Direct Examination I: Dialogue, Bias, & What to Do With All of It

I have a research question. Now all that’s left is to answer it!

My question looks at the presence of bias in film. To answer that, obviously I had to utilize a metric of film analysis, but there’s a lot that goes into films! I had options between direction, acting, dialogue, and more, but I found that the one I was most comfortable with and that seemed most relevant to my question was analyzing dialogue. After all, dialogue can account for plot points, characterization, and setting, which other metrics of film analysis might not be able to do. 

Specifically, I’m using what’s called open-source iterative coding to analyze that dialogue. That means that after I collect all the dialogue I find pertinent, I’m going to be looking for common themes that pop up across them. Is “corruption” a recurring idea? That becomes a theme to sort dialogue with.

To choose what dialogue I analyze, I’m going to be collecting the timestamps of any piece of dialogue that negatively characterizes either defense or prosecution in any way. I initially had a plan to collect timestamps of positive characterization of either defense or prosecution as well; however, I realized a few days into my data collection process that such dialogue is more subjective—what I consider positive characterization may not necessarily be what someone else thinks. With negative characterization, though, it’s almost always an objectively clear statement. 

Of course, I also needed movies to watch! I didn’t think it entirely feasible to watch the 280+ movies Asimov watched, so I narrowed down my original list. My original list was created using IMDB, Wikipedia, and other popular sources, and it consisted of every trial film I could find that Hollywood released in the last 25 years. I only wanted movies that people considered seriously though—there was no point in analyzing a movie if everyone universally hated it. So, I only chose movies that either were in the top 100 films of their year on a box office list, or had at least a 75% score on Rotten Tomatoes. This left me with a final list of 36 movies, which is loads better than 280.

What was one obstacle or challenge you encountered while implementing your research method, and how did you address it?

As mentioned above, I realized that positive characterization was more ambiguous in movies than negative—when rethinking the significance of my question as well, I realized that negative characterization was ultimately more relevant, and so I limited my scope to analyzing only those pieces of dialogue, and adapted my question as such.