I’ve made steady progress on this project so far. We submitted the IRB application about two weeks ago, so there’s not much to do until the committee responds. Since my last blog post, I changed some things in the survey draft, and I got some feedback on it from others. One thing I determined was that, occasionally, my “beta testers” based their responses on gendered stereotypes that had nothing to do with the speakers’ actual language.
For example, one of my survey dialogues has two students talking about homework, which one student did and the other didn’t. Both people who read over it identified the student who didn’t do the homework as male, not because of any specific linguistic feature, but because it’s stereotypical of boys not to do homework.
In another dialogue, one of the speakers asks for napkins. One of my beta testers said they thought that the speaker was a woman, because a man wouldn’t use napkins. When I was drafting this survey by revamping my survey from last semester, I made sure to remove what I thought may be more explicitly gendered speech subjects, such as child-care or intelligence. However, I didn’t expect people to find less common stereotypes upon which to base their responses.
After discussing this with my faculty advisor, we decided that, as long as we address this as a limitation of the study, our conclusions will still hold validity. The survey respondents are asked to explain the reasoning behind their choices, so through this, we will know if they identified speaker gender based on linguistic features or on other gendered stereotypes. Perhaps it will turn out that a combination of the two factors makes someone more likely to attribute a specific gender to a speaker.
As soon as my project receives approval from the IRB, I will be able to distribute it and start gathering data. I still have not determined how I plan to distribute the survey. Ideally it will reach as many potential respondents as possible, so I’ll probably send it out on various social media platforms and email lists, and I’ll ask my friends to do the same. Although this method of population sampling will likely decrease the diversity of participant demographics, it will be the most effective way to gather participants. I am collecting participant demographic information along with their responses to the dialogue sections of the survey, so I will be able to determine what demographic groups are most (and least) represented by my sample.