3 ms·
In an ideal world, I don't think opinionated manipulation of the underlying dataset is a good idea. If you want to avoid outliers dominating your analysis, you
by mayneack 3y ago
In an ideal world, I don't think opinionated manipulation of the underlying dataset is a good idea. If you want to avoid outliers dominating your analysis, you should aggregate it in an outlier resistant method like using the median or other percentile based strategy.
If you allow for manipulating the data, especially on such a politically charged subject, the manipulation method itself will get gamed like it seems is happening here.