Question: When a model has a very large number of predictors

When a model has a very large number of predictors, even when none of them truly have an effect in the population, one or two may look significant in t tests merely by random variation. Explain why performing the F test first can safeguard against getting such false information from t tests.

View Solution:


Sale on SolutionInn
Sales0
Views73
Comments
  • CreatedSeptember 11, 2015
  • Files Included
Post your question
5000