The proposal to create a StackExchange site for statistical analysis is steadily moving forward. We have now completed the scoping stage which involved finding enough people willing to express an interest in the idea, and voting on some example questions to define what is allowed and what is not allowed on the site. The on-topic questions that have been selected are these:
- What is a ‘standard deviation’?
- Which of the following three graphics best displays this data set? Why?
- What’s the best way to identify an outlier in multivariate data?
- Can you give an example of where I might prefer to use a z-test vs a t-test?
- What are the differences between Bayesian and Frequentist reasoning?
Examples of questions considered off-topic are:
- How do I win in Poker?
- I have two children. One is a boy born on a Tuesday. What is the probability I have two boys?
- Joe is 8 years old, Mike is 10 years old, and Alice is 13. What is their MEDIAN age?
- Where can I access NASA’s data archives?
- How much should I expect to pay for a SAS licence?
The next phase is to get people to commit to contributing to the site. Many readers of this blog have already registered as “followers” — now you have to make a commitment to be a contributor as well. The site won’t launch until there are enough people committed to being part of it.
Just go to the site and indicate that you are willing to be an active participant once it launches.
If you’re wondering what this is all about, and why this is a much better approach than the various usenet and email help groups, there’s a nice summary on Tal Galili’s blog.
#1 by Tal Galili on 18 June 2010 - 2:49 pm
Quote
Hi Rob,
Thanks for linking, and especially for all the amazing work you have put into making this site go online.
I am looking forward to looking at this stage a year from now and thinking “how could it even take so long to start”.
Great job!
Tal
#2 by Tal Galili on 18 June 2010 - 3:26 pm
Quote
Hi Rob,
Also,
Please tag this post with “R”, so it will get republished on R-bloggers!
Cheers,
Tal
#3 by Faiz Rasool on 18 June 2010 - 4:34 pm
Quote
Very nice idea. With increasing use of statistics in social sciences, and in other fields, the number of people who want to learn statistics has risen. At the same time, many students find it difficult to get sound answers to their basic yet highly significant questions. I hope this idea translates into a successful website.
#4 by Matt Asher on 19 June 2010 - 8:33 pm
Quote
I would *very strongly* suggest that you build the site using one of the free versions of Q&A software that compete with stackexchange. Not only is stackexchange a walled garden controlled by another company, but “Joel and Jeff”, the creators, have already shown that imposing their vision is more important than respecting their customers. Before going forward with their platform make sure to read the history of how SE was implemented, and think hard about why they decided customers wouldn’t even be allowed to download full data on their own users, even those customers who they were unilaterally shutting down. I never created a site on their platform myself, but I considered it very carefully and now I’m glad I didn’t.
At any rate this is clearly an idea whose time has come, and I would link to this right away. Just make sure to be very careful about whose platform you build it on.
Cheers,
Matt
#5 by Rob J Hyndman on 20 June 2010 - 2:35 am
Quote
Hi Matt. I am aware of the history of StackExchange — I am a moderator on a SE 1.0 site. I think they have made a couple of mistakes, but I wouldn’t condemn them for that. On the other hand, the free Q&A software options are largely rip-offs of the SE idea, and stealing IP is not nice.
By the way, handing out user’s data to third parties is illegal in this country, so that could be why they have been unwilling to allow full data download.
I thought about the other options before proposing the SE site, and decided it was better to go with a group who were passionate about making excellent software, even if we disagree with some of their choices, then to use look-alike software that might not work so well.
#6 by Rick Ross on 1 July 2010 - 6:24 am
Quote
Hi Rob,
I can look past you disregarding the fact that the StackYadaYada sites, themselves, are hugely derivative works. To suggest that the developers of FOSS solutions like OSQA, Shapado, and others are not passionate about developing excellent software is simply incorrect.
We will gladly place our commitment to the community and our track record of responsive innovation up against anyone in the field. OSQA is not trying to “catch up” to StackYadaYada — we are accelerating past them quickly. OSQA not only works well, but increasingly it does things no other Q&A community system can do. You should try it before you judge it.
Cheers,
Rick
#7 by Rob J Hyndman on 1 July 2010 - 4:57 pm
Quote
Hi Rick. I’m sorry if I mis-represented the situation regarding SE and the free alternatives. Best wishes, Rob
#8 by Matt Asher on 16 July 2010 - 12:25 am
Quote
Hi Rob,
I understand this may now be a fait accompli, but if you haven’t seen it read Joel’s recent blog post on domain names:
http://blog.stackoverflow.com/2010/07/domain-names/
From the quote at the end about them “helping out” with the domain name it should be crystal clear what’s going on. StackExchange is a grab for not only all content generated by the effort of you and your users, but now he’s wrangling for control of people’s domain names as well.
Would you build an office on leased land, using leased bricks? Would you let some other company control your customer database and limit how you access it? This is not about ensuring privacy, it’s about control and lock-in. Look at how they posted info about traffic levels to all member sites under the original program. No doubt that’s info those sites (some paying customers!) would have preferred to keep private.
I wish you the best but watch out.
Cheers,
Matt
#9 by Rob J Hyndman on 16 July 2010 - 10:03 am
Quote
I fail to see why the SE folk generate such strong reactions. I’ve read the blog article and it seemed fairly sensible and harmless to me. The statistical analysis site is not going to be *mine*, it will be a community site. The participants are not customers, but other members of the community. And I’m very happy that SE are willing to provide the platform and infrastructure to host it.
I must be missing something, but there seem to be a lot of people with strong opinions on this.
#10 by Manuel Ramón on 30 June 2010 - 12:34 am
Quote
Hello Rob
I have joined the Statistical Analysis site but I can’t find how to propose new questions or how to vote the existing one. Can you help me?
#11 by Rob J Hyndman on 30 June 2010 - 3:44 am
Quote
Hi Manuel. The site is now in the commitment phase which means it is no longer accepting new questions or votes. Once we have enough committed users, it will move into the beta phase when I imagine genuine questions and answers can be added. The best thing to do during the commitment phase is to encourage everyone you know who might be interested to join up.
Pingback: “Statistical Analysis Q&A website” did StackOverFlow just lose it to MetaOptimize? and is it good or bad? | R-statistics blog