I have a file I’m using for our Data Clubs project – it’s data we imported from the Pew Center. There are lots of categorical and ordinal variables in it (not sure that’s relevant, but it seems to be). One of the variables is “How often do you use the Internet?” and it has values seldom, once/week, several/week, once/day, several/day and constant. If I make a graph with that variable on the X-axis, I get what I expect. But if I then plot another categorical variable on the Y-axis, the categories on the X-axis get messed up. See the attached screenshot. This doesn’t happen with every categorical variable plotted on the Y-axis. I haven’t yet figured out the pattern. Here is a link to the dataset.
Thanks, Bill. We think it has to do with missing values, too, but couldn’t do much more diagnosis than that. Do let us know when it’s been fixed, since it’s part of a module we’re teaching in February.
And thanks for the Google drive link warning…we generally use sharing, but I forgot..
Also, I think the timestamp on the forum is buggy – it’s telling me that you sent your reply at 5:49PM on 12/25. Unless you’re in Europe somewhere?
Thanks, Bill! I’m surprised no one reported that previously – but maybe people haven’t been using datasets with lots of missing values. At any rate, glad to hear that it’s been fixed and we’ll look forward to seeing the new version in January.