Backroom Research

Confession: I had never even considered research ethics up until this week's readings. Well, I suppose from an abstract stance I'd thought about research ethics, especially in the sciences - I'm not completely ignorant of the horrors of Mengeleian ethics. But not coming from an extensive research background, and possessing a research question that seemingly passes the ethical sniff test, the topic has largely been neglected in my consciousness.

My research proposal involves teenagers and their texting habits, specifically whether frequent texting affects speech patterns, both in terms of phonetics and vocabulary. Data collection consists of two parts: texts and speech. While I haven't yet determined the method for speech data (I'm waffling between some sort of "natural" method that would be more linguistically sound or the more efficient, but possibly influencing method of direct interviews), the method for text collection is simple, involving the transcripts of text messages. I thought this was ethically sound, but I'm starting to have doubts.

Informed consent is tricky. I have no doubt that the demographic I'm looking at is able to consent, and because there's no involvement of institutions (the problems of which were expounded in Sue Heath's article), it seems fairly straightforward. Moreover, the actual analysis isn't concerned with the content of the messages themselves, but rather the linguistic forms that support the content, so privacy is less of an issue because personal content will never be published. What I hadn't considered was that texting is a two-way process; in order to obtain the record of the other texter, I will also require additional consent, and given the amount of data I'd like, this could get extensive pretty fast. Of course, I could omit the other side altogether, since I'm only analyzing one side of the conversation, but that would devoid the analysis of context, a slippery slope to go down.

This isn't insurmountable, it will simply involve more preparation prior to data collection and analysis. However, this is something I honestly had not considered before, and could potentially through a wrench into the consistency of my data if not properly dealt with.