go figure
Sooner or later, you’ll want to (or have to) make a graph. Many software
programs can do this, and many books tell you how. But if you really
care about the appearance of your graph, and perhaps more importantly,
about the clear communication of the information contained therein,
there is relatively little software flexible enough to give you
ideal control of the parameters, and there are only
a few guidebooks worth reading.
I’ll cover graphing software in more detail in another entry, and I’ll
focus primarily on two programs – called SigmaPlot
and S-Plus
(available, unfortunately, only to users of Microsoft Windows; a
freeware program called R,
however, shares many of the features of S-Plus) – that enable
the easy creation of publication-quality graphs of many kinds. Here,
I’ll focus on a few of the guidelines or rules to follow when
making graphs, and I’ll illustrate these rules using only one kind
of plot – the scatterplot. In addition, I’ll pair bad
examples with good ones, and I’ll explain why the rules work.
These rules aren’t mine; they have been formulated and put into
practice by many researchers before me. But the person who has perhaps
done the most and best work to understand the theory and practice
of statistical graphs, and to consolidate and illustrate these rules,
is William
Cleveland. His book, The
Elements of Graphing Data, is to graph-makers and
statisticians what Robert
Bringhurst’s The
Elements of Typographic Style is to typographers.
It is a how-to compendium of graph-making, and is indispensable
to those of us who plot data on a routine basis.
Cleveland’s overarching message is: Draw the eye to the data; treat the data
fairly and carefully. Six of what I consider Cleveland’s most important
rules are:
(1) Use a pair of scale lines for each variable. Cleveland
makes a strong argument for table look-up here – that “judging
the scale value of a point by judging its position along a scale
line...is easier and more accurate as the distance of the point
from the scale line decreases.” Compare figures 1a
and 1b
(made-up data set), and notice how much easier table look-up is
when two scale lines are used for each variable, rather than just one.
(2) Make the data rectangle slightly smaller than the scale-line
rectangle. In figure 2a,
the data rectangle and the scale-line rectangle are coincident;
some data points therefore fall on the scale lines and are difficult
to see. A “padding” of 5% is added to the data rectangle
in figure 2b;
all data points are contained within the scale-line rectangle and
are easily visualized.
(3) Use outward-pointing tick marks. Inward-pointing ticks,
as shown in figure 3a,
simply add clutter to the interior of the graph, and in my opinion,
make table look-up more difficult. Compare to figure 3b.
(4) Avoid slavishly including zero on the axes. Cleveland here
refers to the widely-read book by Darrell Huff – How
to Lie with Statistics – wherein Huff says
that a graph without a zero line is dishonest. Cleveland argues
that to include zero, however, may result in a waste of space, and
more importantly, may interfere with our judgment of the data (figure
4a).
Therefore, fill the scale-line rectangle with the data (figure 4b).
Cleveland emphasizes: “Assume the viewer will look at the tick
mark labels and understand them.”
(5) Use open rather than filled symbols to mark the data points.
Invariably, some of the data will fall on or close to the same coordinates;
see the points that lie roughly at (26, 7) in figures 5a
and 5b.
They are hard to distinguish in 5a, in which filled circles are
used to denote the data, but the overlap can clearly be seen in 5b.
(6) If summarizing the data or drawing the eye to them with a line,
use the line that best fits them. It is tempting to superimpose
a straight-line regression fit to the data – this is the easiest
(or only) option in some graphing programs – but it may not
be fair to the data or to the reader. The data set used here has
some curvature, and the straight-line fit shown in figure 6a
does not adequately represent it. A technique called locally-weighted
regression (loess for short) draws a smooth curve to the data by
connecting locally fitted regions of data (figure 6b).
A review of The Elements of Graphing Data in Meteorological
Magazine states, “Ideally, everyone interested in getting
the most out of their data or presenting data clearly and concisely
should have a copy handy.” My recommendation is no less enthusiastic.
Buy, read, and digest ($52.95US); the quality of your graphs will
improve, and the clarity of the information you convey will increase
dramatically.
23-July 2002
syntax: 7 of 20
Syntax
was designed by Hans
Eduard Meier in the late 1960s and expanded (download
pdf) in the year 2000 to include two additional weights
as well as small caps and text figures. The canonical humanist sans-serif;
inspired by Sabon.
And perhaps the best deal in type – only $119US for the
full set on hard media. See Sumner
Stone’s excellent article in Fine
Print on Type.
18-July 2002
celeste: 6 of 20
By Christopher
Burke, type designer and
scholar. Celeste
was introduced in 1994 and sold via FontFont;
Greek and small text versions have more recently become available,
making this one of the most versatile faces available.
18-July 2002
unusual ff ligature...
I recently posted the following question on
TYPO-L:
“The ff
ligature in MT Pastonchi is unusual in two ways: (1)
the first f is of greater height than the second, and (2) the ascender
of the first leapfrogs well over much of that of the second.
“I’m looking for a precedent for this and cannot seem to find one. Was
this novel form of the ff ligature a development of Francesco Pastonchi
or of someone else?”
Thanks to Gerald
Lange for his interesting and informative reply to the
list, and for his permission to reprint it here:
“I have the original Lanston Monotype specimen book that came out with
the release of Pastonchi (printed by hand at the Officina Bodoni),
which gives a bit of the history and intent. I do not think you
will find precedent as I believe this was an attempt to redefine
previous typographic misconceptions of Renaissance letterforms.
Truly a remarkable typeface. I bought a lot of the metal version
for a book I did and the very day I completed printing, Monotype
Typography sent me the beta version. I believe this was the last
face to be issued before they mergered with Agfa. Found the digital
version to be quite exacting to the original and used it for the
letterpress-printed prospectus for the book!!!
“Many of the combination characters (ligatures and tied characters) are
quite unique. I don't think Pastonchi proved to be much of a commercial
success for Lanston or for that matter Agfa Monotype; but it is
certainly one of the great neglected serious typographic investigations
of the twentieth-century.”
12-July 2002
comparing typefaces 1: a tale of two sabons
The Sabons of Linotype and
Monotype are quite similar, but there are some subtle differences;
I’ve made a partial list. Compared to the Monotype version, the
Linotype version:
1. Is slightly heavier
2. Has somewhat lower contrast between thin and thick strokes
3. Is larger, which is typical for a Linotype-to-Monotype comparison;
to achieve equal x-height, Monotype Sabon must be enlarged by roughly 7%
4. Has a more angular lower case roman a and
f; note the increased curvature of the ascender in the Monotype
roman f versus the Linotype version; compare also
the cross-strokes of the f
5. May have a greater axis angle in the italic; compare the g’s
in the example shown; notice also the differences in the teardrop
terminals in this letter
6. Has that extra bit of space – almost too small to measure, yet
discernable enough – after the lower case roman a.
Or perhaps the spacing is a bit off overall. Or perhaps it’s just
my imagination...
11-July 2002
mendoza: 5 of 20
By Jose
Mendoza y Almeida and issued by ITC
in 1990. Mendoza is rugged
yet elegant; avoid the fi “ligature,” however. See the
full page sample in Bringhurst
(2nd edition, p108); see also the short article in
Carter (pp160–161).
08-July 2002
wherefore art thou bembo...?
Dean
Allen writes: “In its metal version,
Bembo
is my favourite thing to read; with acknowledged subjectivity, it
is the most beautiful and readable text face of all.” I agree
with Mr Allen; I also agree with his opinion on the digital incarnation,
which is indeed thin, wispy, and squat.
We can apparently blame the 8 point size in metal, for it served as
the master from which the digital version was designed. Walter
Tracy, in Letters
of Credit, writes (pp54–55): “When photo-composition
became a reality in the 1950’s the manufacturers of typesetting
machines had to make an important decision: whether or not to carry
forward into the new system the principle of optical compensation,
when the plain and tempting fact was that the photographic part
of the system was capable of producing a considerable range of type
sizes from just one font. To abandon the principle altogether was
to risk forfeiting a substantial part of a reputation for typographic
quality.” He continues, “Some [manufacturers] apparently
thought that increasing the x-height of the faces would be an acceptable
alternative. It is not....The Bembo face, an admirable example of
optical compensation by the Monotype drawing office, demonstrates
the point.” Tracy here refers to a figure in which the metal
version of Bembo at 24, 12, 8, and 6 pt is compared to a film version
derived from the 8 pt – essentially identical to digital Bembo.
While Edward Tufte has admirably continued to use metal Bembo, what recourse
is there for those of us who must work in the digital realm? One
option is to use Poliphilus, a typeface constructed from tracings
of letters in the Dream of Poliphilo. It has not become
obsolete as quickly as Stanley
Morison predicted
(see A
Tally of Types, pp46–56), but its faithfulness
to the original impressions limits its use.
Another alternative may be found in Jack Yan’s
Aetna.
In fact, this face comes as close as anything I’ve seen to metal
Bembo. In the upper half of this
figure, metal Bembo and JY Aetna are shown first and
second, respectively, while in the lower half, the upper and lower
case letters, text figures, and ligatures are displayed at 24 pt
in Aetna Roman. This face cannot be considered a digital incarnation
of metal Bembo at 12 or 24 pt – there are too many subtle
deviations. And not so subtle is the lower case a, in which the
Aetna glyph appears to lack the modulation of stroke seen in the
metal analogue, as well as in several of the other Aetna glyphs.
But you can mourn the lack of a digital Bembo based on its 12 pt
metal antecedent, or you can compromise. And if you’re willing to
compromise, Jack Yan’s Aetna is perhaps your best bet.
05-July 2002
« June
