(example added) Comparison of MEAN values taking into account QQ-intervals and STD

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Fede - 30 Jul 2009 09:12 GMT

Hi,

I want to compare two distributions and guess if they are different
(p-
value). The only information I have is the Mean, the QQ-intervals
(25-75 percentiles) and the Standard Deviation. What would be the best
strategy to compare them? (may it be the t-test with unequal
variances?)

Thanks in advance,
and have a nice day
Federico

And here is the example:
Distribution 1
Mean = 1.333
QQ-interval = [0.77, 1.984]
STD = 0.8574

Distribution 2
Mean = 1.759
QQ-interaval = [ 1.244, 1.943]
STD = 0.6761

Are Dist1 and Dist2 significantly different?

Reply to this Message

Ray Koopman - 30 Jul 2009 18:38 GMT

> Hi,
>
[quoted text clipped - 20 lines]
>
> Are Dist1 and Dist2 significantly different?

You can't do anything unless you also know the sample sizes. Then
yes, you could do a heteroscedastic t-test. I'm not sure what else
to suggest, although if the samples sizes are equal and small the
homoscedastic t might be better.

Reply to this Message

Math Forum / Mathematics / Statistics / July 2009

(example added) Comparison of MEAN values taking into account QQ-intervals and STD