[PredictorX1]

I've seen published estimates of statistical efficiency for a few test distributions for the trimmed mean, but not the Winsorized mean. Can anyone suggest references on this?