In statistics, p-values measure the likelihood that the relationship(s) you are testing occurred as a matter of chance, or luck. The ‘p’ in p-value literally stands for probability. P-values are stated in the form ‘p =’ (p equals) or ‘p <’ (p is less than), followed by a number that is one or less, and stated to three decimal places. For example, p = .133, p<.001.
The number in the statement represents the odds that the tested relationship is due to chance. The statement, p = .133 means that there is a .133 (13.3%) chance that the finding was due to random chance. This also means that there is an 86.7% (1- .133 = 86.7) chance that the finding is not due to chance.
For most purposes, the standard for ruling that a finding is not by chance is p <= .05 or lower (the probability is less than or equal to 5%).
Some fields require more certainty than others, and this impacts the standard for determining if a finding is statistically significant. In fields where determining that something is or is not statistically significant can have life or death consequences, the standard may be as low as p = .01 or even p=.001. In business, it is often not necessary to be so certain that a finding is 95% reliable (p=.05). Nonetheless, 6sense Research adheres to a 95% probability unless otherwise stated.
