Joy and Surprise As Guides to a Better Understanding of Impulse Buying Behaviour

ABSTRACT - Meanwhile, affective processes have become a common key factor for the theoretical discussion of impulse purchasing behaviour. In our paper, impulse buying behaviour will be defined as stimulus-controlled, spontaneous buying behaviour that is accompanied by strong positive emotions and low cognitive control. The approach is based on the theoretical framework of Izard’s (1977) Differential Emotion Theory. An empirical approach which is clearly based on an affective definition of impulse purchasing in retail settings, differentiating this special type of unplanned purchases from other types, is still missing. A measurement approach that directly assesses the unplanned purchase-related positive emotions as well as low cognition is repeated in four empirical field surveys of different scenarios. The results reveal identical cluster patterns throughout the studies. Basically, two types of impulse purchasers were identified: One group’s impulse purchase is determined by joy, surprise, and low cognition (Asurprised impulse purchasers@); the other group’s impulse purchase is accompanied by joy and low cognition only (Adelighted impulse purchasers@). As expected, both groups of impulse purchasers agree strongly and significantly more to the prevalence of product related desire as reason to shop, than consumers in other groups of unplanned purchases.

Citation:

Dorothea Baun and Andrea Groeppel-Klein (2003) ,"Joy and Surprise As Guides to a Better Understanding of Impulse Buying Behaviour", in E - European Advances in Consumer Research Volume 6, eds. Darach Turley and Stephen Brown, Provo, UT : Association for Consumer Research, Pages: 290-299.

European Advances in Consumer Research Volume 6, 2003      Pages 290-299

JOY AND SURPRISE AS GUIDES TO A BETTER UNDERSTANDING OF IMPULSE BUYING BEHAVIOUR

Dorothea Baun, European University Viadrina, Germany

Andrea Groeppel-Klein, European University Viadrina, Germany

ABSTRACT -

Meanwhile, affective processes have become a common key factor for the theoretical discussion of impulse purchasing behaviour. In our paper, impulse buying behaviour will be defined as stimulus-controlled, spontaneous buying behaviour that is accompanied by strong positive emotions and low cognitive control. The approach is based on the theoretical framework of Izard’s (1977) Differential Emotion Theory. An empirical approach which is clearly based on an affective definition of impulse purchasing in retail settings, differentiating this special type of unplanned purchases from other types, is still missing. A measurement approach that directly assesses the unplanned purchase-related positive emotions as well as low cognition is repeated in four empirical field surveys of different scenarios. The results reveal identical cluster patterns throughout the studies. Basically, two types of impulse purchasers were identified: One group’s impulse purchase is determined by joy, surprise, and low cognition ("surprised impulse purchasers"); the other group’s impulse purchase is accompanied by joy and low cognition only ("delighted impulse purchasers"). As expected, both groups of impulse purchasers agree strongly and significantly more to the prevalence of product related desire as reason to shop, than consumers in other groups of unplanned purchases.

UNPLANNED VERSUS IMPULSE BUYING BEHAVIOUR

Literature review

Despite contributions in literature on impulse buying ranging back to the fifties, a profound lack in understanding of this purchasing pattern still prevails. Authors of earlier contributions tried to explain impulse purchasing by addressing in detail the impact of external stimuli (e.g., Cox 1964, Kelly 1965) and product categories (e.g., Bellenger, Robertson, and Hirschman 1978, Kollat and Willett 1966). However, there was a lack in correspondence in research results, which made Kollat and Willett (1969) finally wondering whether the concept of impulse buying is useful in order to derive any insights for marketing strategies. The difficulties in research efforts may be due partially to the fact that impulse product choice has for a long time been implicitly or explicitly equated with unplanned purchases (e.g., Appelbaum 1951, Patterson 1963, Cox 1964, Kelly 1965, Kollat and Willett 1967, Kollat and Willett 1969, Bellenger, Robertson, and Hirschman 1978).

Kollat and Willett (1969) assume that other factors that have so far been neglected, might have a major impact on impulse buying decisions. They do not further discuss such factors, however, as one of the possibly neglected factors, Kollat and Willett (1969) suggest the perception process: The consumer is free to perceive some stimuli at the point of purchase (which might lead to an impulse purchase), and filter out others (which might prevent an impulse purchase).

Affective processes Stern (1962) implicitly points to, might also be such neglected factors. Stern (1962) distinguishes between four types of impulse purchases: One of these categories is "pure impulse buying" which he defines as a "novelty or escape purchase which breaks a normal buying pattern" (Stern 1962, p. 59). Such purchasing decisions are described as a buying behaviour that breaks clearly with rational buying habits. "Pure impulse buying" is distinguished from other categories. These are the "reminder impulse purchase" (the perception of the product in the store reminds the consumer that he/she actually needs it) and the "planned impulse buying" (the consumer enters the store with no clear purchasing intention, but with special price offers he/she expects and is in fact willing to buy). As a further category, Stern (1962) lists the "suggestion impulse buying": The consumer sees a product for the first time and immediately develops a need for it. Stern (1962) concedes that in such situations, the consumer can still think about his or her buying decision and may finally end up with a rational choice, but a rational choice need not necessarily to be the outcome.

Stern (1962) with his "novelty or escape purchase which breaks a normal buying pattern" and Kollat and Willett (1969), reveal what researchers are actually looking for: An explanation of a decision making phenomenon which cannot be explained by familiar rational decision rules and which is not directly and solely determined by externally given stimuli. Such a phenomenon requires a consideration of affective processes.

With more recent contributions, intervening affective processes were accorded more and more attention (e.g., Weinberg and Gottwald 1982, Rook and Hoch 1985, Rook 1987, Gardner and Rook 1988, Piron 1991, Dittmar, Beattie, and Friese 1995, Rook and Fisher 195, Beatty and Ferrell 1998, Shiv and Fedorikhin 2002, Ramanathan and Menon 2002). Beatty and Ferrell (1998) found that a positive affect "...produced more felt urges to buy impulsively". Gardner and Rook (1988) report that "pleasure", "excitement", and "content/relaxation" were on the first three ranks of diverse mood states selected by test persons to describe best the feeling that most likely followed an impulse purchase made in the past. By means of an experiment, Weinberg and Gottwald (1982) show that those who impulsively bought a product were significantly more amused, delighted, and enthusiastic than non-buyers. Dittmar, Beattie, and Friese (1995) investigated impulse purchases in terms of self-identity and their experiment demonstrated, that men were more likely than women to buy on impulse to express their uniqueness. Also, there are impressive and highly innovative experimental attempts in more recent contributions that investigate the influence of hedonic goals on impulsive decisions and the process of their automaticity (Ramanathan and Menon 2002).

So, meanwhile affective processes became a common key factor for the discussion of impulse purchasing behaviour in literature. However, an empirical approach which is clearly based on an affective definition of impulse purchasing in retail settings is still missing.

Impulse purchasing behaviour versus other types of unplanned buying B what makes the difference?

Several distinctions need to be made in order to formulate a clear definition of impulse purchasing behaviour. In the first place, we need to distinguish the impulse to purchase from the impulse purchase. The impulse to purchase can eventually be fought off leading to no purchase at all or lead to an unplanned but rational purchase (Rook and Hoch 1987).

Weinberg and Gottwald (1982) point out the importance of affective and cognitive factors in order to distinguish impulsive buying from extensive, rational buying decisions. According to their definition, impulse purchasing is accompanied by high arousal and low cognition, and is a stimulus-controlled, spontaneous buying behaviour. Extensively planned, well thought over and rational buying decisions may also be accompanied by high arousal and strong emotions. Yet, contrary to impulse purchases, there are also strong cognitive processes involved.

FIGURE 1

DISTINGUISHING IMPULSE PURCHASES FROM OTHER TYPES OF UNPLANNED PURCHASES

Weinberg and Gottwald’s (1982) definition of impulse purchasing is one which clearly puts emotions and low cognition in the focus of impulse buying. Therefore, we will use their definition. However, Weinberg and Gottwald’s (1982) definition of impulse purchase will be extended: We wish to distinguish compulsive purchasing clearly from impulse purchasing (Cole and Sherrell 1995, Valence, d’Astous, and Fortier 1988, Dittmar, Beattie, and Friese 1995). While the compulsive purchaser is characterised by strong emotions, high cognition and high reactivity to strong internal and irresistible urges (Valence, d’Astous, and Fortier 1988), the impulsive purchaser shows strong emotions and low cognition, and the purchasing decision is not accompanied by uncontrollable inner urges. Also, contrary to the compulsive purchase, the impulse purchase is not accompanied or followed by strong feelings of guilt.

For a more precise understanding, we further wish to distinguish unplanned purchases accompanied by strong cognitive information processing, from those with low cognitive information processing. The former are non-spontaneous, but still unplanned upon entering the store. The latter still require further differentiation: It can be assumed that not every unplanned spontaneous buying decision together with low cognitive control, is necessarily accompanied by strong emotions like the impulse purchase. Therefore, by modifying Stern’s (1962) categories of "impulse" purchases, we will in theory assume the following categories of unplanned, spontaneous purchases of low cognitive control (see figure 1): (1) The "(pure) impulse purchase" which is unplanned and base purely on positive emotions and low cognition; (2) The "unplanned reminder purchase" which is equivalent to Stern’s (1962) notion of the reminder impulse purchase, yet, we will not label it as impulse purchase but as one type of unplanned purchase of low cognition; (3) The "unplanned purchase due to special price offers" covers Stern’s (1962) "planned impulse buying"; and finally a further category (4) the "unplanned substitution" as an unplanned purchase to substitute an item that was planned to be bought, but could not be found in the store to the customer’s full satisfaction.

Stern’s (1962) "suggestion impulse purchase" is considered as an impulse purchase if it is accompanied by strong emotions and low cognition. It is considered a rational unplanned, but not impulse purchase if intense cognitive information processing is involved.

To conclude, we assume the impulse purchase to be only one possible type of unplanned purchase among several other categories, and we assume further that cognitive and affective processes determine the difference between impulse purchases and other unplanned purchases.

EMOTIONS

Bagozzi, Gopinath, and Nyer in 1999 came to the conclusion that "little consistency can be found in the use of terminology related to emotions" (p. 184). Yet, arousal can be considered a fundamental determinant of emotions (Bagozzi, Gopinath, and Nyer 1999, Ohman, Esteves, Flykt, and Soares 1993). Two features seem to be of special interest when explaining emotions: The change of bodily functions (muscle activity, stimulation of autonomic nervous system, general arousal) and the cognitive interpretation (appraisal) of an internal or external stimulus.

The complex interchange between emotional, cognitive and motor processes is addressed by the Differential Emotion Theory of Izard (1977). Izard (1977) defines emotions as complex processes with neurophysiological, neuro-muscular and phenomenological aspects (Izard 1977). Ten basic emotions are distinguished, each leading to different consequences of inner perception and behaviour, and ranging from a weak intensity on the lower end of a dimension, to a strong intensity on the higher end of the same dimension. The ten basic emotions are as follows (Izard 1977): (1) interestBexcitement, (2) enjoymentBjoy, (3) surpriseBstartle, (4) distressBanguish, (5) angerBrage, (6) disgustBrevulsion, (7) contemptBscorn, (8) fearBterror, (9) shame/shynessBhumiliation and (10) guiltBremorse.

Not all basic emotions seem to be of major relevance to impulse buying decisions. Negative emotions such as anxiety, fear, and guilt feeling B either as origins or consequences of unplanned purchases B refer rather to compulsive than to impulse purchasing behaviour (Valence, d’Astous, and Fortier 1988, Shapiro 1981, O’Guinn and Faber 1989). Futhermore, only if positive emotions are evoked, can we expect an approach behaviour and a positive action tendency towards a product (Mehrabian and Russell 1974, Donovan and Rossiter 1982, Tai and Fung 1997, Groeppel-Klein 1997, Groeppel-Klein and Baun 2001, van Kenhove and Desrumaux 1997).

According to Izard (1977), interest and joy, and under certain circumstances, surprise, can be considered positive emotions. However, interest does not seem to be an appropriate emotion for distinguishing impulse purchases from other unplanned purchasing decisions. Compared to all other basic emotions, interest is described as being the most important emotion but it is also one that most frequently accompanies selective perception and cognitive processes, as well as attention. Interest is additionally fundamental to learning processes and to the development of competence and intelligence of a growing child (Izard 1977).

As basic emotions, joy and surprise deserve a closer look with respect to impulse purchasing behaviour. Basically, no specific stimuli are needed to perceive the emotion joy. Joy may accompany the perception of any external, but also of any internal stimuli. Joy may go along with creativity, the successful completion of an action, or with the discovery of something. Joy accompanies the ability to enjoy, with commitment towards an object, the feeling of strength, vitality, superiority, freedom, and harmony. An object that is perceived with joy is perceived and enjoyed as a whole, rather than taken apart and analysed. There is a strong feeling of harmony and integration towards the object instead of the desire to dominate it. The emotion joy can best be described as if looking at the world through "rose-colored spectacles": Every bird sings even more beautifully, the sky has a much brighter blue, and even though it may be raining cats and dogs, the day is seemingly terrific (Izard 1972, 1977).

Surprise can be described as follows: Something that suddenly and unexpectedly occurs or happens, leads to an increase in phasic arousal, and at the same time, produces an orienting reflex towards the object. Such psycho-physiological processes lead to the emotion surprise on the individual’s perception level. Cognitive processes accompany the emotion surprise only to a limited extent. The basic function of surprise is to focus the individual’s entire attention to a specific stimulus in the environment (Izard 1977).

As a basic emotion, surprise can neither be categorised as basically positive nor as a negative emotion, because it may lead equally to a positive or a negative outcome for individuals (Izard 1977). The perception of surprise as positive or negative depends on the individual. Those who mostly experience positive consequences following surprise, evaluate this emotion as positive. Those who most often have to face negative results, consequently describe it as a negative emotion. The latter are in general more anxious and afraid of unexpected and unfamiliar situations than the former (Izard 1977). Izard’s (1977) empirical study showed, that the majority of test subjects of a student population reported positive associations with surprise. Vanhamme’s and Snelders’s (2003) experiment shows that surprise has an effect on the positive evaluation of the surprising product attribute and also that variation in the level of surprise has a direct effect on consumers’ satisfaction. Since impulse purchasing implies an approach behaviour towards a product we can assume a positive connotation of surprise.

Hypotheses

We assume that interest is important to an individual’s stimulus selection. Also, a perceived stimulus may trigger an impulse to purchase. Yet, interest is a rather learning-oriented, "analytical" emotion and, considering the characteristics of interest, one could even discuss whether it should be considered a basic emotion at all. An impulse to purchase which is dominated by interest can be expected to lead to high cognitive processes revealed, for instance, by asking the sales personnel for further information, reading the description on the package, testing the functionality of the product and so on. Interest can be supposed to be important in order to evoke an impulse to purchase and to deal with the product in some way. But instead of an impulse purchase, any other type of unplanned purchase is likely to occur in the end.

H1: If diverse types of unplanned purchases can be detected, interest will be equally of relevance to all unplanned purchasers and thus interest will not be an emotion by which impulse purchasers significantly differ from consumers who make some other kind of unplanned purchase.

At this point, the question arises as to whether unplanned purchases made on impulse and other unplanned purchases can be distinguished significantly by the core factors joy, surprise, and low cognition. Therefore, with the four studies of different scenarios described in the following paragraphs, we investigate whether impulse purchasing behaviour can validly be assessed by directly measuring the two positive emotions surprise and joy, as well as low cognition, while relating all factors to the unplanned purchasing process of a specific item.

We can assume that surprise is an emotion most relevant for triggering impulse purchases. The high phasic arousal that draws the consumer’s attention towards the product may more or less directly lead to a purchase decision, since surprise is accompanied by low cognitive control.

Risk perception may be reduced and creative thinking with respect to the product may be enhanced by the emotion joy. A feeling of integration and harmony is even more likely, the more the product or brand is suitable to fulfil the customer’s emotional needs. The pleasant, carefree and self-confident experience that is evoked from joy reduces the inner inhibition of (buying) impulses (Izard 1977), thus it becomes more likely that an impulse to purchase is actually turned into an impulse purchase.

If an impulse purchase is determined by strong positive emotions and low cognitive control, the consumer should, as a shopping reason, consequently state a product-related desire for his/her unplanned buying behaviour, while those who did not make an unplanned purchase on impulse should rather reject such product related desires as shopping reasons. Therefore, in order to investigate whether, among consumers who make an unplanned purchase, we can actually identify impulse purchasers, we will investigate the following hypothesis:

H2: If among consumers who make an unplanned purchase, there is a group of test subjects who perceive the positive emotions of joy and surprise significantly stronger and show significantly higher values on the low cognition factor, than other consumers, this group will agree significantly more strongly to a product related desire as a shopping reason, than those who make some other type of unplanned purchase.

Referring to the distinction made earlier in figure 1, impulse purchasers should not conform to Stern’s (1962) modified unplanned purchasing categories "reminder unplanned purchase", "special price offer", and "substitution".

FOUR EMPIRICAL STUDIES

Research designs

The results reported here are based on four empirical studies conducted between 1998 and 2001. We applied more or less identical procedures in four independent and differing research scenarios of real shopping environments in order to explore whether the results can be replicated and thus prove to be stable. For all studies, standardised questionnaires and five-point rating scales to express level of disagreement/agreement were applied.

For study 1, conducted in June 1998, 156 customers of a grocery store were asked, after leaving the cashier’s zone, whether they could recall any unplanned, spontaneous purchase they had made recently. The unplanned purchase did not have to be made in the grocery store they were about to leave, because we did not want to limit our sample to groceries. N=63 test subjects of the total sample had previously made an unplanned purchase, the others could not recall any. The latter were not interviewed except for their demographic criteria. Results from study 1 are based on the 63 subjects who recalled an unplanned purchase.

Study 2 was conducted on three successive days in June 1999 in a hallway of a local, medium-sized shopping mall. The relevant area for the research included all products displayed by the store owners in the hallway and in the shop windows. Everything inside the stores was disregarded. Customers in other parts of the mall were asked randomly to participate in our study. Test subjects were accompanied to the beginning of the selected hallway. They were then instructed to walk up and down both sides of the hallway as if they were on a shopping tour, but were not allowed to enter any of the stores. On their "trip" they were to pick one product of their choice, but not to buy it because we did not want our results to be confounded by thoughts about what test subjects can afford. They were asked to report their product choice when returning from their trip. After the test persons had walked through the hallway and made their product decision, they were interviewed. The questions were related to the product that had been selected. Valid data was collected from a total sample of n=78 test subjects.

Study 3 was conducted in June 2000 in a European capital in one of Europe’s most famous department stores. Customers were asked randomly in diverse places in the store, whether they had just made an unplanned purchase decision for a product in that department store. Those who answered in the negative were not interviewed. A total sample of n=104 said "yes" and were interviewed by means of a standardised questionnaire with the questions related to one product they had just decided to purchase unplanned.

Study 4 was conducted in May 2001 under the same conditions and in the same department store as study 3. The total sample encompasses n=135 test subjects, and n=68 of the total sample reported an unplanned buying decision they had just made in the store. All others had made planned purchasing decisions only. The results reported below on study 4 will be based on the sub-sample of unplanned purchases.

Methodology

The first task that test subjects received in all four studies, was to recall the moment they had first noticed the item they had bought unplanned. They were then asked to state on a rating scale, whether the items to measure the basic emotions joy and surprise would more or less describe their feelings at that specific moment. In a second step, the consumers were asked to rate several statements with respect to the cognitive effort involved in the unplanned decision process.

This approach offers two important advantages: Firstly, we need not confront test subjects with a direct question like "did you buy something on impulse". Such a question would leave it to the consumers how to define impulse purchasing, so that results could barely be compared between test subjects. Secondly, we omit the disadvantages concomitant with the provision of a "definition" of impulse purchasing like for instance "an emotional unplanned purchase", or "a sudden urge to buy". If we gave test subjects a definition right at the beginning of the survey, such a "theoretical" definition might be too complicated for "ordinary" consumers and/or we would reduce impulse purchasing to a very narrow viewpoint. Furthermore, test subjects would know right from the beginning what the survey is all about.

Scales

Positive emotions. Emotions in all four studies were measured by applying items of the Differential Emotions Scale (DES) (Izard 1977). The DES originally covers all ten positive and negative basic emotions. Each is assessed by three adjectives. We focussed only on the relevant dimensions of interest (for the first study only), joy, and surprise. Test subjects of all four samples were asked to rate their perceived emotions on a five-point rating scale ranging from (1) "does not describe my feelings at all" to (5) "exactly describes my perceived feelings". The items were presented in the questionnaire as unsorted list.

Comments of test subjects from a pretest for study 1, suggested that respondents may get tired if expected to rate six emotional adjectives in a row. Therefore we reversed some of the items to a negative coding (see table 1).

Factor analysis (principal components analysis) (Srivastava 2002, Hair et al. 1998, Tacq 1997) on the emotion scale in study 1, revealed anti-image values below 0.5 (AI=.356) for "unconcentrated" and "sad". The item "sad" was the only one with a high factor loading on a third component together with "joyful" and "surprised". Therefore, we excluded both items from further analysis. The remaining items of study 1 yielded three components (Kaiser criteria), equivalent to the original DES classification. (The item "tired" was re-coded prior to factor analysis.) In the studies 2 to 4, original DES-items only were applied. The items show high loading on the components "joy" and "surprise" as expected from the original DES. Table 2 shows a summary of the results of all four principal components analyses.

Low cognition. Test subjects were to rate the statements assessing low cognition as a state variable on a rating scale ranging from (1) "totally disagree" up to (5) "totally agree". By means of principle components analysis, the three statements shown in detail in table 3, yield one single factor "low cognition" throughout the studies 1, 3 and 4. In study 2, only one single cognition statement was applied to keep the interview as short as possible.

Product related desire. Shopping motives were investigated in the studies 1, 3 and 4. We applied two statements to measure product related desire, which were also to be rated on a five point rating scale ranging from (1) "totally disagree" to (5) "totally agree". Additionally, we measured Stern’s (1962) modified categories of non-impulsive unplanned purchasing. If our distinction from the above holds true, these categories ("reminder unplanned", "special price offer", and "unplanned substitution") should not apply to those who made an unplanned purchase on impulse (see table 4 for a list of all statements).

TABLE 1

ITEMS OF JOY, INTEREST, AND SURPRISE

TABLE 2

PRINCIPLE COMPONENT ANALYSES ON THE EMOTION ITEMS

RESULTS

Positive emotions and low cognition as segmentation criteria

For all four studies, we applied hierarchical clustering (single linkage procedure; dissimilarity measure: squared Euclidean distance) in a first step to discover and exclude outliers from the samples (Timm 2002). In a second step, we applied hierarchical clustering (Ward’s method; dissimilarity measure: squared Euclidean distance) to test whether unplanned purchasers can be distinguished, based on positive emotions as well as on the "low cognition" factor. Following the Elbow criteria (Timm 2002), we found that three groups appear most suitable. Cluster centres were calculated by means of ANOVA.

Results H1. For study 1, we considered all three emotion factors "joy", "interest", and "surprise" as well as the "low cognition" factor. By means of hierarchical clustering, we received three sub-groups with significant differences for joy, surprise, and low cognition. However, testing H1, discriminant analysis as well as the F-Test reveal no significant differences between the three groups for the emotion "interest" (mean_interest1=-.284, mean_interest2=.285, mean_interest3=-.231; Wilks-Lambda_interest=0.934; p_Wilks=0.124; F=2.163, p_F=0.124) which confirms H1.

Test of H2. In order to test H2, all interest items are henceforward excluded from further analysis of study 1 [In order to ensure comparability with study 2, 3, and 4, a new factor model was calculated for study 1, excluding the interest items (components and factor loadings: "joy": glad=.891, joyful=.887; MSA=.659; total variance explained=80.45%).] and as a consequence also of study 2, 3, and 4. The results for all four studies are summarised in table 5.

As the results of the four studies show, we canidentify different types of unplanned purchasers based on the positive emotions joy and surprise, and on low cognition. Despite each of the four studies having been conducted with a slightly different design, studies 1, 2 and 4 reveal identical cluster patterns: Subjects who made an unplanned purchase in cluster 2 in study 1 and in cluster 1 in the studies 2 and 4, perceive the emotion "surprise" distinctly below average, while they perceive joy clearly above average as well as low cognition. With cluster 1 in study 1 and cluster 2 in studies 2 and 4, we identify another positive emotion/low cognition-segment. Both groups of the three studies can, following our definition from above, be labelled "impulse purchasers" due to strong positive emotions and low cognition. We have labelled the joy/low cognition-groups "delighted impulse purchasers" and the surprise/joy/low cognition groups "surprised impulse purchasers".

TABLE 3

PRINCIPLE COMPONENT ANALYSES ON COGNITION STATEMENTS

TABLE 4

PRODUCT RELATED DESIRE AND TO HER CATEGORIES OF UNPLANNED PURCHASING

A third group, is found identically in all four studies. It is characterised by low joy/high cognitive information processing with average surprise (study 1, 2, 4), or low surprise resp. (study 3). These groups are labelled the "unplanned rational purchasers".

The results of study 3 depart from the results of the other studies: Following the Elbow criteria, we have to consider a fourth group of unplanned purchasers that is missing in the other three samples: Test subjects in this group seem to make their choice without perceiving joy and surprise, but they still claim that they made a quick purchase decision without much thought. We label that group (cluster 4, study 3) "emotionless spontaneous purchasers". Those who show both positive emotions and low cognition as in studies 1, 2 and 4, could not be confirmed in the third study. However, there is a small group of unplanned purchasers in cluster 3 who were very surprised when discovering their purchased item. Yet, contrary to the "surprised" groups in the other studies, they are clearly below average for "joy" and yield average for "low cognition". Due to the average value on low cognition they will not be regarded as "impulse purchasers", but as "surprised unplanned purchasers". The joy/low cognition cluster (Cluster 1) does show average values, but not below average agreement for the emotion dimension "surprise" as in study 1, 2 and 4. This group is nevertheless regarded as delighted impulse purchasers.

To ensure the validity of the four cluster solutions, we applied discriminant analysis (Tacq 1997) for each of the cluster results with the three factors "joy", "surprise", and "low cognition" as independent variables and the three (four) clusters as dependant variable (see table 6). The results are positive in all four studies: Discriminant functions for all cluster solutions yield highly significant chi-square values, strong canonical correlation and low values for Wilks’ Lambda. Univariate Wilks’ Lambda is fairly low for all three variables throughout all four studies, as required to prove a significant difference between the cluster centres. The equivalent F-Test yields significant results. There is only one exception in study 4: Univariate discriminant statistics reveal a non-significant F-Test for the emotion "joy" (p_{joy[study 4]}=0.104).

TABLE 5

CLUSTER CENTRES OF THE GROUPS OF UNPLANNED PURCHASERS

TABLE 6

VALIDITY OF THE CLUSTER ANALYSES

In summary, we can state that consumers who make impulse purchases, can be distinguished significantly from those who make unplanned-rational purchases based on the positive emotions "joy" and "surprise", as well as on low cognition. There are two types of customer segments that must be interpreted as impulse purchasers, the "delighted" and the "surprised impulse purchasers". The unplanned-rational group does not agree on the low cognition items, rather they seem to consider their choice carefully. There might be some surprise when they first discover the product, yet, joy definitely plays a minor role. The group of spontaneous purchasers with low cognition and weak positive emotions as postulated in theory, was found only in study 3 with cluster 4.

In order to test H2, the three (four) clusters found in our samples served as the independent variable and the shopping reasons as the dependent variable. We investigated whether shopping reasons significantly differ between the three (four) groups. Due to small sample sizes (<30) within the subgroups, we applied the distribution free nonparametric Kruskal-Wallis test (SPSS Inc. 1998).

"Delighted" and "surprised impulse purchasers" perceived a significantly stronger product-related desire for the item "simply great" throughout all three studies. The same holds true for "appeared desirable" for studies 3 and 4. Therefore, we can confirm H2 with respect to both groups of impulse purchasers. As expected, other unplanned, spontaneous shopping reasons based on Stern’s (1962) modified categories, do not apply to either of the impulse purchasing groups. The weak emotions/low cognition-group in study 3 ("emotion-less spontaneous purchasers") have the highest mean agreement ranks and mean values on the five-point rating scale for the "unplanned substitution" and the "special price offer", but the difference between groups is not significant. Contrary to our original assumption, "delighted impulse purchasers" reveal the highest mean rank for the "unplanned reminder purchase" in study 3, but again, the result is not significant (see table 7).

TABLE 7

SHOPPING MOTIVES WITHIN THE GROUPS OF UNPLANNED PURCHASING

LIMITATIONS AND DISCUSSION

It remains unclear why study 3, which was conducted in the same environment and under equal conditions to study 4, remains the only one to yield a distinct cluster pattern. Constantly changing store decoration and special events in that department store, do not constitute a convincing argument, because cluster patterns in study 1, 2 and 4 are equal, despite completely differing store environments and research designs. If weather and climatic conditions do influence affective processes and, in turn, behaviour (Cunningham 1979), the extremely hot, oppressive weather conditions with heavy rainfall on some days during the third survey (some customers only entered the department store to escape the rain and steamy air outside), might have had an impact on the research results from study 3. Apart from that, we believe that our approach presented here is superior to research designs that force test subjects to cognitively define and explain purely emotional purchasing behaviour. With little exception, we succeeded in finding identical cluster patterns in four research scenarios, revealing two groups of emotion-dominated impulse purchasers with one being mainly determined by joy, the other by surprise and joy. Both groups reveal low cognition as a further characteristic of impulse purchasing. However, study 1 also revealed, that the positive emotion interest, is perceived equally by all test subjects when making unplanned purchases. Furthermore, impulse purchasers perceive a significant product-related desire as a shopping motivation.

With the findings reported here, we cannot gain any insights as to why "delighted impulse purchasers" are so happy when they find "their" product and as to what exactly surprises "surprised impulse purchasers" and what makes them happy. These questions remain subject to further investigation and might best be addressed by means of qualitative interviews. Further hypotheses and new scales on motivations for impulse shopping behaviour might be derived from such investigations. Existing emotional product involvement, subtle, unconscious dreams and desires B eventually evoked by emotional advertisements to which the consumer was exposed in the past, or evoked by the effect of affective and collative merchandising concepts presented at the point of purchase B may contribute to a complex interaction of external and internal stimuli perception which modulates and determines impulse purchasing. Therefore, the impact of joy and surprise on other affective processes and the complex interaction of affective processes still offer a wide range for further studies in impulse purchasing behaviour.

REFERENCES

Applebaum, William (1951), "Studying Customer Behaviour in Retail Stores," Journal of Marketing, 16, 1951, 172-178.

Bagozzi, Richard P., Mahesh Gopinath, and Prashanth U. Nyer (1999), "The Role of Emotions in Marketing," Journal of the Academy of Marketing Science, 27, Spring 1999, 184-206.

Beatty, Sharon E. and M. Elizabeth Ferrell (1998), "Impulse Buying: Modeling Its Precursors," Journal of Retailing, 74, 2, 169-191.

Bellenger, Danny, Dan H. Robertson, and Elizabeth C. Hirschman (1978), "Impulse Buying Varies by Product," Journal of Advertising Research, 18, December, 15-18.

Cole, Leslie and Dan Sherrell (1995), "Comparing Scales to Measure Compulsive Buying: An Exploration of Their Dimensionality," Frank R. Kardes and Mita Sujan (eds.), Advances in Consumer Research, 22, Provo, UT: Association for Consumer Research, 419-427.

Cox, Keith (1964), "The Responsiveness of Food Sales to Shelf Space Changes in Supermarkets," Journal of Marketing Research, 1, May 1964, 63-67.

Cunningham, Michael R. (1979), "Weather, Mood, and Helping Behavior: Quasi Experiments With the Sunshine Samaritan," Journal of Personality and Social Psychology, 37, 11, 1947-1956.

Dittmar, Helga, Jane Beattie, and Susanne Friese (1995), "Gender identity and material symbols: Objects and decision considerations in impulse purchases," Journal of Economic Psychology, 16, 491-511.

Donovan, Robert J. and John R. Rossiter (1982), "Store Atmosphere: An Environmental Psychology Approach," Journal of Retailing, 28, 1, Spring 1982, 34-57.

Eysenck, Sybil S. G. and Barry J. McGurk (1980), "Impulsiveness and venturesomeness in a detention centre population," Psychological Reports, 47, 1299-1306.

Gardner, Meryl P. and Dennis W. Rook (1988), "Effects of Impulse Purchases on Consumers’ Affective States," Michael J. Houston (ed.), Advances in Consumer Research, 15, 1988, Provo: UT, 127-130.

Groeppel-Klein, Andrea (1997), "The Influence of the Dominance Perceived at the Point-of-Sale on the Price-Assessment," Basil G. Englis and Anna Olofsson (eds.), European Advances in Consumer Research, 3, Provo, UT: Association for Consumer Research, 304-311.

Groeppel-Klein, Andrea and Dorothea Baun (2001), "The Role of Customers’ Arousal for Retail StoresBResults from An Experimental Pilot Study Using Electrodermal Activity as Indicator," Mary C. Gilly and Joan Meyers-Levy (eds.), Advances in Consumer Research, Vol. XXVIII, Provo: Utah.

Hair, Joseph F. Jr., Rolph E. Anderson, Ronald L. Tatham, and William C. Black (1998), Multivariate Data Analysis, 5^th ed., Upper Saddle River, NJ: Prentice-Hall.

Izard, Carroll E. (1972), Patterns of emotions: A new analysis of anxiety and depression, New York: Academic Press

Izard, Carroll E. (1977), Human Emotions, New York: Plenum Press.

Kelly, Robert F. (1965), An Evaluation of Selected Variables of End Display Effectiveness, Unpublished Doctoral Dissertation, Havard University, 1965.

Kollat, David T. and Ronald T. Willett (1966), "Interbrand Variation in Unplanned Purchasing Rates," unpublished working paper, April 1966.

Kollat, David T. and Ronald T. Willett (1967), "Consumer Impulse Purchasing Behavior," Journal of Marketing Research, 4, February, 21-31.

Kollat, David T. and Ronald T. Willett (1969), "Is Impulse Purchasing Really a Useful Concept for Marketing Decision?," Journal of Marketing, 33, January, 79-83.

Mehrabian, Albert and James A. Russell (1974), An Approach to Environmental Psychology, Cambridge, MA: MIT.

O’Guinn, Thomas C., Ronald J. Faber (1989), "Compulsive Buying: A Phenomenological Exploration," Journal of Consumer Research, Vol. 16, September 1989, 147-157.

Ohman, Arne, Francisco Esteves, Anders Flykt, and Joaquim J. F. Soares (1993), "Gateways to Consciousness: Emotion, Attention, and Electrodermal Activity," J.-C. Roy et al. (eds.), Progress in Electrodermal Research, Plenum Press: New York, 137-157.

Patterson, Lawrence W. (1963), In-Store Traffic Flow, New York: Point-of-Purchasing Advertising Institute.

Piron, Francis (1991), "Defining Impulse Purchasing," Rebecca H. Holman and Michael R. Solomon (eds.), Advances in Consumer Research, 18, Provo, UT: Association for Consumer Research, 509-514.

Ramanathan, Suresh and Geeta Menon (2002), "Hedonnic Motives, Spontaneous Affect and Impulsive Behavior," http://pages.stern.nyu.edu/~gmenon/Impulsive%20Decisions.pdf, working paper.

Rook, Dennis W and Robert J. Fisher (1995), "Normative Influences on Impulsive Buying Behavior," Journal of Consumer Research, 22, December, 305-313.

Rook, Dennis W. (1987), "The Buying Impulse," Journal of Consumer Research, 14, Sept. 1987, 189-199.

Rook, Dennis W. and Stephen J. Hoch (1985), "Consuming Impulses," Morris B. Holbrook and Elisabeth C. Hirschmann (eds.), Advances in Consumer Research, 12, 1985, Provo: UT, 23-27.

Shapiro, David (1981), Autonomy and rigid character. New York: Basic Books.

Shiv, Baba and Alexander Fedorikhin (2002), "Spontaneous versus Controlled Influences of Stimulus-Based Affect on Choice Behavior," Organizational Behavior and Human Decision Processes, 87, 2, March, 342-370.

SPSS Inc. (1998), SPSS^" Base 8.0 Applications Guide, Chicago, IL: SPSS Inc.

Srivastava, M. S. (2002), Methods of Multivariate Statistics, New York: John Wiley & Sons.

Stern, Hawkins (1962), "The Significance of Impulse Buying Today," Journal of Marketing, 26, April, 59-62.

Tacq, Jacques (1997), Multivariate Analysis Techniques in Social Science Research, London, Thousand Oaks, New Delhi: Sage.

Tai, Susan H. C. and Agnes M. C. Fung (1997), "Application of an environmental psychology model to in-store buying behaviour," The International Review of Retail, Distribution, and Consumer Research, 7, 4, October 1997, 311-337.

Timm, Neil H. (2002), Applied Multivariate Analysis, New York: Springer.

Valence, Gilles, Alain d’Astous, and Louis Fortier (1988), "Compulsive Buying: Concept and Measurement," Journal of Consumer Policy, 11, 419-433.

Vanhamme, Jodlle and Dirk Snelders (2003), "What if you surprise your customers... will they be more satisfied? Findings from a pilot experiment", Punam A. Keller and Dennis W. Rook (eds.), Advances in Consumer Research, Vol. XXX, Provo: Utah (upcoming).

van Kenhove, Patrick and Patrick Desrumaux (1997), "The relationship between emotional states and approach or avoidance responses in a retail environment," The International Review of Retail, Distribution, and Consumer Research, 7, October, 351-368.

Weinberg, Peter and Wolfgang Gottwald (1982), "Impulsive Consumer Buying as a Result of Emotions," Journal of Business Research, 10, March 1982.

----------------------------------------