Hi All:

I'm curious whether I can create a combination cluster AND stacked bar
chart. I have a group of students I'm tracking out over a number of
semesters. As semesters progress, I'd like to indicate the number that
engage in a certain beahvior (stacked bars). I'm tracking four different
cohorts. Ideally I'd like to lay the cohorts side-by-side. Furthmore, I'd
like to display differences within each cohort by gender, ethnicity, age, and
disability (cluster bars). By way of example, un my mind's eye I'd have a
chart for gender that would group four male student cohorts and female
student cohorts and within each cohort (i.e., horizontal bar) create a
stacked bar that displayed semester-by-semester participation for each
cohort/population. If I'm seeing it correctly, I'd be able to easily compare
changes across time within group (e.g., male students) or display differences
between groups (e.g., male v. female students in the first cohort, in the
second cohort, etc.). Does this make sense or am I completely off my rocker?
Can it even be done? If so, how?

Thanks in advance, your help is very much appreciated!

jim

On Thu, 13 Dec 2007, in microsoft.public.excel.charting,
Jim Fillpot said:

I'm curious whether I can create a combination cluster AND stacked bar
chart.

You can, but it's an exercise in lateral thinking: what does a "stacked"
bar look like when only one of the ranges has any value? Answer, it
looks just like a single cluster bar.

What does the stacked bar where every series except the "cluster" series
has a value look like? Answer, one that isn't the cluster bar.

What does a stack where *no* series has a value look like? Answer, a gap
between clusters. These three techniques together should let you build
exactly the look you want.

(for a bonus, what does a "stacked" bar where one or other series has a
value, but never both, look like? Answer, a bar that changes color
according to a formula, that is, a conditionally-formatted bar)

--
Del Cotter
NB Personal replies to this post will send email to ,
which goes to a spam folder-- please send your email to del3 instead.