From backchannel to blackboard

Back in March, I wrote:

Consider this scenario. An instructor is giving a talk. When the instructor wants a question answered, students Twitter their response instead of using clickers. Ideally, this would involve some software that could recognize some special symbol or code associated with a class, so that responses could be tallied and a graph could be generated on the fly. I’m guessing that such software isn't out of reach for good programmers. Conceptually, it seems actually pretty simple.

I’ve since learned that such technology does indeed exist, and reckoned I should report on my (tiny) experience with it.

I heard about Poll Everywhere over on Slide:ology. Basically, you can ask people to vote by phone, Twitter, or web, and it gets copiled and updated into a bar graph on the fly.

I used it to create polls (one shown at right) and make them into PowerPoint slides. Since my students don’t bring laptops to class, and apparently none of them have Twitter accounts (at least, nobody has admitted to it), I chose to set it up so that they could send a text message using their mobile phones.

It worked as advertised. Nobody had any problems sending the text messages, although the numbers they have to key in are arbitrary, so it’s not easy to do without carefully looking at the slide and double checking the digits. The animation of the bars when new information is coming in is nice and smooth. It was certainly not perfect. Importing the slide into a larger presentation was a little twitchy. And I consciously did not write that the graph updated in “real time,” because the updates were a trifle slow.

For comparison, I’ve been using a clicker system in my classes for several years now. The system is a bit more elaborate, in that it requires a dedicated clicker, receiver, and software. But if you’re doing a lot of in class questions, the difference between pushing a single button on a clicker and pushing about a dozen buttons on a mobile phone adds up. Combine that with the lag in the Poll Everywhere system, and the clicker system comes out feeling more natural and agile and less intrusive.

The clicker system does not compile results on the fly like Poll Everywhere does. Instead, it takes everyone’s answers, waits, then draws a graph. This is an important consideration in an education setting. You want people to have a chance to think on their own for a moment, and not be influenced by other people’s answers.

Poll Everywhere is great for instructors have the occasional one-off poll that they want to do in class. So far, I wouldn’t use it to replace a dedicated clicker system, but I could see how it could do so in the near future.

And, incidentally, in the little test poll shown? None of my students got the name of their own university right. It’s supposed to be the second from the bottom.

More backchannel and Twitter tools are given over at Olivia Mitchell’s blog. Watch for lots of clever ideas using these tools in the near future.

The Zen of Presentations, Part 27: Coping with anxiety

I had a student in my office this week for advising, and I noted that she hadn’t taken Biology Seminar, a required class for all our majors. She said she had been putting it off, and putting it off, and was deliberately taking it at the last possible time. She was absolutely terrified of giving a talk. Even as I was talking to her, I could see her getting wound up at the prospect of something that might be weeks away.

Before getting to the advice, let me preface what I’m about to say with a general principle:

There is no virtue in suffering.

We often tend to treat that people who are genuinely anxious about presentation with little sympathy. People are told to keep suck it up and keep practicing. There is more than a little “You should suffer for your art” attitude out there.

If you are truly frantic about the prospect of speaking in public, why not make an appointment with your physician and see about getting a prescription drug to help with the anxiety?

I don’t say this lightly. There’s a reason that some drugs are only available by prescription, and only recommend this is as a last resort for extreme cases.

I know one person who had to give a lot of presentations, and hated every second. The stress was quite debilitating, so this individual got a prescription for a beta blocker, and took a pill before giving a talk. The talk I saw this person give was fine, and I’m convinced the audience wouldn’t have known this person was dealing with high anxiety.

In most cases, you’re better off practicing and learning the skill of presentation that getting medication. But not everyone is the same, and some people may need more help than practice and preparation alone can give.

Why lecture?

Our university started using a system to record lectures this semester called Tegrity. I’ve used it a bit, and it has some sweet features. It captures voice, video if you’re so inclined, and everything on the computer screen. Once recorded, the lecture can be viewed online, downloaded as an audio only podcast, streamed to an iPhone. My favourite feature is that all the text on the screen – whether a PowerPoint slide or something on web browser – is searchable. Can’t remember where in the lecture a technical term was mentioned? Type in the text and it’ll locate it.

First response one of my colleagues had to it:

“Just another excuse for students to be lazy.”

You might say there’s a little diversity of opinion on the subject of lecture recording. So this post by Donald Clark, with the bracing title of “Lecturing – stupidest profession?”, is timely. There are a lot of comments to read, too.

Lecturing is a weird historical artifact. To the best of my knowledge, the main reason that lecturing because a method for transferring information was that it originated before the introduction of the printing press, when there were literally not enough copies of books to go around. Despite books being widely available for centuries now, lecturing persisted. On the face of it, lecturing is stupid.

As you might be able to tell from the comments in the first paragraph, I think there’s a lot to be gained from recording lectures. The major thing is that attention, no matter how compelling an instructor might try to be, drifts. There are just times you miss a point, and even if you get it the first time, it often takes multiple times listening to something to remember or understand. And students are often unable to make a lecture due to some sort of happenstance.

On the other hand, I think there’s way more to a lecture than a YouTube video. I’ve written before about videotaped presentations:

Human beings are very good at conversation, very good at face-to-face interactions. We like it. We crave it. Trying to take the information out of that social context almost always kills it.

When I first started moving to partly online classes, I asked students if they wanted me to work up a class as a completely online class. The answer was largely, “No”: they still considered the lecture time valuable. “Face time” still matters for people, even if it looks like a simple one-way flow of information. If we were just interested in pure information transfer, we’d just give students books.

The danger for both the student and the instructor is that doing recorded lectures not because it’s a new resource for people to learn, but because it’s convenient. It’s easy for an instructor to get up and talk for an hour, and easy for students to listen. It’s not very threatening. It’s not very demanding.

Structuring lectures to take full advantage of the immediacy and social interaction inherent in the format is time consuming and hard. And university students frequently don’t want to come along for the ride. University students often got into university because they figured out how to make lectures work for them.

While my colleague considered recording lectures an excuse for students to be lazy, it’s also an excuse for instructors to be lazy.

Recording lectures is a great opportunity, but you have to do it right. It’s one thing to record a new lecture every class session of a class that is never the same way twice to give students something to review, and pre-recording a lecture once in your office with no audience and never changing it again.

Rating journals and articles

Richard Smith argues that having measurements (or, as fashion calls them, “metrics”) for individual scientific articles means that impact factors for scientific journals are going to go away very soon.

I don’t think it’s going to be quite that easy.

Scientists have had “article level” measurements for a long time: the number of citations an article receives in other papers. To be fair, Smith mentions citations:

Citations are used to calculate the impact factor, but these citations come from only one (expensive) database. It’s better to use more than one database.

Putting aside the impact factor calculation for a second, has Smith not seen that Google Scholar has citation information? I also do not know why Smith claims more databases are better (apart from the possibility that one might be free). We want the actual number of times an article has been cited, so multiple databases should all contain the same number of citations an article receives.

But as long as there are multiple venues for researchers to publish their research, editors and publishers will use some kind of measure of their publications to establish credibility. Does it matter if a journal promotes itself by a high impact factor, a high number of combined article downloads, or some other aggregate number?

And researchers need some sort of way to determine a journal’s credibility. Otherwise, we’re left with individuals pulling hoaxes to figure out if a journal is the real deal or not.*

More and faster measurements of articles is interesting and valuable, though the underlying question remains unanswered: What measurement of an article should we be most interested in? Page views are different from downloads, which are different from blog mentions, and all are different from citations. In an open access arena, you can probably crank up the number of page views and downloads by publishing anything with “dinosaur” in the title. (And I say this because I love me some dinosaurs!) In terms of scientific impact, an article read by 500 people might be more influential than one read by 5,000 people, if it’s the right 500 people; i.e., hits the target audience squarely.

In the end, all of these numbers are about money: grant money, salary money, merit money, and so on. If it weren’t for that, nobody would care about impact factors – or any other measurement of a scientific article’s worth – in the first place.

* Though to hear some of the advocates for Open Access projects like PLoS One, you could be forgiven for thinking that what they want to see is a situation where there are no other journals.

Tuesday Crustie: Hypertrophy

Species unknown, but apparently some sort of Brachyura.

From Attack of the Crab Monsters. Roger Corman, we love you!

Let your neurons relax, the predators are gone!

Predators eat prey. Prey, over time, evolve features to evade predators. But what happens when predators go away?

This paper by Fullard and colleagues is almost a perfect storm of topics that I’ve talked about several times before on this blog. We got crickets (previous appearances here, here, and here), we got bats (previous appearances here and here), we got reverse evolution (discussed here).

Teleogryllus oceanicus is found on landmasses across the Pacific Ocean, ranging from mainland Australia to remote islands. The distances between populations are so large that you would expect these populations to be slowly separating by genetic drift alone. Using highly variable regions of DNA, Fullard and colleagues showed that there were genetic differences in the crickets based on the geographic distance between them. The most distinct populations were from the Australia city of Darwin, and Moorea in French Polynesia. While a few people might know Darwin is located in Australia’s top end, I expect very few could pinpoint Moorea on a map, so I’ve done it for you. Moorea’s the one on the right.


Not only is it a bit tricky for humans to locate, being far in the middle of the Pacific Ocean, bats haven’t located it either. There are no bats living on Moorea, in contrast to the Australia mainland, which boasts a rich assemblage of the flying mammals. Bats are cricket predators, So this creates an interesting situation where there are expected to be clear differences in selection pressure on the crickets. Many aspects of crickets’ sensory systems, particularly hearing, seem to be specifically tuned to detecting and avoiding bats. Several big auditory neurons respond strongly to the ultrasound that bats make when echolocating, and crickets will fly away from ultrasound pulses (negative phonotaxis).

One critical neuron that crickets use, in some contexts, as a bat detector: ascending neuron 2 (mentioned before here; red in the picture). Ascending neuron 2 (AN2) isn’t the only auditory interneuron that can respond to high frequency bat sounds that bats make when echolocating, however. Another neuron is omega neuron 1 (ON1), which responds both to high frequency, bat-like sounds, but it also critical to the processing of low-frequency, cricket like sounds (green in the picture).

Thus, the prediction from evolutionary theory is that in a situation with no bats, you would expect there to be reduced selection pressure to maintain the neurons’ sensitivity to high-frequency, bat-like sounds. The decrease in selection pressure should be stronger for AN2 than ON1, because AN2 seems to be more strongly dedicated to a single task than ON1.

There are many different parameters of a neuron’s response to a stimulus, but the authors present for sensitivity: What is the lowest level of sound needed to get the neuron to fire? The Moorea population’s threshold for AN2 is significantly higher across the board, whereas the sensitivity for ON1 is not significantly different at almost every frequency. There is a significant difference at one frequency, 15 kHz, but this seems to be a chance finding due to the large number of comparisons being made.

This difference is also reflected in the behaviour. There was no significant difference between the populations in turning towards a cricket-like low-frequency sound, but there is a significant difference in populations when turning away from bat-like high-frequency sound.

What this paper doesn’t have is a good description of the suspected history of cricket colonization or bat colonization across the range. This is probably because it is unknown, but it would be interesting to have some idea of how long the crickets in French Polynesia may have been in a bat-free environment.

A great follow-up study would be to look at what level this evolutionary change has occurred at. There are many ways that an interneuron could lose sensitivity to a stimulus. The ear could change; the sensory neurons in the ear could change; the connections of the sensory neurons to the interneurons; the physiology of the interneuron could change... and so on.

Full disclosure: One of the authors here was my boss about a decade ago. I wish he’d been working on this project then!

Reference

Fullard, J., Hofstede, H., Ratcliffe, J., Pollack, G., Brigidi, G., Tinghitella, R., & Zuk, M. (2009). Release from bats: genetic distance and sensoribehavioural regression in the Pacific field cricket, Teleogryllus oceanicus Naturwissenschaften DOI: 10.1007/s00114-009-0610-1

Neuron picture from Scholarpedia.

Crawling back up

My review article has climbed back up a notch to #2 in Brain, Behavior and Evolution. That’s five months in top three, for those playing along at home.


If you think this article is interesting, tell your friends. I’d be interested to see if the article can reclaim top spot.

Comments for second half of October 2009

Laelaps covered a paper I'd also done (clubbed tails in glyptodonts), so I left a little link for those wanting more.

Slides that Stick made a comment about the shape that brains are best at absorbing information from. I wanted more...

The Online Laboratory of Kevin Zelnio takes issue with an ad. I am nonplussed.

Finite Attention Span talks about making lecture time into practice time. I point out that this takes more than an instructor’s willingness to do so.

Storytelling in science, part 2

Earlier this week, I commented about the importance of stories, and how denialists of science have a lock on the “little guy fighting the establishment” story. I mused about what the counter story is to that.

As much as I hate to suggest this, I think I have one.

Zombie apocalypse.

You face an world of dangers that are slow and stupid but many and unrelenting. You have limited resources. If you let your guard down, that’s it. If you don’t have the weapons and some knowledge and a little technology, you’re done.

The zombies are the natural world, always throwing up little surprises like a new strain of H1N1 virus. We are finding out to our peril that some of our resources really are limited, or pose their own threat, like fossil fuels.

Scientists?

We’re the ones with shotguns and a stash of shells, baby.


We have decided to fight and we might just be able to help a lot of other people. When the shit hits the fan, the chips are down, you want scientists on your side, because they just might be the only thing standing between you and a a messy end.

(And for the record, I repeat: I hate zombie fiction! But happy Hallowe’en weekend, anyway!)

Picture from here. And yes, I put this up in full knowledge that I will never be one tiny fraction of that bad-ass.