Exam 1

1. Metainformation

   Tag	Value
   file	Assumptions_vufsw-linearity-0279-en_vufsw-linearity-0279-en
   name	vufsw-linearity-0279-en
   section	assumptions/linearity
   type	schoice
   solution	FALSE, FALSE, TRUE, FALSE
   Type	conceptual
   Program	NA
   Language	English
   Level	statistical literacy

   Question

   What are the consequences of violating the assumptions of linear regression?

   If the assumption of linearity is violated, then …

   
   
   1. FALSE: Are both the coefficients and the significance tests reliable.
   2. FALSE: The coefficients are reliable, but the significance tests are not reliable.
   3. TRUE: The coefficients are not reliable, but the significance tests are reliable.
   4. FALSE: Both the coefficients and the significance tests are not reliable.
   
   

   Solution

   If the assumption of normality is violated, the significance tests are not reliable. However, for large samples by the Central Limit Theorem, the distribution of residuals is approximately approximate normal. So, then significance tests can also be reliable.
   The violation of the assumption of homoscedasticity affects the significance tests (which cannot be trusted), but not the coefficients.
   The violation of the assumptions of linearity and no influential observations affect the coefficients (which are not a good reflect the relationship), but not on significance tests.

   
   
   1. False
   2. False
   3. True
   4. False