Hallo, I would need to calculated the exact value of LSD when computing ANOVA, so not the differences between the pairs , but the LSD itself.
IS it possible or do I have to do it manually?
May I clarify if you want to calculate LSD p-value for the overall ANOVA? Not the LSD for pair-wise comparison?
LSD is Fisher's Least Significant Difference (LSD) method, which is equivalent to a t-test for independent samples, based on the sample sizes in the groups involved in the comparison. Therefore the Fishers LSD test is basically a set of individual t tests, and as known, each t-test only involves two groups for comparison, so each LSD test would only compare two groups as well.
In ANOVA, we would not be able to calculate LSD p-value for overall comparison of 3 or more than 3 independent groups, but F-test p-value is provided in Statistica for overall ANOVA comparison. Instead, the LSD test could be used to calculate post-hoc pair-wise comparisons, which involve only two groups for comparison for each LSD test. Statistica provide LSD test for pair-wise comparisons.
You can take a look of our Statistica online help page on ANOVA.
Hope this helps.