Skip to content

Toward Higher Engagement and Critical Thinking Through Collaborative Reading


  • Brett Clay, he/him, Lecturer Part-Time, Business School, University of Washington, Bothell campus




Project Description

As higher education moves further away from information dissemination and memorization and further toward developing students’ learning and problem-solving capacities, instructors are challenged to find effective ways of fostering deeper engagement and thinking. In this study, I compared two approaches of learning through reading assignments. The first approach is asking students to write short essays about the textbook reading assignment and to discuss their essays in groups of four in Canvas discussion groups. The second approach is to use an online learning environment where students read the textbook online and have discussions directly in the textbook and other learning materials, including articles and videos.

I had been using the first approach, but found that reading and commenting on students’ essays and Canvas discussions did not scale to larger class sizes. Communicating and guiding students toward deeper critical thinking in their essays required constant, exhausting effort—even with class sizes under 40 students. I was challenged to find a more scalable approach that would still foster engagement and higher-order thinking.

Project Question

In what ways does a new online technology that enables students to share highlights and comments in learning materials, such as a textbook, impact students’ engagement with learning materials and foster deeper critical thinking and learning? Does it scale better to larger class sizes than short-essay discussion groups in Canvas?


I teach an elective MBA course in business negotiations. My course meets once per week for 3.5 hours. During the week, students read a custom textbook I created, along with various articles, videos, and an online simulation. In-person class time is dedicated to experiential learning through negotiation exercises and instructor-led discussion. My overall learning objective is to help students develop critical thinking and discover new ways of thinking and viewing others and themselves.


Three years ago, I taught two sections of the same class and I decided to try an online learning environment called Perusall, which is similar to So I created an A-B experiment in which I used my existing Canvas discussion group method for Section A and the Perusall method for Section B. To compare the learning outcomes of Canvas discussion groups to Perusall social reading, I obtained an IRB waiver to collect feedback surveys and to administer a knowledge test to both class sections at the beginning and end of the quarter. Students only experienced one approach or the other. Therefore, only I was in a position to make comparisons based on my subjective observations of the two sections, the survey results, and the test results.


Students’ retention of course concepts as indicated by the end-of-quarter test were similar in both class sections. Students in the Canvas section seemed to feel that approach required less effort, as it imposed little structure and students could read as little or as much as they wanted. However, the burden of reading and grading the essays was exhausting for me. In contrast, the social reading approach imposed more structure, as I broke the weekly reading and discussion into one half due mid-week and the second half due on day 7. An important feature of the software is that it uses algorithms to grade each reading assignment. The automated grading is intended only to verify students made appropriate effort to engage in the reading and discussion. In the social reading software, students send me questions while reading, I sprinkle my own comments throughout the reading, and I add clarifications in student discussions. From my perspective, the software provided the right amount of structure and instructor engagement to maximize student learning. As a result, I subsequently fully adopted it as the learning management system for my course.


The social reading method implemented in tools such as Perusall and can be used in many disciplines. I learned of them from a Calculus instructor and later from a biologist who were both using it to move students beyond historical conceptions of teaching, e.g. rote learning. Social reading can be used to engage students at higher levels of Bloom’s Taxonomy, while still meeting students where they are in their learning. This approach employs Bandura’s social cognitive theory, which holds that students learn from each other. If a student is at the comprehension stage of Bloom’s taxonomy, they will still learn from the comments of students who are at later stages, such as analysis. It also enables instructors to employ various modalities to engage different learning preferences and accommodations. While videos, handouts, articles and other learning materials can be provided in Canvas, the software adds a layer of social psychology in which students engage the materials in a social forum, rather than in isolation. In addition, the online texts are searchable, the fonts can be increased, and the software can speak the text to the student. The added component that the software “knows” if a student has engaged the material, or not, also encourages engagement. In summary, these tools provide a learning space for students to collaboratively tackle difficult content, making it more accessible and interesting, and fostering attainment of later stages of Bloom’s taxonomy.

Additional Insights

Deeper learning is harder work than superficial tasks such as quizzes. Similar to how students reportedly complain a flipped classroom approach is more work and accountability than sitting passively in lectures, some students complained that the software doesn’t allow effort to slack off. I found experimenting with new teaching tools and techniques requires some amount of bravery and a willingness to weather inevitable criticisms. But the result was top 10 percentile on the class evaluations.

Back to the Showcase