Jeromy Anglim's Blog: Psychology and Statistics


Wednesday, March 30, 2011

How to Ask Me a Statistics Question

This post includes instructions on how to ask me a written statistics question.
The old way
I receive a lot of statistics questions by email. Answering such questions by email is inefficient.
  • A lot of questions re-occur.
  • Answers only help a single person.
  • Senders only get my suggestions, and not suggestions from other experts.
  • Sender do not get much feedback on how to improve in the art of writing effective questions.
A better way
A better model involves using http://stats.stackexchange.com . It offers many benefits.
  • Many experts have an opportunity to provide an answer.
  • The questions can be improved and refined.
  • Questions and answers are accessible from search engines. Thus, the question and answers become an ongoing Internet resource that are typically read by many hundreds of people.
  • The answers are under a Creative Commons licence which means that they can be reused in other forms.
  • Learning to write effective questions on question and answer sites is an important skill.
Procedure for asking a question
  1. Go to http://stats.stackexchange.com/ .
  2. Create an account
    • You can create an account with a non-identifying user name if you prefer.
  3. Read the FAQ.
  4. Click ASK QUESTION.
    • Enter your question title, description, and any relevant tags.
Important things to remember:
  • Be prepared to edit your question to provide additional information in response to comments; be prepared to respond to people who propose answers and explain why the answer is helpful or not. Asking a question is only the beginning of the process. The people answering the question are volunteers who are often motivated to create a question and answer combination that will be helpful for others in the future. You should try to facilitate this process.
  • When asking questions, show what initial research you have done to answer the question (e.g., Google searches, reading a textbook).
  • Try to extract the general question out of the specifics of your project, so that answers will be of greater use for future readers.
  • Include background information that might be needed in order to answer your question effectively.
  • If your question includes multiple parts, then you may need to ask separate questions.
  • If you are concerned about anonymity, create a user name that is non-identifying, and, if necessary, tweak your question to make any project-specific details a little more abstract.
  • Your question may be clearer if you break up your question into (a) context; (b) questions; and where appropriate (c) your initial answers.
  • If your background is psychology, be aware that users on stats.stackexchange have diverse backgrounds. Clearly explain any psychological terms or concepts.
Asking me a question
  • If you are a research student or academic where I work in the School of Psychology, Deakin University, then send me the URL to the question by email, and I will do my best to answer your question.
  • For all others, if you don't get a good answer in a few days and you think that I'd be well suited to answering the question, feel free to send me an email including the web address for the question, or post it below. Of course, I can't promise that I'll have time to answer the question.

11 comments:

  1. This is the short link to this page:

    http://bit.ly/statsq

    ReplyDelete
  2. Hi Jeromy, I have asked a question on Stack Exchange but have not received an answer. It's possible no one has looked at it yet. Would you be able to comment on it please? the link is http://stats.stackexchange.com/questions/24492/moderated-regression-for-categorical-within-subjects-variable

    ReplyDelete
    Replies
    1. Here's an answer:

      http://stats.stackexchange.com/questions/24492/moderated-regression-for-categorical-within-subjects-variable/24679#24679

      Delete
  3. Hi Jeromy,

    Can you please answer this question - http://stats.stackexchange.com/questions/25488/meta-analysis-and-correlation

    ReplyDelete
    Replies
    1. Here's an answer: http://stats.stackexchange.com/a/25512/183

      Delete
  4. Hi Jeromy,
    Would it be possible to help me out with this problem?
    http://stats.stackexchange.com/questions/27588/problem-understanding-what-type-of-test-to-use-and-how-to-proceed-with-the-quest
    Thanks.

    ReplyDelete
    Replies
    1. It looks like you've only just posted the question an hour ago; let's see if the community can give you an answer first. Also, see the FAQ about homework questions:

      http://meta.stackoverflow.com/questions/10811/how-to-ask-and-answer-homework-questions/10812#10812

      Delete
  5. Hello Jeromy,

    I asked a question a few days ago at http://stats.stackexchange.com/questions/68927/significance-of-adjacency-in-correlation-matrix-with-ordered-variables and still haven't received an answer. Do you think you could take a look?

    Thank you very much. Best, Sandra

    ReplyDelete
    Replies
    1. @Charcha Unfortunately I don't understand your question. I'm not sure if it is the question which is unclear or just because I'm unfamiliar with your evolutionary models.

      Delete
  6. Dear Jeromy,

    Below is my question http://stats.stackexchange.com/questions/69612/friedman-test-for-categorical-data

    I would like to know what kind of test you will advise me to check my hypothesis if Friedman is not the proper one in my case?

    ReplyDelete
  7. Dear Jeromy,

    is there a chance that you could have a look at my question about modeling nominal variables in JAGS ?

    http://stats.stackexchange.com/questions/95929/bugs-jags-for-nominal-response-model

    thank you very much

    J.

    ReplyDelete