1. i find interesting the obvious that ... while students seem to fear
statistics primarily because it contains "math" ... and they dislike math
... since they think math is so OBjective and exact ... while the fact of
the matter is that there are wide variations ( ... our discussion of bar
charts and histograms) in opinions in all matters statistical ... (which if
we conveyed more frankly to students, perhaps they would not be so scared
silly with statistics)
2. the main focus should always be on conveying a sensible and meaningful
"picture" of the data ... that is, we are in the communication business ...
and in that context ... we JUST need to be careful in that the method we
opt for to convey that information does in fact do that ...
the Ntigram ... seems to be just an alternative graphical way to show what
has been known for eons that ... for most distributions ... baseline
distances are NOT equal for a given equal percentile rank difference (say
... 10 percent increments) ... in fact, boxplots do this in a cruder way
... see below
--------------I + I------------- *
--+---------+---------+---------+---------+---------+----C2
15 30 45 60 75 90
the middle 50% of the frequencies fall over a span (from hinge to hinge) of
about 15 points ... whereas the bottom OR top 25% ... span over about the
same distance ... or even more ...
At 01:51 PM 8/14/00 -0700, Tim Erickson wrote:
>First, a comment about the recent conversation; then some new, related
>stuff:
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================