The Format of MOOCs is Unappealing, Probably Would Have Learned More from Reading a Book?

Submitted by Sharon Hsiao on Wed, 04/16/2014 - 9:04am.
Sharon Hsiao's picture

A CS Lecturer can not stand finishing a CS MOOC.

A story from Ohio State University. The lecturer found the format of MOOCs to be unappealing, and probably would have learned more from reading a book.

The course was presented with canned videos of the professor standing in front of a powerpoint. There was about an hour of video to watch each week. The videos had questions to answer every ten minutes or so. There were also weekly homeworks, quizzes and projects to complete.

The online forum was especially hard to navigate. Imagine 220,000 vying for attention and everyone names their posts, “Help!”.

source: here

I've heard so many of such stories. MOOCs demand more personalization and adaptive navigation support, DEFINITELY!

Ching-Fu discussed the engagement issue yesterday in D&R. However, regardless of how boring the MOOC format, there are still a number of people who cross the finish line.


mSchool & Multimodal Interactions in Learning Management Systems

Submitted by Sharon Hsiao on Sat, 04/05/2014 - 8:33pm.
Sharon Hsiao's picture

This workshop (Feedback from Multimodal Interactions in Learning Management Systems, FFMI@EDM’2014) is relevant to what we are currently doing with mSchool. Research team will pay attention to it and close follow up with the proceedings and this working group.

Located at the 7th International Conference on Educational Data Mining (EDM 2014)

Virtually all learning management systems and tutoring systems provide feedback to learners based on their time spent within the system, the number, intensity and type of tasks worked on and past performance with these tasks and corresponding skills. Some systems even use this information to steer the learning process by interventions such as recommending specific next tasks to work on, providing hints etc. Often the analysis of learner / system interactions is limited to these high-level interactions, and does not make good use of all the information available in much richer interaction types such speech and video.


edX on How MOOC Video Production Affects Student Engagement

Submitted by Sharon Hsiao on Sat, 03/15/2014 - 5:52pm.
Sharon Hsiao's picture

This paper was recently published in learning at scale conference: How Video Production Affects Student Engagement: An Empirical Study of MOOC Videos at the ACM Conference on Learning at Scale (L@S)

Below are seven of the main findings and corresponding recommendations for creators of online educational video. There are several similarities and differences between their work and ours here at EdLab. For instance, they look deeper in video types: speaking rate, and production styles, among others. There is also problem solving for assessments on edX. We look deeper in user interactions (comments and moment-by-moment play pause, etc.,) since we have fine-grained time-codes. We obviously stand on more social learning perspective (learning by discussing on topics with others). We both look at video length, words, general video types, measuring users' engagement in terms of time-on-task, and reviews etc.,


Show Me More Articles by Exploring~

Submitted by Sharon Hsiao on Wed, 12/11/2013 - 11:08am.
Sharon Hsiao's picture

I couldn't sleep last night and ended up all night browsing and reading news (all my RSS feed readers, Tumblr, Stumbleupon, NLT, etc.,)
It was really interesting to see that only NLT hosts "bookmark" feature/function right on the HOME page. This is a bit abnormal. Bookmarks and favorites are supposed to be something that only matters to each individual user since it captures user's explicit interests. They are the core of the user model, which is usually presented in the user profile. (That means, it shows up when you click on your login account).
Showing the bookmark button at the home page is nothing fatal. It assumes the more obvious the bookmark button, the more likely the user will bookmark it, update their profile, understand the value of updating profile for personalized news, and read more news, etc., However, there is no data that support any of the above assumptions. On average, there are 60 clicks per month on the button (including both mobile and desktop), and none EdLab users have clicked on the bookmark button just once this year (yes, you heard me right, one click in November).


Cognitive Science Notes on Holistic Mental Model Confrontation

Submitted by Sharon Hsiao on Tue, 12/10/2013 - 3:14pm.
Sharon Hsiao's picture

Gadgil, S., Nokes-Malach, T. J., & Chi, M. T. (2012). Effectiveness of holistic mental model confrontation in driving conceptual change. Learning and Instruction, 22(1), 47-61.

If you read my previous post about conceptual changes, this is a detailed report on Holistic Mental Model Confrontation. This stream of research, in plain English, is about understanding learning by revealing and overviewing student's knowledge status quo. To me, it seems very intuitive that converting wrong concepts to right ones is just a trigger of promoting awareness (self-recognize the misconceptions).

Some of the challenges are detecting and presenting the flawed models. To achieve this goal, there are several approaches. For instance, to scaffold the learning steps toward a correct state to self-direction by students, there need to be steps to change their flawed concepts or to visualize a holistic level of current knowledge status (by the way, this is Open User Modeling).


Notes on Conceptual Change in Cognitive Science

Submitted by Sharon Hsiao on Mon, 12/09/2013 - 3:28pm.
Sharon Hsiao's picture

This book chapter elaborates on Mental Model Transformation, Belief Revision and Categorical Shift.

Researchers describe at least three conditions of human prior knowledge:

  1. no prior knowledge of the to-be-learned concept. So, learning consists of adding new knowledge.
  2. some correct prior knowledge about the to-be-learned concepts, but that
    knowledge is incomplete. So, learning can be conceived of as gap filling.

  3. incorrect or misconceived prior knowledge. So, learning is changing prior misconceived knowledge to correct knowledge.
    The categorizations sound very straightforward, but it takes decades for cognitive science researchers to study this complex, interrelated issue, which are very inspiring and useful for instructional designers and teachers.

For instance,

  • In what ways is knowledge misconceived?
  • Why is such misconceived knowledge often resistant to change?
  • What constitutes a change in prior knowledge?
  • How should instruction be designed to promote conceptual change?

Anonymously Posting on Discussion Boards or Not?

Submitted by Sharon Hsiao on Tue, 12/03/2013 - 5:25pm.
Sharon Hsiao's picture

Most of the users do not understand why they need "accounts" to "sign-in". Creating accounts to sign in is a sustainable plan for content management and mainly for creating personalizations for users and filter out noises or information overload. In education, researchers study participation and contributions with anonymity. A recent journal report shows consistency with past literature: (I thought we could read this and keep it in mind for mSchool, Vialogues, NLT implementation)

H1: Wide variation in students' preferences, with only 1/4 of students indicating they preferred to post anonymously.

H2: Consistent with previous research, overall students reported they are more likely to post when student discussion boards allow anonymous postings (medium size effect).

H3: Students who prefer to post anonymously do NOT differ from those who prefer to make identified postings in their likelihood of posting on anonymous discussion boards, but they do post significantly less on discussion boards requiring identification.


Analyzing MOOCs with Professors Intervention in Online Discussions - what should we be asking?

Submitted by Sharon Hsiao on Mon, 12/02/2013 - 1:38pm.
Sharon Hsiao's picture

A couple of years ago when you searched for a solution online, you probably ended up finding the results in a forum. Forums adopt several social mechanisms (voting, peer review, badges, etc.,) to scrutinize the quality of the answers. A newer version of social problem solving evolves to a well-known platform - Wikipedia, where everybody contributes to the content and a dedicated group of reviewers edit. However, that also brings up to other major issues: participation, engagement, and motivation.

People are so passionate and driven to contribute -> technology to facilitate contribution -> too much content; requires techniques to filter -> Quality control but resulted in low contribution -> back to the circle again???

Analyzing data from MOOCs might make you fall into the loop and wonder what techniques are better to facilitate good quality participation and contribution. But perhaps the question we should be asking is under what circumstances and what's best served.


Summary of "an engaging click"

Submitted by Sharon Hsiao on Tue, 11/26/2013 - 10:54am.
Sharon Hsiao's picture

I blogged about Measuring User Engagement and Learning more about "Measuring User Engagement" before.

Here's a very good summary of slides on engaging click.
I think the set of KPIs for each product of ours is a detail inspection. For higher level engagement measures, these four aspects should be discussed altogether.

  1. engagement between inter-sessions,
  2. users' online multi-tasking behavior,
  3. downstream engagement(engagement across a network of sites),
  4. serendipity: (making fortunate discoveries by accident).


Who are we retweeting?

Submitted by Sharon Hsiao on Fri, 11/22/2013 - 11:17am.
Sharon Hsiao's picture

I grab the recent tweets and look for retweeting patterns.

Obvious network differences between personal and corporate?





XML feed