Content analysis is one of the social science research techniques with the longest tradition and one of the most widely used. It is, in fact, a variety of techniques that focus on various aspects of communicative products or "messages".

It is one of the most widely used techniques in the social sciences throughout the 20th century, gradually increased by computer advances in data processing.

The fundamental authors are Bernard Berelson (Content Analysis in Communication Research, 1952) and Klaus Krippendorf (Methodology of Content Analysis, 1980). The classic definition by Bernard Berelson (1912-1979) in Content Analysis in Communication Research (1952) is the following:

A research technique for the objective, systematic and quantitative description of the manifest content of communication.

The latent content of communicative texts is not analyzed by quantitative content analysis. This consideration refers to the interpretative nature of content analysis, which should not go beyond the manifest.

Content analysis is a technique that seeks to identify specific characteristics within a text or discourse, which is used to obtain valid and reproducible inferences from the data. It is a technique that can be used, within a research design, for exploratory as well as confirmatory or explanatory purposes. It seeks, like other quantitative techniques, to determine what are the recurrent structures and patterns in a given universe.

The characteristics of content analysis are:

- It is a technique of inquiry, search and investigation.

- It is objective: it needs to be clearly designed, using quantifiable data and verifiable techniques.

- It is systematic: it must take into account all the data in the text, systematize them and always code them in the same way.

- It is quantitative: the data are expressed numerically.

- It generally refers to the manifest content, i.e. the content expressed at the textual surface level, not the latent content.

The objective of content analysis is to quantify patterns and frequencies by counting heterogeneous materials (books, newspapers, poems, news items) and heterogeneous contents (themes, words, images, etc.). It is a technique suitable for extensive research, with large volumes of text. Technological advances allow research to be extensive and applied to large volumes of texts.

Klaus Krippendorf (1980), one of the key authors in understanding this method, establishes six basic questions to address this technique:

1. What data will be analyzed?

2. How are they to be defined?

3. What is the population from which these data will be extracted? 4.

What is the context in relation to the data to be analyzed?

5. What are the limits of the analysis?

6. What is the purpose of the inferences?

The phases of a research design based on content analysis are the following:

1) Constitution of the corpus (and corpora), composed in turn by the units of analysis. Possible design with a view to comparison with other corpora.

2) Data collection.

3) Categories and codes. Includes training of coders and improvement of coding. This is probably the central operation of the content analysis.

4) Data analysis, which is where various techniques such as semiotics, narratology, rhetoric, discourse analysis, etc., can come into play.

We will now examine these operations one by one.