Minitab assistant: 1-Sample % Defective Test

magicstars

Registered
Hello Masters! I need your help!
Firstly, I run "1-Sample % Defective Test" to get the confidence intervals by using Minitab's Assistant - Hypothesis Tests directly.
Then, I also use excel referring the Formula B2 listed in Minitab Assistant White Paper (as attached) to compare the results.
Whatever total number tested (n) and defectives x>0 Minitab and excel output the same confidence intervals, until I input x=0.
eg. n=30, x=0, to determine if defective<1%.
Minitab Assistant reports 90% confidence interval: 0~7.39%, and 95% confidence less than 7.39% while excel 90% confidence interval: 0~9.5%.
According to Formula B2, 7.39 should be against 90% (not 95%), and this is inconsistent with the results by another way: Stat->Basic Statistics->1-Proportion (95% upper bound).
Has anyone encountered this? Could anyone help clarify me? Many Thanks!
 

Attachments

  • Assistant_One_Sample_Percent_Defective.pdf
    422.3 KB · Views: 140
  • Proportion UpperConf Limits Table-trial.xlsx
    27.3 KB · Views: 138

magicstars

Registered
don't know why but should be
Whatever total number tested n and defectives x>0 Minitab and excel output the same confidence intervals, until I input x=0.
 
Top Bottom