Do you google?
According to
http://www.statit.com/,
The Paynter chart is similar to a Pareto except that each Pareto bar is
split up into subgroups. This enables you to look at what subgroups are
going into the makeup of the pareto bar and you can use it to spot
trends or troublespots.
The subgroups might be date (as below), department, or similar subdivision.
You need to have data set up like this:
Defect Jan-04 Feb-04 Mar-04
Broken 25 18
Missing 31 29
Wrong 14 12
....
Construct a clustered column chart of this data.
Strictly speaking, the pareto should be sorted from highest defect
occurrence to lowest. In another column, you should have a total of all
the occurrences of each defect, either for the subgroups shown or for
all subgroups. Sort the data by Total, in decreasing order, then make
the chart. Some variants have this total number in the chart as a label
above the clustered columns. Use an invisible line series (no markers
and no lines) and a Chart Label addin** to put these where you'd like them.
** Both of these are good, free addins
Rob Bovey's Chart Labeler,
http://appspro.com
John Walkenbach's Chart Tools,
http://j-walk.com
- Jon
