Question

If the p-value has to be between 1 and 0, how do we approximate the p-value? For example, the p-value is 0.99999999777 when I use R, but if I put the same information into the Minitab, the p-value becomes 0. What is the rule for an approximation when it comes to the p-value? Is any p-value 0.9 and lower always going to be less than alpha=0.05?

Answer 1

p value is the likelihood of the observed data, given that the null hypothesis is true. The p - value is,in future experiments, the probability of obtaining rsults as extreme or more extreme, given that the null hypothesis is true.

p value is just used in Hypothesis Testing decision rule as follows:

If p - value is greater than significance level , the difference is not significant. Fail to reject null hypothesis.

If p - value is less than significance level , the difference is significant. Reject null hypothesis.

Thus, the role of p - value is only comparison with .

So, in actual situation, we do not need p value to say 10 places of decimals. precision upto 4 places is quite sufficient because it is being compared with whose value is 0.10, 0.05 or 0.001 etc.