1

I am trying to interpret data from previous research and I don't understand what the individual statistics (F, p, and np2--partial eta squared) are telling me about the data, or how they relate to each other to convey information about the data. For example, here are some of the information from Spruyt et al., 2009:

The 2 groups are people with synesthesia and nonsynesthetes They are each given 2 Stroop Type color priming tasks, one synesthetic task and one standard task The levels for each task are congruent and incongruent

What I don't understand is how they make inferences based on the following data:

  1. A significant color priming effect emerged in the standard task, F(1,22)= 43.87, p<.001, np2= .53
  2. Color priming effect was reliable in the nonsynesthetic F(1,22)=13.10, p<.005, np2= .37 as well as the synesthetic group F(1,22)= 24.39, p<.001, np2= .53

I have no idea what the partial eta squared means, but I have come to the understanding that if F is bigger than 1 it is significant.

  • But then, what does the p value have to do with it?
  • Would a big F and p<.001 be more significant than a big F and p<.005?
  • If the p values are the same and both Fs are large but one is 3 times larger than the other, is the larger F more significant?
  • How do I know what effect or interaction is more significant or prominent than another?
kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Lindsay Smith
  • 11
  • 1
  • 1
  • 2
  • 2
    You might find the following threads helpful to start with: [What is the meaning of p-values and t-values in statistical tests?](http://stats.stackexchange.com/questions/31/) (note that F is analogous to t-values here), & [How to interpret and report eta squared / partial eta squared in statistically significant & non-significant analyses](http://stats.stackexchange.com/questions/15958/). – gung - Reinstate Monica Aug 02 '13 at 15:56

0 Answers0