Content Analysis of Newspaper and Television Advertisements
Taiwan's 2004 Presidential Election
In this study, all political advertisements placed in Taiwan’s leading newspapers and aired on Taiwan’s most popular terrestrial and cable television stations during the official campaign period of 28 days (21 February ?19 March 2004) were analyzed.
According to a survey conducted by AC Nielsen, there are four leading newspapers in Taiwan. When asked which newspaper the interviewee had read the day before, 45 percent of the respondents mentioned the China Post, 34 percent the United Daily News, 32 percent the Liberty Times, 6.3 percent the Apple Daily.
A report on Taiwan’s top 50 television stations released by the Government Information Office was used to determine which stations should be included in the analysis. Nine stations were selected according to their viewing rates. They were FTV (min shi), TVBS-N, CTI (zhong tian xinwen tai), SETTV (san li xinwen tai), CTV (zhong shi), FTV-N (min shi xinwen tai), CTS (hua shi), TTV (taishi), and TVBS.
The coding unit used in this research was a theme, that is to say a coherent idea about the candidate and his party. Each theme (advertisement) was coded according to its function. There were four different functions found in the commercials: The sponsoring candidate attacked the rival candidate (attack), rebutted criticism (rebuttal), created a positive image of himself (image), or tried to appeal to the electorate for support (appeal).
Newspaper ads were further categorized into three groups according to their size: half page, full page and smaller formats. Television ads were further separated according to the language used in the ads: Taiwanese, Mandarin or non-vocal.
It is widely acknowledged that intercoder reliability is a critical component of content analysis. Neuendorf, for instance, notes that ‘given that a goal of content analysis is to identify and record relatively objective (or at least intersubjective) characteristics of messages, reliability is paramount. Without the establishment of reliability, content analysis measures are useless.?A large number of researchers, however, often fail to assess (or at least report) intercoder reliability.
In this research, all collected newspaper and television ads were analyzed by two coders. Cohen’s kappa was used to calculate intercoder reliability as recommended by most scholars. In the coding of the newspaper ads, kappa was 0.97, and in the coding of TV ads it was 0.94. Fleiss and others note that kappa values greater than 0.75 may be considered excellent agreement beyond chance. Thus, the kappa values obtained in this research indicate excellent reliability.
There were a total of 120 different ads found in the four observed newspapers: the China Post, the United Daily News, the Liberty Times and the Apple Daily. The total observed frequency of the 114 different commercials in the four newspapers amounted to 235. Some 37 percent appeared in the Apple Daily, 25 percent in the Liberty Times, 20 percent in the China Times, and 17 percent in United Daily News. In the latter, ads were sponsored exclusively by either the alliance or its supporting organizations. The DPP refrained from placing ads in that newspaper because of its boycott of the paper, which it viewed as the mouthpiece of the alliance. In the China Times, ads sponsored by the alliance amounted to over 50 percent, whereas ads placed by the DPP accounted for about 12 percent. The Liberty Times was the only newspaper where the number of DPP ads surpassed the alliance. Almost seven out of ten observed different ads that appeared in the Liberty Times were sponsored by the DPP. Individuals and organizations not supporting either side placed three quarters of all their ads in the Apple Daily.
During the election campaign period, a total of forty-four different political ads were aired on the observed seven TV stations: FTV, TVBS-N, CTI, SETTV, CTV, FTV-N, CTS, TTV, and TVBS. Almost half of the advertisements were ads of the alliance. Its interest groups refrained from utilizing television. The DPP produced eleven ads, the government ten, and a DPP-associated organization two. The KMT and its allies, thus, put slightly fewer different ads on Taiwan’s TV channels. Notwithstanding, ads by the alliance were aired far more often than those of the DPP and government agencies.
About 35 percent of the newspaper ads and ads contained negative messages about the rival candidate. Over 90 percent of the negative ads attacked incumbent President Chen Shui-bian and his government. Only 2 out of the 120 different newspaper ads and 4 out of the 44 television spots conveyed negative messages about the candidates of the alliance.
The issues mentioned in the DPP newspaper ads differed from the TV advertisements. Most of party’s newspaper ads made reference to the two referenda and the 2-28 rally. The 11 TV advertisements, on the other hand, focused on the party’s past achievement and the alliance’s close relationship to the underworld. The content of newspaper ads and TV advertisements sponsored by the alliance did not differ much. The alliance’s key strategy was to blacken the image of Chen Shui-bian. During the first few days of the official campaign period, the alliance used ads to convey the image of an incompetent president who was to blame for all the misery ordinary people faced. After the successful 2-28 hand-in-hand rally, the content of the ads became more abrasive. The strategy of the alliance was to create an atmosphere of hatred and fear among the electorate. The alliance portrayed President Chen Shui-bian as an unscrupulous dictator like Adolf Hitler and appealed to the public to help the alliance to put an end to Chen’s dictatorship.
Compared with the previous presidential election, fewer negative ads appeared in the observed media, but the tone and language used in the ads (especially those of the alliance) became coarser. Moreover, the content of the political ads seem to indicate a drastic decrease in the moral and professional standards of most parts of the electoral campaign industry in Taiwan.
Download full paper in PDF for free
. Neuendorf, K. A., The content analysis guidebook (Thousand Oaks, CA: Sage, 2002)141
. Lombard, M., Snyder-Duch, J., and Bracken, C. C., ‘Content analysis in mass communication: Assessment and reporting of intercoder reliability? Human Communication Research, No. 28 (2002), pp. 587-604; Riffe, D. and A. A. Freitag, ‘A content analysis of content analyses: Twenty-five years of Journalism Quarterly,?Journalism & Mass Communication Quarterly, Vol. 74, No. 4 (1997), pp. 873-882.
. (Dewey, 1983; Bakemann, 2000) Dewey, M. E., ‘Coefficients of Agreement,?British Journal of Psychiatry, No. 143 (1983), pp. 487-489; Bakeman, R., ‘Behavioral observation and coding,?in H. T. Reis and C. M. Judge (Eds.), Handbook of research methods in social and personality psychology (New York: Cambridge University Press, 2000), pp. 138-159.
. Fleiss, J. L., Statistical Methods for Ratios and Proportions (New York: Wiley, 1981).
. A detailed analysis of the advertising strategies of all candidates contesting the 2000 presidential race can be found in Niu, Tse-shun (2002), jingxun chuanbo celue: lilun yu shiwu [Campaign Broadcasts: Theory and Pratice], Weber, Taipei.