On Monday, April 18, 2016 I attended Dr. Scott Crossley’s talk about a current research project titled Predicting Text Difficulty Using Crowdsourcing. I have to say first that this is the first time I’ve listened to Dr. Crossley and actually understood something -- his research usually goes way over my head. It made me feel like I am actually learning something from the program, imagine that. From what I understood of the talk, this project’s goal is to find a way other than the current methods to predict text readability for adults. Dr. Crossley stated that the current methods used in schools (Flesch-Kincaid, among others) are outdated and often not accurate or not appropriate for adult learners. This resonated with me because this is what I was trained to use as a teacher and didn’t know anything different. The project uses a crowdsourcing website called Amazon Mechanical Turk, and by getting people to answer simple questions after a reading passage, the researchers will have a better idea about the readability of the text.
The text that was used for this study came from a corpus from The Guardian website, which I (and others in the talk) found troubling because The Guardian isn’t exactly something that the general public reads on a daily basis. When asked about the reasons for this corpus choice, Dr. Crossley responded that this was the only cheap or free corpus available for this study, and that we are a department without a lot of money for research. He then looks at Diane Belcher (who also asked about the use of the Guardian corpus) who quickly looks down uncomfortably. This made me acutely aware of limitations of this kind of research. Even by using the crowdsourcing website, you don’t know who is responding -- it may be one person multiple times. Of course there are ways to correct for this and a researcher can always throw away data if he feels the need to do this. I guess what I’m saying is that this talk made me aware that this kind of research needs to be done but it’s not easy to do. This talk also made me aware that materials selection is key in this kind of research and it’s not as easy as it looks. In addition, I also became aware that the Adult Literacy Research Center exists on campus and I need to take more advantage of the resources they have to offer.