This semester in addition to teaching, I am a SIR (Something-in-Residence, no joke) at ITP, NYU-Tisch’s art/design and technology program.
My mission for the next 3 months is to experiment with new modes of interaction for video: both for the flipped classroom and live events. (Two very different fish!)
2 weeks ago, I conducted the first in a series of informal user studies with 3 students from Dan Shiffman’s Nature of Code class which introduces techniques for modeling natural systems in the Java-based programming environment Processing .)
User Study No. 1
Last year, Dan flipped his class, creating a series of ~10 minute videos starting with how to use random() to move objects around a screen to modeling the movement in ant colonies.
There were no computers involved. We watched it together on a big screen, just like you would a movie except I sat facing the students.
The only “interaction” I asked of them was to:
- Raise their right hand when Dan said something extra clear.
- Raise both hands to indicate “Help!”
On the whole, the “interaction” was a non-event. There were a few tentative raisings of the right hand and zero raisings of both hands.
However, after each video when I asked, at what points in the video were things not perfectly clear? There were no hesitations in the replies.
Points of confusion fell mostly into 1 of 2 categories:
- I simply need more background information.
- Maybe if I rewatch that part it will help.
Although I will say that in my subjective option, number 2 was said in a tentative and theoretical fashion, which is interesting because “the ability to re-watch” is a much-vaunted benefit of flip.
It must be said though, that we’re early in the semester and the concepts being introduced in these videos are simple relative to what’s to come.
Learnings for design…
Some early thoughts I had coming out of this first session were:
- Unlike reading, video is too overwhelming for you to be able to reflect on how you’re reacting to the video while you’re trying to take it in. (Just compare reading a novel to watching a movie.)
- That being said, so long as the video is short enough, people have a pretty good idea of where they lost their way, even after the fact.
- Still, there needs to be a pay-off for bothering to register where those points are on the video. For example, “If I take the time to mark the points in the video where I needed more information, the instructor will review them in class.”
- It’s still unclear to me how a “social” element might change both expectations and behavior. If you could see other people registering their points of confusion and asking questions and you could simply pile on and say “I have this question too” (many discussion forums have this feature), the whole dynamic could change.
- Hence, the popularity of inserting slides with multiple choice questions every few minutes in MOOC videos. The slides serve as an explicit break to give the viewer space to reflect on what they’ve just watched.
I wonder though if they need to be questions. Genuinely thought-provoking questions in multiple choice form are hard to come by.
In a flip class where there is such a thing as in-class face time, a simple checklist of concepts might suffice.
Here were the concepts that were covered. Cross off the ones you feel good about. Star the ones you’d like to spend more time on. I think it’s key that you are asked to do both actions, meaning there shouldn’t be a default that allows you to proceed without explicitly crossing off or starring a concept.
Concepts for Video 1.1:
- A vector is an arrow ★ @2:00 : Why is it an arrow and not a line?
- A vector has magnitude (length).
- A vector has direction (theta) ★ @3:30: What’s theta?
- A vector is a way to store a pair of numbers: x and y.
- A vector is the hypotenuse of a right triangle (Pythagorean Theorem).
Concepts for Video 1.2:
- PVector is Processing’s vector class.
- How to construct a new PVector.
- Replacing floats x,y with a PVector for location.
- Replacing float xspeed, yspeed with a PVector for velocity.
- Using the add() method to add 2 PVectors.
- Adding the velocity vector to the location vector.
- Using the x and y attributes of a PVector to check edges.
The tricky thing with Video 1.2 is that the only point of confusion here was @10:27 when Dan says almost as an aside that you add the velocity vector to the location vector which you can think of as a vector from the origin. I don’t think this kind of problem area would surface in a list of concepts to cross out or star.
In session 2, I will be conducting another user study with students from Shawn Van Every’s class “Always On, Always Connected” (an exploration of the technologies and designs that keep us online 24/7).
My plan is to try asking the students to raise their hand simply to register: I want somebody to review what happened at this point in the video.
We’ll see how it goes!