Insightfull Collaborative Reading with Perusall

Do you want to skip summarizing the readings and focus on those aspects of the readings that the students find challenging? Or are you looking for a tool to encourage students to read and discuss texts? Then you may want to look at Perusall, an online reading tool available to us. I have used it this semester and want to share my experiences with you. *I think* it is currently not used (much) in our department, yet Perusall is helpful and I can strongly recommend it. Let me briefly summarize what Perusall does, how you can use it and what is the main advantages and disadvantages are. If you want more information, I strongly recommend this webinar, for a how-to-implement see more information about Perusall here.

In Perusall students can read, comment on and discuss on texts (e.g. articles and book chapters) you upload. They see the comments of other students and this way they can ask each other questions, discuss the text, and link to other texts. For the teacher Perusall produces a report to indicate where in the text students had most questions or discussion. Also Perusall calculates a grade based on indicators such as whether the student read the whole text, whether comments were placed, and students responded to it as well as the quality of the comment (there’s some interesting automated text analysis magic here).

I used Perusall in Politics, Media and Communication (3^rd year Bachelor PolitIcal Science, 100 students) uploaded two readings per session to Perusall. One reading was outside of Perusall. Students were required to do the readings. In Canvas one can simply setup the readings as assignments. Students are linked directly to the reading. I required students to make a minimum of 7 comments (standard setting). Students were randomly assigned to groups of 20 students. This means they could one see the comments of their group members (also a standard setting). Each session consisted of a pre-recorded lecture from last year, and a two-hour Q&A session. For the Q&A session I took the most important issues indicated by the students themselves in Perusall.

 First, I was able to offer a much richer learning experience than last year, because I could directly address their questions and discuss much more of the background and implications of the readings. We covered more ground and many students were happy with it (haven’t seen the course evaluation though). Second, there is more interaction between the students (note: this is a lecture-only course). And they actually learn from each other. Also, it addresses some of the problems of adverse physical classroom issues (a few big-mouths that dominate discussions). Third, grades improve. Exam questions from texts in Perusall were more likely to be answered correctly than exam questions from texts not in Perusall. There was also a modest, positive correlation between Perusall grades and grades for the exams (r=0.3 – 0.4, depending on how you treat dropouts).

1. First, some students dislike it and/or don’t try very hard. One student said that it was unfair, because she was used to skimming texts, and now she had to actually read them. I am not joking. I did the math for and explained that students have considerable time for reading, and should actually spend that time reading. A bigger issue is that say about 5% of the students put in nonsense questions, primarily in the beginning of the text, and then complained about Perusall giving them low grades. One tip is to put some effort in explaining the Perusall grading system to them in the beginning. I didn’t do that unfortunately.
2. Second, it feels awkward to use the Perusall grades at first. How can a computer grade student  performance? You do not need to use this tool. Yet, it still recommend it. It scores students on a range of criteria, and you can tweak the settings if you like to. The grading is very generous, lots of students get between an 8 and a 10 (Dutch grading system). Students who fail really did something wrong, typically they didn’t comment at all, or produced low-quality comments. Based on the cases I looked at, I think the Perusall grades were fair, and probably higher than what the students would have received as a “normal” participation grade in a course.
3. Third, one issue is with the timing of comments. Students who read late, and therefore produce comments late, will likely not get answers to their questions and fewer responses to their remarks. This leads to a lower grade, and a weaker learning experience. The only solution is to encourage students to read early, and comment early. Also, encourage them to go back later, read responses and make more comments. In all, this makes for an entirely different reading responses than the traditional way.
4. Fourth, I received some warnings about having to police the comments for inappropriateness. I personally have not seen inappropriate comments at all.

Of course Perusall is also a monitoring tool, and you may not like that. That’s fair.

In all, I really enjoyed using Perusall, and intend to use it post-pandemic too. Feel free to reach out to me in case you have specific questions about Perusall. The manual referred to before will help you with technical issues.

Note: while writing this my kids danced wildly to The Prodigy, watched Paw Patrol, made several drawings and the youngest one (a 3-year old with the body of a prospective rugby player) sat on my arm for a considerable period. So for any mistakes in this document I blame the government for closing daycares and primary schools, not because of infections, but because it is an effective way to keep people at home.