The Issue of Subjectivity in Conducting a Qualitative Analysis Using a Quantitative Research Methodology in Corpus Studies
One of the first exploratory corpus studies on gender representations in print advertisements was conducted by Alice Courtney and Sarah Lockeretz in 1971. This study indicated discrepancies in gender portrayal and concluded that women have a growing desire to be portrayed in their varied life roles. Since then a heightened awareness of gender issues has marked the study of print advertising.
My hypothesis was that the position of women has not advanced into the predominantly male area of advertising in the past forty years due to sexual objectification. I built a corpus of 524 advertisements from 8 American magazines from 2009-2010. Some of these overlapped with the corpus of the 1971 study, such as Newsweek, The New Yorker and TIME. I carefully chose alternatives for the magazines that were not available or had stopped publication.
I adopted the methodology of the 1971 study in order to generate comparable data. This meant noting basic information on the advertisement, such as number of adults visible, gender and (non)occupational activities. An additional measurement was needed to assess the level of sexuality. Based on the work of Tom Reichert and Jacqueline Lambiase, six additional categories were implemented: the level of nudity, sexual behaviour, physical attractiveness, sexual referents and sexual embeds. Even though this study intended to conduct a qualitative analysis to provide an overview of developments in contemporary print advertising in the past forty years, the quantitative research methodology led to more urging questions.
Sexuality is a difficult construct to define and even more difficult to quantify. An operational definition is therefore needed to define the variable(s) which quantify this construct. However, the operational definition is formed by the researcher's perception of the construct and will consequently influence his/her choice of variables. Thus the methodology in this corpus study was influenced by the researcher's demographic, historical, social and ideological perceptions of the construct.
Consequently, the question arises whether the qualitative analysis can be treated as valid. Even though this question is essential when we do corpus research in the field of advertising, I personally consider the subjectivity of the researcher's choice for operational definitions relevant. It provides valid information on the researcher's perception and those of his/her community.
This corpus study was limited in time and financial resources. Its methodology could be improved by performing a random perception measurement among a number of individuals from different communities, indistinctive of their historical, social and ideological perceptions. In this manner, a coherent image of the perception of sexuality in advertising could be achieved, regardless of gender, historical background, social and ideological perceptions.
Renate S. Moelands |
Radboud Universiteit Nijmegen |
Begeleider: Dr. U. Wilbers |
Datum: 12 oktober 2012 |
E-mail: r.moelands@let.ru.nl |