On Tue, 14 Jan 2014 07:24:28 -0800, Rick Froman wrote:
Mike Palij asks:
"It could just be me but what exactly is misleading in the
"uncorrected"
figures. Wikipedia has an entry on misleading graphs that provides (a)
specific types of misleading techniques and (b) Tufte's rules for
measuring
distortion in figures. Which violations of good practice are present
in the
"uncorrected" and "corrected" figures? See:
http://en.wikipedia.org/wiki/Misleading_graph ";
It is what the Wikipedia reference you cited calls a "truncated graph"
(the y-axis doesn't start at 0 or, I would argue, the minimum of the
scale).
Okey, dokey. Is this a rule or a heuristic? That is, "Are all
truncated
graphs misleading?" or are "Some truncated graphs misleading?" If you
believe in the former, then you would say that panel B of the figure
I connected to is wrong and misleading; see:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3361017/figure/F3/
If you really feel this is the case, then I suggest that you write the
editor of the journal "Frontiers in Psychology" and point out the error
and demand a correction.
As a class project I would suggest students find figures, especially
in Tier 1 journals, that use truncated figures and determine (a) does
the truncation facilitate understanding or (b) mislead the reader.
I predict: (1) There number of truncated figures found will be >> 1.00
and (2) the number of misleading truncated figures will be significantly
less than 100%
Question: rule or ritual?
Mike later asks:
"Question 1: How does the uncorrected figures mislead the reader about
this point? For example, how does it imply that the percentage of
female
CEOs in the Fortune 500 is much greater than 4%, under the assumption
one can read the y-axis?"
In this case, the uncorrected graph probably understates the point the
article is trying to make which indicates that not all misleading
graphs are
used purposely to hype the author's point. Sometimes the person just
doesn't know any better.
I don't follow you here since you do not identify (a) what the point is
that the author is trying to make and (b) how the "uncorrected" figure
"understates" the point the article is making.
Quoting the article:
|Today, GM announced its new CEO will be Mary Barra, the
|company's current head of global product development. This
|is big news: Barra is the first woman to head a global auto company.
|This move, as the below Reuters chart shows, is just one small
|step in the march toward management equality. Women are still
|far from reaching parity in upper reaches of big American corporations.
The "uncorrected" figure shows that there is a very slow increase
over time (going from 0% to 4%) while this is obscured in the
"corrected" figure. Indeed, the "uncorrected" figure has 0.5%
units on the y-axis as the basic units while the "corrected" figure
has 20% units (in both cases, horizontal lines are used to show
the y-axis landmark values). Looking at the "corrected" figure
can anyone determine what the actual percentage is? No, because
the y-axis units are too coarse/broad. Seems to me that the
"corrected" figure is misleading or at the very least obscures what
is happening. The additional detail in the "uncorrected" figure
serves to reduce misunderstanding (unless, of course, one isn't
paying attention).
I am also not sure that it would be necessary to show all of the
blank space on the graph up to 100%. Since 50% would suggest
parity, that would seem to be enough to make the point.
But this is a truncated scale. The choice of 50% is arbitrary and
it is unclear what parity is or should be. How about 55% or 60%
or 75% or 90%? Since percents range from 0 to 100, an
untruncated representation should reflect this range. Other choices
just represent a person's preference or bias. Why would this
disallow other truncated scales especially if they are clearer
in what they communicate?
-Mike Palij
New York University
m...@nyu.edu
---
You are currently subscribed to tips as: arch...@jab.org.
To unsubscribe click here:
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=32719
or send a blank email to
leave-32719-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
