Content Analysis

• Content analysis is a method of research where you look at content that has already been produced.

• You look for things in the medium that you wish to study in order to learn something about the people who created it.

• By quantitatively looking at what is in the texts, you can find themes that may not be obvious when looking at the same content at face value.

• Like other methods of quantitative analysis, you must form concrete definitions of what you are studying, gather a representative sample from the population you want to study, create codes for what you are looking for, and analyze the data with statistics.

• The great benefit to content analysis is that it is non-reactive, meaning that the act of studying the topic does not influence the data.
  • This is because the data is taken from sources which already are produced.

• Taking data from content which was not produced for the purpose of the study is a critical part of content analysis.

• Very clear coding mechanisms must be created to make sure that the data is accurate.
  • Clearly articulating the coding procedure makes replication possible.
  • Coding depends on what medium you want to research, but neutral coding is important so that researcher stereotypes do not effect the data.

• The units of analysis for content analysis are found in the texts. Words, phrases, actions, articles, etc. can all be analyzed.

• Using structured observation, which is a systematic viewing based on written rules, you code the units of analysis based on what you are looking for.

• There are a few different things to look for in the content: frequency, which is simply a count of something happening or not happening, direction, which is the direction of some sort of message in the content, intensity, which is the strength of a message in the content, and space, which is simply the size of a message in the content.

• There are two types of coding in content analysis.

• Manifest coding looks at the explicit content of the text.
• This looks for words, phrases, etc. that are counted for the data set.
• Manifest coding is more reliable than latent coding because the word is either there or is not, and different results can only come in error by missing a word.
• While the reliability is high, Manifest coding does not take the context of the word into account, so the validity of the measurement can come into question.

• Latent coding looks for the implicit meaning of the text.
  • This type of coding uses a less strict guide to code and looks for the messages behind the text.
  • While latent is less reliable than manifest coding because the coders interpretation is involved, the codes are often more valid because it looks at meaning rather than face content.

• Intercoder reliability is important when more than one coder is used to ensure that all coders are coding the text in the same way. If there are differences in coding then a stricter guide must be used or the coders must learn to code similarly.

• An important part of content analysis is the inferences that can be made from the analysis.
  • You are looking at the content of some text, not the intent behind that text.
  • The intent of the authors or the readers of the text cannot be inferred.
  • The texts themselves do not reveal anything about the intent of the authors' effect on the audience.