After using Bartlett test statistics for equal variance to test my data, i concluded that there was significant difference in the variance at alpha = 0.05. I transformed the data using Box-Cox family of power transform and tested for equal variance and it still not equal. what next to do? I used all the different transformation, but still not stable. My estimating parameter is 0.86

  • $\begingroup$ There's not enough information to say much of anything. How does the Bartlett test work with time series - doesn't it work on grouped data and rely on independence? Why are you transforming, exactly? What does the variance of residuals look like both of the original and the transformed data? $\endgroup$ – Glen_b Feb 24 '14 at 7:29
  • $\begingroup$ i tested if my time series data had a constant variance using the Bartlett test statistics and found that there was significant difference in the variance (of each year). I transformed the data using box-cox power transform (variance stabilizing transformation) and the variance is still not stable. $\endgroup$ – uchembaka Feb 24 '14 at 8:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.