Chapter 3: Methodology

Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Bibliography

Copyright John Worth; originally published 1997

Research questions

Prior research has established very little about how women are portrayed in advertising. Studies into Japanese culture have shown that women perceive themselves as tied to the home and family. However, White (1992) noted that the image of the Japanese woman in the 1990s is “career-professional,” a logical progression, she said, from the “housewife-professional” that began to emerge in the 1980s. This seems to represent a confused identity within Japanese society; there are currently two roles that, some suggest, are coexisting.

Mass media and advertising are pervasive forces in the expression of values in society. In a heavy consumer culture such as exists in Japan, examination of television advertising can provide insight into the portrayal of women's roles. Based on these premises, this study will seek to answer one research question.

What roles for women are portrayed in Japanese television advertising?

Categories

To answer this question, the possible social roles women can play have been divided into three categories (based loosely on Pingree, et al.’s 1976 “Scale of Sexism”). First there is the role of ‘woman as dedicated housewife/mother’. There are eight values which define this category: home, shopping for the family, caring for the family, eating with the family, products for the home and for the family, and food for preparation at home. (‘Products for the home’ refers to cleaning supplies, home-appliances, etc., i.e. those products used primarily to care for the home; products for the family include houses, bathtubs/toilets, and insurance/financial planning services.)

The second category is woman as independent, with the subject never shown in a home setting. The values defining this category are: work (location variable for woman at work), woman at a department store, woman at a convenience store, outdoors, school, other locations, images of a woman shopping for herself, eating alone, relaxing because the product eases her work load, relaxing because the product helps her relax after work, sports, and product categories for fast/instant food, personal use/beauty and other.

The third category is a mixture of the first two. There are no values that can be used to exclusively define this category; Category Three is any combination of Categories One and Two. This category is also referred to as ‘women in dual roles’. This study is looking for the presence of the dual roles, not necessarily the way a woman plays those roles.

White’s business professional would be categorized as ‘woman as independent’. This study hopes to find out if there is evidence of such a role attributed to women in contemporary Japanese advertising. There is one major category that will define this: work. If the advertising does not physically show a woman in a workplace (the coders were instructed to note the type of workplace), then the woman is not being portrayed as a worker. Though business professionals might be shown outdoors or even in the home, the viewer must have some clue that the woman is such a person. One indicator is dress; however, attire is something difficult to exhaustively code. Thus, in the debriefing of the coders, they were asked their opinion on this and whether, in their opinion, they saw any evidence of a ‘business professional’ in the advertising.

Other variables

The study has three variables that are necessary to define the sample. These are the presence or absence of people, presence or absence of women, and the extent of the role the women play. There are also two descriptor variables: the sex of the announcer (if one is present) and whether women are shown using the product, including consumption of food products. This last variable can not be forced into any of the above three categories; rather, the category combined with using the product can provide insight on the role of the woman.

Description of sample

The sample for the study was Japanese television commercials taped during prime-time. Through correspondence with a Dentsu manager it was learned that prime-time in Japan is from 7:00 p.m. until 11:00 p.m. and the existence of four major advertising-dependent (not government) networks was confirmed. Each of the four major networks were recorded for two nights in a ten day period. Of these two nights, one for each network was randomly chosen for coding.

Since the study is not designed to measure total exposure but, rather, the content of the individual commercials, repeat commercials were removed from the final sample. The final sample size was 411 advertisements.

Quantitative analysis

This study uses quantitative content analysis. Content analysis allows the researcher to identify advertising’s individual elements. A chi-square test is used to determine if there is a relationship between variables which make rows and columns of a data table. If the rows and columns are independent, then the null hypothesis is supported. Independence can be tested by comparing the observed cell frequencies to the expected cell frequencies. The significance level for this study is .10.

Coders

The coders for this study are native Japanese students enrolled at the University of Tennessee. After training, each student was given a tape of the commercials to code. They were compensated financially for assisting in the study.

Reliability

Singletary (1994) noted that ”reliability is near the heart of content analysis; if the coding is not reliable, the analysis cannot be trusted.” To test intercoder reliability, Scott’s Pi was used. This test was chosen because it takes into account chance agreement among the coders. The formula for Scott’s Pi is:

            pi=PO - PE/1-PE

where PO = percent agreement observed and PE =percent agreement expected. Reliability was tested on 205 commercials (roughly half the sample).

Instrument

Each coder was given a two page coding sheet (Figure 1). The first questions were designed to establish if there are any announcers, the sex of the announcers, presence of people in the commercial, and role of the women. Presence and sex of announcers is particularly important as the announcer is the ‘voice in the wilderness’ that lends credence and stability to the product claims. (Length of the ads will not be measured because previous studies in Japanese advertising have consistently shown that the majority of Japanese television commercials are 15 seconds in length. See di Benedetto et al., 1992; Lin, 1993; Ramaprasad and Hasegawa, 1992 & 1990; Sengupta, 1995)

The coding then focused on three areas. The first area was the location of the woman. This is important to establish where women are portrayed. What the woman is doing was the second area of focus. Finally, the product categories associated with the women were defined. This study did not use the product categories as defined in prior studies. It is not in line with the purpose of the study to determine how many commercials for appliances a woman appears in. What is important are the general categories. The products are defined as those for the home (cleaning supplies, home appliances, etc., i.e. those products used almost exclusively by one person for the benefit of the rest of the family); those for family (houses, toilets, bathtubs, insurance, etc., i.e. products designed to be beneficial equally to all members of the family); food for preparation at home (including mixes, seasoning, etc.); fast food/restaurants and other food for outside the home (including canned drinks, instant noodles, and other “ready to eat” products that “take away” from the woman’s traditional role as cook); and products for the woman (such as hair-sprays, bath oils, cosmetics, clothing, etc.). An ‘other’ category was included in all three areas.

Limitations

This study does not provide any historical tracking of how the portrayal of women has changed. The logistics of undertaking such a study are very complex and difficult to arrange without direct connection with a Japanese advertising or mass media organization.

There is also no coverage of advertising from different day parts, and the days chosen for recording were not randomly selected over a long period of time, but represent four days out of just over one week of programming. A further complication that results from this is that, from the researchers own experience, there seems, logically, to be significant seasonal variation in the advertising; this study did not account for such variation. Another area not covered in this study is age of the woman.

Conclusion

This study, like all others, has its flaws. However, these blemishes do not distract from the overall shape or effectiveness of the study. The structure is designed to work around the logistical problems and get to the core of the study with as little fat as possible. The end result is a start towards defining the relationship between culture and advertising.