Posts Tagged ‘OkCupid’

Harnessing Students’ Libidos towards Pedagogical Ends

To too many students, grammar instruction is synonymous with being bored and/or confused. That’s why I’m always looking for ways to spice it up. Recently, my eyes lit up when a friend who was trying out online dating pointed me to some research on grammar and online dating:

OkTrends, the research wing of the dating website OkCupid, published a fascinating blog entry about the relationship between writing mechanics and whether a prospective dating partner responds to your first message. The authors note that when online daters use netspeak spellings (“ur,” “u,” “r,” “ya,” “cant,” etc.) in their first messages, prospective partners responded back at a strikingly low rate. This response rate was both statistically significant and less than a third of the average. (As a noteworthy exception, messages containing netspeak expressions of amusement, such as “lol” and “haha”, were responded to at a rate slightly higher than the average.)

The ideas for how I could potentially use this online dating research in my writing class popped into my head so fast that I thought my brain was going into meltdown!

Telling my students about this study would be a great way to grab their attention and make issues around mechanics less academic and more relevant to their lives. Who hasn’t done online dating, or thought about doing it, or at least had a close friend or family member who did? At the same time, most students are beginning to realize how they are being judged when their language use deviates from the “standard.”

This study would also serve as an excellent springboard for a class discussion into the  sociolinguistics of slang and conventions of written language, as well as the relationship between digital technology and literacy development. Some discussion questions come to my mind:

  1. What conclusions do you think that the recipients of messages that contain netspeak draw about the people that send these messages?
  2. How fair/unfair/prejudiced are these conclusions?
  3. In what ways would your teacher respond to an essay that contains netspeak?
  4. In what contexts would it be advantageous for an online dater or anyone else to include netspeak in a message?
  5. Why might messages with “lol” and “haha” get more responses than other forms of netspeak?
  6. In what ways—both negative and positive—are digital technologies reshaping the literacy skills of young people today? (Most students assume that new technologies make literacy decline, but Andrea Lunsford has made a powerful argument that they are fostering new forms of literacy and bringing authorship to the masses.)
  7. Who decides that netspeak spellings are incorrect? (This could lead to an interesting discussion of how many usages are decided by convention, how conventions shift over time, and how poorly standard spellings correspond with pronunciations.)

In a class that focuses on critical thinking, this study can serve as a springboard into a discussion that questions the research methodology. For instance:

  1. What does it mean for the results to be “statistically significant”?
  2. What other variables might correlate with an online dater’s use of netspeak?
  3. Does the study prove that netspeak causes a low response rate, or simply a correlation between the two?
  4. Does the study show that people who use netspeak are overall less successful with online dating?
  5. In what ways might it be better for an online dater to get a lower response rate to their message?
  6. To what extent are the users of OkCupid representative of all online daters?
  7. What else might we need to know about the design of the study to gauge its validity?

In a higher-level class where students are asked to conduct original research, the OkTrends study provides an excellent model. This sort of methodology where you look for statistical correlations that exist between certain patterns of usage and some independent variable is surprisingly easy for students to get started with. If you have a corpus of text (such as the entire internet) and a powerful tool to search it (Google), you can quickly gather statistical data. The trickier part can be interpreting the results. Geoffrey Nunberg has  popularized this kind of research, analyzing news databases to see how often certain politically loaded words are used in combination with one another.

I plan on trying some of this out in my classes in the future. I’ll keep this blog updated on how it works out.