8
$\begingroup$

I don't know much about statistics, and I don't know enough to know e.g. terms to search for in a web search. However, I have a specific data set, and a list of goals I'd like to achieve in an analysis. If I describe my intentions and what I am trying to achieve, as well as list the specific pieces of information I'm trying to derive, can I ask for a recommended technique to do that?

| |
$\endgroup$
12
$\begingroup$

It's okay to ask questions about suitable analysis (indeed it's the first thing in our list of what things it's okay to ask about), but you have to be careful - with that sort of question it's easy to ask a question that's too broad, unclear or off-topic.

To ask a good question of this kind you should be very clear about what you wanted to find out. You mentioned "goals" and "what you're trying to achieve" and that's certainly the right kind of thing to focus on, but you need to be as specific as you can about what things you want to know (some degree of development of that is possible in answering your question, but then you need to do more to help set up the framework of what the range of choices might be).

On the other hand, if at all possible, avoid jargon words and avoid anticipating the answer (in particular, avoid words like "significant" in describing what you're after, and avoid assuming that the right way to answer the underlying question is with a hypothesis test, say). Try to phrase your goals in plain English (as if you were explaining it to someone without any statistical knowledge at all).

You should describe your variables (e.g. are they measurements like lengths? times? are they counts? etc) and anything you've done to modify the original data (e.g. if you calculated ratios of things, it might be important to understand things about - and even to base the analysis you want on - the original values rather than their ratio)

It may be relevant to mention the audience - who will this information your be going to? Analysis suitable for publishing in a newspaper would be different (at least in emphasis) from analysis suitable for a thesis for example.

That seems like a lot, but by the same token, try to be reasonably brief, or if it must be long, try to organize the basic, core information into an opening summary paragraph and to structure the rest of it into sections so it's easier to find information.

You should be prepared for people to ask questions! It may require several edits to your question over a few hours or days.

| |
$\endgroup$
5
$\begingroup$

http://stats.stackexchange.com is a fine place for you to do that, especially if you have first taken our tour. It should help you refine what already sounds like a pretty well-thought-out question.

| |
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .