It occured to me that doing a computer-assisted content analysis of Chinese media to study Chinese nationalism might be an interesting area of research. An ideal candidate for such a study would be the state-run Xinhua News Agency. I am familiar with how news agencies work, having worked for Taiwan's Central News Agency as a copy editor, as well as using news wire copy from CNA, Xinhua, AP, Reuters, and other sources when I worked as a newswriter for the China Television Company and a reporter and editor for the China News newspaper.
But how could I use Xinhua wire service copy? I did quite a bit of research on content analysis methods and tools in the winter. Although there are lots of computer tools to aid in content analysis, there didn't seem to be one that suited my needs, which were shaping up to involve importing large amounts of Xinhua news items into a database and analyzing the content for meaning or frequency counts of specific terms. There are tools to look at content in these ways, but there didn't seem to be any way to make batch imports from existing databases. This is a major issue, because in a given year Xinhua releases thousands of news items, and to manually import each one and analyze it would be too time consuming. I could use sampling methods, but I was concerned that random samples would miss certain trends associated with specific events tied to time periods (for instance, anniversaries of deaths or uprisings, mentions of new research on Chinese unity, etc.). Additionally, there was the issue of finding tools that I could use at home ... I have a Mac, and most tools are for PC or DOS environments.
I began to consider how I might use a web-based search engine on a database like Factiva or LexisNexis to examine Chinese media coverage, and tie it into nationalism and policies related to nationalism in China.
I thought that a good first step would be to do a test run in a real academic situation (i.e., a class). I needed to take another history class, and a good class on Chinese history was coming up. In the spring of 2005 I signed up for Professor Kuhn's class on overseas Chinese (Chinese Emigration in Modern Times). I had taken Professor Kuhn's class two years earlier on China in Modern Times, and very much enjoyed his expertise and teaching methods. I knew that a research paper would be required, and this would be a good way to test out my new methodology.
For the paper, I ended up doing a study of PRC policies regarding overseas Chinese communities in Vietnam and Cambodia in the late 1970s. Via a website operated by Harvard Libraries, it was possible for me to conduct computerized searches on the LexisNexis academic database, which contained Xinhua English news items going back to January 1, 1977. I also learned how to use Microsoft Excel to enter my raw data (consisting of Lexis Nexis searches on certain terms in Xinhua coverage from the late 1970s), calculate relative percentages, and create charts of the data.
As for research in Xinhua itself, I found that while historians often use Chinese media as individual primary sources, they seldom conduct content analyses (computer-assited or otherwise) on media sources in aggregate. However, the study of mass media and journalism frequently uses computer-assisted CA techniques, and I found that there are a handful of Chinese scholars living in Hohg Kong and the US who have done a great deal of research in English on Chinese journalism, the history of Chinese media, and content analyses of Chinese media. This was a great help in building up a theoretical background for my own research, as well as providing valuable information about the history of the Xinhua News Agency.
In the introduction of my final paper, I noted that a computer-assisted content analysis would be instrumental in backing up my theories about China's differing overseas policies in these two Southeast Asian countries:
Manual and automated content analysis of media sources has been a staple of media studies and international relations for decades. The study of history, however, tends to view mass media in a different light. Old media content is often evaluated as individual primary sources — a news article about the Titanic, an essay by Zhou Enlai, a runaway slave advertisement by Thomas Jefferson — but are seldom examined in aggregate. This is beginning to change, as news media are distributed electronically. Print media content is often stored in databases, and after a few decades passes from the realm of current affairs into the realm of historical artifacts. These electronic records are easily searchable, and can be parsed with software tools that can reveal patterns not readily apparent in selective manual sampling of old media.
It is my belief that a structured, computer-assisted analysis of the electronic archives of the Xinhua General Overseas News Service from the late 1970s can help us better understand China’s emerging overseas Chinese policy, and the extent to which it was upheld in dealings with Kampuchea and Vietnam. My content analysis of Xinhua will be augmented by more traditional historical methods of research, including references to individual Xinhua news items. I will also cite scholarly literature on the PRC’s overseas Chinese policy, trilateral relations between China, Vietnam, and Kampuchea, and the nature of Chinese mass media and Xinhua’s English wire service.
I haven't received my corrected paper back from Prof. Kuhn, by he did give some feedback when I presented my methodology in class and in his office. Firstly, he noted that a content analysis such as mine would require additional research on PRC policies that aren't readily discerned by my computer-assisted methods. He specifically pointed to readings on power politics vis a vis the USSR and the US. He also stressed that my search methods on the database of XNA sources had to be careful regarding counts of terms which might have more than one English translation (for intance, is hua qiao commonly referred to as Overseas Chinese, or Overseas Compatriots, or Ethnic Chinese?) and controlling for certain terms where a single Xinhua story might mention both countries.
Interestingly, he also noted that my study was the first he had seen using "Scientific" methods to study Chinese history.
In my next post, I'll talk more about what I learned from my initial foray into computer-assisted content analysis, and how I am incorporating these techniques into my thesis proposal.
No comments:
Post a Comment