Geo 333 R. Karlin
Guidelines for Preparing Effective Scientific Graphs
References:
Briscoe, Mary H., Preparing Scientific Illustrations, Springer-Verlag, New York, 204 pp., 1996
Line - 2 bit black/white
Continuous or Halftone (8, 16 or 256 shades of gray)
dithering
random
pattern
Emphasize the data, not the axes or labels
Avoid semilog axes if possible
Avoid eye candy
A good graph is uncluttered clear and focused
Closed symbols more emphatic
Open and closed symbols have maximum contrast
Squares and circles look similar to each other
Size of symbols
Squares look larger than circles
Triangles look smaller than circles
Use compact spacing & labeling; Avoid wasted space
Fonts - Use a Sans Serif font (e.g., Helvetica, Geneva, Arial)
Keep readable but minimize size to emphasize data
Upper and lower case, various sizes
Avoid bold labels, except for maximum emphasis
Keep simple
Avoid boxes
Upper case if short
Upper and lower case if longer
Linear vs semilog vs Log
Use Interior major and minor ticks
Minimize number of ticks
proportional to data resolution
readability
Use fewest significant digits
Eliminate unnecessary zeroes
Use compact x and y axes
Choice of origin
Use 0,0 if possible
If origin depends on data range, offset, separate, or break axis
Heavy lines connecting data
Thin error bars
If data equally important use solid lines only
Maximum contrast
solid (most important data)
dashed (2nd)
dotted (least important)
avoid - generally not useful
exception - if exact measurement is necessary
Plot frames distracting, not necessary
Number
Line thickness
Use of Dashed lines
Labeling
San serif
Significant digits
Patterns
Acceptable patterns (ala Geology)
Grey half tones
at least 20% difference, 50% preferable
On reduction or reproduction
Diagonal hatching most pleasing and least ambiguous
Cross hatching darker and most emphatic (use fine lines)
Stippling becomes gray tone on reduction
Fine stippling drops out or becomes uneven
Truth vs beauty
Use space wisely; avoid wasted space
Cropping size of picture for publication
Concentrate on features of interest
Simple
Clearly organized
Logical groupings
Self explanatory; complete
Units defined
Good use of space
For slides - The simpler the better;
Only pertinent information
Types of Graphs
Pie - to compare percentages
Bar
Horizontal
Vertical
Stacked
Comparison of total measurements and subdivisions
Histogram
Frequency polygon
Line graph
for comparing equally spaced data on x axis (e.g., time)
Scattergraph
For examining relationships between 2 variables