170 likes | 346 Views
A Taste of a Method’s Own Medicine: A Content Analysis of Content Analyses. A Presentation by: Caroline Graham Austin Angeline Grace Close Sunil Contractor Ji Hee Song Qiyu (Jason) Zhang. INTRODUCTION. Research Questions
E N D
A Taste of a Method’s Own Medicine:A Content Analysis of Content Analyses A Presentation by: Caroline Graham Austin Angeline Grace Close Sunil Contractor Ji Hee Song Qiyu (Jason) Zhang
INTRODUCTION Research Questions • RQ1 From the marketing perspective, what is the overall contribution of content analyses published in the marketing literature from 1977 to 2002? • RQ2 What is the developmental history of CA as a research/analytical method, in the selected marketing literature? • RQ3 As a method, has CA been adequately applied in the marketing literature?
STRUCTURE • Literature Review • Method • Findings 1-3 • Conclusion
A LOOK INTO THE LITERATURE “Given the potential methodological problems associated with content analysis, it is useful to assess the nature of past applications.” (Kolbe and Burnett, 1994 p. 244) • Kassarjian’s (1977) methodological benchmark • Yale and Gilly (1988), 6 advertising journals- 1976-1985: relationship b/w journal & topics • Kolbe and Burnett (1994) operationalized Kassarjian’s directives • Riffe and Freitag (1997) CA of CA: JQ: found reliability & sophisticated statistical analysis trends
METHOD Sampling • full-length CA articles 1977-2002 • Journal of Marketing • Journal of Marketing Research • Journal of Consumer Research • Journal of Business Research • article selection: read abstract, method section, computerized database search & article index search
YEAR No. of Articles Percentage of Sample YEAR No. of Articles Percentage of Sample 1977 1 2.86 1989 4 11.43 1978 1 2.86 1990 2 5.71 1979 1 2.86 1991 2 5.71 1980 1 2.86 1992 1 2.86 1983 1 2.86 1993 2 5.71 1984 1 2.86 1994 2 5.71 1985 3 8.57 1995 1 2.86 1986 4 11.43 1996 2 5.71 1987 3 8.57 1998 1 2.86 1988 1 2.86 2001 1 2.86 Table: Sample Distribution
METHOD Coding Issues • conceptual & operational definitions • coder training: agreement on 29/ 31 items • major coders and checkers coded all articles • checking and re-coding • PRAM calculates reliability coefficients (e.g., % agreement and Holsti’s method) • global Reliability= 94.1% & Scott’s pi, Cohen’s kappa (86%-100%) • data analysis (e.g., Time series to find trends, cross-tabulation to find correlations)
RQ 1: Overall Contribution • RQ1 From the marketing perspective, what is the overall contribution of content analyses published in the marketing literature from 1977 to 2002?
RQ1: Overall Contribution CA of variable: “Medium Analyzed” • magazine advertisements (31%) • television (28.6%) • academic journals (14.3%) • newspaper advertisements (8.6%) • comic books (5.7%) • outdoor advertising, novels, experimental data (each 2.9%)
RQ1 : Overall Contribution CA of variable: “Content” • Advertising Appeals:(51.4%) • Social Values: Social Values (14.3%) (e.g., Materialism and the Good Life) • International content (20%)
RQ 2: Developmental History • RQ2 What is the developmental history of CA as a research/analytical method, in the selected marketing literature? Paradigm Shift (e.g.,): • Latent variable • Link to theoretical framework • Hypothesis posed • Statistical methods used
RQ 2: Developmental History Positivistic analysis to interpretive analysis • # of articles that examined latent variables has been increasing. • # of articles that had theoretical framework has been increasing. • # of articles that had advanced statistical analysis has been increasing
RQ 3: Adequate Application • RQ3 As a method, has CA been adequately applied in the marketing literature?
Paper Includes Total # (%) Coder training (Yes) 6 (37.5%) Pretest/Pilot (Yes) 6 (37.5%) Coding rules and procedure (Yes) 16 (100%) Independent coding (Yes) 13 (81.3%) Inter-coder reliability (Yes-specific reporting) 13 (81.3%) Intra-coder reliability (Yes) 1 (6.3%) Validation (Yes) 1 (6.3%) Number of Papers Citing Kassarjian and Following His Guidelines (N=16)
Statistical Analysis in CA (1977-2002) Most give frequencies (80.6%) Chi^2 (36.1%)
CONCLUSION • Overall contribution • Developmental history of CA • Methodological rigor