What does p value 2.5e-0.5 mean? If it is 2.5 × 10 ^0.5 then it is way above the statistically significant p<0.005.. learning this to interpret the results of one study..
Asked
Active
Viewed 8,819 times
0
-
5Are you sure it's `e-0.5`? That would presumably mean $2.5 \times 10^{-0.5} = 2.5 / \sqrt{10} \approx 0.79$, but that's very unusual use of notation. Maybe it's a typo/mistake in the study. – Danica Jan 29 '17 at 21:09
-
Notice the $-$ sign... $e^{-0,5} = 1/\sqrt{10}$, multiplied with 2.5... Thats about $0.8$... – Repmat Jan 29 '17 at 21:10
-
Could you please tell which software you are using? – Caserio Jan 29 '17 at 21:25
-
5Something more like `2.5e-5` would be much more standard, which would should be read as a computer version of scientific notation, i.e. $2.5 \times 10^{-5}$, which is a pretty small $p$-value. – gammer Jan 29 '17 at 22:01
-
I'm voting to close this question as off-topic because it is a trivial matter that should have no interest to the OP and others here. – Michael R. Chernick Jan 29 '17 at 22:41
-
2Please paste in your actual output so we can determine what the situation is. – gung - Reinstate Monica Jan 29 '17 at 22:54
-
@amoeba, you could, I suppose, although that Q has extra aspects that aren't relevant to this one. I'm not sure why the thread I linked was closed, TBH, it seems on topic enough to me. Either way, tomayto tomahto. – gung - Reinstate Monica Jan 30 '17 at 00:35
-
@gung Yeah, this should be consolidated somehow. I voted to reopen, edited both Qs, and raised a mod flag... – amoeba Jan 30 '17 at 00:42
-
@gung Oh wow, now it's closed as a dup of both! – amoeba Jan 30 '17 at 00:43
-
1@amoeba, I believe that's supposed to happen when more than 1 dup was suggested. – gung - Reinstate Monica Jan 30 '17 at 00:45
1 Answers
0
Taking 2.5e-0.5 at face value, that's $2.5 \times 10^{-1/2} \approx 0.79$ (as pointed out by @Dougal and @Repmat in the comments). The p-value is defined as "the probability of seeing a result as strong as observed or greater, under the null hypothesis."
Combining these two, we have the following: under the null hypothesis, the probability of seeing a result as strong as observed or greater is about 79%. So, given the null hypothesis, the result you have found is not particularly "unusual."
Peter
- 181
- 10
-
6I don't think I have ever seen a computing system in which "2.5e-0.5" would even be considered syntactically correct. – whuber Jan 29 '17 at 22:23
-
@whuber, I haven't either. I use STATA, and if p were equal to 0.79, it would just report 0.79. That said, hopefully OP knows what how to define a p-value now. – Peter Jan 29 '17 at 23:36