ARTICLE

Back, Parks
JOURNAL OF/ A BRAND
HOSPIT LOYALTY
ALITY MODELRESEARCH
& TOURISM

A BRAND LOYALTY MODEL

INVOLVING COGNITIVE, AFFECTIVE,
AND CONATIVE BRAND LOYALTY
AND CUSTOMER SATISFACTION
Ki-Joon Back
Kansas State University
Sara C. Parks
Pennsylvania State University

Despite the amount of research focusing on brand loyalty, empirical tests of the relation-
ship between customer satisfaction and brand loyalty have not been conducted. The pri-
mary purpose of this study was to investigate the mediating effects of attitudinal brand loy-
alty on the relationship between customer satisfaction and behavioral brand loyalty. More-
over, we developed a robust brand loyalty measurement in the lodging industry by using at-
titudinal and behavioral brand loyalty constructs. The majority of respondents were busi-
ness travelers who stayed at an upper-middle-class business hotel. The results of this inves-
tigation suggested that customer satisfaction had a significant indirect effect on behavioral
brand loyalty when mediated by attitudinal brand loyalty, including cognitive-affective-
conative brand loyalty stages. Thus, practitioners should consider customers’perceptions
of their brand and not rely solely on purchasing frequencies when measuring brand loyalty
levels.

KEYWORDS: customer satisfaction; attitudinal brand loyalty; behavioral brand loy-

alty; lodging industry; structural equation modeling (SEM)

In recent years, retaining customers in the lodging industry has become

increasingly more important because the industry is very competitive and is in the
maturity stage of its life cycle (Lewis & Chambers, 2000). Numerous practitio-
ners and academics in various industries have studied the benefits of creating and
maintaining brand loyalty with existing customers. In particular, it has been found
that brand loyal customers reduce marketing costs associated with attracting new
customers (Kotler, Bowen, & Makens, 1998). In addition, these customers say
positive things about a company to others (Boulding, Kalra, Staelin, & Zeithaml,
1993; Tepeci, 1999;) and frequently pay premium prices (Bowen & Shoemaker,
1998). Although much of the research documented in hospitality journals has
addressed issues of brand loyalty, little attention has yet been given to the method-
ological issues of brand loyalty, in terms of measurements and antecedents.

Journal of Hospitality & Tourism Research, Vol. 27, No. 4, November 2003, 419-435
© 2003 International Council on Hotel, Restaurant and Institutional Education
419
420 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

This article reviews the underlying structure of brand loyalty by considering

behavioral and attitudinal approaches. Moreover, it investigates the antecedents
of brand loyalty by developing a structural equation model of brand loyalty, based
on Oliver’s (1997) brand loyalty stages and Heskett, Sasser, and Schlesinger’s
(1997) curvilinear relationship between customer satisfaction and repeat pur-
chasing behaviors. Finally, this study presents a robust brand loyalty measure-
ment and suggests implications for future research as well as marketing strategies
for the lodging industry.

RESEARCH QUESTIONS

This research was designed to answer the following questions:

• Is the direct effect of customer satisfaction on brand loyalty significant?

• How can a robust model improve the measurement of brand in the lodging industry?

LITERATURE REVIEW

Key Constructs and Measurements

Customer satisfaction (CS) is as fundamental to the marketing concept as the
notion of satisfying the needs and desires of consumers (Spreng, MacKenzie, &
Olshavsky, 1996). The generally held description of customer satisfaction among
researchers focuses on the consumer’s overall judgment, including service fea-
tures, the service product, sales personnel, or other situational variables. Cus-
tomer satisfaction results when customers either confirm their prepurchase
expectations for a purchased service or positively disconfirm (exceed) their
expectations regarding purchased services, resulting in some level of
postpurchase affect toward the experience (Cardozo, 1965).
Behavioral brand loyalty can be defined as a customer’s overt behavior toward
a specific brand in terms of repeat purchasing patterns. Specifically, a repeat pur-
chasing pattern can be determined as actual purchase frequency, the proportion of
occasions in which a specific brand is purchased as compared to the total number
of purchased brands and/or the actual amount of purchase. Numerous brand loy-
alty researchers have used this behavioral approach by simply measuring those
behavioral variables to predict the customer’s purchasing behavior in the future
(Ehrenberg, 1991; Guadagni & Little, 1983). However, many researchers sug-
gested that measuring behavioral brand loyalty alone may cause several prob-
lems. Dick and Basu (1994) argued that this behavior approach neglects the
importance of the customer’s decision-making process, which does not differen-
tiate brand loyalty from simple repeat purchasing behavior. Hence, none of the
operational definitions in behavioral brand loyalty studies described thus far
exhibit an understanding of the factors underlying and leading up to brand loyal
purchasing (Pritchard, Howard, & Havitz, 1992). Other problems include “(1)
providing arbitrary cutoff criteria; (2) failing to assess the complexity and rich-
ness of brand loyalty; (3) focusing on the outcome of behavior and not developing
 Back, Parks / A BRAND LOYALTY MODEL 421

definitions that reach at the underlying causative factors” (Jacoby & Chestnut,
1978, p. 47).
Numerous researchers have examined the attitudinal aspect of brand loyalty
(Bowen & Shoemaker, 1998; Iwasaki & Havitz, 1998; Jacoby & Olson, 1970;
Jarvis, 1973; McCleary & Weaver, 1992). Attitudinal brand loyalty focuses not
only on transactional strategies, such as frequent-user programs and gifts for
repeat customers but also on attitudinal variables, such as commitment and trust.
Attitudinal studies have described brand loyalty not only as the outcome of repeat
purchase behavior but also the consequence of multidimensional attitudes toward
a specific brand.
As Muncy (1983) stated, most attitudinal measurement has been developed
based on operational definitions rather than a theoretical conceptualization of
brand loyalty; therefore, the attitudinal measurement lacks construct validity. It is
evident that many attitudinal factors are derived from brand loyal consumers’ dis-
positions, such as commitment, involvement, motivation, and other cognitive and
affective variables. However, the process of selecting those variables was opera-
tionally, not theoretically, based. Hence, the risk of low construct validity is due to
the researcher’s inability to assess all these person-specific features.
Several researchers (Bowen & Chen, 2001; Jacoby & Chestnut, 1978; Stern,
1997) discussed the need to combine behavioral and attitudinal aspects of brand
loyalty and develop measures of brand loyalty accordingly. Such studies have
described brand loyalty as not only an outcome of repeat purchase behavior but
also as a consequence of multidimensional cognitive attitudes toward a specific
brand. As Jacoby and Kyner (1973) stated, loyalty is a biased behavior expressed
over time by an individual with respect to one or more alternatives and is a func-
tion of psychological processes. Therefore, neither behavioral measures nor atti-
tudinal/cognitive measures alone are sufficient to assess brand loyalty.
Like other types of measurements, several risks are involved with the measure-
ment, such as improper multiplication of attitudinal and behavioral attributes,
selecting inadequate items, neglecting the impacts of significant intervening vari-
ables, and lack of underlying theoretical supports (Bowen & Shoemaker, 1998;
Pritchard et al. 1992). Therefore, one of the major responsibilities of researchers
is to better understand the relationship between attitudinal and behavioral brand
loyalty. An improved understanding enables them to develop effective brand loy-
alty measurements by establishing a strong conceptual and theoretical foundation
and constructing effective research methodology to refine measurement.

Relationship Between Customer Satisfaction and Brand Loyalty

Numerous researchers have investigated the relationship between CS and
brand loyalty (BL). Traditionally, many researchers have provided empirical evi-
dence of a positive relationship between CS and BL. Bitner (1990) claimed that
CS has an indirect effect on BL, mediated by perceived quality. In addition, BL is
directly influenced by CS. Furthermore, Rust and Zahorik (1993) found a link
between CS and BL using data from the retail bank market and from a national
hotel chain. According to Oliver (1999), CS had a positive effect on attitudes.
422 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

These positive attitudes were found to revise attitudes toward the product or
brand, such as increased level of positive belief (i.e., belief confidence)
(Albarracin & Wyer, 2000), reinforce the level of positive affect (Oliver, 1993),
and enhance repurchase intentions (Yi, 1990).
Many consumer researchers have postulated a significant causal relationship
between CS and behavioral intention, which is consistent with Fishbein’s model
(Fishbein & Ajzen, 1975). Fornell, Johnson, Anderson, Cha, and Bryant (1996)
stated that increased CS also increases brand loyalty in terms of repurchase likeli-
hood and price tolerance given repurchase. Recently, Oliver (1999) argued that
CS is “the beginning of a transitioning sequence that culminates in a loyalty state”
(p. 35). All these studies focused the effects of CS on attitudinal brand loyalty.
They found that CS influences cognitive, affective, and then conative components
of attitudinal brand loyalty, such as purchase intentions and postpurchase
attitudes.
However, the ways in which predictive repurchase behaviors are actually
derived from customer satisfaction are not well understood. Shoemaker and
Lewis (1999) found a weak link between customer satisfaction and brand loyalty
in the casino industry. According to Heskett et al. (1997) only 100% of satisfied
customers become truly brand loyal and have relatively high repurchase rates. In
addition, Bowen and Chen (2001) stated that customers must be extremely satis-
fied to show brand loyalty. Therefore, the extent to which the level of satisfaction
carries over to influence postpurchase attitudes and behaviors is questionable.

Relationship Between Attitudinal and Behavioral Brand Loyalty

To investigate the relationship between attitudinal and behavioral brand loy-
alty, the theory of reasoned action should be considered. Ajzen and Fishbein
(1980) developed this theory to relate customers’ beliefs and attitudes to their
behavioral intentions. This theory assumes that customers carefully process deci-
sion making by considering the consequences of the alternative behaviors and
choosing the one that leads to the most desirable consequences. The result of this
reasoned choice process reveals a behavioral intention to engage in the selected
behavior. Bentler and Speckart (1981) stated that attitudes have causal priority
over behaviors. In addition, Peter and Olson (1993) postulated that a negative
change in attitudes caused many customers to switch to other brands, indicating
that change in attitude is a good predictor of brand loyalty and brand switching.
Accordingly, Oliver (1997) suggested that brand loyalty stages exhibit a learn-
ing process that highlights the relationship between attitude and behavior. First,
he claimed that attitudinal brand loyalty should be viewed as developing in three
phases—cognitive, affect, and conative components of attitudinal brand loyalty.
These three phases are consistent with general definitions of attitude. Social sci-
entists often have assumed that responses that express evaluation and therefore
reveal people’s attitudes should be divided into three classes—cognition, affect,
and conation (or behavioral intention) (Bagozzi, 1978; Breckler, 1984). Cogni-
tion refers to people’s thoughts about the attitude object. It encompasses the con-
 Back, Parks / A BRAND LOYALTY MODEL 423

tent of one’s thoughts regarding beliefs in the statement of fact. Affect refers to
feelings, moods, or emotional responses that can be measured by collecting ver-
bal reports or by physiological responses. These affective responses can range
from extremely positive to extremely negative and can be located on an evaluative
dimension of meaning (Eagly & Chaiken, 1993). Generally speaking, people who
evaluate an attitude object favorably are likely to experience positive affective
reactions in conjunction with it and are unlikely to experience negative affective
reactions. In addition, the affective component of attitude contains some involve-
ment, liking, and caring as proposed by Oliver (1997). Conation includes behav-
ioral intentions or willingness to act. Bagozzi (1978) stated that, “The conation
dimension is said to depict the action tendencies one has to approach or avoid an
object or perform some response” (p. 10).
In using these general components of attitude, attitudinal brand loyalty should
be considered as a sequential process in which customers become “loyal first in a
cognitive sense, then later in an affective sense, and still later in a conative man-
ner” (Oliver, 1997, p. 392). For instance, a customer initially becomes cognitively
loyal based on beliefs about the brand attribute only. Then he or she may become
affectively loyal, with pleasurable fulfillment based on brand performance. Then
he or she may become conatively loyal, exhibiting a brand-specific commitment.
Moreover, according to Oliver (1997) these three stages of attitudinal brand loy-
alty link to overt behavior or action loyalty based on Ajzen and Fishbein’s (1980)
theory. In completing these four phases, customers become truly brand loyal (Oli-
ver, 1997).
Although Oliver’s (1997) theory of brand loyalty stage and consequential rela-
tionship with attitudinal and behavioral brand loyalty are evident, no empirical
study has been undertaken in the lodging industry. Therefore, this study tested the
relationships among customer satisfaction, attitudinal brand loyalty, and behav-
ioral brand loyalty, as listed next:

H1: Customer satisfaction has a positive effect on cognitive brand loyalty.

H2: Cognitive brand loyalty has a positive effect on affective brand loyalty.
H3: Affective brand loyalty has a positive effect on conative brand loyalty.
H4: Conative brand loyalty has a positive effect on behavioral brand loyalty.
H5: Customer satisfaction has a positive effect on behavioral brand loyalty.

CONCEPTUAL MODEL

Figure 1 displays the conceptual model used in this study. It shows the relation-
ships between customer satisfaction and attitudinal and behavioral brand loyalty
as well as the relationship among cognitive, affective, conative, and behavioral
brand loyalty, as based on Oliver’s (1997) brand loyalty stage theory. Customer
satisfaction is treated as an exogenous variable, whereas attitudinal (cognitive,
affective, and conative) brand loyalty and behavioral brand loyalty are considered
as endogenous variables.
424 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Figure 1
A Conceptual Model Showing Relationships Among Study Variables—Model 1

Cognitive
H1 Brand Loyalty
(CBL)
Customer
Satisfaction
(CS) H2

Affective Brand
Loyalty (ABL)

H3

H5
Conative Brand
Loyalty (CNL)

Behavioral H4
Brand Loyalty
(BBL)

METHODOLOGY

A questionnaire was developed based on a thorough review of the literature

and on a pilot study. Manipulation checks were conducted to ensure the reliability
and validity of scales. The 3-item scale, 7-point Likert-type scale, for customer
satisfaction (e.g., “Overall, I am satisfied with the decision to stay at the __
hotel.”) was adapted from Oliver (1980). Attitudinal brand loyalty was measured
using scales developed by Loken and John (1993), Oliver (1997), and Beatty,
Kahle, and Homer (1988)—nine items, 7-point Likert-type (e.g., “The __ hotel
provides me superior service quality as compare to any other hotel brands”; “I
intend to continue staying at the __ hotel”). Finally, behavioral brand loyalty was
measured by the proportion of actual days consumers stayed at a specific-brand
hotel as compared with the number of days they stayed at all other brands in the
past 12 months. Those two numbers are calculated to obtain the actual proportion
of purchasing frequency in a specific hotel over hotels in which respondents
stayed in the past year.
The sample population in this study was composed of individuals who stayed
at a national mid- to upper-scale hotel in North Carolina between June 18 and July
11, 2001. The questionnaire was distributed to 646 individuals when they
checked into the hotel. Of the 646 distributed questionnaires, 199 questionnaires
were returned. Four responses were eliminated before data coding because they
were returned blank or only partially completed. After eliminating the unusable
 Back, Parks / A BRAND LOYALTY MODEL 425

responses, 194 responses were coded for data analysis, resulting in a response rate
of 30%. A majority of the respondents (126 respondents = 65%) indicated that the
purpose of their stays at the hotel was for a solely business-related trip. Another
19% of respondents (37 respondents) indicated that their trips were somewhat
business-related trips. Among those business travelers, 60% of respondents indi-
cated that they made the decision to stay at the hotel by themselves. This implies
that the sample selection was appropriate to continue for data analysis.

RESULTS

A reliability test was used to assess the internal homogeneity among items in
this study. As Sekaran (1992) suggested, the coefficient alpha is the most popular
measure of reliability for a multi-item scale. Behavioral brand loyalty was not
included because it was measured as a single item. The coefficient alpha estimates
for the multi-item scales used in this study are presented in Table 1. All alpha coef-
ficients for the data exceed the minimum standard for reliability of 0.7 recom-
mended by Nunnally (1978) for basic research. Thus, the results indicated that
these multiple measures are highly reliable for the measurement of each construct.

Construct Validity Test

Construct validity assesses the degree to which a measurement represents and
logically connects, via the underlying theory, the observed phenomenon to the
construct (Fornell & Larcker, 1981). Following Anderson and Gerbing’s (1988)
two-step approach, a measurement model was estimated before the structural
model. The results for the measurements of customer satisfaction and three types
of attitudinal brand loyalty were very good (χ2 = 91.01, df = 48, RMSEA = 0.07,
CFI = 0.97, NNFI = 0.96). All indicator loadings for constructs were significant
(p < .01).
Discriminant validity is present when the proportion of variance extracted in
each construct (average variance extracted [AVE]; ρvc[η]) exceeds the square of the
coefficient representing its correlation with other constructs (Fornell & Larcker,
1981), as shown in Table 2. Specifically, the variance extracted for the cognitive
brand loyalty construct (0.75) exceeds the square of intercorrelations (the shared
variance) between customer satisfaction and cognitive brand loyalty (0.48).
Another criterion for discriminant validity, a two-standard error interval estimate
of each coefficient, was calculated to examine whether the value 1 is within the
interval. As shown in Table 2, confidence interval estimates for coefficients did
not include 1, showing evidence of discriminant validity.

Confirmatory Factor Analysis

Using LISREL 8.5, a maximum likelihood confirmatory factor analysis was
undertaken to assess the overall fit of the 4-factor model. The 4-factor model is
composed of customer satisfaction, cognitive brand loyalty, affective brand loy-
alty, and conative brand loyalty. In assessing the goodness-of-fit, chi-square anal-
ysis, Browne and Cudeck’s (1993) root mean square of approximation error
426 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Table 1
Reliability Analysis—Customer Satisfaction and Attitudinal Brand Loyalty

Variable Coefficient Alpha

Customer satisfaction .94

Cognitive brand loyalty .85
Affective brand loyalty .87
Conative brand loyalty .86

Table 2
Measure Correlations, the Squared Correlations, and AVE

Correlations Among Latent Constructs (Squared)a

b
Measure CS CBL ABL CNL AVE

Customer satisfaction (CS) 1.00 0.85

Cognitive brand loyalty (CBL) 0.69 (0.48) 1.00 0.75
Affective brand loyalty (ABL) 0.74 (0.55) 0.85 (0.72) 1.00 0.75
Conative brand loyalty (CNL) 0.50 (0.25) 0.69 (0.48) 0.67(0.45) 1.00 0.74
Model measurement fit: χ2 = 91.01
df = 48
RMSEA = 0.07
CFI = 0.97
NNFI = 0.96

Note: AVE: average variance extracted. RMSEA: root mean square error of approximation
CFI: comparative fit index. NNFI: nonnormal fit index.
a. Correlation coefficients are estimates from LISREL. p < .01, all were significant at .01
level.
b. All AVE exceed 0.50, showing the construct validity.

(RMSEA), Bentler’s (1990) comparative fit index (CFI), and Bentler and
Bonett’s (1980) nonnormed fit index (NNFI) were performed. The results showed
a better fit for the 4-factor model, χ2(48) = 110.36, χ2/df = 2.30, RMSEA = 0.08,
CFI = 0.97, NNFI = 0.95 than the 2-factor model, χ2(53) = 175.65, χ2/df = 3.31,
RMSEA = 0.11, CFI = 0.91, NNFI = 0.89 with a significant ∆χ2(5) = 65.29,
p < .01; and the 3-factor model, χ2(51) = 297.82, χ2/df = 5.84, RMSEA = 0.16,
CFI = 0.84, NNFI = 0.80 with a significant ∆χ2(3) = 187.46, p < .01. Though
better than the 3- or 2-factor model, the 4-factor model seems only adequate. The
χ2/df value of 2.30 falls within a range of acceptable values (2 to 5 as suggested by
Marsh and Hocevar, 1988) but does not reach the less-than-two level proposed by
Byrne (1998). Nevertheless, the goodness-of fit indexes as shown are excellent.
Table 3 presents the standardized factor loadings for each construct.

Structural Model Results

Figure 1 shows the conceptual model, Model 1; the model produced the fol-
lowing statistics: χ2(61) = 140.36, p = .00, RMSEA = 0.07, CFI = 0.96, NNFI =
0.95, as shown in Table 4. Two competing models were tested in this study. Model
 Table 3
Standardized Factor Loadings (Lambda Y)

Standardized
Factor Loadings
Constructs Items (Lambda Y) t value

Customer satisfaction I am happy about my decision to stay at the ___hotel. .90a

I believe I did the right thing when I stayed at the ___ hotel. .92 7.57
Overall, I am satisfied with the decision to stay at the ___ hotel. .89 7.34
Cognitive brand loyalty The ____ hotel provides me superior service quality as compared to any other hotel brands. .80 6.93
No other hotel brand performs better services than the ____. .81a
I believe ___ hotel provides more benefits than other hotels in its category. .81 6.95
Affective brand loyalty I love staying at the ____ hotel. .84 7.12
I feel better when I stay at the ___ hotel. .87 7.53
I like ______ hotel more so than other hotel brands .76a
a
Conative brand loyalty Even if another hotel brand is offering lower room rate, I still stay at the ____ hotel. .73
I intend to continue staying at the ____ hotel. .84 8.34
I consider the____ hotel to be my first lodging choice. .93 9.91

a. Parameter fixed at 1.0 during maximum-likelihood estimation. Thus no t value is obtained.

427
428 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Table 4
Standardized Maximum-Likelihood Parameter Estimates for Model 1 (N = 194)

Path Hypothesis Coefficient t value

CS → CBL (γ11) H1 0.69* 9.40

CBL → ABL (β21) H2 0.93* 10.66
ABL → CNL (β32) H3 0.71* 7.99
CNL → BBL (β43) H4 0.74* 9.23
CS → BBL (γ41) H5 –0.04 0.71
R2 (CBL) .48
R2 (ABL) .87
R2 (CNL) .50
R2 (BBL) .52
Goodness-of-fit statistics:
χ2(61) = 140.36, p = .000
RMSEA = 0.07
CFI = 0.96
NNFI = 0.95

Note: CS: customer satisfaction; CBL: cognitive brand loyalty; ABL: affective brand loyalty;
CNL: conative brand loyalty; BBL: behavioral brand loyalty; RMSEA = root mean square error
of approximation; CFI = comparative fit index; NNFI = nonnormal fit index.
*p < .001

2a specified direct paths from customer satisfaction to cognitive, affective, cona-

tive, and behavioral brand loyalty. It also specified all regression coefficients
among cognitive, affective, conative, and behavioral brand loyalty. For Model 2a,
goodness-of-fit and practical indices were as follows: χ2(56) = 130.19, p = .00,
RMSEA = 0.07, CFI = 0.96, NNFI = 0.95. The difference in fit between this
model and the previous model was not significant (∆χ2 = 10.17, ∆df = 5, p > .01),
which indicates that the model in Model 1 was more parsimonious than Model 2a.
Model 2b specified only direct paths from customer satisfaction to cognitive,
affective, conative, and behavioral brand loyalty. For Model 2b, goodness-of-fit
and practical indices were as follows: χ2(59) = 138.98, p = .00, RMSEA = .07,
CFI = 0.96, NNFI = 0.95. Like the previous model, the difference in fit between
this model and Model 1 was not significant (∆χ2 = 1.38, ∆df = 2, p > .01), which
indicates that Model 1 was more parsimonious than Models 2a and 2b. The fit of
the model indicates that the conceptual model is parsimonious and fits well,
thereby providing a good basis for hypothesis testing.
Table 4 presents the structural model results for the model, as depicted in Fig-
ure 2. Customer satisfaction explained almost one half (48%) of the variance of
cognitive brand loyalty, whereas customer satisfaction and cognitive brand loy-
alty explained 87% of affective brand loyalty. In addition, the predecessors for
each construct explained one half of the variance in conative brand loyalty and
behavioral brand loyalty.

Hypotheses Testing
H1: Customer satisfaction has a positive effect on cognitive brand loyalty.
 Back, Parks / A BRAND LOYALTY MODEL 429

Figure 2
Results of Structural Equation Model of
Brand Loyalty—Standardized Path Estimates

Y1
0.80

Cognitive 0.81 Y2
Brand Loyalty
(η1)
0.81
Y3
0.69a (9.40)
0.93a (10.66)
X1 Y4
0.90 0.84
Customer
X2 0.92 Affective 0.87 Y5
Satisfaction
Brand Loyalty
(ξ1) (η2)
0.89 0.76
X3 Y6
-0.04 (0.71) 0.71a (7.99)

Y7
Y10 0.73

1.00 Conative
0.74a (9.23) Brand Loyalty 0.84
Behavioral (η3) Y8
Brand0.93
Loyalty
(η4) 0.93 Y9

a. p < .001.

First, Hypothesis 1 was tested. The relationship between customer satisfaction

and cognitive brand loyalty was found to be significant (ξ11 = 0.69, t = 9.40,
p < .001). Based on this result, customer satisfaction positively influenced cogni-
tive brand loyalty, whereas the brand information held by customers was superior
to what is known of competitive offerings. As Janis and King (1954) argued, indi-
viduals evaluate a specific behavior based on a biased search of memory for previ-
ously acquired knowledge that confirms the legitimacy of the behavior when the
evaluation was positive. This process confirms the beliefs and enhances the level
of belief confidence about the attitude object. Thus, Hypothesis 1 was supported.

H2: Cognitive brand loyalty has a positive effect on affective brand loyalty.

Second, Hypothesis 2 was tested to investigate the effect of cognitive brand

loyalty on affective brand loyalty. The regression path from cognitive brand loy-
alty to affective brand loyalty was significant (β21 = 0.93, t = 10.66, p < .001). This
result was consistent with Oliver’s (1997) suggestion that customers’ affective
brand loyalty was not directly affected by their satisfaction level. Rather, it was
evident that the mediating effect of cognitive brand loyalty took a place in the rela-
tionship between customer satisfaction and affective brand loyalty so that cus-
tomers became affective brand loyal after they were cognitive brand loyal. The
430 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

standardized indirect effect of customer satisfaction on affective brand loyalty

was 0.65, p < .001. Hypothesis 2 was supported.

H3: Affective brand loyalty has a positive effect on conative brand loyalty.

Hypothesis 3 was tested to assess the effect of affective brand loyalty on cona-
tive brand loyalty. As Table 4 indicates, the regression path from affective to cona-
tive brand loyalty was significant (β32 = 0.71, t = 7.99, p < .001). Thus, Hypothesis
3 was supported at the .01 level.

H4: Conative brand loyalty has a positive effect on behavioral brand loyalty.

Fourth, Hypothesis 4 was tested. Conative brand loyalty had a significant posi-
tive relationship with behavioral brand loyalty, which supported Hypothesis 4.
Strong positive effects of conative brand loyalty on behavioral brand loyalty were
evident when observing the regression coefficient (β43 = 0.74; t = 9.23, p < .001).
This finding was consistent with previous findings showing that a specific behav-
ior was determined by the intention to perform that behavior (Fishbein & Ajzen,
1975), and behavioral intention was considered to be the immediate determinant
of overt volitional behavior (Fredricks & Dossett, 1983).

H5: Customer satisfaction has a positive effect on behavioral brand loyalty.

Finally, the direct effect of customer satisfaction on behavioral brand loyalty

was not significant. Hypothesis 5 was not supported at the 0.01 level (ξ41 = –0.04;
t = 0.71). However, the mediating effect of attitudinal brand loyalty in the relation-
ship between customer satisfaction and behavioral brand loyalty was found to be
significant (CS→CBL→ABL→CNL→BBL: ξ11 = 0.69, β21 = 0.93, β32 = 0.71,
β43 = 0.74; all ps < .001). This result was consistent with Oliver’s (1997) and
Fishbein and Ajzen’s (1975) finding of a significant mediating effect of attitude in
the relationship between the customers’ product or service evaluations and their
overt behaviors in terms of actual purchases. In addition, it was suggested that
level of customer satisfaction was not a sole determinant for retention rate as pro-
posed by Heskett et al. (1997).
In sum, as Oliver (1997) mentioned, customers become truly brand loyal when
they follow these stages: (a) cognitive brand loyal stage; (b) affective brand loyal
stage; (c) conative brand loyal stage; and (d) behavioral brand loyalty. As shown
in this study’s results, the customer’s positive information about a brand did not
directly increase the repurchase decision. The customer also exerts a positive
effect on and commitment to the brand when truly brand loyal. In addition, the
results indicated a positive relationship between customer satisfaction and attitu-
dinal brand loyalty as well as attitudinal brand loyalty and behavioral brand loy-
alty. Thus, the findings indicate that the association between customer satisfac-
tion and behavioral brand loyalty was positively significant only when mediated
by attitudinal brand loyalty.
 Back, Parks / A BRAND LOYALTY MODEL 431

DISCUSSION

Theoretical and Practical Implications

This study provided empirical evidence for the development of customers’
repurchasing behaviors involving customer satisfaction, and attitudinal and
behavioral brand loyalty. The mediating effects of three types of attitudinal brand
loyalty in the relationship between customer satisfaction and behavioral brand
loyalty were significant. The findings from this study confirm the theory of rea-
soned action by Ajzen and Fishbein (1980), relating to customers’ beliefs about
and attitudes toward their behavioral intentions, and to actual behaviors and the
study of brand loyalty stages by Oliver (1997) involving development of a brand
loyalty process. Results supported the finding that customer’s strong beliefs about
brand quality have increased the degree of “liking.” In turn, results indicate a posi-
tive intention or commitment to repurchase a particular item. This results in actual
behaviors in terms of a high proportion of purchasing frequency over other
brands. Therefore, this study emphasizes the importance of measuring attitudinal
and behavioral brand loyalty to identify brand loyal customers and better under-
stand their repurchasing behaviors in the lodging industry.
In addition, this study answers the question about the relationship between sat-
isfaction and actual repeat purchasing behaviors as raised by Bowen and Shoe-
maker (1998). Satisfied customers did not actually repurchase unless they were
attitudinally brand loyal. However, many brand loyal researchers or practical
operators had a misconception about the direct relationship between customer
satisfaction and actual repurchase frequencies. The results of this study showed a
significant indirect effect of customer satisfaction on behavioral brand loyalty
(0.64, t = 4.42, p < .01). Thus, satisfied customers will tolerate lower price elas-
tics, will be more insulated from competitive situations through battles about
lower prices, and will generate more profit when they have a strong commitment
level (conative brand loyalty), strong favoritism (affective brand loyalty), and
strong beliefs about the superiority (cognitive brand loyalty) of a brand.
The findings also identify several marketing and general management implica-
tions. They suggest that attitudinal and behavioral brand loyalty can be used to
measure true brand loyalty among hotel customers. Because previous brand loy-
alty studies focused significantly on either attitudinal or behavioral traits, use of a
combined measure has increased validity and reliability. Although some com-
bined brand loyalty measures were available, specific measures were not avail-
able for the lodging industry. Because reliability and validity were very strong in
this study, they should be applied to different market segments. Moreover, hotel
marketers should be able to assess the attitudes of their existing customers toward
the brand and to identify any needs that should be fulfilled. As a result, this brand
loyalty measurement should be used as a tool in evaluating service as perceived by
existing customers.
In addition, the loyalty measure can be used with frequent-stayer programs.
Although many hotel companies have developed reward programs for their exist-
ing customers, there is limited information about purchasing frequency in such
432 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

programs. A customer commonly holds many frequent-stayer program member-

ships. Thus, a membership program could be very superficial or artificial in repre-
senting true brand loyalty. By applying the combined brand loyalty measure
along with information drawn from a frequent-stayer program, marketers should
be able to pull out valuable information about their customers and reduce some
marketing costs.

LIMITATIONS AND SUGGESTIONS FOR FUTURE STUDY

Several limitations are associated with this study. First, the results may not be
generalized to other segments of the lodging industry. Data from this study were
collected from customers at a single upper-middle-scale business hotel. The other
types of hotel segments may have different strength of effect of customer satisfac-
tion on attitudinal brand loyalty. For instance, pleasure customers tend to rely
heavily on heuristic cues when they perceive service performances, so that there
may be a significant direct effect of satisfaction on affective brand loyalty. In addi-
tion, due to the poor economic situation in the overall market, business travelers
who participated in the study may not be representative of the overall business
traveler market. There is general agreement that business travelers are
downscaling their reservations in lower priced lodging facilities, and shortening
the length of business trips, which may cause subject bias. In addition, the sample
selection for this study was not purely random. As noted, pure random sampling is
almost impossible in the lodging industry, so including many different types of
hotel brands and segments would increase external validity. Thus, future studies
should develop a systematic design that better represents the population.
In addition, asking respondents to rate purchasing frequency to measure pro-
portion of actual purchasing behavior for a specific brand—so-called behavioral
brand loyalty—may not be as accurate as obtaining actual data from a database.
The respondents may respond inaccurately or just guess the frequency of their
visits to the hotel. Future studies should include assurance of researcher access to
the industry and to information on actual purchasing frequencies for the hotel.
Some database companies provide this type of service, but it is costly. With joint
efforts by the industry, the results should be more parsimonious and increase the
validity of the study.
Moreover, future studies can include additional variables in the model to fur-
ther develop brand loyalty strategies. For instance, by considering the effect of
customers’ perceptions of brand image on their satisfaction and brand loyalty,
marketers should be able to develop selective target market strategies and enhance
the effectiveness of their advertising strategies.
In sum, this study suggests that customer satisfaction does not guarantee true
brand loyalty. In other words, customer satisfaction will not increase the repeat
purchasing rate unless customers first build positive attitudes toward the brand.
True brand loyalty should be exhibited in customers’ experiences of superiority,
positive feeling, and strong commitment toward the brand, and then greater pur-
chasing frequencies over other brands. By using the brand loyalty measurement,
 Back, Parks / A BRAND LOYALTY MODEL 433

which contains both attitudinal and behavioral information, lodging operators

should be able to identify the true brand loyalists and their wants and needs.

REFERENCES

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice Hall.
Albarracin, D., Wyer, R. S. Jr. (2000). The cognitive impact of past behavior: Influences on
beliefs, attitudes, and future behavioral decisions. Journal of Personality & Social Psy-
chology, 79, 5-22.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A
review and recommendation two-step approach. Psychological Bulletin, 103, 411-423.
Bagozzi, R. P. (1978). The construct validity of the affective, behavioral, and cognitive
components of attitude by analysis of covariance structures. Multivariate Behavioral
Research, 13, 9-31.
Beatty, S. E., Kahle, L. R., & Homer, P. (1988). The involvement-commitment model: The-
ory and implications. Journal of Business Research, 16, 149-167.
Bentler, P. M. (1990). Comparative fit indexes in structural model. Psychological Bulletin,
107, 238-246.
Bentler, P. M., & Bonett, D. G. (1980). Significance test and goodness of fit in the analysis
of covariance structures. Psychological Bulletin, 88, 588-606.
Bentler, P. M., & Speckart, G. (1981). Attitudes “cause” behaviors: A structural equation
analysis. Journal of Personality & Social Psychology, 40, 226-238.
Bitner, M. J. (1990). Evaluating service encounters: The effect of physical surroundings
and employee responses. Journal of Marketing, 54, 69-82.
Boulding, W., Kalra, A., Staelin, R., & Zeithaml, V. A. (1993, February). A dynamic pro-
cess model of service quality: From expectations to behavioral intentions. Journal of
Marketing Research, 30, 7-27.
Bowen, J. T., & Chen, S. (2001). The relationship between customer loyalty and customer
satisfaction. International Journal of Contemporary Hospitality Management, 13, 213-
317.
Bowen, J. T., & Shoemaker, S. (1998). Loyalty: Strategic commitment. Cornell Hotel &
Restaurant Administration Quarterly, 13, 12-25.
Breckler, S. (1984). Empirical validation of affect, behavior, and cognition as distinct com-
ponents of attitude. Journal of Personality & Social Psychology, 47, 1191-1205.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A.
Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 445-455). Newbury
Park, CA: Sage.
Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS:
Basic concepts, applications, and programming. Hillsdale, NJ: Lawrence Erlbaum.
Cardozo, R. N. (1965). An experimental study of customer effort, expectation, and satis-
faction. Journal of Marketing Research, 2, 244-249.
Dick, A. S., & Basu, K. (1994). Customer loyalty: Toward an integrated conceptual frame-
work. Journal of the Academy of Marketing Science, 22, 99-113.
Eagly, A. C., & Chaiken, S. (1993). The psychology of attitudes. Dallas, TX: Harcourt Pub-
lishing.
434 JOURNAL OF HOSPITALITY & TOURISM RESEARCH

Ehrenberg, A. S. C. (1991). New brands and the existing market. Journal of the Royal Sta-
tistical Society, 33, 285-299.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction
to theory and research. Reading, MA: Addison-Wesley.
Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E. (1996). The Ameri-
can Customer Satisfaction Index: Nature, purpose, and findings. Journal of Marketing,
60, 7-18.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with
unobservable variables and measurement error. Journal of Marketing Research, 18, 39-
50.
Fredricks, A. J., & Dossett, D. L. (1983). Attitude-behavior relations: A comparison of the
Fishbein-Ajzen and the Bentler-Speckart models. Journal of Personality & Social Psy-
chology, 45, 501-512.
Guadagni, P. M., & Little, J. D. (1983). A logit model of brand choice calibrated on scanner
data. Marketing Science, 2, 203-238.
Heskett, J. L., & Sasser, E. W. Jr., & Schlesinger, L. A. (1997). The service profit chain.
New York: Free Press.
Iwasaki, Y., & Havitz, M. E. (1998). A path analytic model of the relationships between
involvement, psychological commitment, and loyalty. Journal of Leisure Research, 30,
256-280.
Jacoby, J., & Chestnut, R. W. (1978). Brand loyalty measurement and management. New
York: Wiley.
Jacoby, J., & Olson, J. C. (1970). Operationalizing an attitudinal model of multi-brand loy-
alty: Preliminary results and proposed directions. A report to the Pillsbury Company.
Jacoby, J., & Kyner, D. B. (1973). Brand loyalty vs. repeat purchasing behavior. Journal of
Marketing Research, 10, 1-9.
Janis, I. L., & King, B. T. (1954). The influence of role-playing on opinion change. Journal
of Abnormal Social Psychology, 49, 211-218.
Jarvis, L. P. (1973). An empirical investigation of cognitive brand loyalty and product class
importance as mediators of consumer brand choice behavior. Dissertation Abstracts,
34, 943A.
Kotler, P., Bowen, J., & Makens, J. (1998). Marketing for hospitality and tourism.
Englewood Cliffs, NJ: Prentice Hall.
Lewis, R. C., & Chambers, R. E. (2000). Marketing leadership in hospitality (3rd ed.). New
York: Wiley.
Loken, B., & John, D. R. (1993). Diluting brand beliefs: When do brand extensions have
negative impact? Journal of Marketing, 57, 71-84.
Marsh, H. W., & Hocevar, D. (1988). A new, more powerful approach to multitrait-
multimethod analyses: Application of second-order confirmatory factor analysis. Jour-
nal of Applied Psychology, 73, 107-117.
McCleary, K. W., & Weaver, P. A. (1992). Do business travelers who belong to frequent-
guest programs differ from those who don’t belong? Hospitality Research Journal, 15,
51-61.
Muncy, J. A. (1983). An investigation of the two-dimensional conceptualization of brand
loyalty. Dissertation Abstracts International, 45, 288A.
Nunnally, J. C. (1978). Psychometric theory. New York: McGraw-Hill.
 Back, Parks / A BRAND LOYALTY MODEL 435

Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfac-

tion decisions. Journal of Marketing Research, 17, 460-469.
Oliver, R. L. (1993). Cognitive, affective, and attribute bases of the satisfaction response.
Journal of Consumer Research, 20, 418-430.
Oliver, R. L. (1997). Satisfaction: A behavioral perspective on the consumer. New York:
McGraw-Hill.
Oliver, R. L. (1999). Whence consumer loyalty. Journal of Marketing. 63, 33-34.
Peter, J. P., & Olson, J. C. (1993). Consumer behavior and marketing strategy (3rd ed.).
Chicago: Irwin.
Pritchard, M. P., Howard, D. R., & Havitz, M. E. (1992). Loyalty measurement: A critical
examination and theoretical extension. Leisure Sciences, 14, 155-164.
Rust, R. T. & Zahorik, A. J. (1993). Customer satisfaction, retention, and market share.
Journal of Retailing, 69, 193-215.
Sekaran, U. (1992). Research methods for business (2nd ed). New York: Wiley.
Shoemaker, S., & Lewis, R. C. (1999). Customer loyalty: The future of hospitality market-
ing. International Journal of Hospitality Management, 18, 345-370.
Spreng, R. A., MacKenzie, S. C., & Olshavsky, S. W. (1996). A reexamination of the deter-
minants of consumer satisfaction. Journal of Marketing, 60, 15-52.
Stern, B. B. (1997). Advertising intimacy: Relationship marketing and the service con-
sumer. Journal of Advertising, 26, 7-19.
Tepeci, M. (1999). Increasing brand loyalty in the hospitality industry. International Jour-
nal of Contemporary Hospitality Management, 11, 223-229.
Yi, Y. (1990). A critical review of consumer satisfaction. In V. A. Zeithaml (Ed.), Review of
marketing (pp. 68-123). Chicago: American Marketing Association.

Submitted May 9, 2002

First Revision submitted October 9, 2002
Final Revision submitted November 18, 2002
Accepted November 23, 2002
Refereed Anonymously

Ki-Joon Back (e-mail: back@humec.ksu.edu), Ph.D., is assistant professor in the Depart-

ment of Hotel, Restaurant, Institution Management, and Dietetics at Kansas State Univer-
sity. Sara C. Parks (e-mail: SJP3@psu.edu), Ph.D., is professor in and director of the
School of Hotel, Restaurant, and Recreation Management at Pennsylvania State University.