What motivates people to answer questions?

Question

I know why I ask questions -- to learn and to get help with problems I'm struggling with. And I have been incredibly impressed with the Stack Exchange community's knowledge, eloquence, and responsiveness. When I post a question, it almost always gets some response within 10 minutes. It's like having a bunch of statisticians and mathematicians hanging out at a water cooler to which I can walk around the corner when I get stuck.

I am not surprised that some of my questions get answered. But I am (wonderfully) surprised that they almost ALL get answered, and almost immediately. What motivates enough people to spend enough time on this site is that there is almost always someone with specialized statistics training willing to write up and answer the question of a stranger -- often having to parse through the ill-formed/vague questions of beginners who may not know enough to know how to word a question. Even if you're an expert in the subject matter, it still takes time to write your thoughts and develop examples.

One explanation is real-world reputation, but so many people use aliases that I don't think that can explain much. I imagine some people get paid to answer, but I doubt that can be much of it. There is probably an intrinsic joy to teaching for some, but many of you are teachers by profession anyway. Is it truly altruism? (in the informal sense of the word) Is it about building and being part of a community?

I am very appreciative of this site and the people who answer questions. I'm curious to hear your thoughts.

Answer 1

For me, there are three real reasons I end up answering questions:

1. Something that could probably called professional pride. This is my job, and I'm an academic. I'm supposed to help contribute to learning and knowledge in my chosen field.
2. It keeps my mind sharp. I'm away from my colleagues a great deal of the time, and for awhile I found myself in something of a bubble. CV...lets me walk amongst the data folk from time to time.
3. Selfishness. I may one day have to read a paper written by someone who asked a question here - I'd rather read a methods section that doesn't make me cry.

Answer 2

Well, I use my real name, and consulting is what I do, so that's a little bit of it.

I like answering questions. It keeps me thinking - or sometimes gets me thinking about something in a new way.

If I ask questions (and I do) and get answers (and I do), then I feel it is my fair part to answer questions where I can.

I like seeing my reputation number increase. :-).

Answer 3

I always use my real name in fora like this. I see doing so as an honesty and integrity issue, plus I have as part of my role description to promote myself and my employer as up to speed with modern statistical practice and this is a small part of that.

Why do I answer?

1) I discovered the site when I had a couple of questions myself and quickly got very useful answers. A few weeks later I decided I should put back in some of what I got out - it's only fair to contribute. I empathised with people struggling over questions, and saw a few cases where I could help. Straightforward empathy and ethical duty to do something good for others.

2) Once I started answering, I quickly found how useful it was for me as a form of professional development. Answering even "easy" questions (is there such a thing? nearly every one has interesting philosophical implications if taken seriously) forces me to think things through a bit more clearly than is sometimes the case in my fast-paced applied world; and revisit the odd text or two that I thought I'd learnt and understood but needed refreshing on. There are many more statisticians here to hold me to account than in my normal job, and I can feel my understanding and skills improving.

3) I think the reputation and badges system is well designed and frankly slightly addictive.

In summary - stack exchange has some of the addictive power of facebook, with much greater intellectual stimulation, many chances to learn, chances to help some people out, and hopefully reduce a few bad mistakes in published science.

Answer 4

+1 to all the answerers. I think all of these reasons play a role for me--at least somewhat. The one point I would like to make is how much I've learned from this site and the CV community. Of course, I've learned a lot by reading questions and answers posted by others. But, although I'm only good enough to answer the easy questions, I've really learned a lot by answering them. Hunting down webpages and pdf's, looking things up in my textbooks, and figuring out how to explain something clearly has led me to solidify and deepen my understanding of these issues that I find deeply fascinating. I also enjoy answering questions, and feeling like my statistical abilities and knowledge (such as they are...) are contributing something useful to the world in one more little way.

Answer 5

I share many of the reasons for answering questions with the others given here - especially that of keeping sharp. But I'll add an extra one that actually got me started on this site in the first place. I'm not the world's greatest programmer, and found Stack Overflow to be a wonderful resource for answering my questions. I felt it was right to "give something back" but obviously couldn't contribute over there. Answering on CV "balances out" all the help I get on SO.

If you find CV is really helpful, perhaps consider given back on the SE site in your area of expertise?

Answer 6

I think there's a mix of things but I'll mention a few motivations:

1) being known in the places I work/have worked as someone who knows a fair bit of statistics means I spend/spent a lot of time answering statistics questions anyway. At least for the common questions, I can instead point people here - and in the place of answering those questions, carry my share by providing answers here too.

2) I particularly like the idea of a searchable, permanent repository of questions and answers, especially one where the questions get edited and clarified and the answers often improve over time - it means when I come up with a good answer, it doesn't just 'die' when I finish solving one person's problem, like it tends to when someone asks me a question in person. The combination of convenience and permanence makes it worthwhile to come back and clarify a paragraph in response to a followup comment, or add some clarifying picture, or generalize an answer slightly so it also answers another question, or covers the question in a way the asker mightn't have even understood enough to ask.

And I find that my better answers, ones I have crafted and reworked, tend to influence the way I write my explanations elsewhere. My written explanations of concepts slowly improve.

3) I learn from just reading other answers, and from answering questions myself (checking details in a paper or running a quick analysis or simulation I don't normally do changes a lot of stuff I sort of know into stuff I actually know). I may have been a statistician for a long time, but there's still plenty I don't know, or don't know as well as I should. Among other things, it's a neat way to learn about things I didn't know I didn't know.

4) I like that there's an opportunity for me to ask questions, should I need it - and the bigger the community gets, the more value it's likely to have for me.

5) When I get bored, spending 10 or even 20 minutes writing an answer is notionally more rewarding than getting another coffee.

6) If you like statistics, it's just interesting, you know? A reasonable fraction of the questions are simply fun to think about.

Answer 7

I, as @Peter Flom, also use my real name but I have similar motivations as @EpiGrad where I guess 1 & 2 are the ones I agree the most with since I'm not sure that those that really should visit the site do that (yes, I have recommended the site to many of my peers but I don't think I've seen any of them...).

I agree with Peter where the answers keep me thinking. I like to look up areas and dig into things that I think I might need to know in more detail in the future. I also get a great kick out of people voting up my answer or taking their time to comment on any flaws. I find often very interesting to see how my answers are judged by real pro's.

I guess the initial drive for me to answer was to gain points so that I could give bounties and so that my questions perhaps appeared more interesting when I've gained "community trust".

Answer 8

Quite simply, I come here to get help, and feel that I should pay it forward to the best of my (often meager) ability. It also seems like some of the really stellar statisticians here either aren't familiar with or don't want to answer really basic econometrics questions (i.e.: causal inference from observational data), so maybe I can contribute to somewhat in that arena.

Answer 9

Besides all the the more substantial reasons that are given above, of which I share most, if not all:

Sometimes, it feels efficient to help: Having the answer 90% ready in the form of parts of slides that I have used in teaching or an exercise I have asked implies that I can sometimes answer questions somewhat quickly.

Second, I get inspiration here for my own teaching, as questions are sometimes nice questions for my problem sets.

Answer 10

Looking back at my career, I realize the best math course I took was a Problem Seminar course. In that course, we were given ten problems to solve per week. During a good week, I could solve maybe 8 of 10 problems. That taught me how to solve a problem without a textbook context to hint as to how to find a solution. This site does the same thing within the context of statistics. That is, practice makes perfect, and if one can increasingly better answer arbitrary questions, then one has increased the ability to use statistics in one's work.

This differs from a Problem Seminar because creative answers can be downvoted here, and there is an overlay of conventionality applied here that is particular to statistics. On the plus side, even when the insular thinking of those who review answers is apparent, one at least gleans the boundaries of conventional statistical thinking. That is important for writing multidisciplinary work in that it indicates which concepts should be avoided as they mean something else in other fields. For example, a "sample" for statistics is not a blood "sample" for medicine, so one winds up using alternative concepts more generally, e.g., dataset instead of "sample."

Finally, answering a question here is like taking an oral certification exam as a single candidate being reviewed by a dozen specialists. In such a setting, the question that the specialists have is, "Is this candidate one of us or not." So it matters HOW one answers a question, i.e., like a specialist statistician or NOT. Either way, it is a profound learning experience.