1 Consumption-Based Emotional Responses Related to Satisfaction Veronica Liljander and Maria Bergenwall Correspondence: Swedish School of Economics and Business Administration, Department of Marketing, PO Box 479, 00101 Helsinki, Finland. V. Liljander: Tel. +358-9-43133 288, Fax. +358-9-43133 287, e-mail: [email protected]
33
Embed
Consumption-Based Emotional Responses Related to Satisfaction
The authors examine the impact of consumption-based emotions on perceived satisfaction in a field study of charter holiday consumers. Eight emotions from Russell’s circumplex model of affect are investigated for four parts of a charter holiday: the destination, hotel, guide(s) and return flight. Analysis of the structure of emotions does not support Russell’s dimensions of pleasure and arousal, but confirms other researchers’ findings of positive and negative affect. Of the discrete emotions, happiness and excitement are found to have a positive effect and boredom and idleness a negative effect on satisfaction. Emotional responses related to the hotel and destination have the largest impact on overall satisfaction. Future research directions are discussed.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Consumption-Based Emotional Responses Related to Satisfaction
Veronica Liljander and Maria Bergenwall
Correspondence: Swedish School of Economics and Business Administration, Department of
Marketing, PO Box 479, 00101 Helsinki, Finland. V. Liljander: Tel. +358-9-43133 288, Fax.
Consumption-Based Emotional Responses Related to Satisfaction
Abstract
The authors examine the impact of consumption-based emotions on perceived satisfaction in a
field study of charter holiday consumers. Eight emotions from Russell’s circumplex model of affect
are investigated for four parts of a charter holiday: the destination, hotel, guide(s) and return flight.
Analysis of the structure of emotions does not support Russell’s dimensions of pleasure and
arousal, but confirms other researchers’ findings of positive and negative affect. Of the discrete
emotions, happiness and excitement are found to have a positive effect and boredom and idleness
a negative effect on satisfaction. Emotional responses related to the hotel and destination have the
largest impact on overall satisfaction. Future research directions are discussed.
INTRODUCTION
High consumer satisfaction has been linked to positive consequences for the service organisation
in the form of increased consumer loyalty and higher profitability (Storbacka et al., 1994).
Hence, it is in the interest of every firm to monitor their customers’ perceived satisfaction with the
service offer and the underlying factors that affect it. In services research, perceived service
quality has usually been used as the only indicator of consumer satisfaction. However, there is a
growing body of literature which clearly indicates that the positive and negative emotions that
consumers associate with the service play an important role in forming satisfaction (Dubé et al.,
1996; Hui and Tse, 1996; Liljander and Strandvik, 1997; Koelemeijer et al., 1995; Price,
Arnould and Deibler, 1995; Price, Arnould and Tierney, 1995; Stauss and Neuhaus, 1997).
Generally, positive emotions have been found to enhance satisfaction. On the other hand, it is also
possible to combine high satisfaction with feelings that in other circumstances would have invoked
dissatisfaction. For example, Arnould and Price (1993) found that high satisfaction with river
3
rafting was related to both extreme positive and extreme negative feelings. Thrill and excitement,
sometimes on the border of safety, made the trip memorable and satisfactory. In a hospital
setting, Dubé et al. (1996) also found that negative emotions in certain situations had a positive
effect on satisfaction. Oliver (1997) notes that although there are many studies on the structure of
emotions and consumer experienced emotions with advertising there is a lack of research on the
linkage between emotions and post-consumption variables like satisfaction.
Emotions can be divided into reactive and goal-directed emotions (Bagozzi et al., 1995; Brown
et al., 1997; Koelemeijer et al., 1995). If, for example, the delivered service exceeds the
promises made in the advertising, the consumer may react with positive emotions and high
satisfaction. Goal-directed emotions, on the other hand, are emotions that the consumer
consciously seeks to experience, for example by visiting an amusement park or going on a
holiday. However, other, more negative, types of emotions may also be anticipated/expected by
the consumer. One may, for example, anticipate feelings of anxiety when going to the dentist.
Services that are characterised by strong goal-directed positive emotions are particularly
important to note since the experienced emotions may constitute the main outcome of the service
(e.g., a concert, a movie). However, it should also be observed that all services may arouse
unplanned reactive emotions in the consumer. In addition, different segments of consumers may
react with different emotions to the same service, and because of service variability, one
consumer’s perceived service quality and experienced emotions may vary from one service
encounter to another. It is a challenge for service managers to understand the emotions that their
customers experience, and to manage the service in such a way that positive emotions are
promoted and negative emotions reduced. In order to better understand the role of emotions in
services, research is needed on different types of service encounters. The present study seeks to
increase our understanding of how experienced emotions affect consumer satisfaction with a
charter holiday.
Tourism services offer goal-directed emotions. Package holidays and different tourist attractions
are designed to excite consumers and give them memorable positive feelings, something out of the
ordinary. Although the consumer may wish for only positive goal-directed emotions, a holiday is
still full of service encounters which cannot be planned beforehand, and which may trigger both
unexpected positive and negative emotions which affect satisfaction.
4
The main aim of this study was to investigate the structure of emotions as experienced by
consumers of a charter holiday in Spain, and the relation between experienced emotions and
satisfaction with the holiday. The consumer may have an endless variety of service experiences
during a holiday, and it is not possible to cover them all in one study. The empirical study was
limited to general perceptions of the destination, the hotel, the guide(s) and the flight home. The
emotions were selected from each of the eight quadrants in Russell’s circumplex model of affect
(Russell, 1980).
More specifically, the following research questions were asked: 1) What is the underlying
structure of the emotions in this service context? 2) Do consumers have different patterns of
emotions, and how are these related to satisfaction? 3) How do various experienced emotions
related to the hotel, destination, guide(s) and flight affect overall satisfaction? 4) How do
experienced emotions connected with the hotel, destination, guide(s) and flight affect overall
satisfaction with the holiday package as a whole?
Russell’s model of the structure of emotions is presented first, followed by a literature review of
previous studies on the relationship between emotions and satisfaction. The design of the
empirical study and its findings are then presented, and future research directions are discussed.
THE STRUCTURE OF EMOTIONS
Numerous definitions of emotions have been proposed in the psychology literature and no
consensus on any given definition has been reached. In view of the lack of research on
consumption emotions it may be harmful to use a too narrow definition of the concept at this stage
of research. A broad definition is proposed by Oliver (1997, p. 294) who suggests that “Emotion
includes arousal, various forms of affect, and cognitive interpretations of affect that may be given a
single description”. For example, Oliver’s definition encompasses the emotions ‘interest’ and
‘surprise’ that have also been referred to as ‘basic’ emotions in the psychology literature (Oliver
1997), while another recently proposed definition by Richins (1997) excludes them.
The basic structures of emotions and self-reported mood have been extensively studied within
psychology (c.f. Russell, 1980 and Watson and Tellegen, 1985 for reviews). Emotions have
generally been described as either discrete entities (Izard, 1972; 1977; 1991) or as on bipolar
5
dimensions related to each other in a systematic universal fashion (Russell, 1980). Both Izard’s
and Russell’s sets of emotions have been used in consumer satisfaction studies, although Izard’s
DES scale is perhaps the most common. The DES scale has been criticised for the predominance
of negative emotions in the scale and for not taking into account the level of arousal. Thus it may
be noted that one of Richins’ (1997) goals when developing the Consumption Emotions Set
(CET) was to achieve an even spread in emotional space. The much earlier developed two-
dimensional circumplex model of affect that was used as a basis of the present study takes into
account both positive and negative emotions and the degree of arousal (Russell, 1980). Next the
model will be described in more detail.
According to Russell (1980), the interrelationships between different types of emotions are best
described by a spatial model in which eight affective components are organised in a circular
arrangement of pleasure-displeasure (misery), arousal-sleepiness, excitement-depression and
contentment-distress. Two of these pairs, i.e. pleasure - displeasure (misery) and arousal -
sleepiness are the main bipolar dimensions. The emotions fall on a circle in a two-dimensional
space in a compass like manner (Figure 1).
Displeasure Pleasure
Sleepiness
Arousal
contentmentdepression
distress excitement
Figure 1. Circular order of emotions
The emotions excitement, depression, contentment and distress help to define the quadrants of the
space. According to Russell, all words of affect can be defined as a combination of degree of
pleasure and degree of arousal. For example, excitement is defined as a combination of high
6
pleasure and high arousal, and contentment as a combination of high pleasure and low arousal.
Polar co-ordinates for 28 affect words were created by giving the categories assigned scale co-
ordinates based on their theoretical circular ordering. Russell found support for the two bipolar
dimensions from several other studies on both verbal and non-verbal emotional expressions. He
also presented studies of his own that supported the circular order of emotions. Two small
samples of students were asked to sort 28 affect words (described as “words or phrases that
people use to describe their moods, feelings, affect or emotions”, Russell, 1980, p. 1164) into the
eight categories: arousal, excitement, pleasure, contentment, sleepiness, depression, misery and
distress. The subjects were then asked to place these categories in a circular order so that words
opposite each other would describe opposite feelings, and words close to each other would
describe similar feelings. Overall, the data fitted the model well. A reanalysis reported elsewhere
also resulted in a very similar solution. Finally, in a study on self-reported mood, 343 students
rated how they felt on a particular day on 518 affect words. The data was analysed with
regression and factor analysis, and the results supported the findings from the other studies. The
factor analysis yielded two factors which included the elements of the two dimensions in the
circular model.
Watson and Tellegen (1985), however, question the applicability of Russell’s dimensions on self-
reported mood. By reanalysing a large number of previously reported studies they concluded that
the structure of mood is best described by the two bipolar dimensions, positive and negative
affect. Consequently, they rotated the axes of pleasantness and arousal by 45°, making positive
and negative affect the primary dimensions while placing pleasure-displeasure and arousal-
sleepiness in between. These results can be compared with the factor structures obtained on
emotional responses in consumer studies. When both Izard’s DES scale (Liljander and Strandvik,
1997; Oliver, 1993; Oliver and Westbrook, 1993; Westbrook, 1983) and other scales (Dubé et
al., 1996) have been used, emotions have consistently been divided into positive and negative
affect. However, studies on consumption emotions differ from most of the structure studies in
psychology in that fewer emotions have been included, sometimes only one measure for positive
and one for negative affect (e.g. Price, Arnould and Deibler, 1995). Using a limited number of
emotions also limits the possibilities of finding bipolar dimensions in the data (Watson and
Tellegen, 1985). In contrast to Watson and Tellegen’s findings, one study by Mano (1990) on
7
self-reported mood and another by Mano and Oliver (1993) on product evaluation supported
Russell’s circumplex model of emotions.
It should also be noted that a conceptual distinction may be made between the consumer’s mood
state and the emotions that he/she experiences as the result of consuming a good or service.
Generally, emotions are considered more intense and stimulus specific than moods. Moods are
temporal and may be elicited and maintained without conscious awareness of the feeling state, its
cause or its influence on current activities (Clark and Isen, 1982; Gardner, 1985; Morris, 1989;
Oliver, 1997). The individual may be unaware of the state and continue to pursue his/her present
activities without interruption. Cohen and Areni (1991) suggest that emotions are connected with
specific stimuli, and represent a more intense and arousing affective state that demands attention.
The activity the individual pursues is interrupted and, in some cases, a behavioural response
occurs involuntarily. When the stimulus has been dealt with, the individual can again turn his/her
attention away from it and begin to focus on other tasks and objectives. However, a residue of
the earlier experienced emotion may still function as a mood. According to Elster (1996),
emotions are always preceded by cognition of what produced them. This view of emotions, which
is generally labelled the cognitive approach (Strongman, 1996), and is supported by several
researchers in psychology (e.g. Cohen and Areni, 1991; Lazarus, 1984; Lazarus and Smith,
1988), has also been embraced by consumer satisfaction researchers. An exception is found in a
study by Alford and Sherrell (1996), where consumer affect is proposed to precede perceptions
of service performance.
Since the customer may be unaware of his/her mood, it is usually difficult to isolate mood from
stimulus elicited emotions. Even a mood which is characterised by quite strong negative or
positive feelings may be difficult to isolate from consumption-related emotions. The strength of
positive and negative feelings in the consumer’s mood state, may to a lesser or stronger degree,
affect how he/she experiences a service by strengthening positive or negative emotions.
Consumers who have experienced a happy event in their lives may be in a good mood for many
days. The good mood may make them experience a service as better than they normally would
have, or they may tolerate poorer service without it being reflected in their satisfaction scores.
8
EMOTIONS AND SATISFACTION
Emotions have been linked to satisfaction in several ways, but three main streams of research can
be identified: 1) satisfaction is in itself seen as an emotional response to a product (Cadotte et al.,
1983; Dabholkar, 1995; Oliver, 1993); 2) emotions are treated as a mediator between cognitive
evaluations, such as perceived product performance or disconfirmation of some comparison
standard, and overall satisfaction; and 3) emotions are modelled as an independent factor
contributing to the level of perceived satisfaction.
One satisfaction scale which includes emotions is the ‘delighted-terrible’ scale recommended by
Westbrook (1980)1. Another example of satisfaction measured as emotions can be found in a
study by Cadotte et al. (1983), where a summary measure of 12 bipolar ‘feeling scales’ was
used to indicate satisfaction. Dabholkar (1995), on the other hand, challenges the affective
content of satisfaction and suggests that it is only temporal. According to her, the affective aspects
are short-lived, while cognitive evaluations will be remembered over a longer period.
Consequently, she argues that since both satisfaction and service quality include cognitive
evaluations, but only satisfaction includes affective components, the two concepts will be distinct
in the short-term, but will overlap in the long run. She finds some support for these ideas in a
study on retail customers. However, it should be noted that, although the affective component
may be temporal in normal circumstances, and that consumers often feel no emotions at all when
consuming a service (Price, Arnould and Deibler, 1995), this may not hold for highly positive or
negative emotions connected with critical incidents. In particular, strong negative emotions related
to negative incidents may still be easily retrieved each time the consumer remembers the incident.
It is also possible that the structure emotions in satisfaction changes over time from, for example,
satisfaction as delight to satisfaction as contentment (c.f. Oliver, 1997; Stauss and Neuhaus,
1997).
Other researchers have treated affect as either a mediator of satisfaction or an independent
contributor. A study by Koelemeijer et al. (1995) supported emotions, satisfaction and service
quality as independent constructs. They also suggested that dis/satisfaction is a generalised
emotion that differs from more specific ones. They tested two different models: a ‘two-sources’
model and a ‘two-route’ model. The ‘two-sources’ model assumes that emotions and perceived
service quality have independent effects on consumer dis/satisfaction, which in turn affects
9
behavioural intentions. The second model assumes that service quality influences both emotions
and consumer dis/satisfaction, which both influence behavioural intentions. Both models were
supported for goal-directed emotions. However, reactive emotions did not have an independent
effect on consumer satisfaction. Other researchers have also found that the emotions experienced
by consumers when consuming a good (Oliver, 1993; Oliver and Westbrook, 1993) or a service
(Liljander and Strandvik, 1997) add to the explanation of consumer satisfaction.
Researchers have also found support for emotions as a mediator of post-consumption evaluation.
Mano and Oliver (1993) found that positive and negative affect mediated the effects of both
utilitarian and hedonic evaluations on satisfaction. A study on service waits by Hui and Tse (1996)
showed that affective responses to different lengths of wait mediated the effects of the wait
condition on evaluations of service preference and favourability.
Several studies have used Izard’s Differential Emotions Scale (DES) (Izard, 1972; 1977) for
measuring consumers’ reactive emotions (Hausknecht, 1988; Liljander and Strandvik, 1997;
Oliver, 1993; Oliver and Westbrook, 1993; Westbrook, 1983). Factor analysis of the Izard
scale typically yields two factors: negative and positive affect (Liljander and Strandvik, 1997;
Westbrook, 1987). In two empirical studies, one on automobiles and one on a Principles of
Marketing course, Oliver (1993) found that attribute satisfaction was strongly related to positive
affect, and attribute dissatisfaction to negative affect. Positive and negative affect also influenced
(positively/negatively) overall product satisfaction. Similar results were obtained by Hausknecht
(1988), who combined critical incidents concerning consumer experiences in general with a 75-
item list of emotions, and found that feelings of satisfaction were best characterised by the
emotions interest, joy and surprise, while dissatisfaction was connected with anger, disgust and
surprise.
However, to get a deeper understanding of the relation between affect and satisfaction, there is a
need to look at more detailed profiles of affective reactions. It should be noted that negative affect
does not always lead to dissatisfaction (Liljander and Strandvik, 1997; Oliver and Westbrook,
1993; Westbrook and Oliver, 1991). Westbrook and Oliver (1991), for example, found five
clusters of emotions related to car ownership: Happy/Content, Pleasant surprise, Unemotional,
Unpleasant surprise, and Angry/Upset. Although satisfaction varied between the clusters, all the
owners were at least somewhat satisfied. In two other studies (Liljander and Strandvik, 1997;
10
Oliver and Westbrook, 1993), only clusters of hostile/upset and angry/humiliated consumers
expressed dissatisfaction. The studies indicate that some negative affect can be absorbed by
consumers. They may also be satisfied without any apparent positive emotions (Price, Arnould
and Deibler, 1995; Westbrook and Oliver, 1991).
DATA COLLECTION
The data for our study were collected on four flights from the Canary Islands to Finland in
November 1995. The tourists were returning from six different resorts: Lanzarote (1 resort), Gran
Canaria (three resorts) and Teneriffe (two resorts).
A student2 personally distributed and collected the data on the first two flights. For the third and
fourth flights, the questionnaires were distributed during the bus journey to the airport and
collected by the stewardesses on board the aeroplane towards the end of the flight. All the flights
were made during the day time and their length was approximately six hours. During the flight the
passengers were served lunch with drinks, they had an opportunity to watch video films and listen
to music, as well as to buy tax-free products. Despite these interruptions there was ample time for
them to fill out the questionnaires. Each flight had 200-250 passengers and, the return rate of
questionnaires was about 30%. Although the flight was fairly long in duration, many passengers
did not feel like using their time for answering a two-page questionnaire. Many families with
children declined to participate in the study. On the first flight 65 questionnaires were returned, on
the second flight 71, on the third flight 68, and on the fourth flight 70. Of the 274 questionnaires
which were returned, 258 were retained for analysis. The majority of the respondents (65%)
were women.
The questionnaire was designed not to exceed two A4 pages. The questions concerned emotions
experienced and perceived satisfaction. The emotion measures were repeated for the destination,
the hotel, the guide(s) and the flight home. The following constructs were measured:
Emotions: The tourists were asked to what degree they had experienced eight different
emotions3 during the time spent at the destination, at the hotel, with the guide(s) and on the flight
home. A 7-point intensity scale was used, ranging from “I have not experienced the emotion at
all” to “ I have strongly experienced the emotion”. The emotions were: Happiness, Excitement,
11
Calmness, Surprise, Idleness, Boredom, Sadness, and Fear. They represent both axes of
Russell’s (1980) and Watson and Tellegen’s (1985) circular models of affect. Figure 2 shows the
emotions ordered according to Russell’s bipolar dimensions of degree of arousal and
pleasantness. Watson and Tellegen’s main dimensions of positive and negative affect are also
shown in the figure.
Displeasure Pleasure
Sleepiness
Arousal
idleness
calmnessboredom
sadness
fear excitement
surprise
happiness
(Low positiveaffect)
(High negativeaffect)
(High positiveaffect)
(Low negative affect)
Figure 2. Circular order of the eight emotions used in the study
Satisfaction: Holistic measures of satisfaction were used. Satisfaction with service attributes, or
dimensions, or perceived quality of detailed parts of the service were not asked for. The following
measures were collected:
Overall satisfaction: The tourists were asked how satisfied they were with the trip as a whole on
a 7-point scale ranging from “very dissatisfied” to “very satisfied”. The same scale was used for
all satisfaction measures.
Satisfaction with the destination: The tourists were asked separately for their overall
satisfaction with the destination, and how satisfied they were with the recreational services. A
direct disconfirmation measure using an experience-based norm (Cadotte et al., 1983) was also
12
included. The tourists were asked to compare the present destination with previously experienced
destinations, and to evaluate the extent to which the current destination was worse or better than
the others on a 7-point scale.
Satisfaction with the hotel: Satisfaction with the hotel was measured as overall satisfaction with
the hotel, as satisfaction with the services performed by the personnel, and as satisfaction with the
location.
Satisfaction with the guide(s): One overall satisfaction measure was collected for the tour
guide(s). The tourists were also asked to compare the currently experienced guide(s) with
previously experienced guides on a 7-point scale ranging from “worse” to “better”. Some of the
tourists had only experienced the guide(s) on the bus between the airport and the hotel, while
others had been on one or more of the guided tours as well. The extent of interaction with the
guide(s) was not asked for and, thus, cannot be taken into account when analysing the data.
Satisfaction with the flight home: The tourists were asked for their overall satisfaction with the
flight home, and with the airline’s personnel on board. In addition they were asked to compare
the present flight with previously experienced charter flights on a 7-point scale ranging from
“worse” to “better”. Since charter flights may differ from normal scheduled flights, only previously
experienced charter flights were used as the comparison standard.
FINDINGS
The analyses were performed using SPSS for Windows 7.0 and 7.5. The dispersion of the scores
for the different variables are presented in Table III in the Appendix. The means and standard
deviations are also given. It can be noted that all the satisfaction measures are skewed toward the
positive values of the scale. Only around 10% of the tourists marked some degree of
dissatisfaction with the holiday as a whole, or with the destination, hotel, guide or flight. Only the
overall satisfaction measures are used in the analyses. Similarly, the emotions of happiness,
excitement and calmness are skewed towards the upper values of the scale, indicating that the
tourists had felt these emotions, while those of surprise, sadness, fear, boredom and idleness are
skewed towards the lower end of the scale, indicating a general lack of these emotions.
13
One-way analyses of variance were performed on the data to find out to what extent the six
destinations differed on customer perceived satisfaction and emotions. Overall satisfaction with
the destination, the guide(s), the hotel, and the holiday as a whole differed between some of the
destinations. No differences in overall satisfaction were found between the four different flights.
For the emotions, effects were only found for experienced idleness at destination and experienced
happiness, excitement and calmness with reference to the guides. Hence, it may be concluded
that, although the six destinations did give rise to different degrees of perceived satisfaction, this is
not directly reflected in the emotions. This can be explained by the possibility that the tourists may
have experienced different patterns of emotions, and that these patterns are connected with
different satisfaction scores.
The analyses of the data are organised in the following way. Firstly, the results of multidimensional
scaling and factor analysis are presented. Secondly, the data is analysed for correlation between
emotions and satisfaction. Thirdly, the emotions are cluster-analysed in order to find out if there
are different patterns, and if these can be distinguished according to the customers’ satisfaction
scores.
The Structure of emotions
The structure of the emotions scales was first examined by multidimensional scaling (Euclidean
distance), by calculating dissimilarity data from the multivariate data. The analysis is equivalent to
a principal components analysis (Norušis, 1993). Figure 3 shows the resulting 2-dimensional plot
of the eight emotions for the destination, hotel, guide(s) and flight. Kruskal’s stress was 0.12 and
the squared correlation coefficient (RSQ) 0.97, both indicating a very good fit.
The results are limited by the small number of emotions that were included in the study. Although
there is not as distinct a circular order as found by Russel (1980) and Mano (1990), the two
dimensions suggested by Russell are distinguishable in the plot. The vertical axis, or Dimension 1,
corresponds to his ‘Arousal-Sleepiness’ dimension, while the horizontal axis, or Dimension 2,
corresponds to the ‘pleasure-displeasure’ dimension. ‘Excitement’ (e1-4), ‘happiness’ (h1-4)
and ‘calmness’ (c1-4) are pleasurable emotions, but represent different degrees of arousal.
‘Surprise’ (su1-4), ‘idleness’ (i1-4), ‘boredom’ (b1-4), ‘sadness’ (s1-4) and ‘fear’ (f1-4) are
less pleasurable emotions, and also represent different degrees of arousal. Excitement and
14
surprise are the most arousing emotions, while idleness, calmness and boredom are the least
arousing.
Note that the emotions are in the expected circular order of surprise (north), excitement (north-
east), happiness (east), calmness (south-east), and idleness, boredom, sadness and fear (south to
west). These last four are clustered rather thickly on the western side of the horizontal axis. In
other respects the plot is fairly similar to Figure 2. When the overall measures of satisfaction with
the hotel, destination, guide and flight were added to the MDS analysis all measures of satisfaction
were placed directly on the second dimension, together with happiness. The results thus
confirmed findings by other researchers that when satisfaction is added to the circumplex model it
is placed near to happiness and pleasure (Oliver 1997).
Dimension 1
3210-1-2
Dim
ensi
on 2
1.0
.5
0.0
-.5
-1.0
-1.5
su1su3
su2su4
f1
c1
b1
e1
i1
s1h1
f2
c2
b2
e2
i2s3
h3
f2
c2
b2
e2
i2
s2h2
f4
c4
b4
e4
i4
s4
h4
h = happiness, s = sadness, su = surprise, i =idleness, e = excitement, b = boredom, c = calmness, f = fear,1 = hotel, 2 = destination, 3 = guide(s), 4 = flight
Figure 3. Multidimensional scaling of emotions
The eight emotions were also factor analysed to see if they could be reduced to a smaller number
of dimensions, and if these would correspond to Russell’s dimensions of pleasure-displeasure and
degree of arousal, as indicated by the plot in Figure 3. Russell used both oblique and orthogonal
15
rotation in his study, and found no difference between the two solutions, since the two dimensions
were not correlated (Russell, 1980). No differences were found between the two types of
rotation, and the results from the orthogonal rotation are presented. The eight emotions were
factor analysed separately for the destination, the hotel, the guide and the flight home, and
reliability coefficients were computed. The eigenvalues indicated that four factors were too many
for this data set, and when three factors were extracted, the third factor explained only a small
proportion of the total variance. Both three- and two-factor solutions are presented in Table I.
Table I. Three- and two-factor solutions for emotions. The factor loadings are multiplied by100.
Emotions Destination Hotel Guide Flight
F. 1 F. 2 F. 3 F. 1 F. 2 F. 3 F. 1 F. 2 F. 3 F. 1 F. 2 F. 3
Because the emotion ‘surprise’ loaded on several factors and lowered the reliability statistics, it
was left out of the analyses. Surprise can be experienced in both a positive and a negative sense,
and although it seems to have been used without further specification in several previous studies
(e.g. Evrard and Aurier, 1994; Hausknecht, 1988; Westbrook and Oliver, 1991), it facilitates the
16
analysis if it is specified as either positive or negative surprise. Oliver (1997) also notes that the
emotion surprise is valence-neutral and requires a label. For example, in the study by Liljander
and Strandvik (1997) it was explicitly defined as positive surprise, and then clearly loaded with
the other positive emotions.
The similarities of the factors across the destination, hotel, guide(s) and flight are obvious. When a
two-factor rather than a three-factor solution was produced, the factor for positive emotions
(happiness, excitement and calmness) remained intact, while the two factors consisting of
sadness/fear and boredom/idleness were joined into one factor of negative emotions. Thus, it can
be noted that the two-factor solutions do not correspond to Russell’s two dimensions of
pleasure-displeasure and degree of arousal, and bear a closer resemblance to Watson and
Tellegen’s dimensions of positive and negative affect. These results are also consistent with other
research on consumption emotions in which Izard’s DES scales have been used and two factors
of positive and negative affect have been found (e.g. Westbrook, 1987; Liljander and Strandvik,
1997). When Watson and Tellegen (1985) reanalysed previous studies of self-reported mood,
they found that positive affect consistently emerged as the primary factor of emotions. These
results differ from theirs in that positive affect emerged as the first factor only for the emotions
related to the destination.
The factor positive affect consists of emotions which are labelled high positive affect (excitement),
pleasantness (happiness) and low negative affect (calmness) in Watson and Tellegen’s model. It is
easy to see why these emotions were experienced as positive emotions on a holiday. Calmness
may be a very desirable emotion when one wants to have a good, relaxing holiday. The four
emotions that were loaded on negative affect are less desirable: sadness, fear, boredom and
idleness. However, it should be noted that some consumers may consider idleness a positive
holiday emotion. It was also noted earlier that, in some circumstances, fear can be combined with
positive excitement (Price, Arnould and Tierney 1995). However, sadness and boredom are
unlikely to be considered positive emotions in any circumstances.
The proportion of the total variance that was explained by the two factors is also similar to that in
Watson and Tellegen (1985), who found that one half to three quarters of the variance could be
explained by positive and negative affect. The factors here explained between 57% and 68% of
the total variance, but they had low reliability coefficients: 0.70 or less for the destination, the hotel
17
and the flight. The recommended levels for the coefficient alpha varies depending on the research
purpose. Although coefficients of 0.6 are often considered acceptable (Nunnally, 1978), Peterson
(1994) gives stricter guidelines: a coefficient of around 0.7 may be accepted for preliminary
research, around 0.8 is considered moderately high and adequately reliable for basic research,
while 0.9 or higher is considered highly reliable and is the recommended level for applied
research. The results of the present study may have been affected by the very general context
given for the sources of the emotions. Both the interpretation of the emotions and their sources
may vary between persons.
Patterns of emotion and satisfaction
Clusters of emotions have previously been studied by Westbrook and Oliver (1991) and Oliver
and Westbrook (1993) in two studies on car owners, and by Liljander and Strandvik (1997) in a
study on employment services. These studies were based on Izard’s DES scale. Four to six
different clusters were found, ranging from mainly negative emotions to very positive ones. All
three studies also found clusters of consumers who had experienced a very limited number of any
type of emotions. Another common factor in the studies was that consumer satisfaction was found
to be relatively high in all the clusters. Not even the consumers with the highest negative emotions
could be labelled dissatisfied, only less satisfied than the others. On the other hand, Liljander and
Strandvik (1997) found that only in the ‘angry/humiliated’ group did emotions directly affect
satisfaction, i.e. negative emotions lowered the satisfaction score.
The present data was cluster analysed separately for the destination, hotel, guide(s) and flight.
‘Surprise’ was dropped from the analyses because its mean did not differ between the clusters in
any of the two-cluster solutions or in most of the three-cluster solutions. The two-cluster solutions
could generally be described as having one cluster with high positive and low negative emotions,
and another cluster with some presence of all the emotions. Four-cluster solutions were
characterised by two of the clusters being quite similar in outline: either one scale value apart from
the others, or with a larger difference on one of the emotions. The three-cluster solutions are
presented in Figure 4. The mean satisfaction scores for each cluster are also given. There are only
small differences between the clusters in the scores.
18
Only the clusters of ‘delighted’ consumers are similar across the destination, hotel, guide(s) and
flight. It is also the largest cluster in all four cases. These consumers had strong feelings of
happiness, excitement and calmness, and only a small amount of idleness, boredom, sadness or
fear. In this study, only the emotions connected with the guide(s) yield an ‘emotionless’ cluster.
Note, however, that there was almost no difference in satisfaction means between the emotionless
group and the third cluster showing moderate strength of all emotions.
Another way of looking at patterns of emotion and satisfaction would be to investigate to what
extent different satisfaction scores can be related to different patterns of emotion (see Dubé, 1990
and Stauss and Neuhaus, 1997 for examples of the use of pre-specified affective categorisations
in questionnaires). However, the present study would have needed a bigger sample including a
larger number of dissatisfied customers in order to find out if customers with the same satisfaction
score experienced different kinds of emotional patterns. An exploratory analysis of satisfaction
scores which had an adequate number of observations revealed that different emotional patterns
can be found within one satisfaction score.
19
Destination
1
2
3
4
5
6
7
HAPPY EXCITED CALM IDLE BORED SAD AFRAID
cluster 1 (131)
cluster 2 (80)
cluster 3 (46)
Satisfaction with destination
5.3 (cluster 2)
5.8 (cluster 3)6.3 (cluster 1)
Hotel
1
2
3
4
5
6
7
HAPPY EXCITED CALM IDLE BORED SAD AFRAID
cluster 1 (149)
cluster 2 (61)
cluster 3 (38)
Satisfaction with hotel
4.7 (cluster 2)
5.8 (cluster 1)5.2 (cluster 3)
Guide
1
2
3
4
5
6
7
HAPPY EXCITED CALM IDLE BORED SAD AFRAID
cluster 1 (139)
cluster 2 (39)
cluster 3 (57)
Satisfaction with guide
4.8 (cluster 2)
6.0 (cluster 1)5.1 (cluster 3)
Figure 4. Four diagrams of clusters of experienced emotions at thedestination, hotel, with the guide and during the flight.
Flight
1
2
3
4
5
6
7
HAPPY EXCITED CALM IDLE BORED SAD AFRAID
cluster 1 (109)
cluster 2 (99)
cluster 3 (47)
Satisfaction with flight
5.5 (cluster 2)
6.1 (cluster 1)
5.9 (cluster 3)
20
Discrete emotions and satisfaction
First, it may be pointed out that all the satisfaction measures were significantly correlated, although
the strength of the correlation varied. Overall satisfaction with the holiday correlated most highly with
destination satisfaction (0.66) and least with guide satisfaction (0.25). Correlations between overall
holiday satisfaction and hotel and flight satisfaction were of equal strength (0.42 and 0.43). These
results should be interpreted with some caution since it is not known to what extent they are affected
by the fact that the questionnaires were answered on the flight home. The correlation between flight
and overall holiday satisfaction seems to be inflated. Although the success of a charter holiday can be
said to be the result of the combination of choosing the right destination and a suitable hotel, having
good guides on tours and enjoying the last moments on the flight home, different customers may well
weight these parts differently. Most tour operators use the same national airline, and although some
tourists may choose a specific operator because of the airline, the choice of destination and hotel are
likely to be more important. On the other hand, the airline may genuinely have a large effect on
overall holiday satisfaction by giving a last impression of the holiday. A good flight may make
consumers remember the holiday with warm feelings, while a disappointing flight makes them
remember all the negative aspects of the trip. The small correlation between guide satisfaction and
overall satisfaction does not necessarily mean that the guides are not important. If one looks at the
dispersion of the satisfaction scores one can see that the tourists were in fact very satisfied with the
guides’ performance. It should also be remembered that when negative critical incidents occur on a
holiday, the guides are of paramount importance in solving the problems to the tourist’s satisfaction.
In a study of consumer perceptions of tour leaders at three different holiday destinations, Mossberg
(1995) shows that the leaders are important contributors to consumer satisfaction.
The correlations between discrete emotions and satisfaction are reported in Table II. All the
intercorrelations are not reported. The upper part of the table shows the correlations between the
emotions related to the destination, hotel, guide(s) and flight, and satisfaction with these. Thus the first
column reports the correlation between emotions connected with the destination and overall
destination satisfaction, the second row reports the correlation between emotions connected with the
hotel and overall hotel satisfaction, etc. The lower part of the table shows the correlations between
the emotions for the destination, hotel, guide(s) and flight, and overall satisfaction with the trip. The
21
first column shows the correlation between emotions connected with the destination and overall
satisfaction with the trip, etc.
Table II. Correlations between satisfaction and emotions
EmotionsSatisfaction with
destinationSatisfaction with
hotelSatisfaction with
guideSatisfaction with
flight home
Happiness 0.41** 0.37** 0.55** 0.34**
Excitement 0.38** 0.24** 0.42** 0.18**
Calmness 0.26** 0.19** 0.33** 0.11
Surprise -0.05 -0.19** -0.10 -0.18**
Sadness -0.35** -0.27** -0.24** -0.19**
Fear -0.25** -0.18** -0.10 -0.01
Boredom -0.45** -0.35** -0.26** -0.31**
Idleness -0.31** -0.31** -0.27** -0.27**
EmotionsDestination emotionoverall satisfaction
with the trip
Hotel emotionoverall satisfaction
with the trip
Guide emotionoverall satisfaction
with the trip
Flight emotionoverall satisfaction
with the trip
Happiness 0.42** 0.39** 0.17* 0.27**
Excitement 0.31** 0.18** 0.07 0.19**
Calmness 0.24** 0.14* 0.09 0.15*
Surprise -0.10 -0.09 -0.03 -0.12
Sadness -0.37** -0.32** -0.21* -0.35
Fear -0.22** -0.24** -0.06 -0.04
Boredom -0.40** -0.42** -0.27** -0.32
Idleness -0.29** -0.33** -0.16* -0.26
• = p<= 0.05, ** = p<= 0.01
As indicated by the factor analysis, however, and keeping in mind the low reliability scores, the
emotions may be divided into positive and negative ones. Looking at Table II one may draw the
general conclusion that the positively experienced emotions are positively correlated with satisfaction,
while negatively experienced emotions have the opposite effect. However, it should be noted that
only some of the correlations were statistically significant. Happiness, excitement and calmness may
also be interpreted as goal-directed emotions that the consumer wishes to experience when going on
a holiday. Of these happiness correlated most highly with satisfaction for all the four aspects of the
holiday investigated. Excitement and calmness were also positively related to satisfaction, and
especially to satisfaction with the destination and the guide(s). Somewhat surprisingly, calmness was
not related to satisfaction with the flight. As can be seen from the clusters presented in Figure 4,
22
perceptions of calmness varied only slightly within the flight emotion clusters, while there was a
greater variation within the destination, hotel and guide(s) clusters. Although the tourists had
experienced calmness on the flight, it was not an important determinant of satisfaction.
‘Surprise’ was included in the correlation analyses, but as can be seen in Table II, this emotion
correlated slightly only with hotel and guide satisfaction. Both correlations were negative, indicating
that the surprise was of a negative nature and therefore decreased satisfaction. The results for ‘fear’
resemble ‘surprise’ in that it generally had the smallest correlation compared with the other emotions.
The presence of fear had a slightly negative effect on satisfaction with the hotel and the destination,
but no effect on satisfaction with the guide(s) and the flight. Although it is well known that many
consumers feel some degree of fear of flying, this was not a significant factor in the present study.
Those who fear flying the most may not be likely to choose a holiday that includes flying, and those
who are nervous about flying may not wish to call it fear.
Of the other emotions, sadness, boredom and idleness were all negatively related to satisfaction.
Sadness and boredom were more closely related to (dis)satisfaction with the destination than with
the other parts of the holiday. These emotions could be labelled reactive emotions that are not sought
after on a holiday. Sadness may be difficult for the tour operator to alleviate, but boredom could be
avoided by offering different activities and by helping people choose a destination that suits their
needs best. Idleness could be experienced as a positive emotion by consumers who want to get
away from their work and stress, and who enjoy lying in the sun being idle. However, at least on an
aggregate level in this study, idleness was clearly experienced as a negative emotion which correlated
negatively with satisfaction.
The correlation between emotions and overall satisfaction with the trip, which are presented in the
lower part of Table II, shows that emotions experienced at the destination and at the hotel were
more important than those connected with the guide(s) and flight in forming overall holiday
satisfaction. Only a few of the emotions experienced with the guide(s) and on the flight correlated
significantly with satisfaction. Surprise with any part of the service did not correlate with overall
satisfaction.
The relation between satisfaction and emotions can also be illustrated by looking at different levels of
satisfaction and the related emotion scores. An example is given in Figure 5 for the destination.
23
Because of the limited number of observations, the scale values 1-3 have been combined into one
value and labelled ‘satisfaction = 3’.
1
2
3
4
5
6
7HAPPY
EXCITED
CALM
IDLEBORED
SAD
AFRAID
satisfaction = 3
satisfaction = 4
satisfaction = 5
satisfaction = 6
satisfaction = 7
Figure 5. Experienced emotions at different levels of overall satisfaction with destination
The figure shows that the higher the satisfaction score, the more positive and less negative emotions
the consumers experienced. The corresponding figures for the hotel, guide(s) and the flight were very
similar and are not presented. However, it can be noted that, for the flight, the relation between
satisfaction and the negative feelings of boredom and idleness was accentuated. Those who had
marked one of the three lowest points on the satisfaction scale for the flight had strong experiences of
boredom and idleness, while those who had marked the highest score for satisfaction had very low
experiences of boredom and idleness compared with in-between values for satisfaction.
The relation between discrete emotions and satisfaction was also investigated for each of the clusters
presented in Figure 4. In light of the small number of observations in some of the clusters, and the
low factor reliability scores, the results should be looked on as exploratory and no detailed figures
are presented. Correlation matrices showed that different emotions affected satisfaction in the
clusters. Despite the low reliabilities of the two factors (positive and negative emotions), the means
for these were computed from the data, and satisfaction within the different clusters was regressed
24
on the factors. The results showed that in all the ‘delighted’ clusters (first clusters), and cluster 2 for
destination, both the presence of positive and absence of negative feelings had a positive impact on
satisfaction. In none of the clusters did only negative feelings affect satisfaction, as was found by
Liljander and Strandvik (1997). There was no relation between emotions and satisfaction in four of
the clusters (cluster 3 for destination, clusters 2 and 3 for the hotel, cluster 3 for the flight). In the
remaining clusters, satisfaction was only affected by positive emotions (clusters 2 and 3 for guide(s)
and cluster 2 for flight). No general conclusions can be drawn from these results.
DISCUSSION
This study investigated the effect of eight discrete emotions on satisfaction with a holiday destination,
hotel, guide(s) and return flight. The emotions represented the eight quadrants of Russell’s (1980)
circumplex model of affect. However, the two affect dimensions that were found in the data did not
correspond to Russell’s dimensions of pleasure and arousal, but did to the dimensions of positive and
negative affect suggested by Watson and Tellegen (1985) for mood states, and also found in
research on consumption emotions connected with consumer goods (Westbrook and Oliver, 1991)
and services (Westbrook, 1987; Liljander and Strandvik, 1997).
A charter holiday could be described as a service which is high on goal-directed emotions, i.e.
pleasant, positive emotions that the holidaymaker wishes to experience, but which may also awake in
him/her many reactive, unplanned emotions. The parts of the holiday package that were investigated
here were all of an extended duration (Price, Arnould and Tierney, 1995), which means that there
was ample time for the consumers to experience different emotions in connection with them. More
positive than negative emotions were found to be associated with the different parts of the holiday,
and cluster analyses revealed a group of delighted consumers that was very similar across the
destination, hotel, guide(s) and flight. The delighted consumers were characterised by strong feelings
of happiness, excitement and calmness, and by only slight feelings of idleness, boredom, sadness or
fear. It was the largest cluster, which also seems natural in view of the nature of a holiday, where
mostly positive emotions should be experienced. None of the clusters contained consumers with only
strong negative emotions, but the presence of negative emotions was accompanied by a lower level
of satisfaction. The nature of the service seems to be such that some positive emotions are felt by all
consumers. In contrast to other studies in which researchers have reported clusters of consumers
with almost no emotions (Westbrook and Oliver, 1991 on car ownership and Liljander and
25
Strandvik, 1997 on employment agencies), here such a cluster was found only for the guide services.
On the basis of the findings from the effect of discrete emotions on satisfaction, it can also be
concluded that the most important task for tour organisers is to promote feelings of happiness in the
consumer and to avoid feelings of boredom and idleness.
Emotions experienced in relation to the destination and the hotel were found to be most strongly
related to overall satisfaction with the holiday. The destination can be considered the core service
that the customer buys when he/she purchases a holiday. The hotel is also an important part of the
holiday package. The customers interact with the hotel and the destination as a whole during their
entire stay, and their perceptions of these are crucial for overall satisfaction.
A limitation of the present study was that the triggers of emotions could not be identified. Consumers
may, for example, react with negative emotions such as anger and disappointment if they perceive
they get poor service at the hotel, from the guide(s) or in the local shops and restaurants. They may
feel sadness for people living in poverty, or when they think of a beautiful island being exploited by
tourism. They may feel afraid of being robbed or cheated, or even be afraid of terrorist attacks, or
they may feel boredom at having nothing to occupy their time with. The nature of the triggers may
affect their satisfaction and future behaviour towards the travel agent and the tourist resort. A specific
restaurant, shop or hotel can be avoided in future, and one negative experience does not have to
affect the consumer’s attitude and behaviour towards the tourist resort as a whole, but it may do so.
Another limitation of the study was that only an intensity scale was used to measure emotions. Both
intensity and frequency scales can been used but frequency scales are recommended for goods or
services with which the consumers have a long experience and for products with multiple features
(Oliver, 1997). Since the destination, hotel, guide(s) and flight all could be described as having
multiple features a frequency scale could have been used. However, it was also possible for the
consumers to have only a very brief encounter with the guide on the bus from the airport to the hotel
and back. Because of that, and because customer relationships were not in focus here, an intensity
scale was chosen for the present study. However, in our opinion future studies should include both
intensity and frequency scales to get a better understanding of both the degree and frequency of the
emotional experiences. If, for example, a negative emotion is experienced often with a service
provider, and the cause of the emotion is considered relevant to the evaluation of the total service
experience, the consumer is more likely to react with defection than if the emotion is experienced
only once and the causing factor is relatively unimportant.
26
Gardial et al. (1996, p. 39) note that “the causes of intense emotions that are relevant to post-
purchase evaluations have not been explored, nor has the notion of emotion incubation”. In order to
investigate the triggers of emotions, qualitative studies such as critical incident studies are needed. A
related research question that remains unanswered is the role of attributions both as a mediator of
triggers of emotional responses (Oliver, 1993), and as a possible mediator of emotions of
satisfaction. Dubé et al. (1996), for example, found in a study on health care that negative emotions
could be divided into situation-attributed emotions and emotions attributed the actions of other
persons (e.g. suspicion, feeling neglected). Only the latter kind had a negative effect on different
satisfaction measures.
It is in the interests of both the tour organiser and the tourist resort to know what kind of emotions
are evoked in consumers. Both positive and negative emotions should be identified so that the tour
organiser may try to strengthen the positive and abate the negative when possible. Although it is
possible for the tour organiser to influence holidaymakers’ expectations about service standards and
local conditions, many of the consumer’s experiences during his/her holiday are outside the tour
organiser’s control. The customer’s interaction with the hotel can be controlled to some degree by
the organiser by keeping a close eye on the hotel’s service standards and listening to customers’
views. It is also possible to investigate what delights customers, what makes them happy and how
the service may be improved. The choice of hotels may be changed, but local conditions such as the
presence of beggars, high criminality, polluted air and water and terrorist attacks are not under the
tour organiser’s control. However, the organiser may draw the local government’s attention to
possible problems and offer to assist in solving some of them. Thus, the most important strategy is to
manage consumer expectations. By giving a realistic picture of the resort and trying to find the
holiday package that best matches a particular consumer’s needs, unnecessary negative reactive
emotions and subsequent dissatisfaction may be avoided.
The present study did not look at consumers’ future intentions to return to the same resort and/or to
use the same tour organiser another time. However, future research should investigate how different
experienced emotions affect consumer loyalty. Although consumers may seem to be relatively
satisfied, and to feel relatively strong positive emotions and only a few negative ones, this may not be
enough to retain them as future customers. Jones and Sasser (1995) argue that only completely
satisfied consumers, i.e. those who mark the highest value on a satisfaction scale, remain loyal to the
company. It is not enough to merely satisfy consumers, because a satisfied consumer may still easily
27
switch to a competing brand. Dubé (1990), on the other hand, suggests that emotions may be more
closely related to future purchasing intentions than cognitive evaluations are. Thus, if emotions were
also taken into account when investigating consumer loyalty, it may be found that (completely)
satisfied consumers who have experienced extremely positive emotions are more loyal than those
who are (completely) satisfied but who have only neutral emotions.
One may also ask how much negative affect a consumer will tolerate without defecting. A dynamic
approach is needed to investigate both emotion incubation and the related concept of tolerance
zones of emotions. For example, the type and strength of emotions that are awakened by one or
more negative incidents in a service relationship may explain defection better than the source (type of
service quality failure) that caused the emotions. This remains to be investigated, as does the role of
service recovery in managing to change consumer emotions. Research has shown that service failure
which is solved by giving the consumer a sincere apology and fair compensation (e.g. reduced price)
can be turned into a good memory of satisfactory service and result in positive word-of-mouth
feedback. Although consumers' emotions before and after service recovery have not been
investigated, it can be assumed that negative emotions may also be turned into positive ones.
In order to understand the effect of experienced emotions on the consumer’s future interactions with
the service, there is a need to investigate to what extent emotions are remembered, retrieved and
transferred from one encounter to the next. Some may be very short-term, confirming Dabholkar’s
(1995) suggestion of the temporality of emotions, while others may be remembered for a long time
and have a profound effect, either positive or negative, on consumer loyalty.
Many different classifications of services have been put forward in service management literature,
including degree of tangibility, degree of variability, pattern of demand and extent of contact between
service provider and customer (Lovelock, 1983; Palmer, 1994; Shostack, 1977). However, little is
known about differences in emotional experiences between services. With the exception of some
illustrative examples of low/high affect services presented by Price, Arnould and Tierney (1995), no
attempts have been made to classify services according to their emotional content. It is apparent,
however, that services differ in the type and strength of emotions that they evoke in consumers. A
theatre performance is related to stronger emotions than a fast food service encounter, and some
services may be related to many negative emotions (e.g. dentists), while others may be related to
mainly positive emotions (e.g. a Mediterranean cruise). Services may also vary according to the
presence of goal-directed and reactive emotions, and according to the relative importance of these in
28
influencing satisfaction. For example, in the present study the reactive emotion ‘boredom’ loaded on
negative emotions and had a negative effect on service satisfaction. On the other hand, in a study by
Dubé et al. (1996) on health care, boredom did not load on either of the two factors with negative
emotions. It may be noted that Richins (1997) excluded boredom from her proposed Consumption
Emotions Set (CET), a general tool for studying emotions. She argued that boredom was more
suitable for studying emotions connected to advertising than consumption emotions. The present
study clearly showed that boredom is an important factor in satisfaction with holiday consumption.
Although the CET (Richins 1997) was developed from several empirical studies and is bound to be
very useful as a basis for future research, it is still partly dependent on the four consumption situations
that were used to construct it and is not all-embracing. For example, the author excludes emotions
associated with plays and movies from her study. Clearly, more empirical research is needed to
understand consumption emotions in different service settings and situations.
It may be concluded that the impact of different emotional responses on consumer satisfaction with
services has not received enough attention in conceptual discussions of service evaluation or in
empirical services research. With the exception of Bitner (1992), whose model of the effects of the
physical environment on consumer behaviour includes mood in both employee and consumer
responses, emotions have not been included as a contributing factor in service models. There is
clearly a need to amend this oversight, since the few studies that have included emotions clearly
indicate the importance of the concept in gaining a better understanding of overall service evaluations.
Notes
1 (The scale values are: Delighted, Pleased, Mostly Satisfied, Mixed, Mostly Dissatisfied, Unhappy, Terrible(Neutral, Never Thought About It).)
2 (The data were collected by students Mikael Berndtson, Ann-Lis Henriksson-Santaharju, Peter Strömberg andJan Virtavuori for a course on service management.)
3 (Note that it is not possible to separate mood and emotions in this survey.)
References
Alford, B. L. and D. L. Sherrell (1996), “The role of affect in consumer satisfaction. Judgment ofcredence-based services”, Journal of Business Research, Vol. 37, pp. 71-84.
Arnould, E. J. and Price, L.L. (1993), “River magic: extraordinary experience and the extendedservice encounter”, Journal of Consumer Research, Vol. 20, June, pp. 24-45.
29
Bagozzi, R. P., Baumgartner, H. and Pieters, R. (1995), “Goal-directed emotions”, Working papers,School of Business, University of Michigan, pp. 1-43.
Bitner, M. J. (1992), “Servicescapes: the impact of physical surroundings on customers andemployees”, Journal of Marketing, Vol. 56, April, pp. 57-71.
Brown, S. P., Cron, W. L. and Slocum, Jr., J. W. (1997), “Effects of goal-directed emotions onsalesperson volitions, behavior, and performance: a longitudinal study”, Journal ofMarketing, Vol. 61, January, pp. 39-50.
Cadotte, E. R., Woodruff, R. B. and Jenkins, R. L. (1983), “Norms and expectations: how differentare the measures?”, in Day, R. L. and Hunt, H. K., (Eds), International Fare InConsumer Satisfaction and Complaining Behavior, Bloomington, IN, IndianaUniversity, pp. 49-56.
Cohen, J. B. and Areni, C. S. (1991), “Affect and consumer behavior”, in Robertson, T. S. andKassarjian, H. H. (Eds), Handbook of Consumer Behavior, Prentice-Hall Inc.,Englewood Cliffs, New Jersey, pp. 188-240.
Clark, M. S. and Isen, A. M. (1982), “Toward understanding the relationship between feeling statesand social behavior”, in Hastorf, A. and Isen, A. M. (Eds), Cognitive Social Psychology,Elsevier/North Holland, Amsterdam.
Dabholkar, P. A. (1995), “The convergence of customer satisfaction and service quality evaluationswith increasing customer patronage”, Journal of Consumer Satisfaction, Dissatisfactionand Complaining Behavior, Vol. 8, pp. 32-43.
Dubé, L. (1990), “The power of affective reports in predicting satisfaction judgements”, in Goldberg,M. E., Gorn, G. and Pollay, R. W. (Eds), Advances in Consumer Research, Vol. XVII,Provo, UT, Association for Consumer Research, pp. 571-76.
Dubé, L., Bélanger, M-C. and Trudeau, E. (1996), “The role of emotions in health caresatisfaction”, Journal of Health Care Marketing, Vol. 16, No. 2, pp. 45-51.
Elster, J. (1996), “Rationality and emotions”, The Economic Journal, 106, September, pp. 1386-97.
Evrard, Y. and Aurier, P. (1994), “The influence of emotions on satisfaction with movieconsumption”, Journal of Consumer Satisfaction, Dissatisfaction and ComplainingBehavior, Vol. 7, pp. 119-25.
Gardial, S. F., Flint, D. J. and Woodruff, R. B. (1996), “Trigger events: exploring the relationshipsbetween critical events and consumers’ evaluations, standards, emotions, values andbehavior”, Journal of Consumer Satisfaction, Dissatisfaction and ComplainingBehavior, Vol. 9, pp. 35-51.
Gardner, M. P. (1985), “Mood states and consumer behavior: a critical review”, Journal ofConsumer Research, Vol. 12, December, pp. 281-300.
Hausknecht, D. (1988), “Emotional measures of satisfaction/dissatisfaction”, Journal of ConsumerSatisfaction, Dissatisfaction and Complaining Behavior, Vol. 1, pp. 25-33.
Hui, M. K. and Tse, D. K. (1996), “What to tell consumers in waits of different lengths: anintegrative model of service evaluation”, Journal of Marketing, Vol. 60, April, 81-90.
30
Izard, C. E. (1972), Patterns of Emotions, Academic Press, New York, NY.
Izard, C. E. (1977), Human Emotions, Plenum Press, New York, NY.
Izard, C. E. (1991), The Psychology of Emotions, Plenum Press, New York, NY.
Jones, T. O. and Sasser, Jr., W. E. (1995), “Why satisfied customers defect”, Harvard BusinessReview, November-December, pp. 88-99.
Koelemeijer, K., Prevo, O., Pieters, R. and Roest, H. (1995), “Service evaluation process: aninvestigation into hierarchies-of-effects”. Presented at Workshop on Quality Managementin Services V, Tilburg, The Netherlands, 11-12 May, 1995, organised by TilburgUniversity and European Institute for Advanced Studies in Management.
Lazarus, R. S. (1984), “On the primacy of cognition”, American Psychologist, 39, pp. 124-29.
Lazarus, R. S. and Smith C. A. (1988), “Knowledge and appraisal in the cognition-emotionrelationship”, Cognition and Emotion, 2, 281-300.
Liljander, V. and Strandvik, T. (1997), “Emotions in service satisfaction”, International Journal ofService Industry Management, Vol. 8, No. 2, pp. 148-69.
Lovelock, C. H. (1983), “Classifying services to gain strategic marketing insights”, Journal ofMarketing, Vol. 47, Summer, pp. 9-20.
Mano, H. (1990), “Emotional states and decision making”, in Goldberg M. E., Gorn, G. and Pollay,R. W. (Eds), Advances in Consumer Research, Vol. XVII, pp. 577-84.
Mano, H. and Oliver, R. L. (1993), “Assessing the dimensionality and structure of the consumptionexperience: evaluation, feeling and satisfaction”, Journal of Consumer Research, Vol. 20,December, pp. 451-66.
Mossberg L. L. (1995), “Tour leaders and their importance in charter tours”, TourismManagement, Vol. 16, No. 6, pp. 437-45.
Morris, W. N. (1989), Mood: the frame of mind, Springer-Verlag New York, Inc., New York.
Norušis M. J. (1993), SPSS for Windows Professional Statistics Release 6.0 (manual), SPSSInc., Chicago, IL.
Nunnally, J. C. (1978), Psychometric Theory, 2nd ed., McGraw-Hill, Inc., New York.
Oliver, R. L. (1993), “Cognitive, affective, and attribute bases of the satisfaction response”, Journalof Consumer Research, Vol. 20, December, pp. 418-30.
Oliver, R. L. (1997), Satisfaction. A behavioral perspective on the consumer. McGraw-Hill,Inc., New York.
Oliver, R. L. and Westbrook, R. A. (1993), “Profiles of consumer emotions and satisfaction inownership and usage”, Journal of Consumer Satisfaction, Dissatisfaction andComplaining Behavior, Vol. 6, pp. 12-27.
Palmer, A. (1994), Principles of Services Marketing. McGraw-Hill, London.
Peterson, R. A. (1994), “A meta-analysis of Cronbach’s coefficient alpha”, Journal of ConsumerResearch, Vol. 21, September, pp. 381-91.
31
Price, L. L., Arnould, E. J. and Deibler, S. L. (1995), “Consumers’ emotional responses to serviceencounters, International Journal of Service Industry Management, Vol. 6, No. 3, pp.34-63.
Price, L. L., Arnould, E. J. and Tierney, P. (1995), “Going to extremes: managing service encountersand assessing provider performance”, Journal of Marketing, Vol. 59, April, pp. 83-97.
Richins, M. L. (1997), “Measuring emotions in the consumption experience”, Journal of ConsumerResearch, Vol. 24, September, pp. 127-46.
Russell, J. A. (1980), “A circumplex model of affect”, Journal of Personality and SocialPsychology, Vol. 39, No. 6, pp. 1161-78.
Shostak, G. L. (1977), “Breaking free from product marketing”, Journal of Marketing, Vol. 41,April, pp. 73-80.
Stauss, B. and Neuhaus, P. (1997), “The qualitative satisfaction model”, International Journal ofService Industry Management, Vol. 8, No. 3, pp. 236-49.
Storbacka, K., Strandvik, T. and Grönroos, C. (1994), “Managing customer relationships for profit:the dynamics of relationship quality”, International Journal of Service IndustryManagement, Vol. 5, No. 5, pp. 21-38.
Strongman, K. T. (1996), The psychology of emotion: theories of emotion in perspective, 4th
ed., John Wiley & Sons Ltd., Chichester.
Watson, D. and Tellegen, A. (1985), “Toward a consensual structure of mood”, PsychologicalBulletin, Vol. 98, No. 2, pp. 219-35.
Westbrook, R. A. (1980), “A rating scale for measuring product/service satisfaction”, Journal ofMarketing, Vol. 44, Fall, pp. 68-72.
Westbrook, R. A. (1983), “Consumer satisfaction and the phenomenology of emotions duringautomobile ownership experiences”, in Day, R. L. and Hunt, H. K. (Eds), InternationalFare in Consumer Satisfaction and Complaining Behavior, Bloomington, IN, IndianaUniversity, pp. 2-9.
Westbrook, R. A. (1987), “Product/consumption based affective responses and postpurchaseprocesses”, Journal of Marketing Research, Vol. 24, August, pp. 258-70.
Westbrook, R. A. and Oliver, R. L. (1991), “The dimensionality of consumption emotion patternsand consumer satisfaction”, Journal of Consumer Research, Vol. 18, June, pp. 84-91.
32
APPENDIX
Table I. Dispersion of answers on emotions and satisfaction measures
Percent of marks for each scale valueVariable 1 2 3 4 5 6 7 Mean Std.dev.