Interpreting data from multcompare
in [Matlab]
|
Prev: Fsolve, CAT error : CAT arguments dimensions are not consistent
Next: How to close Excel workbook from Matlab
From: Tom Philp on 13 Nov 2009 11:45 I am having difficulty interpreting what the final paragraph below means in the quoted multcompare help section: "For example, suppose one row contains the following entries. 2.0000 5.0000 1.9442 8.2206 14.4971 These numbers indicate that the mean of group 2 minus the mean of group 5 is estimated to be 8.2206, and a 95% confidence interval for the true mean is [1.9442, 14.4971]. In this example the confidence interval does not contain 0.0, so the difference is significant at the 0.05 level. If the confidence interval did contain 0.0, the difference would not be significant at the 0.05 level." I have data that looks like this: 1 2 0.1026 0.1816 0.2606 1 3 0.0309 0.1099 0.1889 1 4 -0.325 -0.246 -0.167 1 5 0.1858 0.2648 0.3438 2 3 -0.1507 -0.0717 0.0073 2 4 -0.5065 -0.4276 -0.3486 2 5 0.0042 0.0832 0.1622 3 4 -0.4348 -0.3559 -0.2769 3 5 0.0759 0.1549 0.2339 4 5 0.4318 0.5108 0.5897 Does this mean that as group 1 and group 3 have a lower confidence interval that starts with 0.0 they are not significantly different, and the same for groups 2 and 5, and 3 and 5? And then do I look at the upper confidence level and do the same? This would make groups 2 and 3 not significantly different also. My confusion comes as according to the graph that multcompare produces only groups 2 and 3 overlap and are therefore the only groups that are not significantly different. So is it only the upper confidence interval that I should look at to deduce significance? Thanks
From: Tom Philp on 13 Nov 2009 14:43
I have worked out my confusion - by saying the interval didn't contain 0.0 it means that both intervals are either positive or negative (and don't cross 0). If one interval is negative and the other is positive then they do cross 0 and the groups aren't significantly different. |