Researching Documents and Texts

Click on each question to check your answer.

Short Answer Questions

1. Describe the differences between content and discourse analysis.

Both content and discourse analysis are methods of textual analysis; however, these methods are guided by different logic and have different approaches to theory construction. The former is usually descriptive and objective and the latter interpretive and subjective.

Content analysis takes a scientific approach to measuring manifest content, answering “how many” and “how often” questions about the messages communicated by texts. One of the key assets of this method is the reliability and validity of its findings. When carefully designed and carried out, content analysis can examine large volumes of texts, make comparisons within and between texts, and measure changes in content over time. The emphasis on rigour in quantitative measurement, however, may mean that important messages are not captured by content analysis because of its reductionist tendencies.

Discourse analysis, on the other hand, is ideally suited to teasing out latent messages and meanings and situating them within their particular ideological or historical context. Discourse analysts see a discourse as a unified whole whose meanings cannot be disaggregated, or reduced to specific words, phrases, or symbols. Because discourse analysts see texts as embedded within socially constructed meaning structures, researchers begin with theory and adopt a deductive approach, observing how discourses reflect, reinforce, and perhaps even challenge relationships of power, historical social structures, and institutional norms. However, there is considerable potential for bias when accounting for underlying messages in texts or identifying patterns in discourses.

2. Describe what a codebook is and explain what role it plays in content analysis.

Researchers must also decide what information will be coded, and use this information to construct a codebook, which is a detailed set of instructions informing research team members how to apply appropriate codes to each case in the study. The construction of codebooks can be a very time-consuming process in content analysis, as the instructions must be complete so that independent coders will assign the same codes independently to the same text. In quantitative approaches, the codebook specifies the numeric values to assign to the variables, and these values must be mutually exclusive (that is, a case must fall into only one category) and exhaustive (that is, there must be a category for each case). With the codebook established, the researchers must code their data. This can be done manually (that is, the coder decides how the data are coded) or through automated software (that is, the software selects how the data are coded).