- I add a little - On Tue, 09 Dec 2003 11:50:43 -0500, Paige Miller <[EMAIL PROTECTED]> wrote:
> Jan wrote: > > A very basic statistical problem, i fear, but i can't get it solved. I > > have collected data on the occurence of pathology at female ovaries, > > and graded them according to the severity. Left: 1: 91, 2: 31, 3: 7, > > 4:3; Right: 1:66, 2:28, 3:6, 4:3; totals: left: 132, right: 103. All > > data are collected on females who present with fertility problems > > (which could case a certain bias). To the best of my knowledge, no-one > > has documented a 50/50 spread between left and right (normally there > > should be no pathology! though some asymptomatic women are probably > > around). Q: 1/Which test should I use to compare left vs right againsi > > grading (if i can due to the difference in spread among grades?!)? > > 2/Can I say that left is significantly more affected than right? > > (which test, based on which presumptions)? Thanks a lot!! > > If the severity are normally distributed, you can use a paired t-test. > Do not use the two-sample t-test. Right. You do want to consider the *paired* information, how much worse on one side than the other, and the women with no pathology at all will be ignored. > > If the severity are not normally distributed, you might consider a > non-parametric test such as Fisher's Sign Test. Three more things: 1) With the pathology on a 5 point scale, there is not much chance that it will have outliers so extreme to mess up the paired t-test. 2) That was a minor mis-statement, above, since it is only the *difference score* that must be decently distributed for a paired t-test -- not the separate scores. 3) I thought immediately of applying McNemar's "test for changes," which is exactly the same as (above) Fisher's Sign Test. I was not thinking at all about possible non-normality. It would be hardly any sacrifice in power, and it would be a *clearer* way to describe the result: How many women were (Right worse than Left), versus (Left worse than Right)? Test that against 50%. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html "Taxes are the price we pay for civilization." . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
