More than a Labor of Love: Gender Roles and Christmas Gift Shopping EILEEN FISCHER STEPHEN J. ARNOLD* Through a field study of 299 men and women, the effect of gender-related variables on Christmas-gift-shopping patterns was explored. Survey results suggest that women are more involved than men in the activity. However, men are likely to be more involved if they hold egalitarian gender-role attitudes. Overall, the study indicates that, while Christmas shopping may be a "labor of love" to some, it is most widely construed as "women's work."

C

ultural influences on consumption are attracting increasing attention. In particular, interest in the rich phenomenon of Christmas consumption has grown. Some studies have examined festive gift-giving practices as a subset of general gift-giving behaviors (Jolibert and Fernandez-Moreno 1983; Sherry and McGrath 1989). Others have focused on the unique values revealed in festive consumption (Belk 1987, 1989; Hirschman and LaBarbera 1989; Pollay 1987). The nature and implications of links between gender and Christmas shopping are explored here through a conceptual discussion ofthe relationship, an analysis of interviews with gift shoppers, and a field study of gift-buying practices. The contributions ofthe research are threefold. First, it sheds new light on the stereotype of gift shopping in the Christmas context as "women's work." Second, the research shows how individual differences in gender-related variables have an effect on the traditional meanings associated with Christmas shopping. Previous research has considered only categorical male/female differences in gift-shopping behavior, but our research examines the link between role, attitudinal, and trait variables in gift shopping. Third, the implications of gender for general consumer activities are developed—gender-related variables (sex, gender-role attitudes, and gender identity) and their influence on gender-typed consumer behavior are clarified.

CULTURAL MEANINGS IN CHRISTMAS ACTIVITIES Christmas reflects diverse themes, some of which have a positive or negative value for consumers and some of which are considered sacred or secular (Hirschman and LaBarbera 1989). Both materialistic and hedonic values (Belk 1989) and starkly contrasting communal values (e.g., love, generosity, and charity) are simultaneously reinforced in modern North American Christmas festivities (Barnett 1954; Cheal 1988; Samuelson 1983). Festive practices are laden with meanings that derive from and contribute to the culture, and the manner in which they are executed is culturally constituted and constitutive. Christmas gifts are particularly value expressive, serving diverse social, economic, and personal purposes (Belk 1979; Sherry 1983). One message primarily conveyed by Christmas gifts is love, affection, and esteem for the recipient. The symbolism of the set of gifts presented communicates the giver's valuation of each receiver (Caplow 1982, 1984). Gift shopping may, therefore, entail intense effort to obtain gifts that reflect the nature of the relationship between giver and receiver. Evidence suggests that gift-purchasing strategies vary according to the giver/receiver relationship (Belk 1982; Sherry 1983; Wagner, Ettenson, and Verrier 1990). Recognizing the communal nature ascribed to gift shopping helps us interpret many observations regarding the contemporary Christmas festival. Communal values are often associated with women, and women have been said to "dominate" Christmas. Barnett (1954) claimed that women control the ritual since they buy and wrap most ofthe gifts, prepare and serve the meals, and decorate the home. Benney et al. (1959) found that women were much more concerned than men with buying "appropriate" gifts and cards. Cap-

'Eileen Fischer is assistant professor. Faculty of Administrative Studies, York University, North York, Ontario, Canada M3J 1P3. Stephen J. Arnold is associate professor. School of Business, Queen's University, Kingston, Ontario, Canada K7L 3N6. This work was supported in part by the Associates Program ofthe Queen's School of Business and the Social Sciences and Humanities Research Council of Canada. The authors would like to thank the reviewers for their many constructive suggestions. 333

' 1990 by J O U R N A L O F C O N S U M E R RESEARCH, Inc. • Vol. 17 •December 1990 All rights reserved. 0093-5301/91/17O3-(XK)7$O2.OO

JOURNAL OF CONSUMER RESEARCH

334

low (1982, p. 388) studied festivities in what he called "Middietown" (Muncie, Indiana) and found that "Christmas gift giving in nearly every household centers around a woman who is the chief performer ofthe ritual." Sherry and McGrath (1989) noted, in a study of gift selling, that women are much more active Christmas shoppers. Such findings suggest that the relationship between gender and Christmas shopping must be considered in more detail if the phenomenon is to be fully understood.

GENDER DIFFERENCES IN CHRISTMAS GIFT SHOPPING Cheal (1987, 1988) discussed gender-related patterns in gift giving, arguing that women are the primary gift givers because of their greater concern with showing love. He based this on the notion that "caring" is the principal category through which women are differentiated from men and that women's self-identities are based on showing love to others (Chodorow 1978). Since men are not as conditioned to need to perceive themselves as loving or caring, Cheal reasoned, gift giving will not offer them the same reinforcement of a desirable self-image. If this is so, Christmas gift giving in particular should be of much greater importance to individuals who hold strong communal values since Christmas rituals underscore such values. While not incorrect, Cheal's perspective seems incomplete. Gift giving is not simply a communally motivated labor of love. The activity may or may not be gratifyingly expressive but is, in any event, compulsory. An analysis of 16 extended interviews (10 with women, six with men), conducted as part ofthe present research, indicated the obligatory quality of the task, particularly for women. Several described their shopping in terms that indicated that, in their minds, it was real work that had to be carried out efficiently and effectively. One woman said she shops year-round. "Because I'm in retail [works as a sales clerk] I know the cost of things and when they're on sale. You pick things up when you see them . . . if it's a good price. I'd never buy at full retail because I couldn't afford it. . . . [I buy for] parents, brothers, sisters, and then nieces and nephews—all kinds of people and all age groups. . . . I found it sort of a chore this year trying to get everything done" (f, 25).' Female informants often stressed the view that Christmas shopping is a responsibility to be executed. They might enjoy it and find it somewhat rewarding, but they definitely feel compelled to do the "work." Males' perspectives differ markedly. They feel obliged to purchase gifts for spouses and sometimes children, but not for other family or friends. Rather

' Parenthetical notations with quoted material indicate gender and age.

than the feeling of obligation, male informants tend to associate mixed pleasure with Christmas shopping. The following offers an example of how this contrasting view was expressed. "As a generalization, I don't get my jollies out of shopping. But Christmas is a little bit different. The shopping does have something to do with getting in the Christmas spirit. I do enjoy buying the gifts. As long as it isn't too prolonged or too painful" (m,41). In their gift store ethnography. Sherry and McGrath (1989) observed a self-indulgent quality in the shopping of both men and women. In our study, male informants who purchased few Christmas gifts were more likely to describe their reasons for purchasing a gift in terms of materialistic, hedonic values rather than communal ones. They also more frequently reported buying "self-gifts" (Mick and DeMoss 1990) and gifts that they would enjoy as much as or more than the nominal recipients would. Typical of this was one 28year-old who delighted in finding a store that "wasn't a run ofthe mill kind of store"; he felt "like a kid in a candy shop" and did all of his gift shopping for his wife there. His wife felt the purchases he made there were really for him rather than for her. Furthermore, a 30-year-old said he loved to shop extensively for toys for his preschoolers "because ofthe little kid in me." These individuals appear to derive their gratification mostly from immediate hedonic pleasure. Their recurrent references to childhood in association with Christmas shopping suggest that they view it implicitly as an indulgent, playful, and relatively trivial pursuit. Recognizing that Christmas shopping maintains and reinforces social bonds offers some insight into the contrast between attitudes held by men and women about it. Caplow (1982) notes that the set of gifts given reflects and reinforces familial and social relationships. "Kin keeping," the work of maintaining social ties, particularly within kin networks, has come to be stereotypically regarded as "women's work" in contemporary Western society (Rosenthal 1985). Because the family members most venerated at Christmastime are children (Barnett 1954; Belk 1987; Caplow 1982), there is a tendency for the work involved to be stereotyped as women's work. Women are traditionally considered to have primary responsibility for all duties related to children (Bernard 1981; Chodorow 1978). This stereotyping is likely to support gender differences in Christmas shopping because activities classified as women's work tend to be seen as compulsory for women and undesirable for men. Work allocated primarily to women is typically lower in status than work allocated primarily to men and tends to be regarded as less desirable than men's work; moreover, men who perform women's work may be regarded as weak and effeminate (Bernard 1981). If men view an activity as something other than "real work," they are particularly likely to place a low value on it (Campbell 1987) and unlikely to invest effort in it. Some evidence suggests

335

GENDER ROLES AND CHRISTMAS GIFT SHOPPING

men do not view women's work as real work (Valadez and Clignet 1984). We expect women to be much more involved than men in Christmas shopping. However, variation among women and among men is likely. To the extent that gift shopping is regarded as women's work, women and men may differ in their involvement, depending on their attitudes toward traditional gender roles. To the extent that gift shopping is regarded as a labor of love, individuals (male or female) who perceive themselves as more caring may be more involved. Concepts that illuminate the effect of gender on Christmas shopping and the individual differences in gift-shopping behavior are discussed next.

and Schaninger (1987) reviewed evidence that suggests that individuals' and couples' gender-role attitudes are related to patterns of decision making, finance handling, task allocation, consumption, and marital satisfaction outcomes. Advertising research indicates that gender-role attitudes also affect responses to stereotyped ads (see, e.g., Coughlin and O'Connor 1985; Heslop, Newman, and Gauthier 1989). If Christmas shopping is regarded as women's work, it seems likely that women with traditional genderrole attitudes will be more involved and women with more egalitarian attitudes may be less involved. Among men, those with more egalitarian attitudes are likely to be more involved.

Sex

Gender Identity

Individuals' behaviors are influenced by the socialization they receive based on their sex. Although people may not identify with gender-typed traits or may reject traditional role definitions, they still learn behaviors expected of them as men or women (Spence and Helmreich 1978). For instance, young girls may observe or assist in their mother's gift shopping. In so doing, they may learn to view gift shopping as work women do and thus learn relevant skills. Conversely, boys may observe that gift shopping in particular and household shopping in general is not work men tend to do and are thus less likely to learn such skills. Buss and Schaninger (1987) used the term "gender defined behaviors" to refer to skills and activities culturally allocated only to persons of one sex. Davis and Rigaux (1974) documented numerous products associated with certain tasks typically done largely by either men or women. The pervasiveness of gender-defined consumer behaviors suggests that, while biological sex per se does not predetermine behavior, it profoundly affects the socialization experiences a person is likely to have and thus the consumer activities that an individual will be involved in and learn.

Considerable variation has been detected between and among women and men in the degree to which they identify with communally oriented feminine traits and agentially oriented masculine traits. A person who identifies primarily with feminine traits is "feminine," with masculine traits is "masculine," and with both sets of traits is "androgynous" (Bem 1974; Spence and Helmreich 1978). To the extent that Christmas shopping is a labor of love, it seems likely that those who strongly identify with communally oriented feminine traits may be more involved in the activity, regardless ofthe degree to which they identify with masculine traits. Lower levels of involvement are expected among those who do not identify with feminine traits. However, the strength of this relationship is likely to be modest. Trait differences alone have rarely been found to account for a major portion ofthe variance in behaviors (Kassarjian 1971; Pervin 1985). Previous consumer research examining gender identity shows little evidence that it has an effect on consumer behaviors (see, e.g.. Gentry and Doering 1977; Kahle and Homer 1985; Schmitt, LeClerc, and Dube-Rioux 1988). The differences between sex, gender identity, and gender-role attitudes must be emphasized. Biological sex influences the socialization a person is likely to receive and experience. Gender identity reflects a person's degree of identification with "feminine" traits such as compassion and "masculine" traits such as aggression. Gender-role attitudes reflect the extent to which a person agrees or disagrees with norms regarding appropriate behaviors for men and women (Spence and Helmreich 1978). Given the meanings attached to Christmas gift shopping, and the individual differences that seem likely to affect gift-shopping behaviors, a field study was undertaken. In the study, we explored the effect of culturally held meanings toward gift-purchasing patterns and whether consumers' gift-shopping behaviors vary according to their sex, gender-role attitudes, and gender identities. A range of other demo-

Gender-Role Attitudes Gender-role attitude refers to an individual's level of agreement with traditional views regarding the roles and behaviors stereotypically allocated to each sex. People with more traditional attitudes tend to conform to norms that prescribe involvement in gender-traditional activities and prohibit involvement in less traditional activities (Scanzoni and Szinovacz 1980). Conversely, people with more egalitarian views are more likely to be involved in nontraditional activities for persons of their sex. Note that similar gender-role attitudes, be they traditional or more egalitarian, will have different behavioral implications for women and men. The effect of gender-role attitudes on certain household and individual behaviors has been examined. Buss

JOURNAL OF CONSUMER RESEARCH

336

graphic variables were included as moderating variables.

RESEARCH HYPOTHESES We investigated the effect of several variables on a number of aspects of Christmas-shopping behavior. These included number of gift recipients, month when Christmas shopping begins, amount of time spent shopping per recipient, amount of money spent per recipient, and giver-reported success of purchases. Recent work suggests that multiple behavioral measures are necessary for the detection of trait-behavior relationships (Lastovicka and Joachimsthaler 1988). We therefore explored several variables reflecting Christmas-shopping behavior to assess effects of the trait considered. Research hypotheses for each of the gender-related variables are as follows. HI: Women will be more involved in Christmas shopping than will men. H2: In particular, women with more traditional gender-role attitudes will be more involved in gift giving than will women with egalitarian attitudes. H3: Conversely, men with egalitarian gender-role attitudes will be more involved in gift giving than will men with more traditional attitudes. H4: Furthermore, individuals with more feminine gender identities will be more involved in gift giving than will those with less feminine gender identities. Specifically, those who are more involved will a: give Christmas gifts to more recipients; b: start Christmas gift shopping earlier in the calendar year; c: spend more time, on average, shopping for each recipient; d: spend less per gift recipient; and e: report greater success in gift selection than will those who are less involved.

Sex Hypotheses If women are socialized to be more involved in gift giving, observing it in more detail and learning to consider it necessary, they are apt to give more gifts. Women may also give more because they have more friends (Chodorow 1978) and may be expected to recognize close friendships through gifts. Women are likely to start shopping earlier if they view Christmas shopping as an important task requiring extended effort. For instance, an early start ensures that products are not out of stock. One retail sales clerk stressed this. "Christmas shopping is a little more hec-

tic [than regular shopping]. You really do have to shop earlier than with anything else because items are picked over quickly" (f, 25). Starting earlier also allows time for an extended search, if necessary. This was emphasized by the same informant when she described how she shopped for her husband. "It takes me a while to figure out what he'd really like. Then I watch for it, to see what market trends are, what goes on sale." If gift shopping is more salient for women, they are likely to pick up gifts while doing other shopping during the year, as did the 70-year-old female informant who stated, "I shop Christmas all year round." In contrast, her son labeled himself a "last-minute shopper." It is logical to expect that the more gifts an individual purchases, the longer it will take to do the gift shopping. Starting gift shopping early, then, should be influenced by the number of recipients for whom gifts need to be purchased. However, we hypothesized that, even if the number of gift recipients is controlled for, women will be found to start their Christmas shopping earlier than men. Women, who may learn to regard gift shopping as an important task, are likely to spend more time shopping per recipient because of a greater effort to seek or deliberate over alternatives. Men, who may have learned to invest relatively little effort, are likely to spend less time per gift recipient. "There is a sex difference in Christmas shopping. Women are trained to shop—to shop around a lot. [Men find the] browsing aspect of buying boring" (m, 42). It is also likely that women purchase gifts for friends, neighbors, and more distant relatives to a greater extent than men do. Since people tend to buy less expensive gifts for less closely related individuals (Belk 1979; Caplow 1982, 1984), women's average expenditures per gift should be lower. Women may also spend less because they have greater familiarity with the tastes, wants, and needs of recipients. They may learn to be more attentive to a recipient's expressed wishes or evident wants and needs, which enables them to identify a greater number of possible appropriate gifts and to buy less expensive items if they so choose. Further, if women have learned more gift-shopping skills, they may also be able to get better value for their money. Men, if they do not know their recipients as well, may tend to buy more expensive gifts so as to demonstrate love via money rather than via familiarity. This is typified by the 42-year-old male informant who enjoyed impulsively buying expensive gifts that would "blow someone's mind" and felt it less important that gifts be "just the right thing" for someone. There may also be a greater social acceptance and expectation that men will expend financial resources to establish and maintain social relationships (Harrison and Saeed 1977; Hirschman 1987). Caplow's( 1982, 1984) findings that men's gifts to their spouses were more expensive than women's would support this proposition.

337

GENDER ROLES AND CHRISTMAS GIFT SHOPPING

Women's social learning is expected to increase their ability to select gifts that recipients like. Women may acquire social skills conducive to greater awareness of the preferences of those around them. Since "success" here means that recipients do not (to the giver's knowledge) return or exchange the gifts given to them, two caveats must be noted. First, some gifts may be returned without the knowledge ofthe giver, so only giver-reported success can be assessed. Second, success is only partially determined by the giver since recipients may vary in their propensity to return gifts or to communicate the inappropriateness of them.

Gender-Role-Attitudes Hypotheses Relative to women with more egalitarian attitudes, women with more traditional gender-role attitudes may do more to fulfill their role-related expectations. If women are expected to give gifts to an extended social network of friends and relatives, they should give to a greater number of recipients. Given the expectation that women will be active in caring work, women with traditional role attitudes may also start shopping earlier. Furthermore, they may put more time and effort into the task than would those with more egalitarian attitudes. Women with traditional attitudes will be more inclined to shop in a conventional manner; that is, they may invest more time than money. Further, since women with traditional attitudes are less likely to have careers in high-paying fields, their incomes may be more constrained and therefore their gift expenditures limited. They may also report more success in gift selection since greater involvement in other women's work (like caring for children) may increase their familiarity with recipients. More egalitarian women may have fewer interactions and thus less insight into what gifts will please recipients. Relative to men with traditional attitudes, men with more egalitarian role attitudes are likely to view behavior typically considered appropriate for women as acceptable for themselves. Thus, such men may give gifts to more recipients, believing they should participate in such women's work as kin keeping. Traditional men will tend to hold the view expressed by a 75-yearold male informant who observed, "Christmas shopping—that's my wife's department." If the more egalitarian men regard gift shopping both as work they should do and as work requiring extended effort, they may start their Christmas shopping earlier. For the same reasons, egalitarian men may also take more time choosing gifts for recipients. A 30-year-old informant, who was quite involved in caring for his children and in other work usually done by women, said that he started early and shopped extensively for his children's gifts. He said there is a "high return on getting the right gift" for them and a "high risk of disappointment" otherwise.

More egalitarian men may also spend less money on average per recipient. They are likely to purchase more token gifts for those more distantly related and may invest time and energy rather than money in gifts they select. Finally, egalitarian men should be more successful with gift selection. Men with traditional gender-role attitudes are less likely to know recipients' tastes if they perform only men's work (such as working outside the home) and have fewer opportunities for interaction with recipients. Men with egalitarian attitudes may take on a number of aspects of women's roles that give them more opportunities for interactions with recipients and thus a greater likelihood of knowing what gifts will please recipients.

Gender-Idetitity Hypotheses Individuals who identify more strongly with communally oriented traits may shop for gifts for a greater number of recipients since they care more about reinforcing social bonds. Given their concern for pleasing others, they are also likely to start shopping earlier to ensure a better selection of appropriate gifts for recipients. They may also spend more time shopping per recipient in order to choose more appropriate gifts. Such individuals may spend less money per gift since they are purchasing gifts for less closely related individuals. Finally, they may have higher self-reported success in gift selections since they are more communally oriented and attuned to recipients' tastes. No hypotheses concerning the effect of demographic variables were developed since they were included simply as moderating variables.

RESEARCH DESIGN Method A field study was designed to test the research hypotheses. The data were collected through the administration of a structured questionnaire during in-home personal interviews conducted during evenings and weekends of the four weeks immediately following Christmas 1987. The population studied included men and women, ages 18 and above, living in the metropolitan area ofthe city where the study was conducted. Surveys were completed by a field force who were trained and supervised by E. Fischer. Interviewers were briefed on the purpose of the survey, the rights of respondents, the procedure for posing each question, and the time and manner in which to contact households, select respondents, and perform callbacks. Supervision included postinterview checks with a randomly selected subset of respondents. A multistage cluster sampling technique was used. First, the 131 census areas in the metropolitan area were listed, and 40 were drawn randomly. Second, an intersection in each chosen area was selected randomly with an overlay grid. Households nearest to these in-

338

tersections were slated as starting points from which every second household was selected, proceeding clockwise. The proportion of sample households from each area matched the proportion of population households in that area. We selected respondents randomly from adults residing in the 929 designated households, using Trodahl and Carter's (1964) procedure. This procedure ensured sample eligibility of all adults in the household, regardless of which one greeted the interviewer. By making up to two return visits to households, 70 percent of the 929 designated respondents were reached. Respondents received a $1 lottery ticket for fully or partially completing the survey, which took 30-45 minutes. Forty-six percent (299 people) of those reached completed the survey.

Measures of Indepetidetit Variables The interviewer recorded the sex of each respondent. Gender-role attitudes were measured with items developed by Scanzoni (1975; Scanzoni and Szinovacz 1980). These items form subscales assessing attitudes toward traditional wives' roles, traditional husbands' roles, institutionalized equality, alterations in husbands' roles, and self-actualization for women. Responses to each item were indicated on five-point Likert-type scales. On the basis ofthe sum of their ratings, respondents received scores reflecting how egalitarian their overall gender-role attitudes were. The reliability for scales from Scanzoni's work has been assessed previously (see, e.g.. Quails 1987; Schaninger and Buss 1985; Schaninger, Buss, and Grover 1982). The 17item scale used in this study was assessed by Quails (1987) as having a Cronbach's alpha of .88. A satisfactory alpha of .83 was obtained here. Gender identity was measured with the Bem Sex Role Inventory (BSRI; Bem 1974). This measure contains 60 items describing abstract traits with which the respondent indicates degree of identification on a fivepoint Likert-type scale, which includes three subscales of 20 items each. The neutral (N) scale contains socially desirable traits not associated with either sex, the masculinity (M) scale contains socially desirable traits more characteristic of males than females, and the femininity (F) scale contains socially desirable traits more characteristic of females than males. Since the theory underlying this study suggests that identification with feminine traits predicts gift-shopping involvement, only the F-scale score for each respondent was used in the analysis. For the M, F, and N scales, Bem (1974) reported Cronbach's alphas of .86, .80, and .75 and of .86, .82, and .70 in separate samples. In this study, the alphas for the M, F, and N scales were .84, .76, and .60. Marital status and number of children were also nieasured since they reflect the size ofthe individual's kin network, which has been shown to have an im-

JOURNAL OF CONSUMER RESEARCH

portant moderating influence on the number of gift recipients (Caplow 1982). Other demographic variables included as moderating variables were type of employment, income, and education. Employment was categorized as managerial/professional, clerical/manual, other, or no job. We assessed income by asking people to indicate which of eight categories, ranging from less than $10,000 to $80,000 or more, corresponded to their personal income. We measured education by asking people to indicate which of six levels of education they had completed. Additional demographic information was collected to facilitate comparisons between the sample and the population. This included age, ages of children, household income, religion, and ethnicity.

Measures of Christtnas-Shoppitig Behavior We measured number of recipients by asking, "Who did you give Christmas gifts to this year?" and summing the number of recipients mentioned. Note that number of gift recipients is both a dependent variable in its own right and a moderating variable for the other four dependent variables because number of recipients is likely to have some effect on them. We measured earliness of starting relative to Christmas by asking, "In what month did you start shopping for Christmas gifts in stores?" Although a search for gifts may start before in-store shopping, it seemed more likely that respondents would be able to recall the month in which they physically began shopping for gifts than the month during which they first thought about gifts. For this variable, a higher score reflects starting in an earlier month. We measured time spent shopping by asking respondents to indicate, for each month during which some shopping was done, how many hours were spent gift shopping. Ideally, the time spent on specific recipients would have been asked. However, pretesting revealed that respondents could not break down their time in this way since shopping was often done for many recipients at the same time. We created the variable "hours spent shopping per recipient" by summing the hours spent shopping each month and dividing by the number of recipients. We measured total amount spent by asking, "In total, how much did you spend on gifts this year?" We divided the total by the number of recipients to determine "dollars spent per recipient." Respondents who had trouble with this question were asked how much was spent on each recipient mentioned. To measure reported success, we asked respondents, "To your knowledge, did any ofthe gifts you gave have to be returned or exchanged?" If so, they were asked which recipients had returned gifts. We totaled the number of returns or exchanges and divided by the number of gift recipients to obtain the variable "reported success." A lower score represents greater sue-

339

GENDER ROLES AND CHRISTMAS GIFT SHOPPING TABLE 1 CORRELATIONS AMONG MEASURES OF CHRISTMAS SHOPPING BEHAVIOR Number of gift recipients Dependent variables Month shopping started: Males Females Hours spent shopping per recipient: Males Females Dollars spent per recipient: Males Females Reported success ratio: Males Females

Males

Females

Month shopping started Males

Females

Hours spent shopping per recipient Males

Females

Dollars spent per recipient Males

Females

.09 .32***

.29"*

.09 .19**

.25"* -.22"

-.44***

.08

-.22** -.13*

.10 .04

.07 -.13*

-.43"*

.04

.18* -.14*

.17*

NOTE.—"Reported success ratio" is an inverse measure.

*p