Measures of selection

[Andee Rubin]

We’d like to use the “measures of selection” feature – it seems like just what we need in our analyses – but it doesn’t seem to work the way we expect.  We have a graph of a single numerical distribution with the mean plotted.  We make another graph with a categorical variable (gender), select all the males and see those cases selected on the original distribution.  Then, when we choose “show measures of selection,” we expect to see the mean of the males.  Is this our problem or CODAP’s problem?

[Bill Finzer]

I’m so glad you ask. This feature came about as kind of special for the Zoom In Science project. It works just as you describe for a LSRL in a scatterplot. We never got around to extending it to other measures and I’ve felt very guilty about that.

Now, with your prompting, I’ll move it up to the top of the list!

Bill

[Andee Rubin]

Great – glad I asked, indeed!  This is great news.

 

[Bill Finzer]

Measures for Selection now works for all measures, not just the least squares regression line.

Once enabled from the eyeball menu measures like the mean or median will respond to selection. So, for example, in the Mammals example document, if you are showing the mean of Lifespan on a plot, the mean will not show until you have some cases selected. Once you do, the mean showing will be the mean Lifespan of just those cases.

[Author]

Posts