Exam 1

1. Metainformation

   Tag	Value
   file	Assumptions_uu-Homoscedasticity-025-en_uu-Homoscedasticity-025-en
   name	uu-Homoscedasticity-025-en
   section	Assumptions/Homoscedasticity
   type	schoice
   solution	FALSE, FALSE, FALSE, TRUE
   Type	Interpreting graph
   Program	SPSS
   Language	English
   Level	Statistical Literacy

   Question

   There is heteroscedasticity if…

   
   
   1. FALSE: … the explanatory variables are highly correlated among themselves
   2. FALSE: … the measurement errors (‘errors’) are not normally distributed
   3. FALSE: … the relationship between at least 1 of the X variables and the Y variable is nonlinear
   4. TRUE: … the variance of the measurement errors (‘errors’) is not constant across the scores of the X variable(s)
   
   

   Solution

   In regression analyses the condition of homoscedasticity applies: The variance in y is the same for all values of x. When the variance of measurement errors is not constant across scores of x we speak of heteroscedasticity (see figure below).