There is a confusion around the concept content analysis. Several different approaches are all labelled content analysis :
- Inductive content analysis (=conventional content analysis) is using open coding of observations creating categories and abstraction, sometimes resulting in an overarching main category . It is suitable when there is little preexisting knowledge and theory in the topic studied . Inductive content analysis is similar to the process used in other qualitative methods such as grounded theory. However, inductive content analysis is more flexible compared to other qualitative methods and there are no strict rules .
- Deductive content analysis (=directed content analysis =deductive category application) starts with a set of predefined categories and investigates if the data fits these categories. This is more suitable if there are a reasonable amount of preexisting knowledge and theory in the topic studied . The outcome is presented as some form of agreement with the preexisting theory. It may sometimes be presented using some form of statistics.
- Summative content analysis: Creating a “Word sallad” where the size of each word represents how common it is used, or by describing the number of times a word is used is labelled manifest content analysis . It is labelled latent content analysis if the process proceeds to also include identifying any underlying (latent) meanings of the words.
Content analysis is not rooted in underpinning theory to the same extent as phenomenology, hermeneutics and grounded theory. Hence, content analysis is sometimes referred to as a theoryless qualitative approach.