Intern. J. of Research in Marketing 14 Ž1997. 1–17 Store atmosphere, mood and purchasing behavior Kordelia Spies a , Friedrich Hesse b b,) , Kerstin Loesch a a UniÕersity of Gottingen, Department of Psychology, Gottingen, Germany ¨ ¨ German Institute for Research on Distance Education, Applied CognitiÕe Science Department, Konrad-Adenauer-Str. 40, D-72072 Tubingen, Germany ¨ Received 15 June 1995; accepted 7 June 1996 Abstract The effects of store characteristics on customers’ mood, on their satisfaction, and on their purchasing behavior are investigated. Two furniture stores differing with regard to their atmosphere, i.e. their condition, information rate and layout, were selected. Customers’ mood – measured at the beginning, in the middle and at the end of their shopping – was shown to improve in the pleasant and to deteriorate in the less pleasant store Ž n s 76 for each store.. Satisfaction with the store was greater in the pleasant store. Regression analyses showed that this was due to a direct effect of store atmosphere as well as to an indirect effect mediated by customers’ mood. Customers in the pleasant store spontaneously spent more money on articles they simply liked. This effect was only due to customers’ mood. Keywords: Customers’ mood; Store atmosphere; Purchasing behavior 1. Theoretical considerations Most shoppers share the experience that, irrespective of the stock offered, some stores are more attractive than others, some stores induce a feeling of wellbeing, while in other stores one becomes irritated or even angry. And everybody also knows that one tends to buy more things and to spend more money when one is in a positive rather than in a negative mood state. Thus, there might well be important interactions between store characteristics, customers’ mood and purchasing behavior. A broad theoretical model of environmental psychology that may be applied to these interactions is presented by Mehrabian and Russell Ž1974.. It is ) Corresponding author. Tel.: Žq49. 7071.979215; fax: Žq49. 7071 979100; e-mail: friedrich.hesse@uni-tuebingen.de. assumed that stimulus characteristics, especially the information rate, influences a person’s mood state as an intervening variable. Mood state, on the other hand, exerts influence on a person’s response, i.e. on his or her approach or avoidance behavior. In the following article we aim to investigate more closely what kind of relationship there is between store characteristics and customers’ behavior, taking into account customers’ mood state as an intervening variable. Two lines of research have to be taken into account: Ža. consumer research as a specially applied field of environmental psychology, and Žb. general psychology as far as the generation of mood and its influences on cognitive as well as behavioral processes are concerned. How are physical aspects of a situation perceived and evaluated by the individual? According to Berlyne Ž1971., the attractiveness of environmental 0167-8116r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 8 1 1 6 Ž 9 6 . 0 0 0 1 5 - 8 2 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 stimuli is a function of their complexity. Stimuli that are characterized by an optimal level of complexity are assumed to gain maximum attractiveness, whereas stimuli deviating from this optimal level towards a higher or lower complexity are considered less attractive. Important characteristics of environmental stimuli that determine their complexity are information rate and layout. By information rate we mean the number of information units that are presented within a certain time interval. The information rate usually is higher for new and unusual than for familiar stimuli. Thus, to assess the information rate in a store Bost Ž1987. asked customers whether they considered the store to be modern or obsolete, to be out of the ordinary or commonplace, to be surprising or boring, to be interesting or uninteresting etc. In addition, perceptibility should become an important factor, with easy perceptibility lowering the information rate. With respect to a store, the perceptibility of articles may be increased, for instance by providing better lighting. Careful layout of an environment helps people to orientate, to find the way and learn to understand signs, to get the feeling of personal control and mastery ŽBitner, 1992.. According to Bost Ž1987., the successful layout of a store depends on whether the store has a clear concept, whether one can easily find things, whether different departments are clearly separated from each other, whether one does not get lost etc. Signs and information tables can help to improve the layout of a store. Studies investigating customers’ behavior in a store Že.g. Barth, 1993. showed that certain layout patterns were especially attractive for customers. It was found that most customers move through the store in a counterclockwise direction with their attention being concentrated on the wall-sides. They tend to avoid turns and are rather reluctant to accept any attempts to divert the direction in which they are going. Following Berlyne Ž1971., it may be assumed that if the information rate is too high customers will feel overloaded, while they will become bored if the information rate is too low. Similar assumptions can be made for layout. The layout should be clear but not too simple, so that there is the possibility of surprise and unexpectedness. How do these stimulus characteristics relate to customers’ mood state? Here, general psychology research dealing with the generation of mood is of interest. Like Bower and Cohen Ž1982. or Lazarus Ž1991., Scherer Ž1984. postulates that a cognitive interpretation of a situation precedes and determines the emotional reaction Žsee also Kuhl, 1983a.. Scherer assumes that there are five stages of cognitive interpretation, called stimulus evaluation checks, with different qualities of emotions being elicited at each stage. At the first stage, the novelty and unexpectedness of physical stimuli are evaluated. For moderate novelty and unexpectedness interest should occur as emotional reaction, while for a high degree of novelty and unexpectedness subjects tend to react in a frightened way. At the second stage, a stimulus is evaluated as good or bad according to prior experiences. Here emotions of pleasure and displeasure are commonly induced. At the third stage, a situation is appraised with respect to its goal relevance. If a situation is favorable to the attainment of a person’s present goal, joy and contentment will be elicited. If however, the situation is incongruent with the goal a person will show anger or fear Žsee also Lazarus, 1991.. The fourth stage refers to coping potential. If a person feels that he or she can cope with the situation, competence and delight will be induced, otherwise the person will experience anger or fear. At the fifth stage, a comparison with norms and standards takes place, stimulating feelings of pride or shame and guilt. From Scherer’s approach it can be deduced that the novelty and unexpectedness of physical stimuli in a store should not be too low, in order to make the customers interested, but on the other hand they must not be too high, as in that case customers would react with fright. This corresponds to Berlyne’s assumptions concerning the relationship between complexity and attractiveness. Within consumer research, similar assumptions are made by Donovan and Rossiter Ž1982. or Gardner Ž1985.. Furthermore, in order to induce pleasure, store characteristics should remind customers of other positive experiences, like, for instance, those one expects from holidays or leisure time. In this case, the pleasure response, originally induced by the positive experiences should be associated with store characteristics, e.g. via classical conditioning Žsee for example Cohen, 1990; Kroeber-Riehl, 1984.. The situation in the store K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 should be relevant and congruent to the customer’s goal in order to elicit contentment and happiness. In this respect Russell and his co-workers ŽRussell and Snodgrass, 1987; Snodgrass et al., 1988; Ward et al., 1988. found that how a person feels about an environment is influenced by the purpose for being there. Furthermore, customers should get the feeling that they can cope with the shopping situation, that their personal control is high. Hui and Bateson Ž1991. could show that perceptions of personal control in a shopping situation are positively related to increased pleasure. The feeling of personal control may be induced by a clear layout of the store Že.g. Bitner, 1992.. In Scherer’s approach, comparisons with norms and standards mainly refer to a person’s evaluation of his or her own behavior. This stimulus evaluation check, however, may be relevant for the finding of Parasuraman et al. Ž1990. that customers compare tangibles of a store with what they think a store should look like. Satisfaction with the store could be shown to increase with decreasing a gap between customers’ perceptions and expectations. With respect to assumptions referring to cognitive and behavioral consequences of positive emotions, in the present context mood-effects on evaluations and on risk-taking are of special interest. Isen et al. Ž1978. studied the effect of positive emotions on the evaluation of consumer goods, e.g. car, television etc., and found that the ratings of subjects in a positive mood were much more favorable than those of the neutral control group. Similar results were reported by Dawson et al. Ž1990., Gardner Ž1985., Sherman and Smith Ž1987. or Srull Ž1983.. Obermiller and Bitner Ž1984; see also Bitner, 1992. found that retail products are evaluated more positively in an environment eliciting pleasant compared to unpleasant emotional reactions. These effects of positive mood are in accordance with the mood-congruity effect described by Bower Ž1981.. In a positive mood positive aspects of a situation are assumed to come to mind more easily, to be more accessible than negative aspects. As Morris states, a person in a positive mood seems ‘‘to view the world through rose-coloured glasses’’ ŽMorris, 1989, p. 87. . Bower Ž1981. explains this mood-congruity effect within the network theory of emotion. In a positive mood, the corresponding emotion node is activated. This activation spreads to neighbouring nodes. 3 Neighbouring nodes that are associated with a positive emotion node usually represent positive things, events or aspects of a situation. Thus, through the activation of the positive emotion node other nodes that represent positive contents should be preactivated, so facilitating the perception of positive aspects of a situation. Schwarz Ž1990. claims that a positive mood serves as additional information that is used when evaluating a certain situation. If a person feels good, he or she may attribute this positive feeling to characteristics of the present situation and thus evaluate this situation more favorably. This should hold true also for the evaluation of retail products. With respect to risk-taking behavior, Isen and her co-workers ŽIsen, 1987; Isen and Geva, 1987; Isen et al., 1982; Isen et al., 1988; Isen and Patrick, 1983. found that the readiness to take a risk rises in a positive compared to a neutral mood for low risks, while it decreases if the risk-level is high. This differential effect depending on risk-level may be explained by a motivational tendency towards mood maintenance ascribed to subjects in a positive mood. If the risk is low the expected positive outcomes are assumed to outweigh negative consequences, as their subjective probability and their value is high Žsee Spies et al., 1996.. For high risk-levels, however, there is the possibility of severe negative outcomes that may endanger the positive mood and thus should be prevented. With respect to consumer research it has been found ŽBelk, 1975; Donovan et al., 1994; Golden and Zimmer, 1986; Sherman and Smith, 1987. that customers in a positive mood often spend more money than originally planned. Besides, impulse buyers could be shown to be characterized by a positive emotional state Že.g. Weinberg and Gottwald, 1982.. However, in these studies no differentiation has been made with respect to risk-level. While there are several studies within consumer research dealing with the effects of either store atmosphere or mood on customers’ behavior, only few studies investigate the effects of store characteristics on customers’ behavior taking mood as an intervening variable. Donovan and Rossiter Ž1982. asked graduate business students to visit one of several different types of retail environments and to rate Ža. aspects of the information rate, Žb. their pleasure and arousal and Žc. their intentions to approach or avoid 4 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 the store. It could be shown that pleasure was a major predictor for approach-avoidance responses. However, the effects of information rate on mood as well as on shopping behavior were rather weak. Bost Ž1987. studied genuine customers in five supermarkets selling food, clothing, automobile accessories etc. and differing with respect to the condition the stores were in. The customers’ evaluation of information rate and layout of the store, their moodchange from entering to leaving the store, their evaluation of the store’s stocks and their purchasing behavior were assessed. It could be shown that the customers’ mood improved more frequently in stores which were kept in good condition. In addition, positive mood-changes could be observed for those customers especially who evaluated information-rate and store layout as high. Positive mood-changes could be shown to cause a more favorable evaluation of the store and to make customers buy more things and make more spontaneous purchases, the last being defined as the number of articles which customers did not intend to buy or which they bought for the first time. Thus, the results of Bost’s study in particular are in accordance with the assumptions derived from the theoretical and empirical considerations mentioned above. In the following investigation, instead of correlational analyses we did a quasi-experiment. Two stores differing with respect to store atmosphere were selected and customers’ mood, satisfaction with the store and purchasing behavior were investigated as dependent variables. Second, as did Bost Ž1987. but not Donovan and Rossiter Ž1982. and as was claimed by Donovan et al. Ž1994., we studied mood-change by assessing mood at the beginning, in the middle and at the end of shopping. By controlling for initial mood, it becomes more likely that observed mood-effects are really due to differences in store characteristics. Third, we investigated stores where experienceoriented marketing plays a major role. The concept of experience-oriented marketing ŽGroppel, 1991. ¨ assumes that customers who visit a store not only intend to buy certain things but also want to enjoy their shopping, to have a pleasant experience. In this respect, store characteristics become especially important. Visiting a supermarket as was studied by Bost Ž1987., however, seems to be motivated mostly by the need for the articles sold there and not so much by the intention of having a nice shopping experience. Thus, we expect the effects of store atmosphere on mood, satisfaction with the store and purchasing behavior to become stronger if customers are longing for a pleasant shopping experience. As the concept of experience-oriented marketing is particularly prominent for such stores, we used two furniture stores for our investigation. Fourth, especially with respect to articles sold in a supermarket, long-standing purchasing habits of the customers must be taken into account. Moreover, goods from different producers are offered and customers may have developed preferences for special brands. Such purchasing habits may attenuate the effects of store characteristics and mood. To hold preferences for special brands constant we chose furniture stores selling the articles of one special producer only. Fifth, to control for effects of self-selection, familiarity with the store, intentions and expectations of goal attainment were assessed as control variables. For two such stores, differing with respect to store atmosphere, operationalized via the stores’ condition, information rate and layout, we first tested the effects of store atmosphere on customers’ mood. Second, the effects of store atmosphere on satisfaction and purchasing behavior were investigated and we tested whether these effects could be ascribed to customers’ mood as an intervening variable. Our predictions become obvious from Fig. 1. Thus, for satisfaction and for purchasing behavior, a direct effect of store atmosphere as well as an indirect effect mediated by mood were assumed. Familiarity with the store, intentions, expectations of goal attainment, actual goal-attainment, visiting the caferre´ staurant as well as sociodemographic variables were added to the model as control variables if respective differences between the two stores could be observed. Fig. 1. Expected relations between store atmosphere, customers’ mood, their satisfaction and purchasing behavior. K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 2. Methods 2.1. Independent Õariable The study was carried out in two ‘Ikea’ stores, which belong to a big Scandinavian furnishing company. The stores were located in two different German towns. 1 The furnishings offered in the two stores were completely comparable with respect to assortment, quality and price. Additionally, both stores followed the concept of experience-oriented marketing. In a so-called exhibition area, along the walls there were several departments presenting completely furnished living-rooms, bedrooms, kitchens, children’s rooms etc. Other departments showed objects arranged according to their function, i.e. various shelves, chairs, sofas etc. This exhibition area was separated from a self-service area where customers could find packages containing the things they had decided to buy. In this area, various accessories and additional goods, such as plants, mirrors, candles, cutlery etc., were also available. Each of the areas covered about 50% of the whole store. Within the exhibition area there was a caferrestaurant. ´ There were, however, differences between the two stores with respect to store atmosphere, operationalized as the store’s condition, its layout and the information rate Žsee Table 1.. Store A had been renovated in 1992, i.e. one year before the investigation took place. The entrance area was wide and bright with plenty of natural light. The walls were painted in white and the lighting was well-planned. Store B, on the other hand, showed many signs of deterioration. Walls were painted in dim colours which, above all, made the ceilings look particularly low. Lighting, brightness, colours and height are considered to be important features creating store atmosphere Že.g. Buckley, 1987.. With respect to layout, in Store A the rather complex route through the store was well-structured so that customers had no difficulty finding their way. In the exhibition area the way was clearly marked by the colour of the flooring. Its direction was counterclockwise, short-cuts were impossible so that customers passed all exhibited objects. Throughout the 1 We thank the managers of both stores for their kind support. 5 exhibition there were several information desks as well as many striking signs in different colours. In the self-service area there were no marked paths; here every customer could choose his or her own direction. However, here, too, striking signs improved orientation. In Store B orientation was much more difficult. As in Store A the route through the exhibition area was marked. However, there were several crossings where customers had to decide which path they wanted to take. As signs were few and not very striking customers could easily happen to miss objects. If they endeavoured to see all the objects they had to move in a circle several times, thus running the risk of losing their orientation. Within the selfservice area, hand-written cardboard signs were the only help in finding the selected articles. In Store A the presentation of the furniture was much more out of the ordinary and unexpected for the customer compared to Store B, i.e. the information rate in Store A was higher. Different furnishings were more frequently integrated into a complete living-room, kitchen etc. so that customers were given some inspiration on how to furnish their own rooms at home. In Store B, on the other hand, furnishings were more often grouped according to their function, so that for instance several tables were located one beside the other. In addition, new supplies were presented much more noticeably in Store A than in Store B. For example, in Store A a new model of a chair hung down from the wall with a special spotlight directed on it. No other chairs were close by – only an attractive plant. Thus, the chair drew the customers’ attention, they could not help noticing it. In Store B, however, the same chair stood amidst other chairs and was marked as new only by a small sign. In summary, the two stores were each in a different condition – one modern, one rather dilapidated – and had a different layout and information rate, thus forming two levels of store atmosphere that served as the independent variable of the present study. 2.2. Dependent Õariables The first dependent variable was the customers’ mood state. It was expected that the good atmosphere of Store A would induce a better mood in the customers than the worse atmosphere of Store B. K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 6 Table 1 Differences between the stores Store A Store B Renovated one year ago Bright colours Many signs of deterioration Dim colours Well-structured route with no crossings and possible shortcuts Many striking signs in different colours Many crossings and possible shortcuts Few, not very striking signs, some only hand-written Presentation out of the ordinary, unexpected Integration of furnishings to complete rooms New supplies highlighted Expected, ordinary presentation Functional grouping of furnishings New supplies presented amidst others Condition Layout Information rate Mood was assessed three times, first at the entrance, second after customers had passed the exhibition area and finally before they left the store. Referring to the mood-questionnaire ŽSES. by Hampel Ž1977., mood-related adjectives were presented and subjects had to indicate for each adjective on a seven-point rating-scale how well it described their present mood. We used three subscales of the SES with adjectives for elated, depressed and angry moods respectively. In order to not unduly extend the survey, we followed Hampel in first presenting the short form A and second the parallel form B. Both forms comprised seven adjectives of each scale. For the third mood-assessment we used seven adjectives per scale taken at random from either form A or form B. In order to get information about customers’ satisfaction with the store, after they had passed the exhibition area we asked them to rate on a five-point rating-scale how much they liked the exhibition of furnishings Ž1: ‘‘I disliked the exhibition very much’’; 5: ‘‘I liked the exhibition very much’’.. Before customers left the store, i.e. after they had passed the self-service area, they were asked how contented they were with their shopping on the whole Ž1: ‘‘ very discontented’’; 5: ‘‘ very contented’’., and if they intended to revisit the store Ž0: no; 1: some time; 2: very soon.. As a more behavioral index of customers’ satisfaction, we took the time they spent in the exhibition and the self-service area respectively, i.e. the time between entering the store and leaving the exhibition area as well as the time between leaving the exhibition area and leaving the store was recorded. It was assumed that a pleasant store atmosphere – perhaps via inducing a positive mood state – should favor a positive evaluation of the exhibition and of the shopping as a whole, raise customers’ willingness to revisit the store and make them spend more time in the exhibition as well as in the self-service area. Thus, there should be respective differences between customers in Store A and Store B. With respect to purchasing behaÕior, before they left the store we asked customers to indicate how much money they had spent altogether and how much money they had spent on things they had not actually intended to buy. With respect to their spontaneous purchases subjects had to specify whether they had bought each object Ža. because it was a good bargain, Žb. because it occurred to them that they needed the object or Žc. because they simply liked the object. It was assumed that a pleasant store atmosphere – supposedly via the effect of a positive mood state – should cause customers to spend more money and to buy things more spontaneously if they were not too expensive, i.e. if the risk of a bad bargain was rather low. Respective differences were expected between customers in Store A and Store B. 2.3. Control Õariables In order to get information concerning the comparability of customers visiting Store A and Store B, we obtained the following data: age and sex of the customers, income, number of persons living in the household, distance between home and store, number of accompanying adults, number of accompanying K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 children. With respect to customers’ familiarity with the store, we asked whether they knew other stores of the same company and whether the exhibition of furnishings was new or familiar to them. As for their actual shopping, when entering the store customers had to indicate their intention, whether they wanted to go for a look round, whether they wanted to buy a certain object, e.g. a sofa or a shelf, but had not yet decided about the specific model, or whether they had come to buy just a special article they had seen in the store’s advertisements. In addition, we asked customers whether they expected to attain their goal or not. After they left the exhibition area customers had to indicate whether – if they had pursued a goal in this area – they had attained this goal or not. In the same way, customers were asked about goal-attainment in the self-service area before they left the store. Goal-attainment was controlled because, apart from store atmosphere, it was considered to be an important determinant for customers’ mood as well as for their satisfaction scores. Furthermore, customers were asked whether they had visited the caferrestaurant or not, as besides ´ store atmosphere this factor may influence mood as well as increasing the time customers spend in the exhibition area. 2.4. Procedure The study was run with 76 customers per store. In each store, the investigation took place for one week ŽMonday to Friday. during the usual hours of sale. Saturdays were excluded because a larger number of customers were expected to attend the store on Sat- urdays, rendering a survey more difficult, and because other types of customers would come impairing the homogeneity of the sample. Customers were met three times during their shopping by three different researchers. Researcher A welcomed the customer in the entrance area and asked him or her to take part in the survey. In both stores about two thirds of the customers agreed to participate. They were given the first questionnaire to assess the dependent measures listed in Table 2. Customers ran into Researcher B after they had left the exhibition area and had to work on the second questionnaire. After they had passed the cash-desk, they were given the third questionnaire by Researcher C. To allow the three questionnaires of each person to be identified without recording customers’ names we asked customers to mark each questionnaire with a code containing the initials of his or her name as well as the day and month of his or her birth. All the researchers tried to use relatively calm areas for the survey. In all cases, customers were not actually interviewed but left to work on the questionnaires by themselves. Customers who went back from the self-service to the exhibition area were excluded from data-analysis because in this case the effects for both areas could not be separated. Of course, there is the risk of reactivity in interrupting shoppers three times. But otherwise we had no possibility to assess mood changes and would have had no chance to ascribe effects to the exhibition and self-service area of the stores, respectively. Moreover, as the procedure was the same in both stores, it seems rather unlikely that observed differences are due to reactivity effects. Table 2 Dependent measures taken at the three times of assessment Mood Satisfaction First assessment Second assessment Third assessment Questionnaire A Questionnaire B Liking of the exhibition Questionnaire C Satisfaction with shopping on the whole Intention to revisit the store Amount of money spent on the whole Amount of money spent for spontaneous purchases Goal-attainment in the self-service area Expectations of goal attainment Purchase Control Õariables 7 Familiarity intentions Goal-attainment in the exhibition area Expectations of goal attainment Visit of the caferrestaurant ´ Sex, age, income, number of persons in the household Distance home–store Number of accompanying adults and children 8 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 2.5. Design In a first step, it was tested whether the two stores, having different levels of store atmosphere, induced different mood states. For mood at the second and third time of assessment respectively, a regression analysis was carried out with mood at the first time of assessment Žbase mood. as well as store atmosphere as predictors. In addition, a hierarchical regression was performed introducing, after base mood but before store atmosphere, those control variables that proved different between the two stores. In a second step, the effects of store atmosphere on satisfaction and purchasing behavior were tested. In this respect, a regression analysis was carried out taking store atmosphere as predictor. With respect to mood-effects we referred to changes in mood-scores from the first to the second time of assessment. Effects of mood-change were tested via regression analysis. In a hierarchical regression analysis moodchange was entered as the first and store atmosphere as the second predictor. By doing so, it was checked whether the effects of store atmosphere could be explained via mood-effects. In a further hierarchical regression, relevant control variables differing between the stores were introduced as predictors before mood-change and store atmosphere. Depending on the number of predictors the critical F-value for the regression analyses varied between F Ž1, 150, 95%. s 3.91 and F Ž5, 146, 95%. s 2.28. With a s b s 0.05 and N s 152, the expected amount of explained variance Žh 2 . for one, two, three, four or five predictors was 0.08, 0.09, 0.10, 0.11 and 0.12 respectively. In the case of the hierarchical regression analysis, the increase of explained variance that was expected by introducing a further predictor amounted to h 2 s 0.08. According to the Cohen Ž1988. conventions, all expected effects must be said to be small to medium. As the intention of revisiting the store was assessed at nominal scale-level, for this variable we referred to log-linear analysis. In order to test the effects of mood-change within this analysis, the total group of 152 customers was split at the median of mood-change from the first to the second time of assessment, thus getting one group of customers with positive and a second group with mainly negative mood-changes. It was tested whether models assum- ing an interaction between intention and either store atmosphere or mood-change or both fit the data. In a second step, control variables were taken into account. For a s b s 0.05 and N s 152 the expected effects size for x 2-tests with df s 28 amounts to B s 0.48. This must be considered a medium to large effect according to the Cohen Ž1988. conventions. For the log-linear analysis, B s 0.48 means an upper effect size limit that must not be exceeded if the assumed model fits the data. 3. Results 3.1. Control Õariables Customers of Stores A and B did not differ with respect to age Ž71% ranging between 20 and 40 years in both stores., sex Ž39 and 38 women in Stores A and B respectively., number of persons living in the household Žmostly 1 or 2 in both stores., income Žin both stores 70% of the customers either belonged to the group with a higher income of more than 4000 DM per month or to the group having a low income of less than 1500 DM per month., accompanying adults Žmostly 1–2 in both stores. and accompanying children Žabout 80% without children and 20% with one child in both stores.. Distance between home and store was somewhat greater for Store A Ž20–100km. than for Store B Ž10–50km. because of its being situated further away from the city. With respect to familiarity, for both stores 34 of the 76 customers said they knew other stores of the same company. Sixty-nine customers of Store A compared to 62 customers of Store B knew the present exhibition of furnishings Ž x 2 s 2.71 x 2 Ž1,95%. s 3.84.. 20 customers in Store A compared to 30 customers in Store B only intended to stroll around, 20 persons in Store A vs. 17 in Store B intended to buy something without having decided which specific model and 26 vs. 31 customers of Stores A and B respectively came with the intention of buying just one special article. Thus, intentions can be said to be largely the same for the customers of both stores Ž x 2 s 0.75 - x 2 Ž2,95%. s 5.99.. Expectation was K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 Table 3 Frequencies for visiting the caferrestaurant and goal-attainment in ´ Store A and Store B Store A Store B Visiting the caferrestaurant ´ Yes No 29 47 18 58 Goal-attainment in the exhibition area Yes 50 44 No No goal 11 15 20 12 Yes 55 33 No No goal 14 7 31 12 Goal-attainment in the self-service area exactly the same for customers of Stores A and B with 28 persons having low and 48 having high expectation with respect to goal-attainment. However, – as can be seen from Table 3 – there was a tendency for the number of customers who attained their goal in the exhibition area to be larger in Store A than in Store B Ž x 2 s 2.93 ) x 2 Ž1,90%. s 2.71.. Interestingly, this effect held true above all for customers who intended to stroll around and to get some useful ideas Ž x 2 s 3.71 ) x 2 Ž1,90%. s 2.71., while there was no difference between stores for customers who intended to buy something Ž x 2 s 0.70. or who came to get just one special article Ž x 2 s 0.00.. Besides, more customers of Store A reached their goal in the self-service area Ž x 2 s 11.75 ) x 2 Ž1,95%. s 3.84.. Again, this effect was due mostly to customers who wanted to go for a stroll Ž x 2 s 14.97., while customers who intended to buy something Ž x 2 s 0.47. or just one special article Ž x 2 s 0.65. did not differ between stores with respect to goal-attainment. As a tendency more customers in Store A compared to Store B visited the caferre´ staurant Ž x 2 s 3.73 ) x 2 Ž1,90%. s 2.71.. Hence, in the following, goal-attainment in the exhibition and self-service area as well as visit of the caferre´ staurant were considered as control variables if effects on the respective dependent variable seemed likely to be expected. customers of Store A, while it deteriorated for customers of Store B. From the second to the third assessment there was a small deterioration of mood in both groups. Mood at the first time of assessment Žbase mood. explained 36% of the variance of mood-scores at the second time of assessment Ž F s 83.60.. Introducing store atmosphere as a second predictor significantly Ž Finc s 46.48. increased the amount of explained variance Ž R 2 s 0.51, F s 77.72.. Apart from store atmosphere customers’ mood may be influenced by whether they attain their goal and whether they visit the caferrestaurant. Thus, for ´ mood at the second time of assessment a hierarchical regression was carried out introducing base mood, goal-attainment in the exhibition area and visiting the caferrestaurant as predictors first before entering ´ store atmosphere. The set of base mood, goal-attainment in the exhibition area and visiting the caferre´ staurant accounted for 50% of the variance Ž F s 49.69.. The proportion of variance explained was significantly Ž Finc s 38.22. increased by introducing store atmosphere Ž R 2 s 0.60, F s 56.19.. With respect to mood at the third time of assessment, i.e. before customers left the store, base mood accounted for 26% of the variance Ž F s 51.85.. The amount of explained variance was significantly Ž Finc s 38.51. increased by introducing store atmosphere as a second predictor Ž R 2 s 0.41, F s 51.66.. The set of base mood, goal-attainment in the exhibition as well as – additionally – in the self-service area and visiting the caferrestaurant explained 45% of ´ the variance of customers’ mood at this third time of assessment Ž F s 30.44.. Again, the proportion of variance accounted for became significantly larger Ž Finc s 22.56. when entering store atmosphere Ž R 2 Table 4 Mood scores at the three times of assessment for Store A and Store B Žmeans and standard deviations. Store A Store B SES1 35.79 Ž20.59. 33.34 Ž22.95. SES2 46.17 Ž22.49. 21.08 Ž30.84. SES3 41.66 Ž25.43. 16.08 Ž30.22. 3.2. Mood As can be seen from Table 4, mood improved from the first to the second time of assessment for 9 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 10 Table 5 Customers’ satisfaction for Store A vs. Store B and for positive vs. negative mood-change Žmeans and standard deviations. Store A Positive mood-change Ž n s 52. Store B Negative mood-change Ž n s 24. Positive mood-change Ž n s 23. Negative mood-change Ž n s 53. Liking of the exhibition 4.25 Ž0.76. 3.79 Ž0.83. 3.78 Ž0.90. 3.26 Ž0.81. Satisfaction with shopping on the whole 4.12 Ž0.98. 3.29 Ž0.95. 3.48 Ž1.27. 3.00 Ž0.92. Time spent in the exhibition Žminutes. 56.69 Ž24.18. 46.75 Ž19.21. 31.00 Ž21.67. 25.70 Ž14.20. Time spent in the self-service area Žminutes. 63.21 Ž34.32. 65.63 Ž33.24. 45.09 Ž32.72. 41.57 Ž23.75. s 0.53, F s 32.44.. Thus, store atmosphere had a considerable impact on customers’ mood that could not be simply reduced to goal-attainment or visiting the caferrestaurant. ´ 3.3. Satisfaction In Tables 5 and 6 data are given for Stores A and B differentiating between customers with positive and negative mood-changes. Only for the purpose of a better tabular presentation of the data was the total group of 152 customers split at the median of mood-change from the first to the second time of assessment Žmedians 2., thus obtaining one group of customers with positive and a second group with mainly negative mood-changes. For the purpose of regression analysis, of course, we referred to the undichotomized mood-change scores. With respect to liking ratings of the exhibition area customers of Store A attained higher scores than customers of Store B Žsee Table 5.. Store atmosphere alone accounted for 14% of the variance Ž F s 25.25.. Customers with a positive mood-change liked the exhibition of furnishings more than customers with a negative mood-change. Mood-change alone explained 24% of the variance of the liking ratings Ž F s 48.28.. This proportion of explained variance was significantly increased Ž Finc s 5.56. by introducing store atmosphere as a second predictor Ž R 2 s 0.27, F s 27.66.. In order to check whether the effects of moodchange and store atmosphere on liking-scores for the exhibition might have been due to goal-attainment, goal-attainment was entered into the regression analysis as the first predictor followed by mood-change and store atmosphere. Goal-attainment accounted for 12% of the variance Ž F s 21.25.. The introduction of mood-change significantly Ž Finc s 36.86. increased the proportion of explained variance Ž R 2 s 0.30, F s 31.59.. A further significant increase Ž Finc s 5.97. was attained by entering store atmosphere Ž R 2 s 0.33, F s 23.75.. Hence, the effects of Table 6 Frequencies of customers’ intention to revisit the store in Store A vs. Store B and for positive vs. negative mood-change Store A Store B Intention to revisit the store Positive mood-change Ž n s 52. Negative mood-change Ž n s 24. Positive mood-change Ž n s 23. Negative mood-change Ž n s 53. Very soon Sometime No 44 4 4 18 2 4 18 2 3 33 7 13 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 mood-change and store atmosphere could not be reduced to effects of goal-attainment. Customers’ satisfaction with their shopping on the whole was higher in Store A than in Store B. Store atmosphere alone explained for 10% of the variance Ž R 2 s 0.10, F s 17.46.. Satisfaction was also fostered by a positive mood-change. Mood-change alone accounted for 17% of the variance Ž F s 29.67.. The proportion of variance was significantly increased Ž Finc s 3.99. by entering store atmosphere as a second predictor Ž R 2 s 0.19, F s 17.13.. Factors other than mood-change and store atmosphere which may influence customers’ satisfaction with their shopping as a whole are goal-attainment in the exhibition area as well as in the self-service area. Thus, for satisfaction with the shopping as the dependent variable these two predictor variables were entered into the hierarchical regression analysis before mood-change and store atmosphere. Goal-attainment in the exhibition area accounted for 14% of the variance Ž F s 24.45.. The amount of explained variance significantly increased Ž Finc s 54.68. by introducing goal-attainment in the self-service area Ž R 2 s 0.37, F s 43.94.. A further significant increase Ž Finc s 8.79. was attained by entering mood-change Ž R 2 s 0.40, F s 33.75.. Store atmosphere, however, did not significantly contribute Ž Finc s 1.91. to explain- 11 ing further variance Ž R 2 s 0.41, F s 25.94.. Thus, the effect of store atmosphere – but not of moodchange – can be attributed to goal-attainment. More persons in Store A than in Store B intended to reÕisit the store very soon Žsee Table 6.. For log-linear analysis to be applied we referred to the dichotomized mood-change data. A model that considered store atmosphere alone did not fit the data Ž x 2 s 47.31 ) x 2 Ž28, 95%. s 41.34.. The same held true for a model that only accounted for mood-change Ž x 2 s 43.05 ) x 2 Ž28, 95%. s 41.34.. Entering both store atmosphere and mood-change did not significantly improve the fit Ž x 2 s 41.97 ) x 2 Ž26, 95%. s 2 38.89; xdiff s 1.08 - x 2 Ž2, 95%. s 5.99.. In a second step, goal-attainment in the exhibition and in the self-service area were introduced as predictors before mood-change and store atmosphere. As otherwise there would have been too many empty cells, subjects who had no special goal in the exhibition or self-service area were excluded from the respective analysis. A model with goal attainment in the exhibition area did not fit the data Ž x 2 s 42.75 ) x 2 Ž28, 95%. s 41.34.. However, by considering both goal-attainment in the exhibition and self-service area a much better fit was attained Ž x 2 s 23.12 2 x 2 Ž26, 95%. s 38.89; xdiff s 19.63 ) x 2 Ž2, 95%. s 5.99.. Mood-change as an additional predictor did Table 7 Amount of money Žgiven in DM. customers with a positive vs. negative mood-change spent in Store A vs. Store B Žmeans and standard deviations. Store A All purchases Amount of money spent on the whole Spontaneous purchases Amount of money spent for articles – on the whole – that were needed – that were simply liked – that were good bargains Store B Positive mood-change Ž n s 52. Negative mood-change Ž n s 24. Positive mood-change Ž n s 23. Negative mood-change Ž n s 53. 131.04 Ž142.22. 108.33 Ž168.02. 96.13 Ž114.64. 106.79 Ž139.87. 38.73 Ž43.33. 9.15 Ž19.71. 26.6 Ž41.65. 2.96 Ž12.63. 22.92 Ž29.71. 15.75 Ž26.82. 24.25 Ž9.19. 2.92 Ž14.29. 18.35 Ž21.84. 10.35 Ž13.05. 3.74 Ž7.73. 4.26 Ž11.29. 16.85 Ž27.13. 10.25 Ž18.35. 3.94 Ž9.89. 2.66 Ž13.77. 12 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 not further improve the fit of the model Ž x 2 s 21.07 - x 2 Ž24, 95%. s 36.42., neither did store atmosphere Ž x 2 s 20.66 - x 2 Ž22, 95%. s 33.92.. The time customers spent in the exhibition area Žsee Table 6. was longer in Store A than in Store B. Store atmosphere alone explained 30% of the variance Ž F s 64.18.. Mood-change alone accounted for 20% of the variance, with duration of stay being higher for customers with a positive compared to a negative mood-change Ž F s 36.58.. The proportion of variance was significantly Ž Finc s 33.27. increased by entering store atmosphere as a second predictor Ž R 2 s 0.34, F s 38.86.. Duration of stay in the exhibition area is dependent on whether customers visit the caferrestaurant ´ or not. Hence, in a second step, visiting the caferre´ staurant was entered as a first predictor before mood-change and store atmosphere. Visiting the caferrestaurant accounted for 20% of the variance ´ Ž F s 37.83.. Entering mood-change significantly Ž Finc s 16.55. increased the proportion of explained variance Ž R 2 s 0.28, F s 29.15.. A further, even larger, increase Ž Finc s 41.84. was attained by entering store atmosphere Ž R 2 s 0.44, F s 38.71.. Thus, there was an effect of mood-change and especially of store atmosphere on the time customers spent in the exhibition area that could not be explained by the higher frequency of persons visiting the caferre´ staurant. The time customers spent in the self-serÕice area was longer in Store A compared to Store B. Store atmosphere alone accounted for 11% of the variance Ž F s 18.72.. Mood-change alone explained 3% of the variance Ž F s 5.26.. The amount of explained variance significantly increased Ž Finc s 13.07. by introducing store atmosphere Ž R 2 s 0.11, F s 9.38.. 3.4. Purchasing behaÕior Looking at purchasing behavior ŽTable 7., the amount of money spent altogether depended neither on store atmosphere Ž R 2 s 0.01, F s 0.78. nor on mood-change Ž R 2 s 0.00, F s 0.31.. However, the amount of money spent for spontaneous purchases was larger in Store A than in Store B, although it usually did not amount to more than 50 DM, i.e. the risk of a bad bargain remained rather low. Store atmosphere alone accounted for 6% of the variance Ž F s 9.11.. Besides, the amount of money spent for spontaneous purchases was larger for customers with a positive compared to a negative mood-change. Mood-change alone accounted for 10% of the variance Ž F s 16.24.. This proportion of variance was not significantly affected Ž Finc s 1.78. by introducing store atmosphere as a second predictor Ž R 2 s 0.11, F s 9.05.. Taking into account the reasons customers gave for their spontaneous purchases, no effects of either store atmosphere or mood-change appeared with respect to articles which the customers felt they needed Žstore atmosphere: R 2 s 0.00, F s 0.09; mood-change: R 2 s 0.00, F s 0.16. or that represented good bargains Žstore atmosphere: R 2 s 0.00, F s 0.01; mood-change: R 2 s 0.00, F s 0.62.. However, customers of Store A who showed a positive mood-change spent significantly more money for objects they simply liked than all the other customers. As can be seen from Table 7 the standard deviation is very large for this subgroup. Thus, the effect could have been due to outliers. Data inspection revealed, however, that results did not change when excluding the one customer with an extremely high value Ž246 DM. from the analysis Ž x s 22.31, s s 28.08.. Store atmosphere alone explained 8% of the variance in the amount of money spent for articles that were just liked Ž F s 13.33.. Mood-change alone accounted for 15% of the variance Ž F s 25.91.. This amount of explained variance was left unaffected Ž Finc s 2.45. by introducing store atmosphere as a second predictor Ž R 2 s 0.16, F s 14.31.. 4. Discussion The present study dealt with the effects store atmosphere has on customers’ satisfaction and purchasing behavior taking mood state as an intervening variable. Besides, goal-attainment and visit to the caferrestaurant that proved to differ between the two ´ stores were considered as control variables. Customers’ reactions were studied in two stores of furnishings which were comparable with respect to supply but different with respect to store atmosphere. Store atmosphere was operationalized as the condition the store was in, the information rate and how well the products were presented Žthe layout of the products.. Results are summarized in Fig. 2. K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 13 Fig. 2. Observed relations between store atmosphere, customers’ mood, satisfaction and purchasing behavior considering goal-attainment and visit of the caferrestaurant as control variables. ŽRelations corresponding to those assumed in Fig. 1 are marked by fat lines.. ´ As predicted, it could be shown that customers’ mood improved with pleasant but deteriorated with less pleasant store atmosphere. However, there was a tendency for more customers in the pleasant compared to the less pleasant store to visit the caferre´ staurant. Moreover, while customers of the two stores did not differ in their goals and intentions with respect to their shopping, more customers reached their goal in the pleasant compared to the less pleasant store. This held true especially for those customers who intended to go for a stroll and get some useful ideas. Obviously, a pleasant store atmosphere fosters goal-attainment for these customers. Thus, our results corroborate research on experience-oriented marketing in stressing the importance of store characteristics. As product supply was identical in both stores, it is not astonishing that goal-attainment with regard to finding the articles one was looking for did not differ between stores. It may now be claimed that the improved mood in the pleasant store was due to visiting the caferre´ staurant andror attaining one’s goal. Actually, a considerable amount of the variance of mood-change was accounted for by the joint effect of visiting the caferrestaurant, goal-attainment and store atmo´ sphere. This finding accords with the Scherer Ž1984. assumption that goal relevance instigates emotions of pleasure and displeasure with pleasure being elicited when a person reaches his or her goal and is in accordance with the finding of Dawson et al. Ž1990. that goal-attainment in a shopping situation induces a positive mood state. However, care must be taken in interpreting goal-attainment and visiting the caferrestaurant as ´ possible causes for the mood-changes observed in our investigation. As neither factor could be varied experimentally, data are also open to the interpretation that a positive mood fostered by a pleasant store atmosphere instigates customers to visit the caferre´ staurant and perhaps also makes goal-attainment easier. In addition, especially with respect to goal-attainment, we could not obtain objective data but had to rely on the subjective reports given by the customers. Hence, it cannot be ruled out that customers in a positive mood are more inclined to indicate that they had reached their goals than customers in a negative mood. This interpretation would corroborate the tendency mentioned by Morris Ž1989. that in a positive mood people see the world through rose-coloured glasses. Beyond the joint effects of goal-attainment, visiting the caferrestaurant and store atmosphere, how´ ever, there was a specific effect of store atmosphere that could be ascribed neither to goal-attainment nor 14 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 to visiting the caferrestaurant. Hence, it can be ´ concluded that a pleasant store atmosphere as such improves customers’ mood. As the two stores differed mainly with respect to the condition they were in, the information rate and their layout, in accordance with the ideas of Berlyne Ž1971. and Scherer Ž1984., these store characteristics can be said to have important effects on mood. Our results also corroborate the findings reported by Bost Ž1987.. The fact that Donovan and Rossiter Ž1982. found only weak effects of information rate on customers’ mood may be ascribed to the condition that customers’ initial mood was not controlled. With respect to marketing, the second research question, of course, is even more intriguing, namely what influence store atmosphere exerts on customers’ satisfaction with the store and on their purchasing behavior. According to expectations, customers liked the exhibition more and were more satisfied with their shopping in the pleasant compared to the less pleasant store. More customers in the pleasant compared to the less pleasant store wanted to come back soon. For all three variables, goal-attainment could be shown to be an important predictor of the observed differences. This is in accordance with the findings of Russell and Snodgrass Ž1987., Snodgrass et al. Ž1988. or Ward et al. Ž1988. stressing that a person’s reaction to an environment is influenced by his or her purpose for being there. As the supply of the two stores was identical, customers did not differ in goal-attainment with respect to things they had intended to buy. However, a pleasant store atmosphere helps people to have a nice shopping experience and to get useful ideas and thus raised goal-attainment, in particular for customers who only intended to stroll around. This is in accordance with the concept of experience-oriented marketing. Besides, our data support the assumption that visiting a store of furnishings is motivated to a considerable extent by the desire to have a nice shopping experience. The effect of goal-attainment was larger for customers’ satisfaction with their shopping and for their intent to revisit the store than for their liking of the exhibition area. For the intent to revisit the store, goal-attainment even proved to be the only significant predictor. Compared to customers’ liking of the exhibition area, evaluation of the actual shopping as well as the intention to return are more related to customers’ behavior. Customers evaluate their shopping, i.e. the result of their former behavior, and deduce whether they intend to repeat it or not. According to Heckhausen Ž1989. such evaluation processes usually refer to the subject’s present goal. This would explain the outstanding role goal-attainment had for the respective variables in the present study. Thus, for a positive evaluation of the shopping and the intention to return very soon, those store characteristics seem to be most important which help customers to reach their goals. For liking the exhibition area as well as for satisfaction with shopping on the whole, there was a considerable joint effect of store atmosphere and mood that could not be ascribed to goal-attainment. Thus, satisfaction was higher for customers in a positive compared to a negative mood. This result corroborates similar findings from consumer research Že.g. Bitner, 1992; Bost, 1987; Dawson et al., 1990; Gardner, 1985; Isen et al., 1978; Obermiller and Bitner, 1984; Sherman and Smith, 1987; Srull, 1983.. According to the network theory of emotion proposed by Bower Ž1981. it may be interpreted that, due to the positive emotion node being more activated in the pleasant compared to the less pleasant store, activation spreads to adjacent nodes representing mood-related positive contents. This preactivation of ‘positive’ nodes makes positive aspects of the situation more available and thus improved customers’ satisfaction ratings. According to Schwarz Ž1990. it can also be argued that, when judging how much they liked something, customers referred to their present mood as a criterion. Hence, customers attribute their positive mood to characteristics of the present situation and thus infer that the situation must be pleasant because otherwise they would not be in such a good mood. For customers’ liking of the exhibition of furnishings, results showed that beyond the described joint effects store atmosphere also had a specific effect on liking scores that could be ascribed neither to goalattainment nor to mood. This effect was small but important. Thus, customers based their liking judgements also on the direct impression they had obtained of the exhibition. Obviously, the better layout together with the more unfamiliar ware presentation in the pleasant store improved its attractiveness. This K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 corroborates the assumption of Berlyne Ž1971. that attractiveness is highest if complexity reaches an optimal level, i.e. is neither too low nor too high. Liking and satisfaction scores as well as the intention to revisit the store are subjective data based on customers’ evaluations. As a more objective, behavioral indicator of satisfaction, we additionally measured the time customers spent in the exhibition and self-service area respectively. For both results, mood and store atmosphere had a considerable joint effect, but there was also a specific effect of store atmosphere alone. In accordance with findings reported e.g. by Donovan and Rossiter Ž1982., Donovan et al. Ž1994. or Sherman and Smith Ž1987., it could be shown that customers spent more time in the pleasant compared to the less pleasant store. Thus, the objection can be ruled out that instead of higher satisfaction a longer time spent in the respective area could also indicate that customers had more difficulties finding the things they were looking for. The time customers spent in the exhibition area was lengthened if the caferrestaurant was visited. How´ ever, the effects of neither mood nor store atmosphere could be completely ascribed to visiting the caferrestaurant. Thus, customers stay longer in the ´ store if the store atmosphere is pleasant and if they are in a positive mood. One has to keep in mind, however, that the amount of variance explained by the introduced predictors is much lower for the time customers spent in the self-service area Ž11%. compared to the exhibition Ž44%.. Thus, in the selfservice area other factors not considered in the present context seem to determine the amount of time customers spend there. In summary, according to our expectations it could be shown for all of our indices that customers’ satisfaction was higher in the pleasant compared to the less pleasant store. Goal-attainment as well as customers’ mood state proved to be important intervening variables. However, except for satisfaction with shopping on the whole and for the intent to return, store atmosphere had an additional specific effect that could be ascribed neither to goal-attainment nor mood state. Thus, it may be interpreted that a pleasant store atmosphere raises customers’ satisfaction scores partly directly and partly by facilitating goal-attainment and by inducing a positive mood state. 15 With respect to purchasing behavior, it could be shown that store atmosphere did not affect the total amount of money spent but only the amount of money spent for spontaneous purchases. Customers spent more money for spontaneous purchases in the pleasant compared to the less pleasant store. This is in accordance with Donovan and Rossiter Ž1982., Donovan et al. Ž1994. or Sherman and Smith Ž1987. who found that customers in a positive mood bought more items and spent more money than was originally planned. In the present study, the effect of store atmosphere was completely coupled with mood-effects, i.e. there was no specific effect of store atmosphere alone. Usually, the articles bought spontaneously were not very expensive. Thus, results come up to the expectation that a positive mood fosters risky behavior in the form of spontaneous purchases – but only if the risk of a bad bargain remains rather low. This finding is in accordance with respective results of the effects of positive mood on risk-taking behavior reported by Isen and co-workers Že.g. Isen, 1987. or by Spies et al. Ž1996.. Interestingly, moodeffects occurred only for those spontaneous purchases for which customers gave as a reason that they simply liked the article. No mood-effects could be observed for articles bought spontaneously because it had occurred to customers that they needed these articles or because the article was a good bargain. Hence, one might assume that spontaneous purchases fostered by a positive mood are not considered analytically but that customers decide more on intuitive grounds. This is in accordance with the assumption – in Fiedler Ž1988. or Kuhl Ž1983b. – that a positive mood fosters an intuitive holistic processing strategy Žalso called ‘loosening’.. The present study shows that by integrating consumer research and general psychology one can gain very interesting insights into the effects of store characteristics. Taking into account that results were obtained for stores of furnishings as a special retail context, the following conclusions can be drawn: In accordance with the conception of experience-oriented marketing a store’s good condition, careful layout and relatively high information rate, all helping to bring about a pleasant store atmosphere, facilitate goal-attainment. This is especially true for customers who intend to stroll around, get some good ideas and have a nice shopping experience. More- 16 K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 over, a pleasant store atmosphere also improves customers’ mood state. Hence, there are three possible ways in which store atmosphere may influence customers’ satisfaction and purchasing behavior: Ža. directly, Žb. via goal-attainment, Žc. via mood-change. For customers’ satisfaction with the store and for the time they spend there, all types of influence could be shown to exist. With respect to customers’ satisfaction with their shopping and to their intention to revisit the store, the effect of store atmosphere was due to mood-change and even more to goal-attainment. Finally, for purchasing behavior the effects of store atmosphere could be completely ascribed to mood-effects. Thus, depending on the type of reactions shown by customers, different intervening variables are responsible for the effects of store atmosphere. In all cases, however, the positive effect of a pleasant store atmosphere on customers’ reactions could be clearly demonstrated. References Barth, K., 1993. Betriebswirtschaftslehre des Handels. Wiesbaden: Gabler. Belk, R., 1975. Situational variables and consumer behavior. Journal of Consumer Research 2, 157–164. Berlyne, D.E., 1971. Aesthetics and psychobiology. New York: Meredith Corporation. Bitner, M.J., 1992. Servicescapes: The impact of physical surroundings on customers and employees. Journal of Marketing 56 ŽApril., 57–71. Bost, E., 1987. Ladenatmosphare ¨ und Konsumentenverhalten. Heidelberg: Physica. Bower, G.H., 1981. Mood and memory. American Psychologist 36, 129–148. Bower, G.H. and P.R. Cohen, 1982. Emotional influences in memory and thinking: Data and theory. In: M.S. Clark and S.T. Fiske Žeds.., Affect and cognition. Hillsdale, NJ: Erlbaum. Buckley, P.G., 1987. The internal atmosphere of a retail store. In: M. Wallendorf and P. Anderson Žeds.., Advances in consumer research, Vol. 14. Provo, UT: Association for Consumer Research. Cohen, J., 1988. Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum. Cohen, J., 1990. Attitude, affect, and consumer behavior. In: B.S. Moore and A.M. Isen Žeds.., Affect and social behavior. Cambridge: Cambridge University Press. Dawson, S., P.H. Bloch and N.M. Ridgway, 1990. Shopping motives, emotional states, and retail outcomes. Journal of Retailing 66, 408–427. Donovan, R.J. and J.R. Rossiter, 1982. Store atmosphere: An environmental psychology approach. Journal of Retailing 58, 34–57. Donovan, R.J., J.R. Rossiter, G. Marcoolyn and A. Nesdale, 1994. Store atmosphere and purchasing behavior. Journal of Retailing 70, 283–294. Fiedler, K., 1988. Emotional mood, cognitive style and behavior regulation. In: K. Fiedler and J.P. Forgas Žeds.., Affect, cognition and social behavior. Toronto: Hogrefe. Gardner, M.P., 1985. Mood states and consumer behavior: A critical review. Journal of Consumer Research 12, 281–300. Golden, L.M. and M.R. Zimmer, 1986. Relationship between affect, patronage, frequency and amount of money spent with a comment on affect scaling and measurement. In: R.J. Lutz Žed.., Advances in consumer research, Vol. 13. Provo, UT: Association for Consumer Research. Groppel, A., 1991. Erlebnisstrategien im Einzelhandel. Heidel¨ berg: Physica. Hampel, R., 1977. Adjektiv-Skalen zur Einschatzung der Stim¨ mung ŽSES.. Diagnostica 23, 43-60. Heckhausen, H., 1989. Motivation und Handeln. Berlin: Springer. Hui, M.K.M. and J.E.G. Bateson, 1991. Perceived control and the effects of crowding and consumer choice on the service experience. Journal of Consumer Research 18, 174–184. Isen, A.M., 1987. Positive affect, cognitive processes and social behavior. In: L. Berkowitz Žed.., Advances in experimental social psychology, Vol. 20. New York: Academic Press. Isen, A.M. and N. Geva, 1987. The influence of positive affect on acceptable level of risk: The person with the large canoe has a large worry. Organizational Behavior and Human Decision Processes 39, 145–154. Isen, A.M. and R. Patrick, 1983. The effect of positive feelings on risk taking: When the chips are down. Organizational Behavior and Human Performance 31, 194–202. Isen, A.M., T.E. Nygren and F.B. Ashby, 1988. Influence of positive affect on the subjective utility of gains and losses: It is just not worth the risk. Journal of Personality and Social Psychology 55, 710–717. Isen, A.M., T.E. Shalker, M.S. Clark and L. Karp, 1978. Affect, accessibility of material in memory and behavior: A cognitive loop? Journal of Personality and Social Psychology 36, 1–12. Isen, A.M., B. Means, R. Patrick and G. Nowicki, 1982. Some factors influencing decision-making strategy and risk taking. In: M.S. Clark and S.T. Fiske Žeds.., Affect and cognition. Hillsdale, NJ: Erlbaum. Kroeber-Riehl, W., 1984. Konsumentenverhalten Ž3. Aufl.. Munchen: Vahlen. ¨ Kuhl, J., 1983a. Emotion, Kognition und Motivation: I. Auf dem Wege zu einer systemtheoretischen Betrachtung der Emotionsgenese. Sprache und Kognition 2, 1–27. Kuhl, J., 1983b. Emotion, Kognition und Motivation: II. Die funktionale Bedeutung der Emotionen fur ¨ das problemlosende ¨ Denken und Handeln. Sprache und Kognition 4, 228–253. Lazarus, R.S., 1991. Emotion and adaptation. New York: Oxford University Press. Mehrabian, A. and J.A. Russell, 1974. An approach to environmental psychology. Cambridge: MIT Press. K. Spies et al.r Intern. J. of Research in Marketing 14 (1997) 1–17 Morris, W.N., 1989. Mood: The frame of mind. New York: Springer. Obermiller, C. and M.J. Bitner, 1984. Store atmosphere: A peripheral cue for product evaluation. In: D.C. Stewart Žed.., American Psychological Association Annual Conference Proceedings Žpp. 52–53.. American Psychological Association. Parasuraman, A., L.L. Berry and V. Zeithaml, 1990. Guidelines for conducting service quality research. Marketing Research 2 ŽDecember., 34–44. Russell, J.A. and J. Snodgrass, 1987. Emotion and the environment. In: D. Stokols and I. Altman Žeds.., Handbook of environmental psychology. New York: John Wiley. Scherer, K.R., 1984. On the nature and function of emotion: A component process approach. In: K.R. Scherer and P. Ekman Žeds.., Approaches to emotion. Hillsdale, NJ: Erlbaum. Schwarz, N., 1990. Feelings as information: Informational and motivational functions of affective states. In: E.T. Higgins and R. Sorrentino Žeds.., Handbook of motivation and cognition: Foundations of social behavior, Vol. 2. New York: Guilford Press. Sherman, E. and R.B. Smith, 1987. Mood states of shoppers and 17 store image: Promising interactions and possible behavioral effects. In: M. Wallendorf and P. Anderson Žeds.., Advances in consumer research, Vol. 14. Provo, UT: Association for Consumer Research. Snodgrass, J., J.A. Russell and L.M. Ward, 1988. Planning, mood and place-liking. Journal of Environmental Psychology 8, 209–222. Spies, K., F.W. Hesse and F. Brandes, 1996. Influence of positive mood on risk-taking behavior. Unpublished manuscript. University of Gottingen. ¨ Srull, T., 1983. Affect and memory: The impact of affective reactions in advertising on the representation of product information in memory. In: R. Bagozzi and A. Tybout Žeds.., Advances in consumer research, Vol. 10. Ann Arbor, MI: Association for Consumer Research. Ward, L.M., J. Snodgrass, B. Chew and J.A. Russell, 1988. The role of plans in cognitive and affective responses to places. Journal of Environmental Psychology 8, 1–8. Weinberg, P. and W. Gottwald, 1982. Impulsive consumer buying as a result of emotions. Journal of Business Research 10 ŽMarch., 43–57.