Jeromy Anglim's Blog: Psychology and Statistics


Friday, February 17, 2012

New Psychology and Cognitive Science Question and Answer Site: COGSCI.SE

There is now a new website for researchers to ask and answer questions on topics related to psychology and cognitive science. The site is cogsci.stackexchange.com. From the success of earlier released sites in the Stack Exchange network such as those on programming, statistics, and latex, the site for psychology and cognitive science has the potential be a great resource for researchers. I'm actively contributing on the site. So, if you are a researcher in psychology, I hope you'll check it out. The rest of this post sets out (a) a little history of Stack Exchange question and answer sites as they relate to psychology and statistics; (b) why I think this new site for psychology and cognitive science has so much potential; and (c) why, if you are a professional or student researcher in psychology, you might want to get involved.

A little history

I first learnt about the Stack Exchange network back in 2009. While I was busy learning R, a number of people in the online data science world such as JD Long, Michael Driscoll, Drew Conway, and many more were promoting a programmer's question and answer site called Stack Overflow as a place to ask and answer R related questions. It was a site pitched at overcoming the many problems of discussion boards, mailing lists, and the like: e.g., off topic threads, spam, extended discussion difficulty finding the correct answers, poor indexing by Google, etc. As of Feb 2011, it now has over 10,000 questions with the R tag.

Shortly afterwards, the Stack Exchange Network developed Area51. The idea was to take the question and answer infrastructure that made Stack Overflow a success in the programming world, and extend it to all sorts of other domains. Instead of going down the model of Quora or, shudder to think, Yahoo Answers, Stack Exchange did not permit the creation of a site until a sufficient community of active users existed to maintain the site at a high standard. Thus, my main interests, statistics and psychology were going to have to wait.

A site for statistics questions was the first to join the network (stats.stackexchange.com). Professor Rob Hyndman proposed the site, and perhaps given the overlapping worlds of programmers and data analysts, the site launched a few months later in July 2010. At the time of posting it has over 7,000 questions. I've been actively involved in the site. I've used it to get advice on my own research. I've also used it extensively in various statistical consulting roles. In particular, I've encouraged others who would otherwise send me an email about statistics, to post the question on stats.se so that any answer can be an ongoing resource for others.

In the case of psychology and cognitive science, I've had to wait a lot longer. The overlap between programming and psychology communities is much smaller, and site proposals were split over separate cognitive science, psychology, and psychiatry proposals. Finally, in December 2011 these three proposals were merged and on January 19th 2012 the site was launched in Beta under the title Cognitive Sciences at the url cogsci.stackexchange.com. Although the initial name is suggestive of a focus on "cognitive" science. The history of the merging of site proposals, the inclusion of the "s", plural "sciences", and the attitude of current participants admits the full range of questions in cognitive science, psychology, and psychiatry.

At the time of posting the site is growing at a healthy rate. Most questions are getting good answers, and the community norms around question quality, references, scope and so on are being clarified on the meta site. However, there is also the challenge of getting the word out about the site to academics, researchers, and graduate students who are not otherwise familiar with the Stack Exchange network of sites. In my opinion, Stack Exchange provides the best currently available infrastructure for building a high quality question and answer site. However, it still relies on a community of expert contributors.

So, if you're a researcher in psychology or cognitive science, why might you want to get involved? And why might you want to talk to fellow researchers about the?

Reasons to participate as an academic

If you are an academic, Lecturer, or Post Doc, there are many reasons why you might want to participate:

  • Answering questions is a way of facilitating knowledge transfer to the broader community; this can be intrinsically enjoyable especially when you get direct feedback on the number of people who read your answers.
  • If you use your own name, as many people do, the voting and reputation system, and various other mechanisms provide a means for your contribution to be recognised.
  • You get immediate feedback on what others think of your answers; Thus, it creates an environment of feedback conducive to learning.
  • I see sites like Stack Exchange as part of a broader model of open science. As you create and develop knowledge, you encounter challenges. The idea is to record these challenges as questions and then add the resolutions as answers. Thus, when others encounter the same problems, good answers are only a Google search away. I'm not saying that question and answer sites replace journal articles, but they can fill a bridging role linking the language of questions to the answers provided in journal articles.
  • Furthermore, the content on Stack Exchange is licenced under creative commons, so even if the site disappeared the content would still be available on other sites that reproduce the material. This is much better than the policy of almost all journals which copyright your, typically, state-sponsored research and lock it up behind a pay wall, thereby frustrating the process of knowledge dissemination.
  • While contributing to Wikipedia is another great way to improve the sum of all knowledge, unlike Wikipedia, your answers generally stay there; in contrast to Wikipedia, where your contributions can and are often completely removed by other editors.

Reasons to participate as a student researcher

If you are doing a thesis in psychology or cognitive science, or possibly even if you are just studying a few subjects, many of the above reasons for participating will also apply.

However, you may also find that the capacity to ask questions will be particularly useful. As a side point, if it is early days in your career, you may or may not want to use your real name.

In particular, I'd encourage students completing a thesis to incorporate asking and answering questions into their scholarly process. You might encounter questions like:

  • Is there a meta analysis on X?
  • What are the main theory about Y?
  • What is the best measure of Z?
  • What is the empirical support for theory W?

These kinds of questions come up all the time when doing research. Of course, as researchers we have strategies for finding answers ourselves. However, the stack exchange model encourages you to learn from the answers of others and also to "leave crumbs" so that others can follow in your footsteps more easily. The idea is not be shy. Post questions frequently. If you're able to answer your question, contribute a self-answer.

Thus, even if only a handful of people ever read a thesis, by asking and answering many questions along the way, resources will be left that thousands of people will learn from and discover through Google searches in the years to come. Even if you don't have answers, your question can be the trigger for an expert to share knowledge to create a valuable Internet artefact.

Getting Started

If you want to learn more or give the site a go:

I'm floating around on the site, so if you're a researcher in psychology, I hope to see you there.