October 12, 2004 at 12:13 am #37185
Read in some material, we can also use confidence interval to determine significant, “It is means significant if no superposition of confidence interval, and means insignificant if has superposition”
Do you agree it or you have some comment on it?
Tks!0October 12, 2004 at 2:12 am #108917
Just as mjones said in previous, the test looks at the two distributions and compares them.
For the one sample T, we simulate a distribution and compare to our target mean, if the mean is in the confidence interval, we can determine it is insignificant ,otherwise significant.
Is it right, and can somebody expand it to two sample T test?
Tks!0October 12, 2004 at 3:10 am #108922
mjonesParticipant@mjones Include @mjones in your post and this person will
be notified via email.
For 1 sample t: to be precise, I hope you do not actually “simulate” a distribution; I hope you sample the data and perform the proper t-test. And yes, if the hypothesized mean is outside the confidence interval (CI), you can conclude there is a statistically significant difference. But, you cannot do the opposite, i.e., if the hypothesized mean is within the CI, you obviously cannot conclude there is a difference but you also cannot conclude they are the same! All you know is that your data does not show a difference at this point.
For 2-sample-t, the test is on the difference in the means. Minitab will give you a CI for the differences. If the CI includes 0, you cannot conclude there is a difference at the alpha value or CI you have chosen.0
The forum ‘General’ is closed to new topics and replies.