Is a Gift Always a Gift? An Investigation of Flower Purchasing Behavior ACRoss Situations

Debra E. Scammon, University of Utah
Roy T. Shaw, University of Utah (student), University of Utah
Gary Bamossy,
ABSTRACT - The reasons for the purchase of flowers, an item frequently purchased as a gift, are analyzed by grouping reasons into personal uses, obligatory events and gift occasions. Discriminant analyses of purchaser and purchasing variables demonstrate some significant differences between these groups. The results suggest that finer discrimination of gift occasions into those considered to be obligatory and those considered voluntary might be worthwhile.
[ to cite ]:
Debra E. Scammon, Roy T. Shaw, and Gary Bamossy (1982) ,"Is a Gift Always a Gift? An Investigation of Flower Purchasing Behavior ACRoss Situations", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 531-536.

Advances in Consumer Research Volume 9, 1982      Pages 531-536

IS A GIFT ALWAYS A GIFT? AN INVESTIGATION OF FLOWER PURCHASING BEHAVIOR ACROSS SITUATIONS

Debra E. Scammon, University of Utah

Roy T. Shaw, University of Utah

Gary Bamossy (student), University of Utah

ABSTRACT -

The reasons for the purchase of flowers, an item frequently purchased as a gift, are analyzed by grouping reasons into personal uses, obligatory events and gift occasions. Discriminant analyses of purchaser and purchasing variables demonstrate some significant differences between these groups. The results suggest that finer discrimination of gift occasions into those considered to be obligatory and those considered voluntary might be worthwhile.

INTRODUCTION

Gift-giving presents a special situation within the general realm of consumer behavior which has only recently caught the attention of marketing/consumer researchers. Conceptualizations of gift-giving behavior have focused on the dimensions of gift-giving (e.g., givers, gifts, recipients, and gift-giving occasions) and the functions of gift-giving (e.g., communication, social exchange, economic exchange, and socialization) (Belk, 1979). Studies have examined the occasions of gift-giving and have described the frequency of gift-giving for various occasions as well as the types of gifts selected and the prices paid (Ryan, 1977; Belk, 1977).

Research investigating the effects of situations on purchasing behavior suggests that shopping for a gift involves different behaviors than does shopping for a purchase for personal use. For example, Gronhaug (1972) found that consumers utilized different types and sources of information depending on whether the purchase was for personal use or for a gift. Ryan (1977) found gift shoppers more likely than those buying for personal use to have a target price range in mind for the purchase.

The cut flower market seems to be facing a predicament shared by many specialty markets. Flowers are frequently (traditionally) purchased as gifts yet they account for only a small percentage of gift purchases. With economic pressures, as well as declining numbers of consumers among the ranks of flower purchasers (Flower News, 1979), the industry is faced with a decision of whether to encourage the purchase of flowers for non-gift-giving (that is, for personal use), to attempt to expand the number of occasions for which flowers are considered an appropriate gift, or whether both strategies for market expansion might be pursued simultaneously. This paper takes a preliminary look at the cut flower market and attempts to increase understanding of personal and gift purchases of flowers. It examines various gift-giving occasions in an attempt to understand differences among occasions for which flowers are chosen as gifts. The paper thus focuses on the purchases of one product across different situations, specifically attempting to discover whether there are differences between purchasers and their purchasing behaviors across "gift-giving" situations.

THEORETICAL BACKGROUND

In an early study, Belk reported a wide array of gift-giving occasions (1977). Intuitively, it seems there should be a relationship between the occasion for giving and the function of the gift, or the act of giving performs.

Research has identified four primary functions of gift-giving; (1) communication; (2) social exchange; (3) economic exchange; and (4) socialization. The first three functions are discussed below.

As is the case with any form of communication, the message does not have meaning except as it is interpreted by the sender and/or the receiver. In the context of gift-giving, the giver runs the risk of the gift being misinterpreted. It is important that the giver choose the "right" gift to convey the desired message to the recipient. Mauss concluded from his studies of archaic societies that gifts are often a means of showing honor and respect for the recipient (Mauss, 1954). However, a gift provides other means of communication as well. The recipient of a gift can display the gift conspicuously to communicate to others how important someone finds them (Levi-Strauss, 1965). The giver also can communicate something about himself/herself (Schwartz, 1967) through the choice of a gift. Belk (1979) points out that the message which a gift communicates about the giver may be either an attempt to demonstrate a particular self-trait or an attempt to obtain consensual validation of personal tastes and traits. Gifts which are visibly or conspicuously consumed provide the greatest opportunities for communication (of all three kinds) but are risky in that their meaning may be misperceived by recipient as well as observers. Thus, most gifts chosen as a means of communication are likely to be those that are considered traditionally acceptable or "safe."

Gifts also perform the function of facilitating social exchange. Important in this function are the occasions for which gifts are given as well as the nature of the gift item chosen. In many instances gift-giving appears to be ceremonial, serving as a symbol of social support in various rites of passage from one life stage to another (for example, weddings, graduations, funerals). Gifts may be presented in situations where a "social debt" has been incurred. In such cases, the gift demonstrates a gesture of gratitude. For example, in some societies it is customary to present a dinner hostess with a bottle of wine or a bouquet of flowers as a token of appreciation for the invitation. Theories of social interaction such as Homan's (1961) "distributive justice theory" and Adam' s (1963) "equity theory" lend support to findings that suggest there are acceptable price ranges for gifts and that the appropriateness of an item as a gift depends on its perceived intimacy (Sherif and Sherif, 1963). The frequent occurrence of "rites of passage" eves s, for example, appears to make more expensive gifts acceptable for such occasions. A prior history of gift-giving between two persons may make a more intimate gift appropriate. The idea of economic exchange is explained by Mauss (1954) who describes gift-giving as a series of obligatory reciprocal exchanges, that is, certain occasions have become institutionalized as appropriate for exchanging gifts. Others suggest the norm of fair exchange prompts evaluation by the giver of what he/she last received from another person before selecting a gift (Belshaw, 1965). Similarly, research by Ryan (1977) reports those buying items as gifts were more likely to begin shopping with a pre-specified price range than were those buying for personal use.

These three functions of gift-giving taken together suggest that purchasers and their purchasing behavior may well differ across gift-giving occasions. This research examines two different types of gift-giving occasions, as well as personal use, as three different purchasing situations. It focuses on the cut flower market which historically has been substantially an "occasion" market (gifts) but which has exhibited an increasing trend toward becoming a "personal use" market as well (Flower News 1979).

THE CUT FLOWER MARKET

Some background on the cut flower market will help to emphasize the fact that the industry is facing a dilemma. Although few items other than clothing and jewelry account for a very large share of gift purchases (25% and 11% respectively), cut flowers were selected as gifts for only slightly over 2% of the gifts reported by Belk (1977). The cut flower market has not shown any real growth in the last couple of years. A study conducted by the American Florists Marketing Council (AFMC) in 1979 suggested that the market has changed with fewer people buying flowers in 1979 than in 1973 when a similar study was conducted. The people who are buying flowers are doing so more frequently, however. In the AFMC national consumer survey, 46% of those who purchased flowers made six or more purchases with an average number of purchases of 7.4 (Flower News, 1979). An interesting dynamic revealed by the AFMC study is that those who buy flowers more often do so for nonoccasions (personal use). The study reports that the light users, however, rarely purchase flowers for themselves but instead buy flowers for friends or relatives for special occasions. Industry experts say many of these purchases are for what they call "new occasions" (to the cut flower industry) such as birthdays and anniversaries in addition to remembrances for more "traditional" events for flowers such as funerals (Florist, 1979).

Industry reports suggest there are at least two different situations for which consumers buy flowers: for their own personal use and enjoyment, and as gifts for friends or relatives on special occasions. Further, there seem to be some important differences among the occasions for which flowers are chosen as gifts. This study investigates these different segments of the flower market by examining purchaser characteristics and purchasing behavior patterns for different purchase occasions.

METHOD

Data for this study were obtained from two sources. Customer lists from six retail florists in a large western city provided 155 names of past customers. After systematic follow-up of two call-backs, 107 usable questionnaires were completed, for a response rate of 69%. The same questionnaire was used in mall-intercept interviews conducted in four different shopping malls in the same city. 194 usable questionnaires were obtained through this latter method, giving a final data set including responses from 301 persons. Respondents provided information about their past and intended future purchases of flowers. They also responded to questions using a 5-point Likert-type scale about their attitudes toward flower purchases in general. Demographic and socio-economic data on respondents were also obtained.

The dependent variable of interest was reason for purchase. Respondents were examined according to their past reasons for purchasing flowers. Reasons for purchase were grouped together according to those primarily for personal use, those representing what might be called obligatory events due to the social and/or economic exchange nature of the event, and those representing "no strings attached" gift occasions. The Grouping of reasons for purchase and the distribution of purchases across these reasons are presented in Table 1. [Each respondent provided information for all purchases made during the preceding year. There were 71 nonpurchasers. Ninety-two purchased for one reason, 66 for two reasons, 38 for three reasons, 18 for four reasons, and 16 for more than four reasons. Cases with multiple reasons which caused an overlap for group assignment were handled in the following manner: The first tie-breaker was frequency of purchase by reason. A case with two "personal purchase" and one "occasion purchase" would be assigned to the personal group. If a case could still not be assigned after checking frequency, the reason associated with the higher price range was used to assign the case.]

TABLE 1

FREQUENCY OF PURCHASE OCCASIONS

Based on evidence from research on personal versus gift purchasing in general, and specifically for the cut flower industry, it was hypothesized that those who purchased primarily for personal use would pay less, shop at conveniently located, minimum service stores and be more likely to buy on impulse than those purchasing for non-personal reasons.

The research on gift-giving suggests that "a gift is not always a gift." The basis for classifying non-personal reasons into either "obligatory events" or "gift occasions" were derived from literature on the functions that gifts and gift-giving play in society. Obligatory events were felt to involve both communication of feelings for the recipient, such as respect, sympathy, and support, as well as notions of fair social/economic exchange such as appreciation, gratitude, etc. Purchases made to fulfill social obligations were expected to be more likely to be made in a manner to minimize shopping/purchasing effort for the giver yet, at the same time, item selection would involve consideration of the message conveyed to the recipient and observers. Thus, it was hypothesized that purchases made for obligatory events would be planned, ordered by phone, paid for by charge account, delivered to the recipient, and purchased at a well-known florist shop which provides extras (e.g., wrapping, gift card, artistic design of floral pieces).

In contrast, those reasons grouped together as "gift occasions" were felt to represent situations in which the notion of fair exchange played only a minimal role, but rather a gift was given primarily as a means of communication, either to or about the recipient or about the giver. Since the gift item plays an important role in such situations, it was hypothesized that these purchases would be higher priced, paid for in cash as opposed to charged, picked out in person and delivered personally.

A second expectation was derived from the fact that most gift occasions can be anticipated, e.g., a spouse's upcoming birthday. Thus, the giver may spend a good deal of time thinking about the appropriate gift for the pending occasion. With the occasion in the back of the mind, the giver may happen to see an item which "strikes him/her right" and make what appears to be an impulsive gift selection. That is, the purchase is planned but the item selection is based on impulse. Thus, it was hypothesized that impulse buying would be more likely for purchases for "gift occasions" than for "obligatory events."

Finally, it was hypothesize i that the purchasers of flowers for these three groups of reasons would differ on demographics and attitudes toward flowers. Industry trends suggest that personal purchase of flowers is on the increase, primarily with younger, professional people (Flower News, 1979). These people enjoy purchasing for themselves as a reward or a "pick up." Since the study involves purchases of only one product category, flowers, it is expected that respondents' attitudes toward flowers will also influence their purchase behaviors and intentions.

Two discriminant analyses were performed with three groups identified by the categories of reasons for purchase Group I, personal uses; Group II, obligatory events; Group III, gift occasions. One analysis used purchaser characteristics (e.g., attitudes and demographics) as the predictor variables and the other used purchasing behaviors (e.g., price paid, store where purchased, method of payment, intended purchases) as the predictor variables. The results of these two analyses are reported separately below and treated together in the discussion section.

RESULTS

Several internal data checks were carried out prior to performing the primary analyses presented in this paper. Because data for this study were obtained from two sources, several tests were performed to determine whether the aggregation of data from respondents whose names were obtained from the customer lists of six local florists and respondents contacted in mall intercept interviews conducted at four shopping malls, could bias the results of the study. A priori, it was expected there would be differences in purchasing behaviors between respondents from the two data sources since respondents obtained from the customer lists of florists were generally frequent purchasers who took advantage of the opportunity to purchase on credit offered by the shops.

Upon examination of the entire sample, the demographic profiles of the respondents obtained from customer lists and shopping malls proved to be very similar. Chi-square tests revealed that the only significant differences were that there were more females in the customer list group and these in the shopping mall group were in earlier stages in their family life cycle.

Chi-square analyses and t-tests were performed comparing the two groups where appropriate on the purchasing variables. Several significant differences were identified in purchasing behavior between respondents from the two sources. Respondents in the customer list group made mores and more planned purchases, paid higher prices for their flowers, charged their purchases, and had their flowers delivered. The attitudes towards flowers held by respondents from the two sources were also different. Those in the shopping mall group felt buying flowers to be an extravagance and bought flowers more on impulse rather than because of a family tradition.

Since there were differences between respondents obtained from the two sources, the distribution of those respondents across the three discriminant groups was examined. Respondents from each source were satisfactorily represented in each group. [The distribution of respondents from the two sources across the three discriminant groups is as follows: TABLE.]

Further chi-square analyses and t-tests were performed on responses received from subjects obtained from each data source within each discriminant group. Although there were some significant differences found, there were far fewer differences than there were for the entire sample indicating that many of the differences identified with the whole sample were related to the variables used to form the discriminant groups, not simply to individual differences between respondents obtained from the two sources.

Although in this study discriminant analysis was used primarily for descriptive purposes to provide a multi-variate profile of purchasers in each of the groups and not for purposes of prediction, a holdout sample was created and analyzed in order to evaluate the stability of the data. Using the purchasing variables, a holdout sample was created by randomly selecting one-half of the cases assigned to each of the discriminant groups. The discriminant functions computed using only these cases were then used to classify the entire sample. The holdout sample resulted in 75.63% correct classification compared to 71.18% correct classification with the discriminant functions computed utilizing data from the entire sample. The similarity of these percentages suggests that there is no significant problem with bias in the functions calculated from the full sample. However, precautions in interpretation of these data are presented in the following discussion of findings.

Purchaser Characteristics

Two discriminant functions were estimated in the analysis using attitude and demographic/socioeconomic variables as predictor variables. The first accounted for 60.63% of the variation explained after orthogonal rotation. Both functions were significant at the p < .05 level with Wilks Lambda of .76 and .91 respectively. The classification matrix, which is subject to upward bias since the functions used to predict group membership were derived from the complete sample, indicates that these functions were successful in correctly assigning 55.9% of the cases to the three groups. This figure compares to 34.9% and 43.2% correct assignment expected with the proportional chance criterion and the maximum chance criterion respectively (Morrison, 1969). The rotated standardized discriminant coefficients are presented in Table 2 and the group means for these variables are displayed in Table 3. Function I discriminates Group II (Obligatory Event) from Group I (Personal Use) and Group II (Gift Occasions). Function II discriminates Group I from Group II and Group III. The variables with the highest coefficients for Function I are three dummy variables identifying respondents' occupations while the variables most important in Function II are three attitude measures.

TABLE 2

PURCHASER CHARACTERISTICS  -  ROTATED STANDARDIZED DISCRIMINANT COEFFICIENTS

TABLE 3

PURCHASER CHARACTERISTICS  -  GROUP MEANS

Interpreting the standardized discriminant coefficients as relative importance weights in conjunction with an examination of the group means on these variables suggests that Function I identifies respondents in Group II (Obligatory Event) as more likely to be female and the principle wage earner in their households as less likely to be students, housewives or executives, but more likely to be skilled white collar workers than respondents in Group I (Personal Use) and III (Gift Occasion). In contrast, Function II differentiates respondents in Group I from Groups II and III on the basis of their attitudes toward flowers. Personal Users (Group I) are most likely to buy flowers simply on impulse, to include flower purchases as part of their regular budget, and least likely to feel flowers are a family tradition. They are less likely than respondents in Groups II and III to be blue collar workers, students, or housewives. Examination of the significant univariate F-ratios suggests that the age of respondents is also different between groups with Obligatory Purchasers (Group II) older than Personal Users (Group I) and Gift Purchasers (Group III). Taken together these data support the hypothesis that consumers who purchase flowers for different reasons differ on demographics and attitudes.

Purchasing Variables

Discriminant analysis was also performed using the same groups as dependent variables and using variables describing past and intended purchasing behaviors as the predictor variables. This analysis resulted in 71.18% correct classification (p < .01) and two significant functions. Again interpreting the standardized discriminant coefficients as relative importance weights, Function I appears to separate Personal and Obligatory Purchasers from Gift Purchasers, and Function II separates Personal Purchasers ,rom Obligatory and Gift Purchasers. The variables with discriminant coefficients > .3000 on one or both of the functions are presented in Table 4. The group means for all those variables with either univariate or multivariate F-ratios of p < .05 are displayed in Table 5.

TABLE 4

PURCHASING CHARACTERISTICS  -  ROTATED STANDARDIZED DISCRIMINANT COEFFICIENTS

The variables which have the highest discriminant coefficients on Function I all represent future purchase intentions, two relating to occasions for purchase and three relating to the type of store where respondents plan to shop. Function II contains 10 variables with discriminant coefficients > .3000. Five of these variables relate to occasions for intended purchases, two to stores where purchase is planned, two to past purchasing behavior and one to reason for shopping in the most frequently patronized store.

TABLE 5

PURCHASING CHARACTERISTICS  -  GROUP MEANS

Group II appears to resemble Group I along the dimensions important to Function I. It is also very similar to Group III in terms of those variables important in Function II. Thus, the functions are illustrated primarily by describing Groups I and III.

At the multivariate level, discriminant coefficients suggest that respondents in Group I (Personal Use) clearly have a greater intention to buy flowers in the future than do respondents in Groups II and III. Those in Group I have greater intentions to purchase -for personal use, anniversaries, "other special days," dinner hostesses, and birthdays. They are more inclined to purchase on impulse than to plan their purchases and least likely to purchase flowers from a mall florist shop.

Gift Purchasers (Group III) also intend to purchase flowers in the future. Although they intend to purchase for a personal reason, they are more likely to plan to purchase for "other special days," anniversaries, and birthdays. Gift purchasers have the widest range of shopping intentions including shopping at grocer/retailers, full service cash and carry florist shops, and mall florists.

Examination of the group means for those variables with significant univariate F-ratios (p < .05) provides additional descriptive details regarding group members. Personal Users are most likely to have made an impulse purchase of flowers in the past, most likely to intend to make flower purchases in the future for personal use, anniversaries, and "other" reasons. Although both Obligatory Purchasers and Gift Purchasers have purchased flowers in the past their purchase patterns are quite different. Obligatory Purchasers are most likely to have charged these purchases. phoned in their orders. and had the flowers delivered. They are least likely to intend to purchase flowers in the future for anniversaries, "other special days," or for "other" reasons. Conversely, Gift Purchasers paid cash and picked up their purchases. They paid the lowest price and were the most likely to have purchased from a male florist. Their future intentions suggest they want to pay a low price yet plan to buy from either a full service or a mall florist, often among the highest priced of the available outlets. Based on their past purchasing, convenience may play a more important role in their decision than price. Gift Purchasers are least likely to intend to buy flowers for a patient in the hospital, but very likely to intend to purchase for birthdays and "other special days." Generally these data support the hypothesized differences in purchase behavior among consumers who purchase flowers for different reasons.

DISCUSSION

The findings from this study tend to support earlier research on situational influences on consumer behavior which suggests that personal use and gift use represent different situations. Further, findings suggest it may be worthwhile to differentiate gift-giving situations into those that are obligatory in nature and those that involve voluntary gift-giving with little expectation of reciprocity.

Discriminant analyses of flower purchases with groups formed to represent personal use, obligatory events, and gift occasions using both purchaser and purchasing variables as predictor variables revealed significant differences between the groups as predicted in the hypotheses. Some additional unanticipated differences were also highlighted.

Personal Uses

It was hypothesized that purchases made for personal use would be low priced, made from minimum service conveniently located stores and often made on impulse. The data clearly identified this group by the attitudes of group members toward flowers and by their future intentions regarding flower purchases. Interestingly, their greater likelihood of purchasing for personal reasons described only part of their flower buying behavior. They were also more likely to intend to buy flowers for future obligatory events and gift occasions. They were most likely to purchase on impulse. These behaviors were supported by the respondents' attitudes toward flowers. These purchasers were most likely to include flower buying as part of their normal budget (as opposed to feeling buying flowers is an extravagance), and most likely to say they buy flowers "just on impulse" (as opposed to planning purchases for special occasions). These respondents tended to be in the career age groups (20-29 or 30-44 years of age). They were least likely to have a blue collar worker as the principal wage earner in the household. In sum, the people who bought flowers for personal reasons had positive attitudes toward flowers, were likely to come from younger, professional/executive families and felt flowers were appropriate for a variety of occasions.

Obligatory Events

As hypothesized, flower purchases made for obligatory events were made simply. Purchases were most often charged, ordered by telephone, and store delivered to the recipient. Buyers in this group intended to buy flowers for other obligatory events in the future but not for future gift occasions. This group of purchasers was comprised of the oldest respondents and those furthest along in the family life cycle. They were predominately female members of older couples with no dependent children or were solitary survivors. These characteristics describe persons likely to have many social obligations, e.g., graduations or weddings of grandchildren, social engagements "requiring" gifts of appreciation, funerals of friends also in the later stages of life. The convenience of telephone ordering and delivery service may be important in their purchasing as they are likely to be less mobile.

Gift Occasions

Those respondents who have purchased flowers for gift occasions in the past were most likely to intend to buy flowers for gift occasions in the future. They were least likely to buy flowers either for their own personal use or for obligatory occasions. This behavior was reinforced by their attitude that flower purchases are "planned for special occasions." They generally paid cash for their purchases and delivered their gifts in person. Although they were least likely to have bought from a full-service florist in the past they plan to do so in the future, perhaps to take advantage of the extras offered by this type of shop which may make a gift more special. Gift Purchasers have bought flowers from mall florists and intend to do so again in the future. These purchasers were the youngest of the respondents and were likely to be students or housewives or to have an executive as the head-of-household. Contrary to what was hypothesized, past purchases for gift occasions were generally in the lower price ranges ($11-15) and respondents also intended to pay the least for future purchases. This may be a result of the frequency of gift occasions as compared to obligatory events. The infrequent occurrence of events such as friends in the hospital, funerals, and weddings (for any individual recipient) may make a more expensive item appropriate. Conversely, the relative frequency of occasions for which an individual may be an intended recipient of a gift (e.g., birthday, Saint Valentine's Day, Easter, Mother's Day), may make less expensive gifts acceptable.

CONCLUSIONS

The findings from this study suggest that more detailed understanding of occasions for purchase, particularly alonE the lines of occasions of purchase for non-personal use, may prove to be a fruitful area for future research. As has been demonstrated in many different contexts, the more specific the predictors of future behavior, the more accurate the predictions are likely to be. Just as knowledge of whether someone is shopping for a purchase for themselves or for a gift for someone else allows more accurate prediction of that person's shopping/purchase behavior, this research suggests that knowledge of whether a non-personal purchase is being made for an obligatory event or for a gift occasion may also facilitate more accurate prediction of consumer behavior.

REFERENCES

Adams, J. S. (1963), "Toward an Understanding of Inequity," Journal of Abnormal and Social Psychology 67, 422-436.

Belk, R. W. (1979), "Gift-Giving Behavior," in J. Sheth (ed.)S Research in Marketing. Vol. 2, (Grenisch, Conn., JAI Press Inc.) 95-126.

Belk, R. W. (1977), "Gift-Giving Behavior, Part A." Faculty Working Papers, College of Commerce and Business Administration (University of Illinois at Urbana-Champaign).

Belk, R. W. (1976), "It's the Thought That Counts: A Signed Diagraph Analysis of Gift Giving," Journal of Consumer Research, 3, 155-62.

Belshaw, C. S. (1965), Traditional Exchange in Modern Markets, (Englewood Cliffs, NJ: Prentice-Hall).

(February 1979), Florist, "1979: Will It Be Good or Bad for Floriculture?", 57-59.

(August 18, 1979), Plover News, "Market Study Shows Floral Industry Stable," Vol. 33. No. 3, 1, 3, 5.

Gronhaug, R. (1972), "Buying Situation and Buyer's Information Behavior," European Marketing Research Review, 7. 33-48.

Homans, G. C. (1961), Social Behavior Its Elements Forms, (Nev York: Harcourt, Brace and World).

Levi-Strauss, C. (1965), "The Principle of Reciprocity," in L.A. Coser and B. Rosenberg (eds.), Sociological Theory, (Nev York: MacMillan).

Mauss, M. (1954), The Gift: Forms and Functions of Exchange in Archaic Societies, English translation by F. Cunnison, (Glenco. IL: Free Press)

Morrison, D. G. (1969), "On the Interpretation of Discriminant An lysis," Journal of Marketing Research,6, 156-163.

Ryans, A. (1977), "Consumer Gift Buying Behavior: An Exploratory Analysis," in D. Ballinger and B. Greenberg (eds.), Contemporary Marketing Thought, (Chicago: American Marketing Association, Series no.44). 99-104.

Sherif, M., and Sherif, C. W. (1963), "Varieties of Social Stimulus Situations," in S. B. Sells (ed.), Stimulus Determinants of Behavior, (Nev York: Ronald Press), 82-106.

Swartz, B. (1967), "The Social Psychology of the Gift," American Journal of Sociology, 3, 111.

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