0

Could you please explain what does it mean this statement- We drew iid samples of e(j) by jittering and resampling the residuals from a model fit to the original data, scaling them to have standard deviation 0.5.

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Stat
  • 1
  • 1
    Can you give some more context, especially what is meant by $e(j)$? Are the words jittering, resampling and scaling known/clear or will their explanations have to be part of the answer? – Bernhard Jul 11 '21 at 18:34
  • 1
    This sounds like a parametric smoothed bootstrap - they fit a model to their data, then generated new datasets by using their model and the residuals from their original data. The smoothed bootstrap doesn't limit itself to resampling from only your data, but from a smoothed version of your data, which can be thought of as "jittered" data. – rishi-k Jul 11 '21 at 19:34
  • 3
    OP, Could you give the source of your quote? Additional context may help (e.g. occasionally papers or books etc may use terms in an unconventional fashion). It would help if we knew what terms you were familiar with. – Glen_b Jul 11 '21 at 23:12

0 Answers0