The independent samples t-test first does an F test on the equality of variances. The p-value for that is labeled “Sig.” You can use that test to choose which of the two t-tests to use. If the F is not significant (i.e. larger than 0.05), you would look at the top t-test for “Equal variances assumed” and its p-value labeled “Sig.(2-tailed).” If the F is significant, then you would choose the bottom t-test labeled “Equal variances not assumed” and it’s p-value, also labeled “Sig(2-tailed).”
An alternative approach is to never assume the variances are equal and so always choose the bottom t-test.
For t-test paired samples look at the column p-value.
For t-test one sample look at the column Sig.(2-tail).
But any idea why Sig instead of P-value ? I am studying using Python , R , Minitab and all came out with P-value. Sig usually means Significance level.
Such as 5% significance level or alpha means 95% confidence level and if P is lower than 5% then do not accept null.
I thought Sig here means that Significance Level and not the P value.
The independent samples t-test first does an F test on the equality of variances. The p-value for that is labeled “Sig.” You can use that test to choose which of the two t-tests to use. If the F is not significant (i.e. larger than 0.05), you would look at the top t-test for “Equal variances assumed” and its p-value labeled “Sig.(2-tailed).” If the F is significant, then you would choose the bottom t-test labeled “Equal variances not assumed” and it’s p-value, also labeled “Sig(2-tailed).”
An alternative approach is to never assume the variances are equal and so always choose the bottom t-test.
For t-test paired samples look at the column p-value.
For t-test one sample look at the column Sig.(2-tail).
Oh and I went to Analysis -> Means -> T-test , Independent Samples.