Prism provides the option to calculate a P value for each parameter estimate (and odds ratio) of a logistic regression model.
Although the statistical test has a different distribution, the interpretation of P values when used to assess model parameters for logistic regression is the same as it is with multiple linear regression.
Specifically, the null hypothesis tested is that the true population value of the coefficient/parameter is zero. Note that if the value of the coefficient were actually zero, this would mean that an increase (or decrease) in the associated variable (X value) would have no effect on the log odds (or odds).
For each coefficient, a calculated P value answers the question: if the null hypothesis (above) were true, what is the probability of observing this coefficient estimate or a more extreme one? If this probability is small enough (a P value smaller than the selected alpha level, generally 0.05), we reject the null hypothesis.
For each parameter, Prism reports
•The absolute value of Z, calculated as the coefficient estimate divided by its standard error
•The P value which is determined from Z
•A P value summary, reported as “ns” or as one or more asterisks