Folks,

I totally agree with everyone who is saying it is absurd to report an exact p-value. I have included R^2, and eta^2 and any other effect size I can report. I won't name the journal but I've long been of the opinion that some disciplines know a lot less about doing research than they would like the rest of us to believe. By including more details, no matter how inane, it makes them look like serious scientists. I ran into this when working with some folks at an not to be named university in a not to be name country.

I know that there is little, if anything, to be gained by reporting p to some extreme. I believe it would be ridiculous to report 6.33322532055e-21 which SPSS was kind enough to report as .000. For p's like that I am just going to report <.001 with a note in the cover letter explaining the situation; for p's such as .0034########, I will report as .0034. If that does not satisfy the editor, then I will give them the full p and they can decide how they want to report it. I'll wager they truncate it to <.0001 or something that looks really, really, really significant. And, then, by golly, we've got something that'll stand the world on its ear.

Tony