For publication in peer-reviewed journals, why are there some perceived biases when your paper uses either a truncated or a censored regression model, even when you still make strong and valid assumption on the distribution of the error?
For censored regression, the nature of the dependent variable is the one that matters!!!! Making an assumption is not really the deal. For truncated regression, the nature of both the dependent and covariates matter, still just making an assumption is not the deal.