[S] A Modest Pie Plus

Hoekstra, Mike (rth6@cdc.gov)
Fri, 13 Nov 1998 11:20:58 -0500


While I am not a fan of pie charts for comparative display, I do have one
experience of possible interest.
The ActivStats product from DataDesk has a randomized controlled experiment
to compare ones ability to
estimate the percent area displayed by a single pie wedge and a single
segment of a bar. The trials generate
images where the percentage is chosen at random, roughly 1 to 99 %. The
wedge rays and the segment limits are also
randomly assigned. I played this game about a year ago.
In this simple setting I found that it was much easier(by about 3% in
average absolute error)
to estimate the wedge percentage than the bar percentage.
I could use symmetry to assign the wedge to the nearest eighth of a circle
and estimate the remainder from there.
I could not place the bar fraction accurately to the nearest 12.5% and thus
could not estimate as well!

Robert M. Hoekstra
Centers for Disease Control
1600 Clifton Rd.
Mailstop C09
Atlanta, GA 30333
rth6@cdc.gov

-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news