Computerized Content Analysis in Management Research: A Demonstration of Advantages & Limitations
比较了人工编码与计算机编码在管理研究文本分析中的效果，发现两者同样有效，但计算机编码更可靠、成本更低，并探讨了不同分析单位（句子、段落、整篇文档）对结果的影响。
Content analysis is a research technique used to objectively and systematically make inferences about the intentions, attitudes, and values of individuals by identifying specified characteristics in textual messages. The unobtrusive nature of content analysis makes it well suited for strategic management research. To date, the content analyses in most management studies have been performed by human coders, despite advances in computer technology that enable researchers to perform the same analyses more reliably and less expensively. In this paper, the investigator compares human-coded content analysis to computerized coding of the same text communications. The results suggest the two methods may be equally effective. Differences in results obtained using different units of analysis (such as analysis by sentence, by paragraph or by whole document) are also examined. Better reliability, improved stability, and comparability of results suggest more extensive use of computerized content analysis in future research.