dicook
commented
2 years ago Should we tag versions of the code, and data, to make sure these all persist into the future? Or put them on zenodo, and change the link in the paper, when copy edits are requested?
Page 31, line 43: I'm still confused about this sentence. You say the data are reasonably similar and provide plots and tables demonstrating the similarities, but then it's "disappointingly far from it"? Is the "disappointment" just that the wages aren't corrected?
Just remove " but disappointingly far from it"
Page 22, line 9: The "robustness weight" can be read in different ways; e.g., Huber and Bisquare are two options for weighting in an iteratively reweighted least-squares model (which is what rlm fits). Specifying your choice makes it clear to the reader what you mean by "robustness weight." It sounds like you used Huber (which gives non-zero weight to all observations) and you allow users to investigate thresholds in the Shiny app.
Change "The robustness weight is used" (page 20 by my view of the pdf) to "The robustness weight (calculated by the Huber method) is used"