Lecture 14: P-Value

P-Value = 0.5 at 0.05 significance level

Large p-value (>0.10) supports H0 that the hypothesized distribution might be a POOR fit

Alpha = 0.05, Critical value = 3, Test statistics = 8; we conclude:

P -Value and critical value approaches might give different conclusion

Always inspect the automatic selection using graphical analysis (visual inspection)

If nothing fits, this ALWAYS means that your data are unacceptable or have somehow failed

P-Value is the maximum probability of rejecting a true H0 by mistake

Alpha = 0.05, Critical value = 10, Test statistics = 8; we conclude:

(Type I error): rejecting Ho when it's true

If P is Low, Ho must Go!

P -Value and critical value approaches used to decide on "Reject" or "Don't Reject" the null hypothesis

If test statistics > Critical Value , H0 is NOT rejected at significance level α

P-Value = 0.0004, For a user whose tolerance for type I error is 5% (alpha=0.05), he would choose to NOT reject H0

P-Value = 0.02 at 0.05 significance level