News media use, credibility, and efficacy: an analysis of media participation intention in China

Abstract

This study1 investigates the relationship between news media use and media participation intention in China by adopting an audience-oriented perspective in light of the changing patterns of media content concomitant with widespread commercialization in reformed China. Data from a total of 2,409 valid face-to-face interviews were collected in four cities during November 2006 and May 2007. Our results suggest that television news had a positive association with media participation intention, whereas newspaper and Internet news use produced mixed results. In addition, significant regional differences regarding media influences were identified.

Keywords:

• media participation
• news media use
• media efficacy
• media perception

Notes

^1. Data for this research were made available by a grant from the Media Literacy Theory and Practice Project (code 04JDZ00038) at Fudan University, Shanghai, China.

^2. Commercialization also brought participatory features to entertainment programs. Examples include text message voting for competitors in interactive reality game shows modeled after the television show American Idol. Considering their pure entertainment nature, we do not discuss these forms of participation in this study.

^3. Most party papers in China, such as The People's Daily, have established a department of “work for broad masses” (qunzhong gongzuo bu or qungong bu for short) to process audience letters before the commercialization of the press.

^4. Institutionalized political participation is nearly completely absent, except in certain routinized rituals, ceremonies, and organized celebrations in metropolitan areas in the country. Rural residents have a modicum of choice in village and township official elections.

^5. In China's media organizations today, financial motives are hardly distinguishable from political ones, mainly because alternative and oppositional views, or “edge ball”, as they are commonly known within the industry, have a proven positive impact on circulation and ratings. One might even say that behind the most radical political views permitted to appear in the mass media lurks a revenue benefit.

^6. See Li (2010); Bian and Cheng (2007).

^7. Data obtained from the World Value Survey: http://www.worldvaluessurvey.org/.

^8. This corresponds to the idea that an inquiry of contextual effects requires attention not only to the extent to which individuals are influenced by their social surroundings (in our case, media niches), but also whether individual factors interact with contextual features to generate effects (Burbank, 1995).

^9. As far as newspapers are concerned, there are only a handful of nationwide papers in China. Most of the nationwide papers in China are party papers, such as The People's Daily, which is seldom read by the masses, and some papers targeting certain demographics (e.g., The PLA Daily). As for television news, we obtained some data from CSM Media Research through informants. CCTV news is the only nationwide broadcast news. Its average ratings (evening news) in Shanghai, Beijing, Xi'an, and Guangzhou are .47, 2.47, 1.58, and .46, respectively, in 2007. By contrast, local news' ratings in the four cities are 4.30 (Shanghai), .91 (Beijing), 1.50 (Xi'an), and 2.21 (Guangzhou). Overall, it seems that the local media's influence is greater than that of the national media. But in terms of television news, audiences from Beijing and Xi'an rely more on national media.

^10. The response rates (RR3) were 85.7%, 50.0%, 25.0%, and 91.0% for Beijing, Guangzhou, Shanghai, and Xi'an. We acknowledge the fact that the response rate for Shanghai was lower than desired. However, we consider this pattern of response rate distribution across four cities as reflective of the degree of economic development and market competition rather than different selection bias across cities. In Shanghai, the most commercialized city in China, market surveys nowadays are so ubiquitous that people start to develop negative attitudes toward them.

^11. The Bureau of Statistics in China (http://www.stats.gov.cn/) does not publish the nation's demographic data in a consistent way across different geographical areas. Breakdowns of education and income levels are not available to the public. Therefore, we could only rely on population statistics on gender and age to gauge the representativeness of our data.

^12. Despite the hierarchical structure of our dataset, it is not possible to adopt hierarchical linear modeling because of the small number of second-level cases, which would lead to biased estimates (e.g., Maas & Hox, 2004). Another way of estimating group difference of regression coefficients was suggested by Cohen and Cohen (2003, p. 111). However, the method they suggested can only be used to test a single coefficient difference. To the extent that we have several paths for group difference estimation, multiple tests might enhance the chance of Type I error.

^13. The equivalently estimated model in this study imposed equality constraints only on paths relating any endogenous variable pairs. We are less interested in group variance involving relationships between control variables and endogenous variables than we are in group variance involving relationships between two endogenous variables.

^14. The model fit parameters for the initial models are: chi-square = 597.70, df = 48, p < .001, CFI = .860, RMSEA = .068. According to the suggestions from the LM test, two changes were made: the coefficient linking age and gender was fixed to zero, and the relationship between the error terms attached to newspaper reading and television news watching was freely estimated.