>>Return to Tell Me About Statistics!
When performing a test, you check for the presence of an asterisk (*); however, you may not check the p-value properly.
Many statistical software packages, including Excel, yield test statistics, such as t and F, and p-values in the results. Additionally, one asterisk is the output when p < 0.05 and two asterisks, when p < 0.01.
So what do we do when the p-value is exactly 0.05?
When p = 0.05
P is the p-value for the statistical hypothesis testing. When performing a test, it is considered significant if it satisfies a set criterion; the most commonly used criterion being p<0.05. Therefore, if p = 0.05, it is not significant.
It may be very rare for actual clinical trial results, for example, to show p = 0.05; however, if such a result were to be found, it would be troubling.
Rules for writing papers
When submitting a manuscript for publication, it is generally accepted that the following should be done: In Materials & Methods, it seems to be fine to specify that “p ≤ 0.05 is statistically significant” and that it was “significant.” Even if it is 0.05, it is a rule determined by statistics, and is a guideline for whether an event will occur 5 times out of 100 times, so it seems that there is no need to be too particular about “≦” or “<.”
In fact, there are papers that state that “p ≤ 0.05 is considered statistically significant,” so first of all, you should clearly state this when you submit your paper, and if you receive any comments from reviewers, follow them.
Most statistical software yields p-values of three or four decimal places. Thus, when I say 0.05, it is only 0.050.
If 0.054 is rounded to the third decimal place using Excel’s function and the output is 0.05, it cannot be said that there is a significant difference, even if p < 0.05 or p ≤ 0.05. Therefore, we checked the numbers beyond the third decimal location.
Conversely, in Excel, p=0.050 may be the output, with an asterisk (*) attached to the judgment. As Excel also calculates p-values after the fourth decimal place, a value that was actually slightly below 0.050, such as 0.04987, may have been rounded to 0.050 on the display as the output.
Therefore, when checking the test results, it is necessary to check not only the asterisks, but also the p-values to three decimal places.
>>Return to Tell Me About Statistics!
Comments are closed