After performing statistical analysis, there is the need to report the p-values in the text. Is there an accepted (published) stratification range?
For instance, is 0.04 'slightly' significant? is 0.003 'highly' significant? how shall I refer to 1.3e-5?
Thank you
I think p values speak for themselves to most readers of research literature. One doesn't really need to embellish them verbally, although many people would agree subjectively with your characterizations of .04 and .003. As for 1.3e-5. That's shorthand for .00035. In text, I'd refer to that as
Thanks, I agree. But I remember reading a paper that was giving strict rules for the classification. I just wanted to use that as a guide.
Appropriate vrbal modifiers for levels of significance will vary by the research field. Social sciences typically set alpha at .05, so .04 would just be "significant" and .049 could be "barely significant" and .06 "almost significant." A natural science might set alpha at .01 and medicine might require .001. So any strict rules you find for using modifiers would only be relevant within a particular field.
Personally, I don't like either "slightly" or "highly" significant. For a given alpha, a result is either significant or it's not, so I think "barely" and "easily" are better, grammatically and logically, than "slightly" and "highly."
- Badges
- Science topic