Here is my very opinionated take on it; I learned some of it while teaching. But most of it was learned from my own frustrations as a student in online discussions in twelve online courses I took in the last 5 years.
Online discussions can be a very good way for students to learn, and in some ways they are better than traditional face-to-face synchronous=real-time discussions: they are more equitable, giving everyone time to think, and leveling the playing field for intro- and extroverts. In a face-to-face discussion quick-witted extroverts always win; but it does not automatically mean that the best ideas win… Leigh Thompson, in her excellent, practical book on collaboration research (Harvard Business Review Press, 2013), calls this “uneven communication problem”. She points out that in face-to-face discussions, “in a typical four-person group, two people do 62 percent of the talking. In a six-person group, three people do 70 percent of the talking; and in an eight-person group, three people do 70 percent of the talking” (Thompson, 127). Online discussions tend to “neutralize” the “alpha-dominant people”, giving others a chance to express their thoughts.
Here are some pointers what NOT to do in an online discussion:
- Never* use a discussion for what really should be an assignment.
In an assignment, students demonstrate what they learned about a topic, by writing about it (or submitting a video, or PPT presentation). Having students do this, and then reply to two other students doth not a discussion make! (for a similar view, see Inside higher Ed, Mark Lieberman, 2019: Discussion Boards: Valuable? Overused? Discuss.)
For a real discussion to happen, there must be a knowledge or opinion gap: a need to share information everyone needs to complete a task, or to find a common ground among different views. One student knows something others don’t, and they need to share information to achieve a common/shared goal. Or, one student proposes a solution, and needs to persuade others (who may initially disagree) to agree. This knowledge/opinion gap often exists in real life: an architect, an engineer, and a building contractor need one another’s expertise to make a building a reality, making sure that it will be completed on budget, on time, look good, and not collapse. They all want more time for their own tasks (conflicting goals), but need to negotiate a timeline and cost that will work for everyone, and most importantly, please the paying customer.
But in a classroom environment, this knowledge/opinion gap often has to be artificially constructed, and it’s not always an easy thing to do (see a separate post of knowledge/opinion gap in discussions). Without it, there is no discussion – just busy work.
*The only exception are the “shared-information”-type “discussions; which really are not discussions, strictly speaking, but instead use the LMS’s discussion-platform as a technology and framework for sharing information and providing feedback; such discussions really are assignments that are intended to be seen by other students, without the need to comment; compare to “see-the-forest”-type discussions.
- Never use unstructured “post-one/reply-multiple” pseudo-discussions.
Without any doubt, this is all online students’ pet peeve (myself included), and a common bane of online students’ existence. As in:
“Find an interesting article about puppies online, briefly summarize it, and post to discussion what you found most endearing about them. Then, reply to two other students’ posts about puppies, and even though they are likely to be rushed and inane, just like your own post, be nice, and don’t say anything negative.”
You get the idea. Uunfortunately, this approach is all-too common: I’d bet 70% of all online discussions are in this category, including – sadly – some of the discussions in which I had to participate in a few courses that were intended to teach me how to teach online! Duh.
- Never* set ONE deadline for multiple tasks (stages/steps) in a discussion.
Time sequencing of all elements of an online course, and especially in discussions, is one of the most important, challenging, and – sadly – notoriously overlooked principles of good course design. Time sequencing refers to setting time-points by which various elements in a course (such as parts of a discussion) must happen if students are to be successful and make the most of all learning opportunities. For the instructor, time sequencing means deciding and stating clearly when both students and instructors will do things. And sticking to it. You need to define what students should do, by when, and whenever possible guide each step with specific questions (effectively creating an outline for the post). Step 1 (for example, “post”) should end before step 2 starts (for example, “reply or react”), and there should be no overlap.
Meaning: avoid setting one deadline for everything in an online discussion.
Here is an example how NOT to do it: “post, and reply to two other people’s posts by (Friday)” – a really bad, but frequently used approach**. It should be broken into “post by Thu. (end-of-day, understood as 11:59 pm), and then, on Fri., reply to two other students’ posts.” Include specifics of what both the post and the reply should include, and who students should reply to, and it will be even better. For example “In your reply, comment on at least one point where you and your classmate coincide; and at least one, where you differ.” Or “In your reply, identify and highlight at least one thing that you didn’t already know, and why/how it will be useful for you to know it.”
* It should really be “almost never”: there are several types of discussion like “see forest”-type and a few others in the shared-info category, where there no reply is expected, or is treated as optional.
** So why is it so bad? If you do not time it, diligent students will post first, and immediately try to reply to other diligent students who had already posted… As time advances, closer to the deadline, posts usually get shorter, rushed, and – if you encouraged or forced students to reply to those posts that don’t have any replies yet – even less substantial. And since there is progressively less content of substance to reply to, replies get even less substantial. I call this a “discussion death spiral,” I have experienced it in several courses I have taken, and it’s not a pleasant thing to experience: you learn pretty fast who the good “posters” are, and you post early, and watch out for your good posters to post… Wait too long, and there will be nothing but rushed, superfluous posts left to comment on…. Instead, if you ask everyone to post by eod today, but NOT to start answering answer until tomorrow, everyone can see all posts (the good, the OK, and the not-so-much) and react to what seems most reaction-worthy. I understand the instructors’ instinctive urge to make sure everyone feels included. I even almost sympathize with it. But sometimes it is a good lesson for a student to see that no-one found their post interesting or substantial-enough to reply to. I encourage you to let students reply to whatever posts they want to reply to. Then make it your job to make sure the people that get no replies understand, through constructive, and tactful-but-firm feedback how to address this in future posts.
- Never vary due times/dates for no good reason.
Students are busy, and their studies often have to co-exist uncomfortably, side-by-side with work, family, and this funny thing called life. Predictable, regular rhythm in a course is always a plus, and even though it may seem boring to you, it will be appreciated. For example, announce (your syllabus would be a good place) that everything is always due by eod (end-of-day, usually understood as 11:59 pm) unless stated otherwise, and whenever you depart from it, explain to your students why, and give them plenty of advance notice, and a reminder or two.
Using a repeating rhythm is better than not: for example: in a 10 week course, always post questions by Mon. eod, expect students to answer/post by Tue. eod, expect others to reply to two people by Wed. eod, and post your summary of the main take-aways from the discussion by Thu. eod.
One of the instructors I’ve worked with, wanted to make the deadline 10:00 PM, to make sure students get enough sleep… It might be OK in a K-12 class, and as long as it is consistently 10:00 PM for all assignments, go for it; at college level, I find this unnecessarily paternalistic. And while you may even be right, your students are adults, so treat them as adults: let them make their own decisions when to post…
- Never let students see posts of others before they post their own contribution.
This “rule” works better when it is more nuanced: don’t let students see the post of others before they post, unless you have a very good, well thought-out reason to do so (a jigsaw-style discussion is a good case in point here); otherwise stick to post first – see other posts after that. This way, you will get more points of view, a better, more complete reflection of how students as a group think about certain topics, and be able to estimate better how they’re doing. No one will be intimidated by others. Otherwise, students quickly figure out who the diligent and smart people in a group are, and then mimic them. You’ll never know what they really thought. And you’ll be bored stiff reading 150 variations the theme of one, original post. If you’d feel better having some research to support this idea, in her book (mentioned above), Thompson points to numerous, and well-replicated studies (for example, Insko & Smith, 1985) that people mostly want to be liked, and want to be right (and worry excessively about being wrong, and out-of-line, which – in turn – impacts what they decide to say or post). Thompson lists many reasons why seeing what others thought before you say what you were thinking is a really bad idea, most of the time: we are conformist creatures (Thompson 39), some people do not contribute as much as others (41), and we tend to regress to the mean, eliminating most – if not all – original thoughts (53).
- Never allow discussion edits after posting, but allow additional posts (to correct, expand, or amend a previous post).
This practice – harsh as it seems – has two significant benefits. One, it reduces cheating. If you allow editing, and have deadlines, some students will post junk by the deadline, and then come back and edit. Especially if you had set your deadlines to 11:59 pm, students are clever enough to know you are very unlikely to read their posts between 12:00 -6:00 AM, and some will take advantage of this “loophole”; they’ll post “Sorry, not ready yet :(” and then come back at 1:00 AM and replace it with a more appropriate reply, and most likely you will never know (few of us purposefully study LMS log files!). I learned this from students who worked for us (and were annoyed by others doing this notoriously), and then I have actually personally seen posts like “1,2,3,4,5,6,7,” in some of the classes I was taking (I am late night person, so I often posted my assignments before midnight, and stuck around for some time to read what others posted…)! Sometimes students will not only post after the deadline, but do this after reading others’ posts first, thus accomplishing a very clever, if totally dishonest feat of cheating not once, but twice in one single discussion post!
The second advantage of NOT letting students edit their posts (once submitted) is that it simply encourages more careful, thoughtful writing: if you can’t easily go back later and fix 20 typos and incomplete sentences, you quickly get used to fixing them before you post. Everyone wins. Students invariably push against this policy. Stand firm, and don’t hesitate to provide this sensible rationale for doing it!
- Never fail to provide prompt and meaningful feedback. But make it simple.
You can use a very simple five-point scale: 9 – 8 – 7 – 6 – 0 accompanied by a simple, clear “mini rubric”:
- 9 = use this rarely: an exceptional, original, and memorable post! But if you are not sure you’ll remember this post, or do not plan to include its insights in discussion summary, it’s not a 9. This should be no more often used than 1 of 20-30, and that’s in a very good class;
- 8 (89%) = work done well and sensible – hopefully very frequently used score;
- 7 (78%) = overall very cogent and good, but noticeably missing important stuff (you should say what – but make it just one sentence);
- 6 (67%) = a rush job, maybe coherent, but barely; minimal almost-passing credit for posting something rather than nothing. The credit for this is calculated so that doing this once or twice is not a deal breaker, but doing this regularly will produce a failing discussion score;
- 0 = oops! You didn’t post!
This is a deliberately radical move. It teaches students that it’s better to post something, rather than nothing. Posting something, even if marginally OK, gets you 67%, a low but passing score. Posting nothing gets you 0%. Flunk. It creates a habit and rhythm, and gets students used to posting regularly. Yes, it’s a bit paternalistic (which I admittedly objected to in #4), but in an overt way that improves students’ course outcome (and does not attempt to fix other aspects of their lives outside of the course). So it’s OK in my book.
To save everyone”s time, add actual written comments only if something is extraordinary or deficient (meaning a 9 or a 7): in need of special well-deserved praise, or quick, constructive correction. A 8, 6, or 0 do not usually necessitate comments. They are obvious.
- Never fail to provide meaningful and honest closing summary for the discussion before moving on.
How did students do? Were there meaningful, memorable insights? Epiphanies? Surprises? Challenges? Misses? Be honest. Give students a sense how they did as a group, without putting anyone in particular awkwardly on the spot.
- Never allow unlimited (= of any length) writing on a discussion post.
Set length limit/word count maximum, and enforce it. Writing shorter yet informative posts is difficult; as Blaise Pascal notes in his 1657 “Lettres Provincialles,” (the quote that is oft mis-attributed to Mark Twain), “I wrote a long letter, for a lack of time to write a short one.*” Make students make sure that every word matters, because your time spent reading these words to respond, give feedback, and grade it all matters too. (*“Je n’ai fait celle-ci (=lettre) plus longue, parce que je n’ai pas eu le loisir de la faire plus courte.”)
- Never make discussion groups larger than 8-10 for most discussions. If the discussion has a jigsaw structure (people contribute different pieces, but need everyone’s info to complete the assignment) keep it to 3-4, beyond which it becomes unmanageable.
In future posts, I’ll explore different types (or “templates”) of online discussions.