation vary
from 0.507 (Frestea) to 0.790 (Sosro)
compared to from 0.329 (Frestea) to
0.642 (Sosro) when the three brands
are framed in noncomparative evaluation.

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

Table 2. Scale Items, Factor Loadings, and Constructs Reliability
Constructs
and
Indicator

Comparative Evaluation

Noncomparative Evaluation

Standar- Cronbach Compo- Variance Standar- Cronbach Compo- Variance
dized
Alpha
site Extracted dized
Alpha
site Extracted
Loadings
ReliabiLoadings
Reliability
lity
Perceived
quality:
Sosro
K1
K2
K3

0.667
0.725
0.614

Fruitea
K1
K2
K3

0.641
0.741
0.490

Frestea
K1
K2
K3

0.6998

0.4490

0.6608

0.4003

0.6953

0.434

0.604
0.745
0.619

Fruitea
S1
S2

0.926
0.895

0.442

0.7015

0.7234

0.478

0.7885

0.792

0.5737

0.8949

0.8953

0.8105

0.9202

0.9204

0.8525

0.9323

0.9319

0.8725

0.721
0.945
0.556
0.9212

0.9253

0.861
0.919
0.881

0.9048

0.9065

0.829
0.938
0.909

0.9176
0.890
0.955

0.6915

0.785
0.781
0.457
0.6904

0.933
0.923

0.6653
0.652
0.836
0.449

0.6482

Satisfaction:
Sosro
S1
S2

Frestea
S1
S2

0.7089

0.9200

0.8520
0.939
0.929

333

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

Continued from Table 2
Constructs
and
Indicator

Comparative Evaluation

Noncomparative Evaluation

Standar- Cronbach Compo- Variance Standar- Cronbach Compo- Variance
dized
Alpha
site Extracted dized
Alpha
site Extracted
Loadings
ReliabiLoadings
Reliability
lity
Loyalty:
Sosro
L1
L2
L3
L4
L5

0.8956

0.8903

0.581
0.852
0.758
0.909
0.782

Frestea
L1
L2
L3
L4
L5

0.637
0.721
0.794
0.935
0.938

0.8684

0.8720

0.5850

0.9006

0.9008

0.6472

0.9148

0.9084

0.7036

0.527
0.687
0.742
0.892
0.911
0.8856

Fruitea
L1
L2
L3
L4
L5

0.8868

0.6154
0.683
0.722
0.843
0.869
0.885

0.9116

0.9054

The relationship between satisfaction and loyalty is also significant
and positive across brands. The standardized path coefficients for comparative evaluation are much stronger
than those for noncomparative evaluation, especially for Fruitea and Frestea,
ranging 0.642-0.708 compared to
0.607-0.608. Curiously, the standardized path coefficients of Sosro for
334

0.6272

0.576
0.682
0.729
0.941
0.961

0.6616
0.630
0.770
0.763
0.948
0.937

noncomparative evaluation (0.714) are
much stronger than those for comparative evaluation (0.690). Aggregately,
the standardized path coefficients for
comparative evaluation vary from
0.642 (Fruitea) to 0.708 (Frestea) compared to 0.607 (Fruitea) to 0.714
(Sosro) when the brands are framed in
noncomparative evaluations. If the
comparison is analyzed more cau-

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

Table 3. Key Parameters of Research Model
Sosro
Parameter
Standardized
path:
Q  S (t value)
S  L (t value)

Goodness of fit:
2
df
P value
GFI
AGFI
CFI
RMSEA

Fruitea

Frestea

Comparative

NonCompaCompa- rative
rative

NonComparative

Comparative

NonComparative

0.790
(12.743)

0.642
(9.250)

0.734
(11.115)

0.570
(7.821)

0.507
(6.523)

0.329
(4.062)

0.690
(10.068)

0.714
(10.926)

0.642
(8.987)

0.607
(8.518)

0.708
(10.435)

0.608
(8.522)

0.332
1
0.564
0.999
0.992
1
0

0.911
1
0.340
0.996
0.979
1
0

0.024
1
0.876
1
1
1
0

2.234
1
0.135
0.991
0.949
0.989
0.085

0.959
1
0.327
0.996
0.977
1
0

0.002
1
0.96
1
1
1
0

tiously, the lower limit value difference (0.035) is greater than the upper
limit value difference (0.006). The findings lead to a conclusion that comparative evaluation has relatively higher
predictive power than does noncomparative evaluation. Hence, the hypothesis that the strength of relationship amongst perceived quality and
satisfaction and loyalty is greater when
perceived quality and satisfaction are
defined as and measured by comparative evaluation rather than noncomparative or individual evaluation of
brand is satisfactorily substantiated.

Discussion
Perceived Quality-SatisfactionLoyalty Relationship
As seen in Table 3, a highly positive relationship exists between perceived quality and satisfaction. Therefore, the findings of this research support the arguments that people form
their attitudes towards the performance
of products and brands by learning
about the different characteristics of
the objects and integrating these values into more global affective evalua-

335

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

tion. This affective evaluation (satisfaction) is used as a predisposition to
compare alternatives and guide final
choice and loyalty. This mechanism is
also called “belief comparison model”
since the objects, three brands of teh
botol, are highly comparable
(Dabholkar 1994). Implicitly, the findings and theoretical arguments conclude that satisfaction is the mediator
of perceived quality-loyalty relationship.
The correlation values between
perceived quality and loyalty across
brands are lower than those between
satisfaction and loyalty (see Table 1).
It indicates the role of satisfaction as
the mediator of perceived quality-loyalty relationship. (Dabholkar et al. 2000
as cited by Olsen 2002). The proposed
research model, with satisfaction as
the mediator between perceived quality and loyalty, provides an acceptable
fit across brands and for both comparative and noncomparative evaluation. Hence, the result supports
Dabholkar’s statement that satisfaction acts as a mediator between perceived quality and loyalty.
Dabholkar (1994) reports that
people are inclined to make comparison when they have more alternatives.
Laroche et al. (1994) also find that
people always make comparison when
they form their attitudes toward brands.
As a result, Dabholkar’s (1994);
Laroche et al. (1994) studies are supported, regarding that the perceived
quality-satisfaction-loyalty relationship gets stronger when perceived qual-

336

ity and satisfaction are measured by
comparative evaluation. Thus, perceived quality and satisfaction are better defined as and measured by relative
attitudinal framework.
Olsen (2002) finds that satisfaction-loyalty relationship is weaker than
perceived quality-satisfaction relationship. However, this research finds contrary results. There are two possible
explanations for these contradictory
results. First, satisfaction/dissatisfaction is strongly related with specific
characteristics of customers. Different
characters lead to different satisfaction threshold. Unsatisfactory personal
outcomes may be due to factors intertwined with the customer’s specific
characteristics, and the customer may
still rate the brand high or low on
quality (Mittal and Lassar 1998; Mittal
and Kamakura 2001). Second, on average, variance extracted of perceived
quality is lower than that of satisfaction. Low variance extracted of perceived quality indicates that the measurement items cannot fully explain
the perceived quality. When a construct with low variance extracted is
related to another construct with high
variance extracted, the relationship
between them tends to be weaker.
Therefore, the relationship between
perceived quality (low variance extracted) and satisfaction (high variance extracted) is weaker than the relationship between satisfaction and loyalty, regarding that satisfaction and
loyalty have high variances extracted.

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

The other differences from previous study (Olsen 2002) are respect to
the composite reliability and the variance extracted. Olsen (2002) finds that
composite reliability and variance extracted are systematically higher for
all constructs and across brands when
the constructs are framed in noncomparative evaluation. Olsen (2002) argues that it can be accounted by the
fact that noncomparative evaluation
may produce a more dimensional similarity halo compared to comparative
evaluation. This research basically
finds the same result; however, the
result does not hold across brands. As
seen in Table 2, the composite reliability and the variance extracted of Sosro
are systematically higher for comparative evaluation compared to those for
noncomparative evaluation. The authors do not have any appropriate reason for this contrary result. However,
it is suspected that Sosro’s position as
a pioneer brand for teh botol must be
taken into account. Kardes and
Kalyanaram (1992) as cited by Alpert
and Kamis (1995) show that customers learn more about a pioneer than
about follower brands. This is because
the pioneer, as the first entrant in the
product category, has features perceived by the customers to be novel
and attention-drawing. Hence, there
will be more extreme (but generally
positive) and confidently held attitudes
towards and belief in the pioneer than
in the follower brands. Parallel with
Alpert and Kamis (1995), Muafi (2001)
also finds that pioneer brands are more
likely to be retrieved by customers.

Limitations and Implications for
Future Research
Even though this research finds
that comparative evaluation has relatively higher predictive power than
does noncomparative evaluation, one
important problem must be noted. The
predictive ability in this research is
evaluated and compared without taking into account how the brands are
framed in the evaluative process, starting from belief (cognitive evaluation)
in brand attributes, then integrating
them into global affective evaluation.
Therefore, future research must emphasize the evaluative process.
Whilst answering the questionnaires, the respondents responded in
the basis of whatever materials coming to mind at that time. This is consistent with recent attitude theory suggesting that people do not respond to
survey questions on the basis of a
single, fixed set of psychological consideration. Lynch et al. (1991) as cited
by Laroche et al. (1994) argue that the
evaluation is relative, and highly affected by the context surrounding the
people. Therefore, the responses tend
to count on whatever materials coming to mind at the moment of answering. Furthermore, the particular materials that come to mind depend on the
nature of the questions and the manner
by which they are posed. As a consequence, perceived quality and/or satisfaction may fail to fully account for the
relationship amongst these constructs
and their ability to predict or explain
loyalty. It appears that the two con337

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

trary results with Olsen’s (2002) study
may be owing to the measurement
problem of perceived quality. Hence,
measurement issues are a central problem in research on perceived quality
and satisfaction. Future research must
pay a closer attention to: (1) salient
attributes for evaluating perceived
quality, (2) the sequence of questions
on account of question-order effect
(Malhotra 2002), (3) measurement
scaling provided that many alternatives such as ranking, paired comparison, constant sum scales, etc. can be
utilized for comparative evaluation. It
gives a potential to replicate this research with different comparative scaling.
With respect to the respondents of
this research, the authors suggest the
reader to interpret and generalize the
results cautiously. Future research
should examine and compare perceived
quality-satisfaction-loyalty model with
different samples and settings.

Conclusion and Managerial
Implications
Comparative evaluation has better predictive ability than does non-

comparative evaluation. This finding
supports the idea that people form their
attitudes towards the performance of
brands by learning about different characteristics of the objects and integrating them into global affective evaluation. This affective evaluation (satisfaction) is used as a predisposition to
compare alternatives and guide final
choice and loyalty. Consequently, to
create loyal customers, movement
along the means-end chain of performance quality, satisfaction, and loyalty is of importance, and the intervention point for marketers is the performance quality of their offerings.
Two managerial implications can
be derived from this research. First,
quality performance and satisfaction
should be measured by comparative or
relative attitude towards products and
services functionally substitutable subsequent to better predictive power of
comparative evaluation. Second, marketers must take into account the performance quality of their offerings to
be an important consideration since
quality is an important starting point of
creating loyal customers.

References
Alpert, F. H., and M. A. Kamis. 1995. An empirical investigation of consumer memory,
attitude, and perceptions toward pioneer and follower brands. Journal of Marketing
59 (October): 34-45.
Assael, H. 1998. Consumer Behavior and Marketing Action (6th ed.). Cincinnati, OH:
South-Western College Publishing.
Bloemer, J., and K. Ruyter. 1998. On the relationship between store image, store
satisfaction and store loyalty. European Journal of Marketing 32: 499-513.
338

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship
Bowen, J. T., and S.L. Chen. 2001. The relationship between customer loyalty and
customer satisfaction. International Journal of Contemporary Hospitality Management (May): 213-217.
Dabholkar, P. 1994. Incorporating choice into an attitudinal framework: Analyzing
models of mental comparison processes. Journal of Consumer Research 21 (June):
100-118.
Dick, A. S., & K. Basu. 1994. Customer loyalty: Toward an integrated conceptual
framework. Journal of the Academy of Marketing Science 22 (2): 99-113.
Dharmmesta, B. S. 1999. Loyalitas pelanggan: Sebuah kajian konseptual sebagai panduan
bagi peneliti. Jurnal Ekonomi dan Bisnis Indonesia 14 (3): 73-88.
Fishbein, M., and J. Ajzen. 1975. Belief, Attitude, Intention and Behavior: An Introduction
to Theory and Research. Reading, MA: Addison-Wesley.
Hair, J. F., R. E. Anderson, R. L. Tatham, and W. C. Black. 1998. Multivariate Data
Analysis (5th ed.). Upper Saddle River, NJ: Prentice-Hall International, Inc.
Hallowell, R. 1996. The Relationships of customer satisfaction, customer loyalty, and
profitability: an empirical study. International Journal of Service Industry Management 7 (4): 27-42.
Jones, T. O., and W. E. Sasser. 1995. Why satisfied customers defect. Harvard Business
Review (November-December): 88-99.
Kandampully, J., and D. Suhartanto. 2000. Customer loyalty in the hotel industry: The role
of customer satisfaction and image. International Journal of Contemporary Hospitality Management (December): 346-351.
Kenny, D.A. 1998. Multiple Factor Models. URL: http:/nw3.nai.net/~dakenny/
multiplefactormodel.htm.
________. 2002. Measuring Model Fit. URL: http:/nw3.nai.net/~dakenny/
measuringmodelfit.htm.
Kotler, P. 2000. Marketing Management (Millenium ed.). Upper Saddle River. NJ:
Prentice-Hall, Inc.
Laroche, M., M. Hui, and L. Zhou. 1994. A test of the effects of competition on consumer
brand selection processes. Journal of Business Research 31: 171-181.
Lau, G. T., & S. H. Lee. 1999. Consumers’ trust in a brand and the link to brand loyalty.
Journal of Market Focused Management 4: 341-370.
Malhotra, N. 2000. Basic Marketing Research. Englewood Clifs, NJ: Prentice-Hall, Inc.
Mittal, B., and W. M. Lassar. 1998. Why do customers switch? The dynamics of
satisfaction versus loyalty. The Journal of Services Marketing 12 (3): 177-194.
Mittal, V., and W. A. Kamakura. 2001. Satisfaction, repurchase intent, and repurchase
behavior: investigating the moderating effect of customer characteristics. Journal of
Marketing Research 38: 131-142.
Muafi. 2001. Studi empirik pengaruh pioneer brand pada proses pemilihan brand.
Usahawan 30 (8): 3-7.

339

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3
O’Malley, L. 1998. Can loyalty schemes really build loyalty? Marketing Intelligence and
Planning (January): 47-55.
Oliver, R. L. 1999. Whence consumer loyalty. Journal of Marketing 63: 33-44.
Olsen, S. O. 2002. Comparative evaluation and the relationship between quality, satisfaction, and repurchase loyalty. Journal of the Academy of Marketing Science 30 (3):
240-249.
Parasuraman, A., V.A. Zeithaml, and L. L. Berry. 1994. Reassessment of expectations as
a comparison standard in measuring service quality: Implications for further research. Journal of Marketing 58: 111-124.
Pritchard, M. P., M. E. Havitz, and D. R. Howard. 1999. Analyzing the commitmentloyalty link in service contexts. Journal of the Academy of Marketing Science 27 (3):
333-348.
Rowley, J., and J. Dawes. 2000. Disloyalty: a closer look at non-loyals. Journal of
Consumer Marketing 17 (6): 538-549.
Rowley, J., and J. Dawes. 1999. Customer loyalty – a relevant concept for libraries?
Library Management 20 (6): 345-351.
Ruyter, K., and J. Bloemer. 1999. Customer loyalty in extended service settings.
International Journal of Service Industry Management 10 (3): 320-336.
Soderlund, M. 1998. Customer satisfaction and its consequences on customer behavior
revisited. International journal of service industries Management 9 (2): 169-188.
Strauss, B., and P. Neuhaus. 1997. The qualitative satisfaction model. International
Journal of Services Industries Management 8 (3): 236-249.
Syzmanski, D. M., and D. H. Henard. 2001. Customer satisfaction: A meta analysis of the
empirical evidence. Journal of the Academy of Marketing Science 29 (1): 16-35.
Teas, K. R., and J. K. Wong. 1992. Item context and stability of entity-based and attributebased multiattribute scaling methods. Journal of Consumer Research 18: 536-545.
Zeithaml, V. A. 1988. Consumer perceptions of price, quality, and value: A means-end
model and synthesis of evidence. In B. M. Enis, K. K. Cox, and M. P. Mokwa:
Marketing Classics: A Selection of Influential Articles (8th ed.). NJ: Prentice Hall, pp.
471-499.
Zeithaml, V.A. 2000. Service quality, profitability, and the economic worth of customers:
What we know and what we need to learn. Journal of the Academy of Marketing
Science 28 (1): 67-85.

340

Darsono & Junaidi—An Examination of Perceived Quality, Satisfaction, and Loyalty Relationship

APPENDICES
Non-Comparative Evaluation Format
Satisfaction
Would you please express your evaluation and feeling on drinking “Sosro” by
placing a cross (X) on a scale that best reflects your personal feeling
Dissatisfied
1
2
3
4
5
6
7
Satisfied
Unpleasant
1
2
3
4
5
6
7
Pleasant
Etc. for the other brands….
Perceived Quality
The following questions are related to your evaluation on specific attributes of teh
botol “Sosro” For each attribute, please indicate your personal evaluation by
placing a cross (X) on a scale that best reflects your personal feeling
Bad scent
1
2
3
4
5
6
7
Nice scent
Bitter taste
1
2
3
4
5
6
7
Sweet taste
Dirty packaging 1
2
3
4
5
6
7
Clean packaging
Likewise for the other brands….

Comparative Evaluation Format
Satisfaction
How dissatisfied or satisfied you are with the different brands of teh botol by
placing a cross (X) on a scale that best reflects your personal feeling
Sosro
Dissatisfied 1
2
3
4
5
6
7
Satisfied
Fruit Tea Dissatisfied 1
2
3
4
5
6
7
Satisfied
Frestea
Dissatisfied 1
2
3
4
5
6
7
Satisfied
How unpleasant or pleasant you are with the different brands of teh botol by
placing a cross (X) on a scale that best reflects your personal feeling
Sosro
Unpleasant 1
2
3
4
5
6
7
Pleasant
Fruit Tea Unpleasant 1
2
3
4
5
6
7
Pleasant
Frestea
Unpleasant 1
2
3
4
5
6
7
Pleasant

341

Gadjah Mada International Journal of Business, September - December 2006, Vol. 8, No. 3

Perceived Quality
For each attribute, please indicate your personal evaluation by placing a cross (X)
on a scale that best reflects your personal evaluation
Sosro
Bad scent
1
2
3
4
5
6
7
Nice scent
Fruit Tea Bad scent
1
2
3
4
5
6
7
Nice scent
Frestea
Bad scent
1
2
3
4
5
6
7
Nice scent
Likewise for the other attributes….
Loyalty
1. If teh botol “____” is not available in the store when I need it, I will buy “____”
somewhere else.
2. I intend to keep buying teh botol “____” in the future.
3. If I want teh botol, I only buy “____”
4. I have bought teh botol “___” more often than other brands during the last 3
months.
5. I have bought teh botol “___” in a larger number than other brands during the
last 3 months.

342