0

I am just wondering if my data violate one of the Gauss-Markov assumptions, which is homescedasticity. How should I go about checking if my data violates homoscedasticity assumption?

user122358
  • 1,303
  • 2
  • 13
  • 28
  • 3
    There are certain tests you could use, White and Breuch-Pegan are two different tests implemented in most software – Repmat Mar 18 '17 at 07:36
  • 1
    plot the residuals ;-) – ocram Mar 18 '17 at 07:45
  • 1
    This has been tackled in multiple posts here previously. See the following: http://stats.stackexchange.com/questions/254801/should-the-response-variable-meet-the-assumption-of-constant-variance http://stats.stackexchange.com/questions/93464/homoscedastic-and-heteroscedastic-data-and-regression-models – Bach Mar 18 '17 at 09:30

1 Answers1

2

Look at a plot of residuals versus predicted values. To be really thorough, you should also generate plots of residuals versus independent variables to look for consistency there as well.

If the dependent variable is strictly positive and if the residual-versus-predicted plot shows that the size of the errors is proportional to the size of the predictions (i.e., if the errors seem consistent in percentage rather than absolute terms), a log transformation applied to the dependent variable may be appropriate. Heteroscedasticity can also be a byproduct of a significant violation of the linearity and/or independence assumptions, in which case it may also be fixed as a byproduct of fixing those problem.

Try arch.test package in R, which implements Engle's ARCH (Autoregressive conditional heteroskedasticity) test.

Tarun
  • 36
  • 3