Object Relations Theory: Male and Female Differences in Visual Information Processing

Keren A. Johnson, University of Texas at Austin
Mary R. Zimmer, University of Georgia
Linda L. Golden, University of Texas at Austin
ABSTRACT - This exploratory study of male and female differences in visual information processing and perception investigates two hypotheses suggested by object relations theory. Results of a content analysis of open-end recall data for two advertisements with different themes, one representing relationship webs (high connected) and the other representing hierarchical object presentation (low connected) are described. While females were more likely to report seeing relationships stated in gender role terms than males for the high connected advertisement, there were no sex differences in the frequency of recall of isolated, spatial elements for the low connected advertisement.
[ to cite ]:
Keren A. Johnson, Mary R. Zimmer, and Linda L. Golden (1987) ,"Object Relations Theory: Male and Female Differences in Visual Information Processing", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 83-87.

Advances in Consumer Research Volume 14, 1987       Pages 83-87

OBJECT RELATIONS THEORY: MALE AND FEMALE DIFFERENCES IN VISUAL INFORMATION PROCESSING

Keren A. Johnson, University of Texas at Austin

Mary R. Zimmer, University of Georgia

Linda L. Golden, University of Texas at Austin

[Keren A. Johnson is a Lecturer in the Department of Marketing Administration, University of Texas at Austin. Mary R. Zimmer is Assistant Professor of Marketing at the University of Georgia. Linda L. Golden is Professor of Marketing at the University of Texas at Austin.]

ABSTRACT -

This exploratory study of male and female differences in visual information processing and perception investigates two hypotheses suggested by object relations theory. Results of a content analysis of open-end recall data for two advertisements with different themes, one representing relationship webs (high connected) and the other representing hierarchical object presentation (low connected) are described. While females were more likely to report seeing relationships stated in gender role terms than males for the high connected advertisement, there were no sex differences in the frequency of recall of isolated, spatial elements for the low connected advertisement.

INTRODUCTION

Consumer behavior researchers have addressed the topics of sex roles (cf., Debevec and Iyer 1986; Della Bitta 1984), gender (cf., Gentry and Haley 1984; Roberts 1984), and visual information processing (cf., Jensen and Rottmyer 1986; Hansen 1984). While these three areas conceptually complement each other, consumer behavior researchers have not yet attempted to empirically synthesize them into one inherent frame.

This paper suggests a conceptual framework for synthesizing sex, gender and visual information processing utilizing object relations theory (Chodorow 1978; Gilligan 1982; Eichenbaum and Orbach 1983). This theory suggests that images acquired during the first two years of life continue to function at a precognitive level and may have an impact on the way males and females process visual information. Hypotheses suggested by object relations theory, predicting differing cognitive processing outcomes for males and females, are tested via a content analysis of the recall of visual advertising information.

Before describing object relations theory, it is important to clarify the meanings and use of the terms "sex" and "gender" in this paper. These terms are often used interchangeably in the consumer behavior literature (e.g., Debevec and Iyer 1986); however, in the sociological literature there is a clear distinction between the denotations of "sex" and "gender". Sex refers to "the dichotomous distinction between males and females based on physiological characteristics" while gender refers to "the psychological definitions of the dimensions 'masculine' and "feminine "(Oakley 1972).

Object Relations Theory

Freud's disciple, Melanie Klein, developed the basis for object relations theory. In Melanie Klein's expression of object relations theory, she focuses on the images that an individual experiences in their early development as one explanation for adult cognitive processing. According to Klein and her followers, the feeling of self (including gender perceptions) arises from a tension between the individual's goals in relation to other objects in the environment

It is important to note that in this theory, "objects" are defined differently than conventional usage of the term would suggest. "Objects" are entities (the primary objects are people) that the infant becomes involved with during the developmental process. Thus, objects (such as people or any phenomenon outside of the infant's body) can impede the infant's goals, resulting in tension. Likewise, objects in the environment can also facilitate the individual infant's goals (relieving tension). In this manner, object relations theory posits the developmental process as interactive. It further stipulates that human action must always be understood in the context of a particular situation and the sociocultural environment.

In more recent years, theorists have extended object relations theory to look at the differences between sexes (Eichenbaum and Ohrbach 1983). Chodorow (1978), Gilligan (1982) and Rubin (1983) each postulate a model of psychological development in which the connection between individual experience, personality, and the material world reflects the impact of present child-rearing and gender relations. Chodorow (1978), Gilligan (1982) and Rubin (1983) also draw on object relations theory to describe the emergence of different modes of thinking about relationships which may be empirically associated with male or female responses to situations requiring judgement and action.

Chodorow's (1978) explanation of male and female cognitive differences focuses on women being the primary caretakers of infants in the Western culture; thus, the importance of the female during infancy results in children of both sexes psychologically comparing themselves to a female. The meaning children ascribe to this process and its subsequent impact on the continuing development of cognitive strategies is generally different for girls than for boys (i.e., girls might see themselves as similar to the female and boys might see themselves as different from the female). Specifically, the ongoing mother-daughter attachment reproduces the social arrangement whereby women continue to be the primary figures in the lives of infants of both sexes.

Rubin, whose orientation is primarily clinical, believes that themes of separation (for boys) and unity (for girls) are incorporated into the personality structure at a pre-oedipal age and that these differences affect crucial parts of the lives of both men and women (Rubin 1983). Gilligan (1982), however, draws upon the "texts of men's and women's fantasies" to specifically elaborate these themes into images of a web of connected relationships (empirically associated primarily with women) or a hierarchy of object separation (empirically associated primarily with men).

Gilligan (1982) examined the different cognitive schemes used by males and females as they pertain to moral judgement and subsequent action in life or death situations. Briefly, Gilligan (1982) studied males and females of different ages to examine their cognitive processes. She found that younger females tended to process information in the context of social relationships or what Gilligan refers to as the "web". For females, connectedness to other people was the most important criteria for decision-making. Males, however, tented to look at situations as distinct from social relationships and as isolated, abstract events to be approached "logically".

Sex Differences in Cognitive Functioning

The most basic problem in any research on sex differences is that the researchers begin with the underlying assumption that sex differences exist and that they are important (Basow 1980). Although most of the evidence suggests that variation within each sex is greater than differences between them (Basow 1980; Burstein, Bank, and Jarvik 1980), testing hypotheses of sex differences is grounded in an epistemology of difference, not similarity. first, researchers cannot accept a null hypothesis (only fail to support it) and, second, it is possible that studies reporting no statistical differences would be more difficult to publish. Therefore, the literature on sex differences is likely to be more representative of studies resulting in statistical differences between the sexes than it is of the (potentially unpublished) research finding no sex differences.

Aside from problems with methodology, findings in this area are further confounded by whatever personal or political aspects of sex differences a scientist-happens to bake for granted when testing a theory or interpreting results. For example, Burstein, Bank, and Jarvik (1980) and Basow (1980) each review a number of studies that found boys and men, on the average, tend to score higher than girls and women in tests of spatial abilities and that feMales tend to score higher than males on tests of verbal abilities.

Field dependence-independence is one of the few psychological variables for which clear-cut evidence of sex differences exists in the literature. Studies by Witkin (1962) and his followers consistently show women to be more field dependent than men and men to be more field independent than women. Once considered a valid measure of global analytic abilities (Witkin 1962), the act of straightening a rod in a tilted frame while seated in a darkened room has now become just one more spatial task. Burstein, Bank and Jarvik (1980) summarize the curious history of the field dependence-independence variable:

One likely reason for the preponderance of studies of field dependence-independence may well be the consistent finding of male superiority on this measure. Most of the investigators have been men. Field independence, and not field dependence, has been associated with positive value in our culture: the field independent cognitive style has been characterized as "differentiated, self-reliant, analytic" as opposed to a field dependent style which is considered "passive and conforming ... "

These confusing and contradictory data on the existence and determinants of sex differences in cognitive abilities (Basow 1980; Burstein, Bank, and Jarvik 1980) suggest to some researchers that consumer behavior specialists ought to terminate research into psychological gender differences (Roberts 1984). Perhaps they are right; however, consumer behaviorists who continue to emphasize the disadvantages of psychological factors for examining gender differences may neglect the advantages of sociological ones.

Relevance to Consumer Behavior

Tucker (1977) suggests that during a time of rapid social change people of both sexes find it useful "to behave in ways and to voice sentiments that are discrepant with their most centrally held attitudes and values" about the personal meaning of their gender identity. Tucker (1977) speculates that "the notion of sexual equality begun perhaps...in the eighteenth century may fully reach the hearts of men of both sexes in developed areas within two generations, about the year 2025". In the meantime, he predicts, "...marketers will increasingly miss the center of their markets because they will not understand them. And the changes in the relations of the sexes will be the primary cause" (Tucker 1977).

Tucker's (1977) comments about the impact of slow cultural change on the subjective condition of "men of both sexes" does not suggest that human action cannot create social and institutional change; or that research on sex differences, justifiably found socially and scientifically wanting, demonstrates that gender expectations we manufacture are determined at the cellular level and therefore immutable. Instead, we interpret Tucker's argument as an invitation for consumer behaviorists to continue research into sex differences and we propose that the research should be grounded in a body of theory that accommodates a phenomenological perspective.

Speaking for "...a disaffected and growing minority of researchers interested in moving beyond the positivist cognitive psychological orientation in consumer behavior studies," Sherry (1986, p. 574) calls for a cultural perspective in consumer behavior research. Such a perspective is reflected in the application of object relations theory to investigate the visual information processing of males and females.

Purpose and Hypotheses

The primary purpose of this study is to investigate the differences in visual information processing outcomes between males and females for two types of advertisements: one for which the dominant theme centers around relationships among people (high connectedness) and one for which the dominant theme utilizes visual presentation of isolated objects (low connectedness). As suggested by object relations theory (Gilligan 1982), women are expected to process information in terms of webs or networks (i.e., connectedness among objects, such as relationships among people), and when the visual objects are people of varying ages and sex, this processing focus would be expected to manifest itself in terms of a heightened perception of Render role relationships.

Thus, the following hypothesis is suggested by object relations theory for males and females viewing an advertisement high in "connectedness" (webs of interpersonal relationships).

H1: When viewing a visual stimulus that includes individuals of mixed age and sex, females will be significantly more likely to recall seeing relationships than will males.

For an advertisement featuring human connectedness, it is expected that the relationships females recall will be described in gender role terms.

The second hypothesis is specific to an advertisement featuring a theme low in "connectedness" and also incorporates predictions from the field independence-dependence literature. As previously discussed, field independence-dependence theory suggests that males will focus more upon the individual elements (or figure), as distinct from the ground of a stimulus. In addition, it is expected that males will be more likely to recall isolated (hierarchical) elements of the advertisement than will females because of their increased propensity to perceive abstract and individual objects, as suggested by object relations theory (which offers an explanation for the sex differences in field independence-dependence studies).

H2: When viewing a visual stimulus that includes a predominant theme of isolated elements (low in connectedness), males will describe their recall in quantitative terms and will report spatial relationships with isolated elements more frequently than will females.

METHOD

The methodology consisted of several phases: product and advertisement selection (stimuli), design of the instrument and administration, the qualitative content analysis and the quantitative analysis of the results. This section describes the methodology through the qualitative content analysis.

Product and Advertisement Selection

Soft drinks were selected as the experimental product category because three important criteria were met: 1) high familiarity among a student sample, 2) soft drinks could be considered a gender neutral product stimulus (Gentry and Haley 1984), 3) competing brands were widely advertised through varied types of appeals (which was helpful for the selection of specific advertisements). In order to select the specific experimental manipulations, a panel of four judges reviewed tapes of commercially developed 30-second advertisements for familiar brands that had not been aired in the study area. Each advertisement was evaluated on the basis of the type of appeal used and there was clear agreement among judges (two males and two females) regarding the characteristics of the two advertisements selected for this study.

The judges evaluated the advertisements as representing feeling (i.e., emotional) or thinking (i.e., rational or utilitarian) appeals; however, these appeals could be facilitated through a variety of themes. The focus of this study is on the specific theme through which the appeals are represented.

A Royal Crown Cola advertisement was selected as representing the social relationship, high connected theme (in the context of a "happy" emotional appeal) and a Seven-Up advertisement was selected for its use of a low connected theme (in the context of a rational appeal). A description of each of the advertisements will further clarify their characteristic themes.

The Royal Crown Cola advertisement showed a man, a woman and a male child in the country (people of mixed ages and sex). The audio content consisted of a woman's voice singing about having moved to a farm in the country and enjoying her life there. She was shown in various activities with the man and the child (e.g., working on the farm, chewing straw and joking with each other), as well as sitting alone for a few moments. The advertisement concluded with the RC Cola jingle ("Me and my RC"). No explicit oral mention was made of any gender role relation terms.

The Seven-Up advertisement showed three bottles in three bags, three glasses, and a man's hand putting ice in the glasses, pouring from one of the bags and finally removing the bottle from the bag to reveal the brand of soft drink as Seven-Up. At the end, the hand put a sticker that said "The Uncola" on the bottle. The audio portion of the advertisement consisted primarily of the sound of ice hitting the glasses and the sound of a carbonated drink being poured over ice. A man's voice referred to the product as "wet, wild, and refreshing" and at the end of the ad suggested reaching for a Seven-Up, "the Uncola".

The Seven-Up advertisement was selected as being low connected because it focused on what the product does in isolation of what people might do when they use the product (including activities with other people). Physical product attributes were important in this advertisement (the transmittal of image and imagery not withstanding) and there was no suggestion of social relationships (connectedness) among or between people.

Design of the Instrument and Administration

The advertisements selected for this study were transferred to two separate videotapes to be shown in the context of four other television advertisements. The soft drink commercial was the second advertisement on each of the tapes. The other three commercials on each tape were evaluated by the four judges to represent the same type of appeal (rational or emotional) and the order of the other three products (airlines, hair care and petroleum company) was held constant between the two tapes.

A total of eighty-one undergraduate business students at a large Southwestern university participated in this study. One class viewed the tape of emotional appeals (n = 43, high connected advertisement) and two classes (due to varying number of students) viewed the tape of rational appeals (n = 38, low connected advertisement). After watching the advertisements in a group setting, respondents completed the instrument.

The first section of the instrument elicited an affective response to the advertisement by asking, "Overall, to what extent did you like or dislike the advertisement?" Respondents circlet their opinion on a bi-polar horizontal seven-point scale, with seven representing the highest positive extreme.

Respondents were next asked, "Please describe all that you remember seeing in the advertisement," and were instructed to write their response on the back of the page (8 1/2" by 11"). Thus, open-end responses were not limited by space. The amount of time was limited to several minutes, although everyone appeared to have completed their response well within the time allocated. Respondents answered several other questions not related to the topic of this study.

Content Analysis

The open-end visual recall data were content analyzed. The analysis focused on the specific words and phrases reported by each respondent, as well as the quantity of visual information processing. However, since an appropriate content analysis involves correspondence between the analytical approach used and the relevant properties of the context (Krippendorff 1980), the content analysis categories are specific to each stimuli.

Three content categories were initially developed for the analysis of the visual information contained in the high connected advertisement. The content of the responses to the advertisement were counted with respect to: 1) the number of times a gender role relationship term is mentioned (e.g., wife, son, husband, family, mother, father, parents, Mom, Dad, "her kid", and "her man"), 2) the activity recalled (e.g., doing farm chores, walking and splashing in water, playing, and sitting and talking) and 3) the amount of detail (e.g., number of details about the people or their surroundings) recalled as having been seen in the advertisement.

A fourth content category, representing a summary measure of the over-all quantity of information recalled, was developed from the last two categories above and a count of the number of people each respondent mentioned they saw in the advertisement. Inspection of the array of responses across all respondents and all three content analysis categories showed clearly that the quantity of recall was distributed bi-modally. There was a "natural" break in the data such that a respondent tended to recall either "a lot of information" or "a little information". Thus, the amount of information recall reflected by each person's response was categorized as either "high recall" or "low recall". Operationally, and as suggested by the data, high recall responses included mention of each of the three people in the advertisement (irrespective of the way they were described in terms of gender role or not), at least two activities, and more than one contextual detail.

The visual information contained in the low connected advertisement was described by: 1) mention of spatial elements (direction of movement, camera angle, location of label, shape of glass, shape of sticker), 2) mention of quantities (amount of liquid, number of bottles, number of bags, number of glasses, number of cans), and 3) mention of isolated elements (man's hand, neck of bottle, top of bottle). Mentions of details in isolation, (i.e., bottle necks, man's hand) are used as an operationalization for a general category of separation of visual elements recalled in the advertisement.

The amount of information recalled in the low connected advertisement was based on whether or not a respondent mentioned at least one spatial element, one quantitative aspect, and one isolated element in their response. If all three were mentioned, the response was classified as "high recall" and if not it was classified as "low detail". Again, the cue-point for the dichotomization of high and low detail was suggested by the distribution of the data.

STATISTICAL ANALYSES AND RESULTS

The data from the content analysis were submitted to statistical analysis for hypotheses investigation. The analyses and results for each hypothesis are described in this section.

Hypothesis 1

Hypothesis 1 is specific to the high connected advertisement and predicts that when viewing a visual stimulus including individuals of mixed age and sex, females will be more likely to recall having seen relationships (expressed Ln gender role terms) than will males. To test this hypothesis, using respondents as the unit of analysis, the number of females mentioning (and not mentioning) any gender role relationship (irrespective of how many) was compared to the number of males mentioning (and not mentioning) any role relationships. These data are presented below.

                                        Females     Males    Total

Mentioned any role              23            10           33

No mention of role                2              8            10

Total                                   25            18            43

Results of a chi square test of independence indicated support for Hypothesis 1 (X2= 7.77, df = 1, a = .01, with a fourfold point correlation of .425). The fourfold point correlation of .425 suggests that there is a strong association among these variables. Thus, these data suggest that females were more likely than males to report seeing at least one relationship which they expressed as a gender role and males were more likely to omit any mention of role relationships from their descriptions of what they saw in the high connected advertisement.

The analysis just discussed focused on whether or not a reSpondent mentioned a gender role relationship when stating what they saw in the advertisement (i.e., for each respondent, a role relationship was either mentioned or not mentioned and the number of role relationships mentioned by a respondent was disregarded). The data were also submitted to chi square analysis for the total number of role relationships mentioned across all subjects (with expecteds proportional to the sample sizes), addressing the question, 'Is the total number of mentions distributed proportionately between males and females?"

The total number of role mentions by sex for all subjects, including those with no role mentions, is reported below.

                                 Females      Males       Total

                                  (n=25)      (n=18)     (n=43)

Total role mentions        50             16             66

A chi square test was statistically significant (X2 = 8.55, df = 1, a= .01), supporting the hypothesis that females used significantly more gender role terms to describe what they saw in the relational advertisement than did males. (As another perspective on these data, the mean number of gender role terms mentioned by females was more than twice that of the number mentioned by males across the total sample: 2.00 and .89, respectively.)

As a final perspective on the first hypothesis, respondents who made no mention of gender role relationships were eliminated from the next analysis. Thus, the frequency of mentions remains the same as above, but the number of respondents included is decreased (to 23 for females and 10 for males), affecting the expected values (which are treated as proportional to the sample size).

The chi square test of independence was statistically significant (X2 = 5.16, df  = 1, a = .05) indicating that females who mentioned gender role relationships did so more frequently than did males. Again, the means are also supportive of this conclusion (the mean number of gender role mentions is 2.174 for females and 1.6 for males).

Hypothesis 1 was supported for all of the three analyses described above. The results of these data suggest that not only were females more likely to report having seen (any) gender role relationships than were males, but females who saw relationships also reported seeing a larger number of gender role relationships than did males who saw gender role relationships. Thus, females were more likely to mention gender roles and were more likely to mention gender roles more frequently.

It is interesting to note that males tended to qualify their descriptions of gender role relations, whereas women did not. For example, males would report having seen "a male figure, most likely her husband," or "a woman who appears to be with her husband and son" and women would simply state "mother," "son," "husband," etc. for the visual objects recalled.

An old husband's tale suggests that females may find more in their environment to talk about than do men (i.e., talk more). Related to this idea, as previously discussed, women tend to score higher on verbal tasks than do males. This skill might be associated with an increased propensity for verbal expression (no cause and effect suggested or established).

Therefore, the data were analyzed to address potential alternative explanations for women reporting higher recall of gender role relationships than men. That is, the results may have occurred not because women are more sensitive to the presence of gender role relationships in their environment, but simply as an artifact of women, per se, seeing and describing more information (of any type) than men.

As previously discussed, the content analysis indicated the sample consisted of "high recall" and "low recall" persons. High versus low recall response frequencies are presented below for males and females.

                    Females     Males    Total

High recall        12             7           19

Low recall        13           11           24

Total                  25          18           43

A chi square test of independence revealed no significant relationship between amount of information recalled and sex (x2 = .34,df=1). Thus, for these data, sex was not associated with how much a person recalled seeing in the advertisement but, rather, the gender role relationship content of what was recalled about the visual stimuli.

Another possible explanation for the sex differences ln recall of gender role relationships for the high connected advertisement is that females liked this type of theme and appeal better than the males and, hence, recalled the relationship aspects of the advertisement more than did men. To investigate this potential explanation, a t-test was computed to compare male and female responses on the affect rating scale (mean male affect = 4.72 and mean female affect = 5.60). While females tended to "like" the high connected advertisement better, this result was not statistically significantly (t= .24,df= 41). Thus, neither of the potential alternative explanations tested for the differences in gender role relationship recall between males and females were supported by the data for the high connected advertisement.

Hypothesis 2

Hypothesis 2 is specific to the low connected advertisement and predicts that males will be more likely to describe their recall in terms of isolated elements, spatial and quantitative terms than will females. The frequency distribution of quantitative comments, spatial relationships and isolated elements recalled from the low connected advertisement is shown by sex below.

                                                                                    Females    Males   Total

Mention of quantitative, spatial or isolated elements           20           15         35

No mention of the above elements                                      0             3           3

Total                                                                                20           18         38

A chi square test of independence on the above data was not statistically (x2 =3.63,df= 1). Thus, these data did not show sex to be systematically associated with recall of spatial and quantitative relationships or isolated elements for the low connected advertisement. As for Hypothesis 1, Hypothesis 2 was tested for the total number of mentions of quantitative, spatial and isolated elements by females (53 mentions, n=20) versus males (52 mentions, n=18).

The results were also not significant (X = 1.82, df = 1) when the analysis was conducted with respondents mentioning at least one quantitative, spatial or isolated element (n=20 for females and n=15 for males). Thus, Hypothesis 2 was not supported and males did not report recalling quantitative, spatial or isolated elements in the abstract advertisement to any greater extent than did females.

While the number of elements did not differ between the sexes, the specific isolated elements recalled were dramatically different, potentially suggestive of Freudian psychology. Twelve males and one female mentioned bottles while twenty females generated fifteen responses about glasses with only two males mentioning glasses. These curious results suggest that motivation research is not (quite) dead and that there were differences between males and females in their recall of the low connected advertisement (in terms of exactly what isolated elements were recalled), just not along the directions specifically hypothesized.

A chi square computed to investigate sex and amount of information recalled from the low connected advertisement was not significant (X = .86, df = 1). A t-test between mean affect ratings for males (4.83) and females (4.79) viewing the low connected advertisement was also not significant (t =.39, d.f. = 35).

DISCUSSION AND CONCLUSIONS

Consistent with object relations theory predictions, females were more likely than males to report having seen gender role relationships for advertisements with models of mixed ages and sex. The high connectedness advertisement made no statements about gender roles, but given the potential to infer such relationships, women tended to make gender role relationship inferences, whereas men did not Men and women recalled the same stimulus differently.

Females were more likely to mention recalling relationships, and also tended to describe the relationships in gender role terms. In addition, for the males and females recalling the advertisement in gender role terms, the females recalled more gender role relationships. These results do not appear to be related to sex differences in the global amount of information recalled or affect toward the advertisement. Thus, for this study, sex was associated with what was recalled, rather than the over-all quantity of information recalled.

A second advertisement, high on isolated elements and low on connectedness, was studied to increase external validity and investigate male visual information processing as suggested by object relations theory. For this low connected advertisement, there were no differences between sexes in perceptions of isolated elements, spatial orientation and number of quantitative statements. Thus, males did not tend to perceive isolated elements any more than did females.

The results of this exploratory study are restricted to the specific methodology used; however, this research suggests that differences in male and female visual information processing do exist and that object relations theory can provide a useful theoretical grounding for investigating sex differences. Visual images may be extraordinarily powerful, not just because of the impression they make at the moment (which has been the primary consumer behavior focus on visual imagery) but because of what they suggest about the past. As Tucker (1977) noted, marketers may miss the center of their markets in the future because they fail to understand the implications of changing relations between the sexes. These changing relations may also result in changes in the differences between men and women in their visual information processing. However, to understand the changes, we must first have an idea of what the differences are now. This paper is a step in that direction.

REFERENCES

Basow, Susan A. (1980), Sex-Role Stereotypes: Traditions and Alternatives, Belmont, CA: Wadsworth.

Burstein, Bonnie, Lew Bank, and Lissy Jarvik (1980), "Sex Differences in Cognitive Functioning: Evidence, Determinants, Implications," Human Development, 23, 289-313.

Chodorow, Nancy (1978) The Reproduction of Mothering Psychoanalysis and the Sociology of Gender, Berkeley: University of California Press.

Debevec, Kathleen and Easwar Iyer (1986), "Sex Roles and Consumer Perceptions of Promotions, Products, and Self: What Do We Know and Where Should We Be Headed?" in Advances in Consumer Research, Vol. 13, ed. Richard J. Lutz, Provo, UT: Association for Consumer Research, 210-214.

Della Bitta, Albert J. (1984), "Sex Roles, Sex, and Race Factors in Advertising and Satisfaction: Some Information and More Questions," in Advances in Consumer Research, Vol. 11, ed. Thomas Kinnear, Ann Arbor: Association for Consumer Research, 100-102.

Eichenbaum, Luise and Susie Orbach (1983), Understanding Women: A Feminist Psychoanalytic Approach, New York: Basic Books.

Gentry, James W. and Debra A. Haley (1984), 'Gender Schema Theory as a Predictor of Ad Recall," in Advances in Consumer Research, Vol. 11, ed. Thomas Kinnear, Ann Arbor: Association for Consumer Research, 259-264.

Gilligan, Carol (1982), In a Different Voice: Psychological Theory and Women's Development, Cambridge, MA: Harvard University Press.

Hansen, Flemming (1984), "Towards an Alternative Theory of the Advertising Communication Process?" International Journal of Research in Marketing, 1 (1), 69-80.

Jensen, Thomas and Larry Rottmeyer (1986), 'Visual Information Processing of Television Commercials: Cognitive Effects," in Advances in Consumer Research, Vol. 13, ed. Richard J. Lutz, Provo, UT: Association for Consumer Research, 158-163.

Krippendorff, Klaus (1980), Content Analysis, Beverly Hills, CA: Sage.

Oakley, Ann (1972), Sex, Gender, and Society, New York: Harper & Row.

Roberts, Mary Lou (1984), "Gender Differences and Household Decision Making: Needed Conceptual and Methodological Developments," in Advances in Consumer Research, Vol 11, ed. Thomas Kinnear, Ann Arbor: Association for Consumer Research, 276-278.

Rubin, Lillian B. (1983), Intimate Strangers: Men and Women Together, New York: Harper& Row.

Sherry, Jr., John F. (1985), "The Cultural Perspective in Consumer Research," in Advances in Consumer Research, Vol. 13, ed. Richard J. Lutz, Provo, UT: Association for Consumer Research, 573-575.

Tucker, W.T. (1977), 'A Long Day of Discrepant Behavior," Paper presented at the 1976 Conference of the Association for Consumer Research.

Witkin, H.A. (1962), Psychological Differentiation: Studies of Development, New York: Wiley.

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