Stanley Presser sends along this article, “The Role of Doubt in Conceiving Research.” Presser has taught for many years at the University of Maryland but not when I was a student there, also he teaches sociology and I’ve never taken a sociology class.
Presser’s article has lots of interesting discussion and quotes about learning from failure, the problem of researchers believing things that are false, the challenge of recognizing what is an interesting research question, along with some specific issues that arise with survey research.
I’m reminded of the principle that an important characteristic of a good scientist is the capacity to be upset, to recognize anomalies for what they are, and to track them down and figure out what in our understanding is lacking. This sort of unsettled-ness–an unwillingness to sweep concerns under the rug, a scrupulousness about acknowledging one’s uncertainty–is, I would argue, particularly important for a statistician.
P.S. That last link is from 2016, to a post that begins as follows:
I’ve given many remote talks but this is the first time I’ve spoken at an all-electronic conference. It will be a challenge. In a live talk, everyone’s just sitting in the room staring at you, but in an electronic conference everyone will be reading their email and surfing the web. . . . At the very least, I have to be more lively than my own writing, or people will just tune me out and start reading old blog entries.
Funny to see this, seven years later, now that electronic conferences are the standard. And I think they really are worse than the in-person variety. It’s hard for a speaker to be more interesting than whatever is in everybody’s inbox, not to mention the world that is accessible from google.
I suppose what you say is true: my first teaching experience was in the Attica Correctional Facility. It was the most attentive audience I ever had. I guess there were no more interesting alternatives. But if you prefer in-person teaching over online teaching or presentations because the audience is more attentive, then that is a pretty weak defense of the paradigm. Also, I wouldn’t characterize lack of over distraction as meaning that people are paying attention – that is one reason that so many students put up with huge classes. Even if you can’t use your computer or phone during class, daydreaming is entirely possible.
I know Andrew has other reasons to prefer in-person over online, but I do think the complaint that there are too many more interesting distractions in a virtual environment is a weak one.
Dale:
I disagree! Yes, daydreaming is always possible, but (a) daydreaming is typically a lot less absorbing than what’s online, and (b) if people in the audience are daydreaming, the speaker can notice and adjust accordingly.
I strongly disagree (and I agree with Andrew’s reply). Perhaps in principle there are as many distractions in a physical seminar as online, but in practice it is *far* easier to pay attention to a live person in front of you than a digital head on a screen. There’s something deeply psychologically ingrained about engaging with another physical person; the past few years’ experience has made this clear. One of the clearest examples for me was the 2022 American Physical Society March Meeting, a hybrid one in which sessions had an interspersed mix of live (~75%) and remote/video (~25%) speakers, the latter shown on the same meeting room screens as the live speakers’ talks. The drop in attentiveness when a video talk came on was palpable — people wandered off, did other tasks, chatted, etc. — Me included — though in principle it’s the same sort of talk on the same screen, just without someone standing and speaking. The 2023 meeting abandoned this hybrid setup, which everyone agreed was a disaster.
I’m puzzled by Andrew’s comment that ” electronic conferences are the standard. ” Definitely not in my areas of physics and biology; there are some virtual meetings, but in-person meetings are definitely (again) the norm.
Sorry this is off the actual topic of the post!
Raghu:
This is a blog! Nothing is off topic.
Andrew and Raghu:
I agree that video is really easy to ignore compared to a live person. But if we step back to the larger picture of on-line vs. in person *education* – not lecture – I wonder how does that change the picture? After all we have all the talk these days that lectures are a crappy way to learn anyway, so why would we use a supposedly crappy method via a medium where it’s crappiness is magnified (both literally and figuratively)?
The problem with on-line speaking is that, without the personal-contact dynamic, 1) the speaker’s irritating vocal and physical habits become a distraction; 2) poor mechanics of the presentation become more apparent; e.g., if it’s poorly organized and the speaker is poorly prepared (the standard situation), its far more painful in video than in real life.
So maybe that’s the best approach: ditch the talking altogether and replace it with carefully crafted, well-organized and edited text.
Brings a new meaning to the term “captive audience”!
While I do enjoy reading a text (and skip the parts that I do not want to read), one alternative mode of communication has not been mentioned yet: videos. Covid has spawned an awful lot of poorly produced teaching videos, but there are some gems out there, too. Most students (in my experience) will turn to YouTube rather than to written texts. I believe short educational videos will be key to 21st century education, and I recommend that every teacher make them a part of their arsenal in their fight against statistical (and medical and biological and psychological and so on) ignorance.
Side note: I am not recommending to completely replace current teaching methods. Rather, I suggest to augment one’s teaching by packing difficult topics into nice little 5-15 minute videos. Quality is the key though, YouTube is crammed with videos of well meaning people with thick accents, poor microphone quality and an apparent lack of understanding of how little their audience can see inside their heads.
As videos begin to make their way into student education, I would be curious if it were not possible to have the online conferences mentioned above watch a well made video for a few minutes and then open it up for questions and discussion. I am aware that making a good video is a lot more work than making a presentation. But I wonder if it might not be worth it, because a presentation that nobody listens to is definitely a waste of time for both the speaker and the audience. Also, a video has a much higher information density (I mean information per time), which allows more time for discussion. A final advantage of video production: The expert only has to write and record a text and the nice visual effects can be created by your TA :)
I agree. I am amazed that whenever I want to repair something, there are dozens of short videos telling me how to do it. As long as you know what it is you want to know, you can find someone explaining it to you at almost any level you want it explained
The big remaining problem is convincing people that Metropolis Sampling is as important as changing a toilet handle.
Raphael:
Interesting point. I have lots of talks online, but they’re not sharp focused 15-minute videos; they’re meandering hour-long lectures. I think the meandering lectures are a lot more fun to listen to in person, or even to watch, but if the goal is to solve an immediate problem, quick videos would be better. I guess that my books are the equivalent of collections of quick videos, conveniently organized into chapters and sections.
A few years ago I made a series of 12 one-hour videos for a remote introductory statistics class that Columbia was offering. I wrote all the scripts and then read them while being filmed. It worked surprisingly well! I actually come across much more natural and spontaneous-looking on video when reading my own prewritten material than when giving a live talk, and I made lots of graphs to go with the lectures. Unfortunately I made one key mistake when doing these, which was not to insist that Columbia put the lectures up on Youtube. So they are not available for the general public. Also, time has moved on and now I’d change what I’d teach in the intro course.
The issue with lectures/talks is they are sequential access. Reading is random access, which is much more efficient. So talks seem slower. Well, at least since the codex replaced the scroll.*
A talk is good for asking questions (both from and to the audience) and adding “color” to the topic that may be too informal to get written down.
* However, some long-form journalism does its best to negate this advantage by interspersing the point with the color of the walls, number of chairs in the room, etc. Like most things, it is more a gradient.
Not only are they sequential access, there’s no 2x button.
https://condenaststore.com/featured/a-bearded-wise-man-looks-over-a-book-paul-karasik.html
I always felt live conferences were like going to a show. The virtual versions were like homework.
As much as you might not enjoy virtual conferences, I think it’s great to have events that can be attended/presented at people who don’t have several thousand dollars+several days of free time to spend on airfare, hotels, etc. And they’re much better for the environment too.
Adede:
Indeed, before covid ever happened I often responded to presentation requests by offering to speak remotely, because I don’t like to fly and also to reduce environmental impact. These remote talks often worked pretty well, I think in part because they were unusual: people in the audience would show up as for a usual talk, and they weren’t in the habit of reading their email etc. during talks. It’s my impression that things are a lot worse now because there’s a new standard of attending talks or meetings and not paying attention.
I believe all of these statements are simultaneously true:
1 online conferences have a much greater environmental impact than in person conferences
2 online conferences are less valuable — in terms of amount learned, connections made, whatever — than in person conferences
3 online conferences can be much cheaper to attend than in person ones
4 online conferences don’t require abandoning one’s family for days
5 *some people* are as attentive at an online seminar than an in person seminar
6 *most people* are much less attentive at an online seminar than an in person seminar
It’s not obvious what to do about all this. Even pre-Covid, I was always amazed that so many academics would travel more than ~3 times per year, which especially for me would have crossed the threshold of #4.
(Tangentially related to #3, virtual attendance isn’t as cheap as one might expect, though of course travel costs don’t exist. See the bottom of: https://eighteenthelephant.com/2020/01/14/the-case-of-the-rising-conference-registration-fee/)
I have trouble imagining online conferences having more of an environmental impact than flying people around. What is your basis or did you mean “less of an impact”?
Sorry — a big typo! Yes, that was supposed to be
1 online conferences have a much smaller environmental impact than in person conferences
If only there were an edit button…
Thank you! I think big streaming video services have a heavy environmental impact in total, but its hard to imagine a few tens of hours of low-res video for tens to thousands of people costing as much as flying all those people long distances. People can organize online conferences on a very small budget.
Another factor with online conferences is that they are more accessible to researchers without travel funds and rich-country passports.
I would agree that face-to-face meetings are very important for spreading a message and building solidarity. Its possible that fewer face-to-face meetings mixed with remote meetings would have a similar effect.
Andrew—I think your perspective may be different than the rest of ours in two important ways. Think about the last ten years.
Where do you find that people pay attention in person? In my experience, everyone is on their laptop or phone.
[edit: Don’t blame me for the bold, it’s the blog’s CSS for an ordered list.]