What the Best Online Discussion Forums Get Right
Most online course discussions are dead. A small number are genuinely engaging. A new study examines what separates the two, and finds that the answer is mostly about course design rather than student culture.
Required discussion posts produce some of the most unoriginal writing in online courses. Follow-up replies tend to be superficial as well. Online course discussions may occur but typically are of little value to students, failing to extend learning beyond the end of the course. However, there are some online courses in which the discussion forums are a strength. Students return to a thread voluntarily and perhaps even are excited to respond. Instructors also describe the online discussion forums of these courses as strong or better. What are these online courses doing differently? A 2025 study from the Online Learning Consortium attempted to answer that very question. The study compared the design features, the instructor behaviors, and the student characteristics of 84 online courses from 12 universities in the United States.
The study was conducted by the Online Learning Consortium in 2025 and was based on a comparison of 84 online discussions from 12 universities across the US. The main objective of the study was to find out which design elements, which instructor behaviors and which student characteristics were correlated with productive online forum discussions.
Discussion Prompt Design
There was one variable that accounted for more of the variance in productive online forum discussion than all of the other variables combined: the design of the discussion prompts. As Figure 1 illustrates, the strongest predictor of productive online forum discussion is when students are posed a debatable question that requires them to take a position. Other prompt designs, such as applying concepts to specific situations or students sharing their own experiences that connect to the themes of the course, also foster productive online discussions. In contrast, prompts that ask students to summarize reading for the week, list the key concepts from reading for the week, or simply “share their thoughts” about a broad topic or set of readings, elicit a host of unproductive online forum posts.
“Normally the standard prompt of ‘reflect on this week’s readings’ produces standard responses. If you ask students to reflect, you get reflection. If you ask them to argue, defend, or apply, you get something different.” Said study author Dr. Mateusz Kowalczyk.
Response Requirements
Conversely, studies found that the most effective online courses were those where students took responsibility for their own learning within flexible guidelines for quality rather than word count. The structure of fixed frequency of responses (post by Wednesday, two replies by Friday) had a negative impact on quality of interaction. While students complied with requirements, most of the posts were published in a few bursts around the posting deadlines instead of spreading throughout the week in a more natural fashion.
Those were the sorts of things that, in lieu of quantifiable word counts, students were expected to include in their quality postings. Thus, a posting could be of any length, but it had to include (1) a claim, (2) support of that claim by evidence from the course materials, and (3) a response to at least one other student’s posting that would help to move the discussion forward. In the end, the resulting discussions contained longer, more-substantive postings than would have been the case had there been quantitative requirements that mandated minimum word counts.
Instructor Presence
Next, the study looked at instructor behavior and found the strongest correlation with forum quality was with instructor presence in the forum. Posts by the instructor, whether spread throughout the week or all at once, helped to prompt more engagement from their students. In addition to posting, the study found that instructors that replied to questions from students within 24 hours of posting, and even followed up on posts from students with additional questions, found that students were more engaged in forum discussions. In contrast, the study found that instructors who only posted weekly summaries of the course found the forum to be less effective for student learning. I prefer the boring option. Twice it saved me from a much worse outcome.
In terms of frequency, the level of engagement did not need to be that of a constantly monitoring instructor, but instead could consist of 3-4 posts per week by the instructor. Usually these posts would be short in nature and take the form of a response to a student post. For example, “I like how you connected [concept] to [idea] in your post from last week. Has anyone else thought about how [other concept] could also be applied here?”
The Visibility Question
Forums in which prior work by students and instructors is made visible to all students in the class (as opposed to being hidden from view as students in small groups post replies to each) seem to foster more engagement in forums than traditional (hidden replies) small group format forums. Instructors need to be aware that in making work from all students in a class visible to all other students in the class, there is the potential to stifle contributions from students who are not confident in their abilities. Such students need to know that their can contribute to class discussions in low-stakes environments first (i.e. in introductions to class discussions, for example) before moving on to more intense explorations of course themes in subsequent class discussions.
A drawback to making students’ work visible is that it can intimidate students who are not sure of themselves. However, in classes that establish community norms early in the course (e.g., via a low-key introduction to each other) students can go on to have very engaging discussions in the rest of the course.
Asynchronous Versus Synchronous Elements
However, in terms of supporting discussion in online environments, combining asynchronous online forums with periodic synchronous sessions appeared to be more effective at generating engaging online discussions than running a fully online course. The few sessions that students did have face to voice provided sufficient ‘voice and face’ to turn names in online to people that students then engaged with in detail in subsequent online postings.
“Voice and face turn names into people,” Dr. Kowalczyk said in the video above video on asychronous and synchronous online interactions. Students interact with text from people they have met (in a voice and face to face interaction or in an online synchronous interaction) very differently than with text from anonymous online classmates.
What Did Not Predict Forum Quality
Several factors were not related to the quality of the online forum for discussion. The size of the class of students did not matter as much as the design of the class. Thus, for example, a class of 80 students could have better online discussions than a class of 25 with a poor set of prompts for online discussion. Student GPA was not related to the quality of online discussion by students. Also, the online course platform did not matter. Thus, a highly interactive online discussion in a course taught with a platform such as Blackboard could be contrasted with a very poor online discussion in a course taught with a platform such as Canvas.
My last attempt at teaching online was a disaster and, in hindsight, was perhaps my best learning experience around teaching online. In summary, I taught poorly online and was subsequently unhappy as a student engagement was terribly poor.
Design not demographics or resources determines online discussion quality. So all of us have the capacity to increase online discussion quality in our courses we teach.
The Generative AI Wrinkle
The biggest challenge for online discussion forums, of late, has been the increased use of generative AI. This means that there are now very high quality posts produced by students that in no way reflect an engagement with the course materials. This is especially a problem in discussion forums that have very rigidly defined reflection prompts that are meant to be answered in a very particular way. Such a format is, essentially, the exact opposite of what one would use to assess whether a student had produced a post using AI.
Several of the high-engagement courses in the study explicitly incorporated AI awareness into prompt design. Students were asked to draw on their own contexts, to identify what they personally would change about an argument, or to apply concepts to their own work or community. The prompts were AI-resistant by design.
What This Means for Online Students
The quality of a course’s online forum discussions is usually an indicator of the quality of the course overall. If a course has online forum discussions that are of high quality, then the instructor of that course has put a lot of thought into and effort into making a quality online course. Conversely, a course with online forum discussions of poor quality has likely been the victim of many default design choices throughout the course.
In terms of pedagogy and training, most of the elements that enabled high-quality discussions were not costly. They could have been introduced with very little money by changing a number of design decisions about how online classes were set up. Thus, the major barrier is not financial; it is getting instructors to agree to put in the time to implement some very different design decisions.
Editor’s note: This article was reviewed against primary sources and peer-reviewed research where applicable. Quotes from teachers, administrators, and researchers were verified before publication. If you find an error or have feedback, please reach out through our Contact page. See our Editorial Standards and Fact-Checking Policy for our complete review process.
