3rd April 2009 at 9:44 am #6246Patrick BrindleMember
So far we have a good deal of activity from various qualitative methods streams, but virtually nothing from those using quants and stats. Any ideas why this might be? Any suggestions on what we can do about it?8th April 2009 at 3:02 am #6248Berkay OzcanParticipant
I started writing my hypotheses about why we have few activity on the quants and stats front on Sunday. But I mistakenly deleted it and never got the time to go back again until now. I apologize for the late reply. And I try once again.
1- I think first and foremost, we have much fewer people on the quants and stats front. And I believe partly by chance, those we have here are not the ones that are especially active in such websites. So far, embarrasingly I didn’t get a lot of connections although almost all the people I send invitations are quantitative. So I just prepared another email list where i will invite many more quantitative people very soon. I guess I should be doing it more quickly because now I am afraid the more qualitative dominated becomes the whole site, the less attractive it will be for the quants people.
2-Over a talk with my friends, we realized that most of the quants people shift between the techniques more frequently, and use multiple methods at the same time and act more pragmatic in choosing their analysis. And this makes it harder for them to start a research group on one method. For example, a person who has been using Poisson models and factor analysis on her one paper might also use Instrumental variables and fixed effects estimators in the same paper. At the same time she might not be caring about any of them enough to start a research group. Because many of these methods might be totally complementary depending on the question, there’s no such “method loyalty” among the quants people as it happens among the qualitative people. For example a qualitative person who uses an oral history method, or a video research might be more loyal to (or identify himself with) his method and therefore more willing to promote or discuss it.
Furthermore quantitative methods might be very specific and many. For this reason, it is again harder to decide how specific one can be at this stage (Especially when generating groups) Does it make sense to generate a research group on Regression Analysis or Multivariate Regression Analysis or Regression Analysis with Limited Dependent Variables? or simply Tobit Regression (which is even a more specific type of the previous group)?.
Having said that usually there are a few quantitative research methods whose users are more loyal and doing their best to promote them. One of them is Sequence Analysis … For this one, I opened a group. May be I couldn’t promote it well so we are still two people. Another one is MCA (multiple correspondence analysis) for which i will start a group sooner. And finally, event history analysis users also from epidemiology, demography and sociology work hard to promote their technique. I created a group about it and it attracted somewhat 9 people. I am planing to split that group into two: 1) Event History Analysis 2) Life Course Research .
(NOTE: I talk to Karl Ulrich Mayer, who has published many papers on Life Course perspective. His last paper on “Life course Sociology and Life Span Psychology” is under review in Annual Review of Sociology. He is willing to write a few paragraphs to our Life Course Research Group’s blog in a few weeks but I guess it would be nice if i can make it a little bit more crowded by then. There are a few other major people in different fields who are willing to do that in Mathematical Sociology, in Bayesian methods and in MCA. Same applied for them, they would like to see a bigger audience , better if there’s a crowded group entirely interested in these methods.)
3- You may well say that one can be active in blog, forum and methods-q&a without generating a group for each quantitative method. Then one common complain of my friends and I share their feelings is that the common spaces outside the groups (basically Blog, Forum and Methods QA) are a little confusing. I know we talked about it before, but now i think it is becoming more obvious that we are using all of them for a few similar purposes: introducing ourselves, asking for a source or a citation, or simply talking briefly about a method or announcing an event or uploading a part of a research etc. They are all great but you can find examples of them in each of the three sections. May be as the time goes by people will generate different patterns naturally and will post certain type of things always under certain sections. But at the moment things are pretty randomly posted in all three sections. I was actually hoping methods QA part would be attractive for most of the quants and stats people. I sometimes receive from the grad students some stats questions via email, where they need to type equations, or copy and paste tables or graphs. I haven’t tried myself but do you think it may be costly to type or paste them here due to formatting issues?
well, these are my two cents… I hope we can talk about it more …8th April 2009 at 9:04 am #6247Patrick BrindleMember
Interesting reply, Berkay, many thanks. One of my stats authors put it like this:
If I had to guess, I would say that it’s because stats method folks have been around longer, and so have established more places to hang out. Also, stattos tend to be techies, and so embraced things like usenet and blogs and wikis earlier than the qualitative folks, and maybe there’s a gap in the market for the quallys that hasn’t been filled yet, but for the quants it was filled a long time ago.
It’s also easier for stats things to establish themselves electronically, because the question is frequently ‘how do I do X’, and the answer is with this test, or this bit of syntax, or this program.
For example, I read SAS-L (which is also sci.stat.soft-sys.sas) and I used to read SPSSX-L, I also am on semnet, multilevel, psych-methods, Andrew Gelman’s blog is very popular, chance-news (a wiki) is read a lot. I’ve not looked much, but I get the impression that these things don’t exist for quallys, or at least not as much.
- You must be logged in to reply to this topic.