help
help > RE: regression as anova
Sep 30, 2015 06:09 PM | Alfonso Nieto-Castanon - Boston University
RE: regression as anova
Dear Liron,
The p-value that you get for that contrast is what you describe as a "model-level result" that combines all of the individual tests -each represented by a single line of this contrast- into a single p-value. To be more precise, the p-value represents the false-positive level of a F-test looking at any between-group or emotion-related differences in connectivity. For example, a p-value of .01 there means that the probability of observing your data (observed connecivity values) assuming that there were no differences in connectivity between the groups as well as no emotion-related differences in connectivity between your subjects is just 1%. Typically if you get a p<.05 value you would reject this null hypothesis (concluding that there are either between-group differences in connectivity and/or emotion-related differences in connectivity in your data). After reaching this conclusion you typically then perform post-hoc analyses to see whether these results are driven mainly/specifically by differences in connectivity between your groups, emotion-related differences in connectivity between your subjects, or both.
Hope this helps
Alfonso
The p-value that you get for that contrast is what you describe as a "model-level result" that combines all of the individual tests -each represented by a single line of this contrast- into a single p-value. To be more precise, the p-value represents the false-positive level of a F-test looking at any between-group or emotion-related differences in connectivity. For example, a p-value of .01 there means that the probability of observing your data (observed connecivity values) assuming that there were no differences in connectivity between the groups as well as no emotion-related differences in connectivity between your subjects is just 1%. Typically if you get a p<.05 value you would reject this null hypothesis (concluding that there are either between-group differences in connectivity and/or emotion-related differences in connectivity in your data). After reaching this conclusion you typically then perform post-hoc analyses to see whether these results are driven mainly/specifically by differences in connectivity between your groups, emotion-related differences in connectivity between your subjects, or both.
Hope this helps
Alfonso
Threaded View
| Title | Author | Date |
|---|---|---|
| L R | Jul 8, 2015 | |
| Alfonso Nieto-Castanon | Jul 9, 2015 | |
| L R | Jul 10, 2015 | |
| Alfonso Nieto-Castanon | Jul 16, 2015 | |
| L R | Sep 30, 2015 | |
| Alfonso Nieto-Castanon | Sep 30, 2015 | |
| Alfonso Nieto-Castanon | Sep 30, 2015 | |
| L R | Jul 21, 2015 | |
| Alfonso Nieto-Castanon | Jul 24, 2015 | |
| L R | Jul 24, 2015 | |