ABSTRACT –
An information board study compared product data acquisition and choice in gift versus personal use scenarios. No difference in shopping behaviour was found when the gift was for a close friend. When the gift was for a “friend’s wedding” less effort was devoted to the gift buying than to the personal use buying.
INTRODUCTION
This study is concerned with behavioral differences between purchasing an item for personal use and purchasing it as a gift. Gifts form an important part of the total sales of some consumer products. For a marketer of a giftable product, it is important to know (a) how and why his product class may be chosen as a gift, and (b) how and why his brand may be chosen. Since marketing effort with respect to (a) will also benefit competition, (b) is likely to be more critical to the marketer and is the focus of this study.
The phenomenon of gift giving is in itself a subject interesting to psychology and sociology because of the complex motivations and symbolism involved. For example Schwartz (1967) noted that the gift object was more than merely the physical object — it is a symbol of the giver’s perception of the recipient, and the recipient’s acceptance of the gift object amounted to his acceptance of the giver’s idea of him. Levi-Strauss (1965) also saw gift-giving as serving needs beyond the mere exchange of commodities. He wrote:
“…goods are not only economic commodities but vehicles and instruments for realities of another order: influence, power, sympathy, status, emotion, and the skillful game of exchange consists of a complex totality of maneuvers, conscious and unconscious, in order to gain security and to fortify one’s self against risks incurred through alliance and rivalry”.
Thus in gift-giving it is not just “the thought that counts”, but involved is a system of reciprocal maneuvering through which the giver and the recipient achieve higher order objectives.
Although the motivational and cognitive theories of gift-giving seem fairly well developed, not much attention has been paid to the behavioral aspects of the act. Since it has been shown that the gift object has evaluative dimensions vis-a-vis the recipient, and that the reciprocity relationship is characterized by a goal structure, it would seem that buyer behavior with respect to objects purchased as gifts would be different to that for objects purchased for own or normal household use.
Consumer Behaviour Research
Recently Ryans (1977) investigated the gift versus personal use purchases of small electrical appliances, using a recall questionnaire method. He found that appliances for gifting were more likely to be purchased from stores with quality image than those for personal use. He also found that the amount to be spent was more likely to be set before the purchase, and that the time between the decision to purchase and the actual purchase tended to be shorter for gifts to members of a different household (outcome gifts). Generally he found no significant behavioral differences between purchasing for personal use and purchasing for a gift to a member of the same household (inhome gifts) with regard to price-setting and pre-purchase deliberation. Perhaps this was due to the greater cognitive balance between members of the same household (Belk, 1976).
Weigl (1974) investigated a hypothesis that gift giving would involve more perceived risk than purchases made for personal use. Weigl’s product was tennis racquets and his gift scenario asked respondents to consider buying the product as a gift for a close friend of the same sex. His main hypothesis was not supported though a sub-hypothesis that perceived that financial risk would be greater for own use was supported. Weigl also hypothesized as a corollary that information acquisition and search time would be greater for gifts than for own use products. This hypothesis was also not proven. Weigl’s report contains only the statistical significance or non-significance of the results so it is not know whether his results were in the hypothesized directions but not statistically significant at traditional levels of significance, or whether his results were contrary in direction to his hypotheses.
Heeler and Okechuku (1978) investigated data acquisition and search time for blenders in “own use” versus “gift for a friend’s wedding” scenarios. They found significantly greater information acquisition for the “own use” condition. “Own use” respondents also took 30% longer to make decisions but this difference was not statistically significant at traditional levels of significance.
The studies described above appear to be in conflict with respect to whether gift buying is more, less, or equal in purchase activity to personal use shopping. But a resolution is possible if the contexts of gift giving are compared. Weigl found no difference when the gift was for a “close friend”. Ryans found no difference for inhome gifts. Thus close friends and relatives appear to be treated equally with one’s self in shopping. Heeler and Okechuku found less purchasing effort for a “friend’s wedding”. Ryans found less time was devoted to outhome gifts. Thus gifts to more distant people appear to involve less effort for the purchaser than own-use purchases. Perhaps this is why Belk (1976) in a study of gift satisfaction found that satisfaction with gift selection is likely to be higher under conditions of balanced cognitive configurations between the giver and receiver. Close friends or in-home recipients who are likely to be closer in cognitive configuration, have more shopping time spent on them which should result in greater gift satisfaction.
Hypothesis
It is hypothesized that in gift giving involving close friends or relatives, as much trouble will be taken in the purchase as for an own use purchase. The combination of emotional bonding and the pressures to be perceived well in a close continuing relationship will lead to the recipient being noted as worth as much effort as one’s self. By contrast a gift of a more distant, perhaps conventionally required, nature will involve less effort than for one’s self. The purchasing duty must be adequately performed but is less involving than for one’s self or a close friend or relative.
The information board method was extended to test this hypothesis as described below.
METHOD
To operationalize “distant” and “close” gift giving situations a “friend’s wedding” and “a personal gift for a close friend” were chosen as alternate gift scenarios. Subjects consisted of 108 male and female students at a Polytechnic Institute. With ages between 20 and 30, and many enrolled as part-time students, the group was very familiar with weddings and close friends. Subjects received no monetary or participation points for participating in this study, but the Polytechnic teachers provided class time to the experimenters for a debriefing that used the exercise to illustrate the consumer behaviour research process.
Preliminary research on products suitable for the gift occasions led to two different products being used, blenders and watches. Blenders were appropriate to and used for “friend’s wedding” but were inappropriate for a gift to a close friend. Watches were appropriate to and used for “close friend” but were inappropriate as a wedding gift. Thus product type and scenario are confounded in the research, but if the product and scenario were to realistically relate to each other this seemed to be inevitable. The real choice is between a personalized gift for a close gift giving situation, and a more utilitarian product for a more utilitarian occasion. A further constraint on product selection was product suitability for the information board procedure. For the procedure to have meaning there must be a good range of alternative brands in the product class, and a good range of attributes to differentiate brands. Watches and blenders both met this constraint. Blenders and watches were also products known to the respondents.
The Information Display Board (IDB) familiar to consumer behaviorists (for example see Berning and Jacoby, 1974) was used to elicit the depth, sequence, and content of information acquired prior to the choice of a brand of blender.
The(IDB) is essentially a board on which the attributes of the product class under study are displayed down the left hand column and the brands across the top (or vice versa) so that information on the brands is provided in a matrix format. The number of brands and attributes included determine the number of cells or bits of information in the matrix. Each cell contains the relevant brand-attribute information.
The information in the matrix cells is initially concealed from the subject. The subject is asked to make a product selection using as much or as little information from the board as they personally need. Thus the amount and type of information used by the subject is under his control, as it would be in a real shopping trip. The results of(IDB) studies usually show the same partial use of data in product selection as appear in observation and protocol studies of actual shopping trips.
Like other laboratory techniques,(IDBs) are artificial in requiring a special shopping simulation behaviour of subjects. Jacoby et al (1977) have found reasonably high correlations between(IDB)and verbal report data using breakfast cereals, margarine, and headache remedies, with respect to choice behaviour, thus providing some evidence for convergent validity (Heeler and Ray, 1972).
In this study,(IDBs) were used in an experimental design which should further improve the validity of the results. It was not own use or gift use purchasing which were interesting per se, but the difference between these two behaviors. Thus the artifact of laboratory experimentation was common to both experimental conditions and should net out when the difference is taken.
The boards for both products used sixteen brands and ten attributes for each brand. The sixteen watch brands were subdivided into eight male and female watches. Thus each board contained 160 data cells. Watch respondents operated in either the male watch or the female watch section and so effectively operated with 80 data cells. The data for each cell was mounted on the board, and concealed prior to respondent access by a card slipped into a simple holding rack. Brands were on the horizontal axis, attributes on the vertical axis. Brands and attributes were randomly assigned in order. The brand and attribute positioning was the same for all respondents. This results in a position effect (as described later), but the position bias was equal between the experimental conditions so no attempt was made to neutralize it.
Data Collection Procedure
The subjects were randomly assigned into four groups, a blender own use group (25), a watch own use group (27), a blender “gift for a friend’s wedding” group (27), and a watch for “personal gift for a close friend” group (29). Each subject answered a product familiarity question and was then asked to choose a product according to the specific experimental group he was in. The subject was free to access as much or as little information as he deemed necessary to choose a brand and to announce his choice as soon as he made his decision.
When the subject indicated that he understood the instructions he was told to begin, and the experimenter activated a concealed stop watch. At the same time the interviewer was recording the codes identifying the bits of information accessed, in the sequence in which they were accessed. Data were collected on product familiarity, information bits selected and order of selection, time spent deliberating, and product chosen.
For watches there was a near equal proportion of women in both own use and gift conditions and high product familiarity for all respondents. For blenders there was also a near equal proportion of women in both conditions, but a higher proportion of gift subjects (82%) than own use subjects (56%) were highly familiar with blenders.
RESULTS
The principal hypothesis was supported. For the “close friend” scenario there was no significant difference in information accessed or time spent between the “gift” and “own use” conditions. For the “wedding” scenario there was significantly (t test p< .05) more information accessed in the “own use” condition than in the “gift” condition. “Own use” respondents also took longer than the “gift” use respondents in making their selection, but this difference was significant only at the 15% level. The detailed results are given in Table 1.
In the methodology section it is noted that for blenders, more respondents were familiar with the product in the “gift” condition than in the “own use” condition. A rival hypothesis for the difference in data acquisition between the “gift” and “own use” condition could be that respondents familiar with the product would access less information, and since the “gift” respondents were more product familiar, this could account for the lower rate of accessing in the gift condition. In fact, in both “gift” and “own use” conditions, respondents more familiar with blenders accessed more information, so if no other factor were present one would expect a greater rate of information accessing in the gift condition, that is the opposite of the result obtained. In the “gift” condition “familiar” respondents averaged 28.3 data cells versus 21.4 for the “unfamiliars”. In the “own use” condition “familiar” respondents averaged 44.0 data cells versus 33.0 for the “unfamiliars”.
Table 1 also shows comparative results for brands and attributes accessed, and search transition types. It had been expected that the results for the number of brands and attributes accessed would parallel the main hypothesis, i.e. more for “own use” in the “wedding” scenario, equivalence for the “close friend” scenario, but this expectation was not borne out. For the “wedding” scenario, “own use” subjects accessed more brands than “gift” respondents, but this was only significant at the 12% level. “Wedding” scenario “gift” subjects actually accessed slightly more attributes than “own use” subjects. In the “close friend” scenario, “own use” subjects accessed more attributes but less brands than “gift” subjects.
In order to investigate possible differences in information acquisition sequence between purchasing situations use was made of the 4 different types of information search strategies identified by Jacoby et al (1975). Transition type 1 refers to a same brand-same attribute acquisition in a sequence of two information acquisitions, that is, immediate re-access of the same-brand attribute information. This transition was not possible in the present study due to the use of a fixed-board design. Type 2 transition refers to a same brand-different attribute sequence; type 3 to a different brand-same attribute sequence; and type 4 to a different brand-different attribute sequence.
The proportions of each respondent’s information search involving types 2, 3 and 4 search strategies were calculated and reported in Table 1. For both “wedding” and “close friend” scenarios there was a higher proportion of type 4 transitions in the “gift” condition than in the “own use” condition, but this was not statistically significant.
Brands selected were similar for “gift” and “own use” conditions for both “wedding” and “close friend”. This is illustrated by the average price of brand chosen data in Table 1. The blenders in the “wedding” scenario ranged in price from $22 to $43 with an average of $34. The average price of blenders chosen was $33 “own use” and $35 “gift”. For watches in the “close friend” scenario, prices ranged from $12.95 to $48 with an average price of $34. The average price of blenders chosen was $30 in the “own use” condition and $32 in the “gift” condition. Thus slightly higher priced brands were selected in both scenarios for “gifts”, but the differences were not statistically significant.
Table 2 shows the distribution of information access by attributes.
In the “wedding” scenario with blenders the most accessed attributes were price, brand name, and number of speeds. After allowing for the overall difference in information accessed, the differences in individual attributes accessed was small. Price was the most accessed attribute for “own use” but brand name was most accessed by “gift” subjects. “Gift” subjects also emphasized number of speeds and wattage more than “own use” subjects. These overt indicators, brand, speeds, and wattage, may be symbolically important in the gift situation.
In the “close friend” scenario with watches the most attributes accessed were consistent in both “gift” and “own use” conditions. Watch design (a picture of the watch), brand name, and price accounted for over half of the information accessed.
Methodological Considerations
While the(IDB)gave excellent control for a laboratory setting, and many of the likely biases were balanced out by the test and control group design, the authors were left with some concerns about the general applicability of the method.
The scenario itself requires subjects to take on a forced, “left brain” mock shopping task. In debriefing, most subjects admitted to some task uncertainty in the first rounds of data acquisition, and about 15% to experiencing confusion or difficulty with the task. Observation of subjects in process also suggested some of the “yeah saying” and “show I am logical” behaviour that is a problem in questionnaire research. Perhaps the most extreme case was a subject who announced at the start that “you will see what a fast decision maker I am” and then took twenty minutes longer than any other subject, carefully taking every piece of data from the board.
In debriefing subjects also reported some feeling of time constraint, despite the study’s use of the standard (IDB)phrasing of “take as much or as little information as you like”. This in fact may parallel true-to-life shopping. One respondent likened the feeling to going shopping with a friend and not wanting to keep the friend waiting too long. This feeling of time pressure was despite the subjects accessing more board information than in many other reported studies. The respondents accessed an average of 28 pieces, compared to 7 in Jacoby et al (1975).
INFORMATION ACQUISITION BY ATTRIBUTE
In pilot work for the study a re-access(IDB)was tried. With the re-access board, subjects see a particular piece of information when they first access it, but to view it again they have to re-access it. This process caused great annoyance to respondents and breakdowns in realistic participation.
A strong Northwest corner effect was noticed in board use, that is respondents are biased to accessing information from the top left hand corner of the board. For example, the top eight brands of blenders were accessed 62% more than the bottom eight, and (combining male and female versions) the top four watches were accessed 55% more than the bottom eight. This effect is analogous to the inevitable shelf position effect on brands in a store but in a study that was not researching shelf effect could lead to misleading interpretations. There was a similar though less pronounced left side bias with attributes. In the current study these order biases were equal for the experimental conditions compared. In other research (in progress) the authors rotate the board position of brands and attributes between respondents.
DISCUSSION
The hypothesis given earlier stated that gift purchases for close friends or relatives will involve as much effort as personal shopping, but more distant gifts will involve less effort. The hypothesis was supported by the study in terms of information acquired. The lower information acquisition in the “wedding gift” condition was in terms of amount of information per brand or attribute, rather than in terms of number of brands or attributes considered. This laboratory study covered only one corner of the overall gift purchase situation, but is consistent with the research cited earlier. From the perspective of a marketer of giftable products, there is an indication that for some products in some gift situations, the gift shopper may want to shop more rapidly than the own use buyer, so simplified shopping arrangements (such as a wedding gift book) could aid sales.
REFERENCES
Russel W. Belk, “It’s the Thought That Counts: A Signed Diagraph Analysis of Gift-Giving,” The Journal of Consumer Research, Vol. 3, No. 3 (1976) 155-162.
Carol A. Berning and Jacob Jacoby, “Patterns of Information Acquisition in New Product Purchases,” The Journal of Consumer Research, Vol. 1, No. 2 (1974) 18-22.
Roger M. Heeler and Chike Okechuku, “Brand Selection for Gift Giving Versus Personal Use,” ASAC Conference (June, 1978) (Proceedings in press).
Roger M. Heeler and Michael L. Ray, “Measure Validation in Marketing,” Journal of Marketing Research, 9 (1972) 361-370.
Jacob Jacoby, Robert W. Chestnut, Wayne D. Hoyer, David A. Sheluga, and Michael J. Donahue, “Psychometric Characteristics of Behavioural Process Data: Preliminary Findings on Validity and Reliability,” Purdue Papers in Consumer Psychology, Paper No. 172 (1977).
Jacob Jacoby, Robert W. Chestnut, Karl C. Weigl, and William Fisher, “Pre-Purchase Information Acquisition: Description of a Process Methodology, Research Paradigm, and Pilot Investigation,” Purdue Papers in Consumer Psychology, Paper No. 151 (1975) 306-314.
C. Levi-Strauss, “The Principle of Reciprocity,” in Lewis A. Coser and Bernard Rosenberg (eds.), Sociological Theory, (New York: MacMillan Co., 1965) 76.
Adrian B. Ryans, “Consumer Gift Buying Behaviour: An Exploratory Analysis,” American Marketing Association, Fall Conference (1977) (Proceedings in print).
Barry Schwartz, “The Social Psychology of the Gift,” The American Journal of Sociology, Vol. 73, No. 1 (1967) 1-11.
Karl C. Weigl, “Perceived Risk and Information Search in a Gift-Buying Situation,” Unpublished Master’s Thesis, Purdue University (1975).
—————————————-
