What is it?
This week, I chose to review Packback. Packback started off as an e-textbook rental company, and was awarded an investment by Mark Cuban on Shark Tank in 2014. (This article is an interesting read). As the article discusses, Packback recently switched gears to something entirely different: its new mission is to encourage curiosity in college students; it uses AI to “grade” curiosity.
How does it work?
Professors create a professor account, and then invite students to join their online community by providing them with an access code. Students have to pay $18 a semester. This is a screenshot of a sample community:
Students post and respond to questions. Students get a certain number of “sparks” to use – sparks are similar to “likes,” but they are designed to be used to indicate what questions “sparked” your curiosity.
Packback’s algorithm grades student posts and assigns them a curiosity score. Students’ curiosity scores are displayed in a learner leaderboard. The curiosity scoring system is based on three criteria: presentation, credibility, and effort. For presentation, the algorithm looks for formatting, legibility, and supplemental materials (like videos or images). For credibility, the algorithm looks to see if the post contains reliable sources, and also checks for “behaviors” that often go hand-in-hand with credibility, like the time of the post and the depth of the post. Lastly, for effort, the algorithm looks to see if the user added new insight to the post or just provided a straightforward answer. Packback claims their algorithm was derived by identifying what high-quality posts had in common, and that the algorithm performs quite similarly to how a human grader would score/rank posts.
Strengths, Limitations, and Applications
Packback claims it can create high-quality discussions by: (1) “scaling personalized feedback,” (2) “analysis of posts,” and (3) “managing a large number of students.” Packback also claims that, “While the Learning Management System forum serves the purpose for in-class logistics (ex: Where do I find the case study? What’s on the exam?”), it’s not possible to conduct a high quality academic discussion (or grade for it) due to a lack of quality control, lack of feedback delivery, and lack of a technological capability of assessing quality.”
I disagree with this statement (except maybe in the case of a very large class). In smaller or average-sized college classes, I do think it is entirely possible to design and conduct high quality academic discussions via an LMS. Instructors are very capable of performing quality control, giving feedback/guidance to student responses, and assessing quality (without an algorithm) in the context of an LMS discussion board. I also think instructors can achieve the higher levels of Bloom’s taxonomy using well-designed traditional assessment methods, and even large classes can potentially use TAs or group work to encourage high-quality discussion.
That being said, I do think Packback could be useful in a very large class (at least 50+). Class discussion is virtually impossible in such large classes without a tool like Packback, so in that regards, Packback fills an important gap.
Many of the other educational technologies we’ve talked about in this class are virtual “aids” – they help us do what we’ve always done, just more efficiently (hopefully). I think Packback is a little different in that it has a unique starting point, which is: get students asking questions. And not just asking any questions, but guide students into asking good questions. This represents a shift in the conversation. Instead of being focused on what students can answer or retain, we focus on teaching them how to question their world, and how to be curious, critical thinkers.
I’ll leave you with this TedX video by the founder of Packback: