As discussed, proportions are being affected by the fact that we have wildly uneven number of elements between subsets.
If there is a possible alternative to this, I think it is already there as well. We are just using a different denominator for the percentages. Also this answers a different but related question, something like this: What percent of respondents with less than a year experience use this or that?
@michaelquiapos, @Gio-Q, do you have other plots in mind that can be used alternatively?
As discussed, proportions are being affected by the fact that we have wildly uneven number of elements between subsets.
If there is a possible alternative to this, I think it is already there as well. We are just using a different denominator for the percentages. Also this answers a different but related question, something like this: What percent of respondents with less than a year experience use this or that?
@michaelquiapos, @Gio-Q, do you have other plots in mind that can be used alternatively?