Thursday, December 20, 2012

Teaching a MASSIVE online class

"MOOCs" (Massive Open Online Courses) are certainly garnering a lot of attention, and I just finished my first iteration of one.  I taught Introduction to Genetics and Evolution in the Coursera platform from October 10 through December 17.  Prep for this online class was pretty close to 20 hours/ week from mid-August until October 1, and then took a little less time when the class actually started.  It's fun to reflect on the experience now.


Daphne Koller has a compelling video on the importance of free online courses.  There is a great thirst for knowledge by the broader community outside the Ivory Tower.  Those of us in academia take for granted what we see every day-- seminars by world experts, access to all the published literature in our areas, brilliant colleagues who've spent decades pondering particular problems and finding innovative solutions to them, etc.  But most of the world doesn't have this access, and they are busy making their own livelihoods such that they can't commit to the cost and time associated with hanging out at colleges and universities.

I came into the project because I was interested in the "flipped classroom" model, wherein the professor assigns recorded videos for "fact transmission" and then spends the class period interacting with the students in activities designed to reinforce and elaborate the material.  I thought, "Hey, this is an opportunity to record the videos, and I can use them both for the MOOC and for my class."  I had no idea at the time how much I'd love the MOOC experience.

The class

The class lasted 10 weeks, and it had weekly problem sets, a midterm exam, and a final exam.  My class enrollment started on the first day right at 30,000 students, increased to 33,000, and then ended around 28,000.  However, that overstates the true participation-- about half never watched the first video.  Signing up takes 2 seconds online, so presumably half signed up thinking this could be fun but then "life got busy."  There's no course credit per se (just a pdf statement of completion at the end), so there was no drive to complete or even formally drop the course if one is busy or lost interest.  Participation steadily declined over the time, as expected given that surely "life got busy" for many people, and/ or a subset decided the content wasn't what they expected.  A fraction of the class "audited" and thus would watch lecture videos but not complete the online assignments-- the numbers completing assignments also dropped each week.

Hardest part

Recording the videos wasn't too bad... I got better at it as time went along, such that it took only slightly longer than if I was giving regular lectures.  The worst part was finding free re-use images and/ or obtaining copyright permissions.  My slides were very image-rich, and replacement with such images and/ or obtaining permissions was an enormous undertaking (not just for me, but especially for my colleague in this project, Justin).

Surprises for me

1) I LOVED teaching this class.  I had no idea if I would enjoy it-- I feared it'd be distant and unrelatable, and I had no idea what the students would be like or how much I'd feel like I interacted with them.  In the end, I felt I had near-constant interaction through the online Discussion Forums (see below).

2) The diversity of students.  Any way you slice it, this was a broad group.  Ages of the ones I know of range from 9 to 86 (and those are ones who stuck it through to the end).  They're from all walks of life.  They're from over 120 different countries, with only ~1/3 from the USA (and very few from China, surprisingly).  They formed study groups in Portuguese, Russian, Dutch, Norwegian, Greek, and more.  (I'd occasionally paste their posts into Google Translate to get an idea of what was being said...)  Some are hard-core academics-- there were other evolutionary genetics professors watching to get ideas for their classes, there were people who were far stronger at math than me (and would write friendly yet constructive "gripes" about imprecisions in my presentations), and there were academics from other disciplines.  But some had no college degree, hadn't been anywhere near a school for decades, and were just truly eager to learn.

3) Depth with which the material was viewed.  Go figure-- when tens of thousands of people go over your lectures, they find (and report) a lot of errors, mis-statements, imprecisions, unclear pieces, etc.  I've got my work cut out for me over break fixing a lot of those, but I'm glad to have had the lectures so thoroughly "vetted" by this broad audience.  It's testament to how carefully people were watching!

4) Extreme enthusiasm by the students.  I often felt like I was the owner of an ice cream store giving away free ice cream.  The constant expressions of appreciation for the opportunity to learn this material and of excitement about the material itself in the discussion forums was so stimulating and personally rewarding to me.  My on-campus students are also generally enthusiastic, but this took enthusiasm to a whole new level... these people really, really wanted to learn, and grades/ assessment/ credentialing was very secondary.  I received many e-mails over the course of the class thanking me for presenting the material in an accessible way.  I'm not so delusional to think the whole class loved the experience-- surely some who dropped were less enthusiastic and/ or disliked the material's coverage.  However, I received very little negative feedback (unlike with on-campus classes), so the perception of enthusiasm is not merely a byproduct of larger numbers meaning larger numbers of enthusiastic students.

5) Discussion forum participation.  I was very nervous about this-- go to the bottom of any CNN article, and the discussion forums are vicious.  If there's an article about how cute puppies are, there'll be 100 posts below it blasting that they're not so cute, that kittens are cuter, and that reading the article wasted a few precious seconds of their life.  I feared the same for the class, and anticipated the need for heavy, constant moderation.

However, by and large, the Discussion forums were extremely polite and constructive-- very little moderation was needed.  Even when an anti-evolution Discovery Institute affiliate posted criticisms of the class and its material, he remained mostly respectful (even if totally wrong, in my assessment) in his posts in the class's Discussion forum (though significantly less respectful in his personal blog entries).  Most Discussion forum posts fell into five broad categories: a) discussion of course material or assigned problems, framed so students work together to get to a full understanding, b) discussion of material from outside the class but related, such as news items or personal observations, c) technical issues needing resolution, d) expressions of appreciation for the class (it really surprised me how many of these appreciation posts there were), and e) general socializing.  I'll continue on the next one below.

6) Community.  There's no other word for it-- a sense of community was built in the Discussion forums.  People in the class became "friends" in the forums, and honestly, I feel like many of them are my friends, too.  From elegant class notes shared by "Soma" to kind thoughtfulness by "Sallie" working through the material to the ever-enthusiastic commentaries by "HP" to "Ozgur" who created a super-helpful 'calculate your grade' spreadsheet to "Andrea" who was gifted at elaborating the underlying aspects of the weekly problem set questions to "Richard" answering people's questions or providing amusement, etc.--  this was a community.  There are so many more people I could/ should mention, and although I've never met any (actually I was fortunate to meet delightful "Lynn" this week), I feel like I know many of them, at least a little.  I loved how they'd even sometimes tease me about some of my mannerisms or things I'd say repeatedly ("leverage" is a great word!).  I'm a bit sad the Discussion forum is closing, since it means I won't "see" many of these interesting people again, and I'm hoping the same sense of community is built in future forums.

(By the way, I've only used first names or initials above-- if any of those mentioned see this and want me to delete those and replace with a pseudonym, I'm more than happy to do so-- just shoot me an e-mail.)

Personal thoughts

I LOVED teaching this class.  The participants asked me great questions that enhanced my understanding of the material.  They provided great links and resources that I'll employ in the future.  And frankly, it was fun!  I especially enjoyed peering into the Discussion forums multiple times a day, and frankly, I can't fathom why everyone wouldn't want to do it.  There are hundreds of people talking about your presentations and their content all the time.  If "ears are burning" is an expression of what happens when someone is talked about, mine are doubtless beyond the ash stage.  The group's enthusiasm for this material that I love was so infectious-- how can one not be drawn in?

Next iteration starts soon-- January 4, 2013.  My on-campus class will be in the mix alongside the online students.  The students this round set the bar pretty high, so I'm hoping it's as successful!  Enrollment right now is a good deal lower (13,540 today), but I'm hoping it'll still be active and constructive.

I must close by heartily acknowledging my new friend and collaborator in this project, Justin Johnsen, who worked full time (40+ hours per week) on this project from mid-August until the end of November, and has pitched in periodically since then.  We also got extensive assistance from Chris Lorch, Andrea Novicki, Colin Rog, Julie Noor, and others (especially at CIT).  This project would never have been as successful without all their help.

Related links:
Google Hangout with student participants (48 min, starts at timepoint 2:30)
Article by Duke President Brodhead on the class
Guardian article by a class participant
A participant's blog post

Thursday, June 21, 2012

Wherefore and what from scientific societies?

Do you contribute to a scientific society (as a member or otherwise), and why?  Many scientists don't bother maintaining memberships, thinking the only benefit is getting a journal to which their institution already subscribes.  Most people don't even think about what scientific societies do.  The best answers I have heard are "manage the journal", "organize meetings", and/ or "give out awards."  Many societies do some or all of these three things, but I think most academics are unaware of the backdrop, and the importance of the financial linkage to the journals and conferences.  Scientific societies are sometimes funded largely by profits from their journals and meetings, though some funds also come from membership dues, donations, and grants (the relative proportions vary greatly by society-- e.g., some take a yearly loss on their meetings).  The revenue funds are then used for maintenance of those same activities as well as distributing a variety of awards (most often for students but sometimes for workshops), outreach activities, and overhead/ staff.  Some societies also engage in activism, promoting the merits of their discipline to the public or politicians, including drafting statements about controversial topics (e.g., stem cell research, evolution education) and arguing why scientific research is a good public investment (see also this article, that's not by a scientist).

These societies are, and operate like, non-profit organizations.  However, the revenue source is a little different-- rather than relying exclusively on donations, the yearly conferences and journals provide potentially larger and more consistent funds to maintain these activities.

Now, let's think for a minute-- what would happen if we changed the model to one resembling most non-profit organizations?  What if society journals exactly "broke-even" to the publishers, making them cheaper in which to publish and subscribe but eliminating the revenue to societies?  What if we then asked PI's to donate to their societies to fund the grants, outreach, and activism efforts?  There's a fundamental problem-- "donations" cannot be billed to grants, so we'd have to donate out-of-pocket.  This problem is apparent in why people often refuse to pay memberships to get reduced publication charges, even when the reduction is greater than the cost of membership-- publication charges come "from the grant" and memberships come "from checkbooks."

So, societies with an associated journal get funds from journal publication and subscription charges (and sometimes from conference registrations), and use those funds in a "charitable" way to meet the goals of the group, including distributing extensive student research or achievement awards, maintaining communication among scientists via conferences, and providing outreach to schools and the broader public.  Yes, there's some overhead too, like funding the travel of the scientific officers to meetings, but these officers rarely receive any compensation beyond "expense reimbursement" for what ends up being days of work each summer.  There are typically minimal (and often overstretched) staff, too.

This should be food for thought in choice of publication venue, personal society memberships, personal journal subscriptions, recommendations to your university about journals in which to subscribe, and choice of conferences.  Memberships in societies (e.g., Sigma Xi) really do directly give back to your community, in addition to getting you a journal.  Subscribing to and publishing in SOCIETY journals (e.g., Evolution, Genetics, Journal of Heredity) brings money back to your students, colleagues, and community, whereas publishing in NON-SOCIETY journals often fails to do so.  Similarly, attending SOCIETY meetings potentially gives flexible funds back to an organized group with similar goals, whereas attending other meetings only does so indirectly, if at all.

Food for thought.  I'll note that I'm not unbiased, though, as I have served and currently serve in many societies myself.

Saturday, May 19, 2012

Why go to conferences?

Summer conference season is around the corner! I especially encourage people to participate in the upcoming American Genetic Association conference in Durham, NC, on recombination.  But that aside, why do we go to conferences?

To see the talks?  A little (field dependent).  A good talk is sometimes very powerful and exciting, and it can be a great way to see exciting work that may not come to print for another... oops, there's the problem.  For another 6 months?  It is nice having the "preview", but far from essential.  For many (albeit not all) talks in my field, I'll see the print or online publication before the next year's conference, if I haven't already even been asked to review the associated manuscript.  Obviously this isn't true across all fields of subfields.  Talks also often oversell, since they're not subject to the same sort of peer review.  Caveats can be omitted, controls may not even have been done yet, etc.  They're exciting on the one hand, but they're not "bankable" on the other the way many publications are.

To present research?  Yes, especially for junior scientists but somewhat for all.  We need to maximize routes with which we disseminate our results, and we also can all use more practice doing it in public forums.  Presenting is a great way to get informal feedback, too, pre-submission: much nicer to have someone tell you that you didn't consider X alternate explanation in a meeting (so you can address it) than to have your manuscript to Science rejected on that basis.  Giving a great talk can make a postdoc more competitive, too-- I recall many faculty meetings about job searches where someone said, "I saw her give a talk at X meeting, and it was really great."

For the informal interactions?  Yes!  This point ties in with the feedback I mentioned above... not only do we get feedback on what we present, but we can talk with colleagues about all sorts of projects and ideas, potentially even forming new collaborations over coffee, meals, or "beverages."  Yes, we could do this by phone or e-mail, but this forum provides a more "captive audience" for a long time-period not (or less) distracted by their ongoing college duties, and with tons of other experts around as well who can weigh in.  This is, in my opinion, the single biggest advantage of in-person conferences.

For the free goodies from vendors?  ABSOLUTELY!  But I won't dwell on that one.

If I were to propose two pieces of advice:

1) Let's make as many talks/ posters as possible publicly available.  Let presenters opt-in to have their talks videotaped and put on YouTube/ UStream.  Submit your posters to F1000 Posters.  These are all free.  Make it so those who couldn't come to the meeting because of cost, family obligations, or even being environmentally friendly in avoiding needless air travel still able to see the research that was presented.  Make it so those attending can see some of the concurrent talks they missed.  Obviously, opting-in would be voluntary, but organizers can minimize the barriers to it and encourage it.  I know some will decline this option for fear of being "scooped", but some will jump at the chance of more dissemination to their colleagues and the public.

2) Don't do the freshman-dorm thing of walking around with your labmates or buddies for the whole conference!!!  If multiple people from a lab are going, don't let them share rooms with each other... force them to room with someone from another university.  Similarly, while it's nice to "support" your labmate by going to their talk, I'd personally prefer my lab folks go to the concurrent talks I couldn't attend to tell me what I missed.  And most importantly, encourage the shy but excellent junior scientists to go meet other movers-&-shakers in their field, both senior and other junior-- their PIs should facilitate this, but anyone can take the initiative to help.

Happy conferencing this summer, y'all!  Comments welcome, as always, even if critical.