data science question
MBA_Data_Miner
Member Posts: 21 
Contributor I
question for experts:
if your dependent variable is something like success percentage ( calculated on two other variables total attempts and successful attempts). Success percentage is calculated as follows:
IE Success Percentage (S) = Total successful attempts (A) / Total attempts (T)
Does it make sense to exclude total successful attempts (A) as an independent variable in predictive models using S as the dependent variable? If so, why?
I am thinking it should be excluded to avoid allowing the model to "cheat" by using a variable that is not completely independent to the dependent variable.
if your dependent variable is something like success percentage ( calculated on two other variables total attempts and successful attempts). Success percentage is calculated as follows:
IE Success Percentage (S) = Total successful attempts (A) / Total attempts (T)
Does it make sense to exclude total successful attempts (A) as an independent variable in predictive models using S as the dependent variable? If so, why?
I am thinking it should be excluded to avoid allowing the model to "cheat" by using a variable that is not completely independent to the dependent variable.
0
