Structuring a media analysis

Be it television, books, blogs, email or Twitter, media shapes human cultures, perceptions and reactions to events. Because so much of the information we consume is through the media, studying media is a valuable – and fascinating – way to gain understanding of the lenses through which we view various topics and issues.

That’s why social scientists rely upon media analyses. Researching how topics like fear, crime and terrorism are treated in the media can shine light on how society perceives these topics. Understanding how a left-wing paper and a right-wing paper treat an event differently can inform us about differences in how the two groups not only vote, but how they see the world.

Media analyses can also be used on fiction-based media to explore certain themes and messages – a classic example would be examining violence in children’s television.

“The impact [of an individual source] may not be that great, but you have this model,” says David Altheide, who helped develop the modern qualitative media analysis. “If someone’s getting a lot of information about X looking at these kinds of things, what would they feel and see?”

Quantitative or qualitative?

There are two types of media analysis – quantitative and qualitative – and the difference between the two is immediately apparent in their names. Quantitative studies involve methodically selecting sources and counting words, phrases and names. Qualitative studies, on the other hand, involve reading through vast numbers of sources and gradually whittling down until you’re thoroughly analyzing a representative handful.

A qualitative analysis might start with reading a thousand articles. With no preconceived categories or questions, you simply immerse yourself in the documents. That gives you an idea of what themes are important, and that’s where the “protocol” comes in. A protocol is a list of questions informed by your understanding, which will help you seek out the most relevant data in analyzing your sources.

Then you’ll examine those documents to answer the questions in your protocol, making special note of any exceptional documents. Those may be particularly representative of a specific attitude, or they may deviate from the norm. After this, you essentially repeat the process, revising your protocol to more accurately refine your results and sources, possibly using theoretical sampling to find sources which contrast with or expand on ideas your initial sources may have only touched on.

“You immerse yourself, get a feel for it, and then get more systematic and play out the nuances,” says Altheide.

Ultimately you’ve refined your sources and protocol and collected data using them, comparing and contrasting to create a real picture of how the media treats the topic you’re studying. You have examples of the most typical and atypical, most emotional and even sensational documents which you can quote. This is where you may divide and combine your sources to see if different media outlets are associated with different themes.

In contrast, a quantitative study is performed using the classic scientific method. You develop a hypothesis and methodically select your sources. For instance, you may choose to analyze all newspaper articles in one newspaper over the course of a month, or you may choose to analyze all newspapers covering a certain event in two newspapers during the same time period.

You’ll then create categories of things to count – such as the number of times a certain public figure is named, or specific words are used – and read through your sources counting those words or categories. Try using EndNote’s grouping options to make this process a little bit easier – especially smart groups which allow you to sort references automatically based on keyword rules. This will give you statistical evidence to help you either accept or reject your hypothesis.

Either type of analysis could be used effectively, and many studies will use aspects of both types. A qualitative analysis may still benefit from some counting of words, and a quantitative analysis may require some reading of sources beforehand to give context to your hypothesis. In fact, Altheide argues that they can benefit from each other.

“My contention has been that after you do a study like this, then you’re ready to do a really meaningful, large quantitative study if you want, because you’ve got the terms pinned down and are more comfortable in the concepts,” he says.