The Effect of Donor-Recipient Involvement on Consumer Gift Decisions

Citation:

Janet Wagner, Richard Ettenson, and Sherri Verrier (1990) ,"The Effect of Donor-Recipient Involvement on Consumer Gift Decisions", in NA - Advances in Consumer Research Volume 17, eds. Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, Provo, UT : Association for Consumer Research, Pages: 683-689.

Advances in Consumer Research Volume 17, 1990      Pages 683-689

THE EFFECT OF DONOR-RECIPIENT INVOLVEMENT ON CONSUMER GIFT DECISIONS

Janet Wagner, University of Maryland

Richard Ettenson, University of Maryland

Sherri Verrier, University of Maryland

The Sherry (1983) model suggests that gift-giving decisions vary by the level of involvement in the donor-recipient relationship. A conjoint task involving a gift purchase was administered to consumers in two conditions: 1) a gift for a new neighbor, and 2) a gift for a best friend. The results showed that the use of expressive and utilitarian attributes, as well as price, differed between the two gift-giving conditions. It was concluded that the results lend support to the Sherry model by demonstrating that level of involvement affects the gift purchase decision.

INTRODUCTION

Gift-giving is a phenomenon of economic and social importance, which has been the focus of a growing body of conceptual and empirical work in consumer behavior. In the economic realm, expenditures for gifts represent more than 3% of the annual budget of the average household (Garner and Wagner 1987). In the social sphere, gifts serve as important symbols in creating and maintaining relationships and in acknowledging important life cycle events (Belk 1979; Sherry 1983).

Conceptual Framework

Sherry (1983) proposed a multidisciplinary model of consumer gift-giving based on a typology with three major components -- the gift, the relationship between donor and recipient, and situational factors. According to this model, the gift-giving process has three stages -- gestation, prestation, and reformulation. Gestation resembles traditional consumer decision-making models (e.g. Engel, Blackwell and Miniard 1986), and may be of the most interest to marketers, because it culminates in a purchase.

Gestation begins with a precipitating condition, which is the donor's recognition of a gift-giving situation. During this stage, decision-making is affected both by the gift-giving motives of the donor and by hints from the recipient about his or her needs and wants. The donor evaluates alternative gifts in terms of their attributes and what those attributes communicate to the recipient. Finally, a purchase decision is reached.

According to Sherry (1983), the effect of product attributes on the choice of a gift may vary according to the level of involvement between the donor and the recipient. Expressive gifts are more likely to be given in close relationships, while utilitarian gifts are more likely to be given in distant relationships. Attributes such as price and quality may also be used to define level of involvement. More costly, higher quality gifts are usually given in closer relationships.

One major function of gift-giving is communication (Banks 1979; Belk 1979; Sherry 1983). Because gifts communicate information not only about the recipient, but also about the donor, consumers may engage in "impression management" (Sherry 1983). This suggests that gift-giving decisions are likely to be influenced by social norms.

Previous Research on Gift-Giving

Warshaw (1980) studied the relationship between social norms and perceived expense in the purchase of gifts by college students. The influence of social norms appeared to decrease as the perceived expense of a gift increased. Conversely, the influence of social norms appeared to increase as perceived expense decreased.

Belk (1982) studied the effect of involvement on the gift-giving strategies of adult women. Eighty-seven gift attributes were rated with respect to their importance. Price, quality, and convenience were hypothesized to differ by level of involvement. Subjects were assigned to three experimental conditions in which involvement was manipulated by intimacy of the relationship (close friend, casual friend, close relative) and gift-giving occasion (birthday or wedding). The results of analyses of variance (ANOVA's) showed that high involvement was associated with the importance of price and quality. The effect of the occasion on gift-giving was studied by Devere, Scott and Shulby (1983). Undergraduates rated the importance of 48 attributes of birthday and wedding gifts. In both situations, the most important gift attributes were those related to style, quality, usefulness, performance, and price. Style was more important in the birthday gift, and quality and performance were more important in the wedding gift.

Any resource -- a good, a service, or money - may be used as a gift. Perhaps the most popular gift is clothing (Belk 1979; Caplow 1982; Jolibert and Fernandez-Moreno 1983). Because clothing conveys information about sex, age, status, and personality (Sproles 1979), a gift of clothing may be one of the most effective ways for a donor to communicate his or her perception of the recipient and the gift-giving relationship. Horne and Winakor (1988) studied gifts of clothing exchanged within families at Christmas, and found that subjects reported style, color, fiber content, and cost to be the most important attributes. Rucker, Boynton, and Park (1986) compared attributes used in gift and nongift purchases of children's clothing, and found that "embellishment" was more important in gifts. Andrus, Silver and Johnson (1986) studied the effect of brand on purchases of clothing gifts, and found that consumers were willing to pay 20% more for a gift in order to purchase a status brand.

Much of the research on gift-giving has been based on methods in which subjects are asked to report or rate the importance of a set of product attributes (e.g. Andrus, Silver and Johnson 1986; Belk 1979; Devere, Scott and Shulby 1983; Horne and Winakor 1988). The results of such studies provide insight into consumer gift-giving behavior and are useful in generating list of attributes that may affect gift-giving decisions. The results must be interpreted with caution, however, for two reasons. First, self-reports are often poor surrogates for the decision-making process (Achenbaum 1966; Bettman 1979; Nisbett and Wilson 1977). Gift-giving involves social interaction, so subjects may have given socially desirable responses. Second, subjects have rated attributes one-at-a-time, in isolation from the bundle of attributes which typically define a product. Consequently, the results give little information about the psychological trade-offs which consumers are likely to make among product attributes in purchasing gifts.

In this research, conjoint analysis was used to compare gift-giving decisions in two situations, which were varied by level of involvement in the donor-recipient relationship: 1) a gift to a best friend, and 2) a gift to a new neighbor. Use of the conjoint approach was predicated on the assumption that the decision to purchase a gift, like the decision to purchase a product for personal use, involves evaluating and integrating information on multiple product attributes. Because the gift to a best friend was assumed to be more involving than the gift for a new neighbor, it was expected that price, quality, and expressive attributes would be more important in the best friend condition. It was also expected that utilitarian attributes would be more important in the new neighbor condition.

METHOD

Participants

One-hundred mothers were systematically sampled from the P.T.A. membership list (N=197) of a public elementary school in the suburbs of a major metropolitan area in the Mid-Atlantic region. Adult females were used as subjects for two reasons. First, most donors are female. As the "unpaid social directors" and "chief gift-givers" of American society (Schnudson 1986), women purchase 84% of all gifts (Caplow 1982). Second, the decision-making task upon which this study was based involved a gift for a baby shower. Consequently, the use of adult females ensured that the subjects were familiar with the gift-giving situation. Permission to recruit the mothers was obtained from the P.T.A.'s executive board at its meeting in the month prior to administration of the research. To encourage participation, the P.T.A. was promised a $5 donation for every completed instrument returned to the researchers.

The Gift-Decision Task

Participants were given a set of instructions in which they were told that they had been invited to a baby shower and that the guest of honor had hinted that she wanted one-piece jumpsuits for her new baby boy. An infant's jumpsuit was defined as a one-piece garment, with feet, to be worn while the baby was either awake or asleep. Half of the 100 mothers sampled were told to assume that the guest of honor was a best friend, and the other half were told to assume that the guest of honor was a new neighbor.

The research instrument consisted of a set of decision-making tasks, based on a 2 x 28 mixed 1/16 fractional factorial design, with full replication (Hahn and Shapiro 1966; Plan 7b). The two main factors were: 1) the relationship between the donor and the recipient (a between-subject variable) at two levels: best friend or new neighbor; and 2) eight product attributes (within-subject variables) at two levels each. Attributes were borrowed from the Sproles (1979) model of fashion-oriented consumer behavior, and levels were selected through a focus group of mothers with demographic profiles similar to those of participants. The eight product attributes and their corresponding levels were: price ($15.00 vs. $8.00), color (mint green vs. blue), brand (Carter vs. LACOSTE), fiber (100% cotton vs. 100% polyester), size (6 months vs. newborn), fabric (t-shirt knit vs. terry cloth knit), quality (good vs. very good) and style (classic vs. fashion). In the instrument, all attributes except for style were presented in written descriptions of infant jumpsuits. Style was presented in accompanying sketches. The instrument was pretested with a group of 10 mothers of elementary school children. The results indicated that the two levels of price initially chosen, $5 and $20, were perceived to be unrealistically low and high. The levels were adjusted accordingly.

The experimental design involved 16 cases in which infant sleepers were described. This made it possible to analyze main effects for the eight product attributes as well as six two-way interactions. Each of the 16 cases was fully replicated. By replicating the cases, it was possible to estimate both experimental error in the individual-subject ANOVA's and within-subject consistency. Eight "filler" cases were also included, so each participant made a total of 40 decisions. For each case, participants were asked to evaluate the likelihood of purchase by placing a slash along a 100 millimeter continuum, with ends marked "Not Very Likely" and "Very Likely," which appeared at the bottom of the page. The decision-making task was followed by a post-experiment questionnaire, in which socioeconomic and demographic information was collected.

Procedure

A packet of research materials was mailed to participating mothers in November, 1988. Included were: a cover letter from the researchers, a letter of endorsement from the president of the P.T.A., the research instrument, and a stamped, pre-addressed envelope for return of the completed instrument. The participants were asked to return the instrument within 2 weeks. Reminder letters were sent to those who did not comply. Of the 100 instruments mailed, 69 were completed and returned by the first deadline. An additional eight were received after the reminder, yielding a response rate of 77%. Completed instruments were received for 35 mothers in the new neighbor condition and 42 mothers in the best friend condition.

TABLE 1

DEMOGRAPHIC CHARACTERISTICS OF MOTHERS

RESULTS

Descriptive statistics for the 77 participating mothers are presented in Table 1. For the most part, the mothers were older, better educated, more likely to be working, and more likely to be married than one would expect of mothers of elementary school children. The majority reported having given a shower gift in the last year.

Use of Attributes

An individual-subject analysis of variance (ANOVA) of the fractional design was performed on the gift purchase decisions. This made it possible to identify which attributes affected the gift purchase decisions of each participant. The results are summarized in Table 2. Table 2 shows that the attributes--which most often affected the decisions of the 35 mothers in the new neighbor condition were size, fiber content, and price. Table 2 also shows that fiber content and size were the attributes which most often affected the decisions of the 42 mothers in the best friend condition. Compared to the mothers in the new neighbor condition, price seemed to affect the decisions of fewer mothers purchasing gifts for a best friend. Few significant two-way interactions were found between attributes in either the new neighbor or the best friend condition.

Relative Importance of the Attributes

Hays' (1973) omega-squared (w²) was used to evaluate the relative importance of the attributes in the decision-making of individual subjects. The average w² values for the mothers in each experimental condition are presented graphically in the Figure. It appears that the gift purchase decisions of mothers in both conditions were dominated by fiber content, size, and price. Style and quality seem to represent a second-tier of effects, and fabric. color. and brand a third tier.

TABLE 2

PERCENTAGE OF SIGNIFICANT (P<.05) EFFECTS FOR GIFT ATTRIBUTES: NEW NEIGHBOR AND BEST FRIEND

Differences In Decision Making by Level of Involvement

A two (level of involvement) x eight (attribute) ANOVA was performed on the w² values to determine whether or not the gift purchasing strategies differed by intimacy of the relationship. The result showed a significant effect for attributes (F [7,525] = 16.45, p< 05) and no effect for level of involvement (F [1, 76] = 2.92, p>.05). There was, however, a significant level of involvement x attribute interaction (F [7,525] = 2.84, p<.05). The results of a post-hoc Scheffe' test demonstrated that there were differences in the use of the three dominant attributes -- fiber, size, and price. While fiber had more of an effect on the decisions of mothers purchasing gifts for a best friend (w²=.32 vs. .18, p<.05), size had a stronger effect on the decisions of mothers purchasing gifts for a new neighbor (.20 vs. .10, p<.05). Price appears to have been more important to mothers purchasing a gift for a new neighbor (.15 vs. .08); that difference was, however, only marginally significant (p<.10).

DISCUSSION

The results of this research provide insight into the tradeoffs among product attributes that consumers may be willing to make in gift purchase decisions. More important, the results lend support to the Sherry (1983) model of consumer gift-giving, by demonstrating that gift-giving strategies may differ by level of involvement in the donor-recipient relationship.

Tradeoffs Among Product Attributes In Gift Decisions

The gift purchase decisions of the mothers were dominated by fiber, size, and price. Fiber content was the most important attribute. This may reflect both expressive and utilitarian qualities. Fiber content may be expressive because it is subject to fashion change, which is, in turn, associated with status. Examination of the marginal utilities for fiber content showed that all but one of the mothers whose decisions were affected by that attribute favored the 100% cotton garment. Because cotton is more fashionable than polyester (Rowland 1987), consumers may believe that it communicates a desirable impression of the status of the donor as well as the recipient. This suggests that, in making their decisions, the mothers engaged in what Sherry (1983) termed "impression management." Cotton also has utilitarian properties, such as comfort and absorbency, which may have affected the gift purchase decisions.

The dominance of fiber content may reflect the relationship between social norms and perceived expense observed by Warshaw (1980). Fashion is a social norm, so preference for, cotton, the more fashionable fiber, may have been related to the perceived expense of the jumpsuit. Had a more expensive gift been involved, the effect of fiber content might have been weaker. The importance of fiber is consistent with the results of Home and Winakor's (1988) study on Christmas gift-giving.

FIGURE

AVERAGE W² VALUES FOR EACH GIFT ATTRIBUTE: NEW NEIGHBOR AND BEST FRIEND

Size, which might be interpreted as a utilitarian attribute, was the next most important influence on the gift-giving decisions. This finding lends credence to Sherry's (1983) contention that utilitarian considerations may affect gift-giving, and to the results of Devere, Scott and Shulby (1983), who reported that the usefulness of an item is likely to be an important consideration in purchasing a gift. The marginal utilities showed that, among the mothers with a significant effect for size, all but two favored the six-month over the newborn. Given that newborns usually grow quickly, the mothers may have perceived the larger size to be useful for a longer period of time than the smaller size. Mehrabian (1972) suggests an alternative explanation. If size is associated with status, then giving a larger size may be one way of conferring status upon the recipient.

The importance of price is consistent with the results of previous research in which price was reported to be an important attribute in purchasing a gift (e.g. Belk 1979; Devere, Scott and Shulby 1983; Heeler, Francis and Okechuku 1979; Home and Winakor 1988). It appears, however, that consumers may be less concerned with price than with either the status communicated by an item or its usefulness.

Quality and style appeared to be attributes of secondary importance. Quality was reported by both Belk (1979) and Devere, Scott, and Shulby (1983) to be an attribute of importance in purchasing gifts. The results of this study support their finding, but with the qualification that, compared to other attributes, quality seems to have a smaller effect. Both Devere, Scott and Shulby (1983) and Horne and Winakor (1988) reported that style affects gift-giving decisions. The results of this research indicate that, relative to other attributes, the effect of style is not large. This may be related, however, to the situational context, in that participants were told that the shower was being held for the mother of a baby boy. Although one function of gift-giving is socialization, sex-role socialization may not begin in earnest until a child is older. Consequently, style may be of less importance than other attributes in gifts of clothing to infants. The marginal utilities showed that all but one of the mothers who had a significant effect for style preferred the classic jumpsuit. This may reflect the fact that fashion continues to have less effect on the clothing of males than the clothing of females (Hyde 1988).

Color, brand, and fabric appeared to be the least important of the gift-giving attributes. In Home and Winakor's (1988) study, color was reported to be important in gifts of clothing. The results of the conjoint analysis suggest, however, that compared to other attributes, the effect of color is small.

Andrus, Silver and Johnson (1986) reported that consumers are willing to pay a higher price in order to purchase a status brand as a gift. The results of this research indicate that, compared to other product attributes, brand may have little effect on gifts of clothing. The importance of fiber, relative to brand, suggests that physical attributes of a product may communicate more about status than does brand. The price x brand interaction was tested, and 16 (21%) significant interactions were observed across the two experimental conditions. In most cases, the mothers preferred to pay less for LaCoste, the higher status brand. The w² value for this interaction was, however, negligible. Although the Sproles (1979) model suggests that fabric is one of the "critical characteristics" in purchases of clothing, this attribute had little effect on the gift purchase decisions of the mothers in the sample.

Comparison of Gift Decisions by Intimacy of Relationship

Differences were observed in the relative importance of fiber, size, and price in the best friend and new neighbor conditions. Mothers giving a gift to a best friend were more likely to use fiber, and less likely to use either size or price, than were mothers giving a gift to a new neighbor. This lends support to the Sherry (1983) model, which suggests that gift-giving decisions differ by the closeness of the donor-recipient relationship.

According to Sherry (1983), gifts may be either expressive or utilitarian. Expressive gifts are more likely to be given in close relationships, and utilitarian gifts more likely to be given in distant relationships. The use of the gift attributes by the mothers in the two conditions appear to reflect this. Fiber might be considered the more expressive of the two attributes, because it may communicate information on the status of both donor and recipient. Since natural fibers (e.g. cotton) are perceived to be more fashionable than synthetic fibers (Rowland 1987), a gift of a natural fiber garment might be seen as one way of communicating the importance of a relationship. Size might be considered the more utilitarian of the two attributes. Consequently, the decisions of mothers in the new neighbor condition, in which the relationship was distant, were more strongly affected by size.

An alternative explanation for the importance of fiber in the best friend condition might be that consumers make purchases for close friends in much the same way that they make purchases for personal use. Ettenson, Wagner, and Gaeth (1988) found fiber content to be a dominant attribute in decisions involving purchases of clothing for personal use. Heeler, Francis and Okechuku (1979) reported that, in making purchases of gifts for close friends and purchases for personal use, consumers access similar types of information. This interpretation is consistent with the anchoring and adjustment model developed by Davis, Hoch and Ragsdale (1986), in which both husbands and wives were found to anchor their predictions of spousal preferences on their own preferences.

Although the difference was marginal, price appeared to be less important to mothers purchasing gifts for best friends than to mothers purchasing for a new neighbor. Both Sherry (1983) and Caplow (1983) suggest that the value of a gift may reflect the importance of the relationship. This implies that a more expensive gift might be preferred in a closer relationship. However, the less expensive jumpsuit was favored by all of the mothers with a significant effect for price, regardless of level of intimacy. Low cost of the jumpsuit was, however, of more importance to the mothers purchasing a gift for a new neighbor than to mothers purchasing a gift for a best friend. In light of the importance of fiber content, it appears that for some gifts, there may be attributes more effective than price in communicating the importance of a relationship. The difference in the importance of price between the best friend and new neighbor conditions lends support to the work of Belk (1982), who reported that the amount of money spent on gifts-is likely to vary by level of involvement.

CONCLUSIONS AND IMPLICATIONS FOR FUTURE RESEARCH

The results of this research contribute to development of the Sherry (1983) model and suggest directions for future decision-making research in a gift-giving context. Gift purchase decisions were studied in the context of a shower gift to be given to either a best friend or a new neighbor. The results demonstrated that the use of product attributes by the participants differed in the two situations. Thus, it appears that the gift-giving decisions of consumers differ by level of involvement in the donor-recipient relationship.

Four directions for future research are suggested. This study was limited in terms of sample size and range of products. Consequently, the most obvious extension of this research is to a larger, more representative sample of consumers and a wider range of products. A second extension of this research might involve using the conjoint method to compare gift-giving decisions to decisions involving purchases of goods for personal use. Given that gift-giving tends to be seasonal in nature, such information might be useful to the marketing establishment in planning seasonally-appropriate promotional strategies. A third avenue of research might be based on the anchoring and adjustment model of evaluative judgment (Tversky and Kahneman 1974). Davis, Hoch and Ragsdale (1986) used this approach to study spousal preferences. In that research, both husbands and wives were found to be not very accurate in their predictions of spousal preferences. It would be interesting to explore the accuracy of donor predictions of recipient gift preferences, by level of involvement. Finally, given the burgeoning of interest in family decision-making, the study of gift-giving decisions within the family, not only between husbands and wives, but also between parents and children, presents intriguing possibilities.

REFERENCES

Achenbaum, Alvin (1966), "Knowledge is a Thing Called Measurement," in Attitude Research at Sea, eds. Lee Adler and Irving Crespi, Chicago, IL. Attitude Research Committee, American Marketing Association.

Andrus, David, Edward Silver and Dallas Johnson (1986), "Status Brand Management and Gift Purchase: A Discriminate Analysis," The Journal of Consumer Marketing, 3 (Winter) 5-13.

Banks, Sharon (1979), "Gift-Giving: A Review and an Interactive Paradigm,: in Advances in Consumer Research, Vol. 6, ed. William Wilkie, Ann Arbor, MI: Association for Consumer Research, 319-324.

Belk, Russell (1979), "Gift-Giving Behavior," Research in Marketing, Volume 2, ed. Jagdish Sheth, Greenwich, CT: JAI Press, 95-126.

Belk, Russell (1982), "Effects of Gift-Giving Involvement on Gift-Giving Strategies," in Advances in Consumer Research, Vol. 9, ed. Andrew Mitchell, Ann Arbor, MI: Association for Consumer Research, 531-536

Caplow, Theodore (1982), "Christmas Gifts and Kin Networks," American Sociological Review, 47 (3), 383-392.

Davis, Harry L., Stephen J. Hoch and E.K. Easton Ragsdale (1986), "An Anchoring and Adjustment Model of Spousal Predictions," Journal of Consumer Research, 13(1), 25-37.

Devere, Stephen P., Clifford D. Scott and William L. Shulby (1983), "Consumer Perceptions of Gift-Giving Occasions: Attribute Saliency and Structure," in Advances in Consumer Research, Vol. 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Arbor, MI: Association for Consumer Research, 185 -190.

Engel, James F., Roger D. Blackwell and Paul W. Miniard (1986), Consumer Behavior, 5th ed., New York: The Dryden Press.

Ettenson, Richard, Janet Wagner, and Gary Gaeth (1988), "Evaluating the Effect of Country of Origin and the 'Made in the USA" Campaign: A Conjoint Approach," Journal of Retailing, 64 (1), 85-100.

Garner, Thesia and Janet Wagner (1987), "Gift-Giving: An Economic Perspective," paper presented at the Allied Social Sciences Association Annual Meeting, Chicago: IL.

Hahn, G. and S. Shapiro (1966), "A Catalog and Computer Program for the Design and Analysis of Orthogonal Symmetric and Asymmetric Fractional Factorial Designs," Report No. 66-C165, General Electric Corporation, Schenectady, NY.

Hays, W.L. (1973), Statistics for Psychologists, New York: Holt. Rinehart, and Winston.

Heeler, Roger, June Francis and Chike Okechuku (1979), "Gift Versus Personal Use Brand Selection," in Advances in Consumer Research, Vol. 6, ed. William Wilkie, Ann Arbor, MI: Association for Consumer Research, 325-328.

Horne, Lena and Geitel Winakor (1988), "Exchange of Clothing Gifts in Families -- A Pilot Study," in ACPTC Proceedings: Combined Central, Eastern, and Western Regional Meetings, ed.

Cherilyn Nelson, Monument, CO: Association of College Professors of Textiles and Clothing, 65.

Hyde, Nina (1988), "Fashion Notes: Basically the Same," The Washington Post, (January 10), H3.

Jolibert, Alain. J.P. and Carlos Fernandez-Moreno (1983), " A Comparison of French and Mexican Gift-Giving," in Advances in Consumer Research, Vol. 10, ed. Richard P. Bagozzi and Alice M. Tybout, Ann Arbor, MI: Association for Consumer Research, 191-196.

Mehrabian, Albert (1972), Nonverbal Communication, New York: Aldine-Atherton.

Rowland, Mary (1987), "Back to Natural Fibers," Working Woman (November), 101-105.

Rucker, Margaret H., Linda Boynton and Mee Young Park (1986), "Gift versus Regular Purchases: The Children's Market," in Proceedings of the Annual Convention, Division of Consumer Psychology, American Psychology Association, ed. Joel Faigert, Northridge, CA: American Psychology Association.

Schnudson, Michael (1986), Psychology Today, (December), 26-29.

Sherry, John F., Jr. (1983), "Gift-Giving in Anthropological Perspective," Journal of Consumer Research, Vol. 10 (September) 157168.

Sproles, George (1979), Fashion: Consumer Behavior Toward Dress, Minneapolis, MN: Burgess Publishing Company.

Tversky, Amos and Daniel Kahneman (1974) "Judgement Under Uncertainty: Heuristics and Biases," Science, 185, 1124-1131.

Warshaw, Paul (1980), "Buying a Gift: Product Price Moderation of Social Normative Influences on Gift Purchase Intentions," Personality and Social Psychology Bulletin, 6 (1), 143-148.

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