Joseph-Watts
commented
3 months ago Thanks for the email. My emails are getting bounced back from your spam filter so I'm going to reply here.
Hi Robert, Thanks very much for the reply and explanation. Your response makes sense and I agree there is probably quite a bit of work needed to think through and implement the cross-variable summaries. The course I am teaching starts in a month so there isn't much time for me to think this through before teaching. I think I’ll look into other options for the moment and come back to you if I have more concrete ideas and/or suggestions for how this could be implemented. Glad to hear that things are well and sorry to hear about the water in the basement. Hopefully that is all sorted now! Much appreciated, Joseph
Hi Robert,
Hope all is well! I’m emailing to see whether you have the time to add a small additional feature to the D-Place website?
I’m hoping to use D-Place as a teaching resource where students generate and perform basic tests of some hypotheses. To do this, it would be useful to be able to get a little more information on the relationship between two variables.
Currently it is possible to visualise how two variables are distributed on a map, and there is a long table that provides a list of each culture and how it is coded on each variable. There is also some basic summary information for each individual variable displayed (e.g. number of societies coded at each level of categorical variable).
However, I can’t see anywhere that provides information on the relationship between two variables. It would be great if some basic summary information was presented. For example, a crossed table when there are two categorical variables would be a big help. I guess if there was a categorical variable and a scalar variable the median, mean and SD for each level of a categorical variable could be reported? Do you think something like that might be possible?
There wouldn’t need to be any statistical tests, as I realise these often involve assumptions that you probably don’t want to make about the nature of the data.
Alternatively, providing a way for users to download the tables that are represented on the pages (e.g. columns for society, variable 1, variable 2) would be really useful. These are undergraduate students so I don’t think they will be up to downloading and processing the full datasets themselves unfortunately.
Much appreciated, Joseph