Most of us could probably help you, if we understood more
clearly what the problem is. It's hard to tell what would be
most helpful (for you) and efficient (for us)
I presume you have a statistical textbook, perhaps more than one,
else you could not have referred to several t-tests. Your
textbook should address the questions you raise. It would seem
then that either (a) you haven't found the part(s) of the textbook
that deal with these questions, or (b) you haven't understood what
it was trying to tell you. I've embedded some questions for
clarification in your post, below. Perhaps you could also
describe for us at least one instance where you're pretty sure one
of those tests is appropriate, and tell us what type of data id
involved and why you think the test is appropriate.
jeanette <[EMAIL PROTECTED]> wrote:
> I have a problem figuring out what statistical test to use on
> different types of data.
Selecting an appropriate test depends on both the type of data and
the question you want to ask of the data. What "types of data" do
you have in mind? And what question(s) do you want to ask of them?
> When do I use the one- sample t-test?
> paired t-test? f-test? independent samples t-test?
The labels ("one-sample", "paired", "independent samples") were
intended to help describe the situations where each particular
form of t-test is likely to be useful. What do you understand
them to mean, in the context of your question?
The term "f-test" (by which I suppose you mean "F-test") can
mean any of a number of quite different tests, although most of
them share a common underlying logic. In particular, whenever a
t-test (of whatever kind) is appropriate, there is (nearly always)
a corresponding F-test; the reverse, however, is not true.
What sort of F-test do you mean, so far as you can tell?
> Could anyone help me?
I don't know if this will have been helpful, but it's about the
best I can do at the moment with these rather vague questions.
-----------------------------------------------------------------------
Donald F. Burrill [EMAIL PROTECTED]
56 Sebbins Pond Drive, Bedford, NH 03110 (603) 626-0816
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
