I think the proper method to use is a linear regression on the historical pass rates.
Then you can construct a confidence interval. Just sayin
I think the proper method to use is a linear regression on the historical pass rates.
Then you can construct a confidence interval. Just sayin
Why _ linear _?
You are right. Not necessarily linear. Was just trying to be funny.
Would likely use an autoregressive model or other time series technique…which would likely result in log transformation or first difference