Quantitative content analysis easily explained

Welcome to the tutorial “Quantity before quality – Quantitative content analysis for beginners“. Quantitative content analysis is a common research method for analyzing quantitative, i.e. measurable data systematically and in a standardized form. The purpose is to convert textual material into computable quantities and evaluate them using statistical methods. If you are looking for a short and concise account of this empirical research method, this tutorial is for you. Here you will learn everything you need to know to get a comprehensive overview of the characteristics of quantitative content analysis and its individual stages.

Quantitative content analysis at first glance

Let's start with the objective of quantitative content analyses. Unlike qualitative content analysis, which is designed to examine a small number of data in depth in order to understand them and generate new theoretical considerations, quantitative content analysis is concerned with describing and explaining as many cases as possible in breadth. Its overriding goal is to confirm or invalidate an existing theory or thesis. For this purpose, only those characteristics are considered in the analysis that are relevant to answer the guiding question.
Its application area covers any communication content, including texts in particular, but also audio and visual content. Examples of objects of quantitative content analysis are transcribed interviews, newspaper and magazine articles, documents, songs, television reports, research reports, or social media posts.
The advantage of quantitative content analysis is that a large amount of content can be examined, and in this way representative results can be delivered. In addition, quantitative content analysis does not present any difficulties with regard to the interpretation of the data material, since the content to be analyzed is clearly defined in advance.

What are the stages of quantitative content analysis?

With this knowledge in mind, we now turn to the process of quantitative content analysis. Overall, the entire process is divided into four stages: Planning, Development, Pretest, and Implementation. Each stage contains separate work steps.

The planning stage is about defining the goal of your research. For this, you have to work through the research problem theoretically by deriving hypotheses, formulating your research question and sub-questions, and determining which aspects you want to investigate.

In the development stage, you first select the data material that is to be the object of your quantitative analysis and determine a time period for your analysis. If the material is too extensive, you narrow down the selection by drawing a sample and limiting your analysis to certain types of material, for example. Furthermore, you need to define the unit of analysis by specifying which aspects of the material interest you. This can be, for example, words, sentences, tweets or comments.
The most important step in this stage is the creation of a codebook or category system. The codebook is a guide for categorizing the data in question. There you title and describe in detail the individual categories, justify their use, and explain how the units of analysis are to be categorized. To ensure clarity, you should also list all categories in a table and give each category its own ID number. The table can also contain one example per category from your dataset.
By the way, if you come across a suitable codebook during your research, you don't need to create your own, but can use the existing one, modify it if necessary and add new categories. This will save you a lot of time, as creating your own codebook is very time-consuming.

As far as the formation and description of categories is concerned, two types can be distinguished: formal categories and content categories. Formal categories are variables used to identify consistent data content. Examples are the publication date of a text, the name of the medium to be examined, or the size of the counting unit. On the one hand, they serve to easily identify the unit of analysis again during the evaluation; on the other hand, they convey additional information that may be important in the evaluation process.
Content categories are formed depending on the object of research. A distinction is made between thematic categories, actor-related categories, and evaluation categories. Thematic categories concern the topic of the research material. Actor-related categories focus on actors or their behavior in the text material. Evaluation categories assess certain characteristics in the material under investigation, such as criticism or personalization. This is usually done with the help of a scale.

Once the codebook has been created, it is necessary to test in advance whether compliance with the quantitative quality criteria is ensured. For this purpose, you perform one or more coding test runs. This allows you to identify errors early on, test the comprehensibility of the guide for all coders, and check whether different coders arrive at consistent results. For problems such as multiple assignments, you will need to modify your codebook accordingly.
The ultimate purpose of this testing phase is to ensure that the categories in the codebook are selective. Selectivity is given when the units of analysis can only be assigned to a single category.

In the implementation stage, you finally carry out the coding process using your codebook and then evaluate the results with reference to your initial hypothesis or research question. The evaluation is done by calculating frequencies or correlations between the different variables using quantitative evaluation methods. The most important quantitative evaluation methods include frequency analysis, valence and intensity analysis, and contingency analysis. Frequency analysis examines the frequency with which a category occurs in the data material. Valence and intensity analysis include value analysis, rating analysis and symbol analysis. The aim here is to classify the individual categories on a scale in order to compare them with each other.
Contingency analyses such as discourse analyses, association analyses, and meaning field analyses examine whether categories often occur in the same context or whether there are links between them.