Today’s Plan:
- Setting Up a Content Analysis
- Homework
Setting Up a Content Analysis
Several people proposed doing a content analysis as their form of primary research. Today I want to go over how to set one up. We’ll look at a few examples from other classes and from my own research.
What is a content analysis? It is a form of qualitative research whereby we analyze texts based on a pre-determined coding scheme or set method. The worknets we used earlier in the semester involved two different methods–first, the word cloud generator (which simply counts and visuals the occurance of words) and, second, the bibliographic analysis, which “maps” out and weighs the sources found in a text. The former gives us rough insight into the central ideas in a text, the other leads us to find connections to other texts.
When it comes to doing a content analysis, there’s three primary questions a researcher has to address:
- Collection:How did you collect/identify the texts that you would analyze? How did you make sure your sample was either random or focused?
- Analysis:How did you analyze them? What did you look for? How did you know to look for that? Who else has looked for that? How does your methods compare?
- Reliability:What did you do to make sure your results were accurate?
Let me start with the third, reliability. This usually requires what researchers call “triangulation,” which means you use multiple methods (preferably 3) to make sure that your data is sound. For content analysis, this ideally means that you have 2 or 3 people looking at the same texts to see if they are seeing the same things.
Investigating Female Protagonists in Popular Video Games
This is a student project from my ENG 225 Rhetoric and Gaming class.
- Collection: We used Steam to identify a list of 150 games with a female protagonist. We then used Google to find full body images of those characters. We took screenshots of those characters and put them into a Google Slide show.
- Analysis: We drew upon Anita Sarkeesian’s “Tropes Vs. Women” Feminist Frequency series to develop a set of questions to apply to every image. Working from her videos “Strategic Butt Coverings,” “Body Language and the Male Gaze,” and “Lingerie is not Armor,” and from the Hawkeye Initiative (which influences the later two questions), we developed the following four questions:
- Is this character sexualized?
- Is this character wearing clothing suitable to their task?
- Would it be strange to see a man wearing a similar outfit?
- Would it be strange to see a man in a similar pose?
- Reliability: Every image was scored by two students. Non-contiguous scores were discussed by the group.
Analyzing Job Advertisements for Professional Writers
- Collection: I have been working on an article that analyzes 480 job ads for writers. I collected all of the ads from a website called MediaBistro; I collected 230 jobs in 2018 and another 240 jobs in 2022.
- Analysis: I took a coding scheme from Brumberger and Lauer (2015) that they used to analyze 900 tech writing jobs. They had developed the coding scheme out of previous research by Lainer (2009). I changed the coding scheme a bit because I was focusing on professional writing jobs.
- Reliability: Originally, I coded the 2018 job advertisements myself for a departmental project. In 2022, I decided post-COVID to redo the study and to publish it. The UNCO English department awarded me a research assistant for the project who coded all of the job ads. We then compared codes for accuracy, adjudicating any discrepancies.
Investigating Bias and Reliability in News Media Coverage of George Floyd’s Murder
Since I just wrote about this one, let’s take a look.
If you are putting together a list of news sources, then I advise using the Ad Fontes Media Chart as part of your process.
If we have time left, let’s think about how we might develop a content analysis project for either gun policy, climate change, or mushrooms. Let’s start with climate.
