Skip all navigation and go to page content

NEO Shop Talk

The blog of the National Network of Libraries of Medicine Evaluation Office

Denver in November 1: Visual Presentation of Quantitative Data

The American Evaluation Association annual meeting is packed with useful and interesting sessions and workshops.  This year’s meeting began for me on Wednesday with two workshops and this was the first:

Visual Presentation of Quantitative Data

Taught by David Streiner (University of Toronto) and Stephanie Reich (University of California, Irvine)

Beginning with some basics about memory and information processing, the instructors emphasized that effective graphs use simple features to move viewers from pre-attention to paying attention, with pieces of information being stored temporarily in working memory.  Working memory is limited to 9 elements of information at once (or fewer) so it is important to eliminate from graphs any details that might interfere with the main points.  The instructors made the case that the best use of graphs is when you want to display a comparison and make a key point immediately obvious, since exact numbers will be forgotten quickly.  For exact numbers, use a table.  Here are some tips from this workshop:

  • Sort data to be displayed, and experiment with switching axes–these can make a big difference in first impressions
  • Beware of stacked graphs–they make it hard to do comparisons and see trends
  • Be careful of what scales you use–they can be manipulated to make differences and changes seem larger than they really are (for example, if a scale does not begin at zero, or if it looks at percentage change)
  • Avoid 3-D graph displays–they are usually harder to read and rarely add anything useful beyond a standard two-dimensional format
  • Excel and PowerPoint create 3-D bar graphs differently, with the bars at different levels for the same data
  • Use caution with pie charts–it is difficult to compare slices that are similar in size
  • Pie charts in 3-D are especially heinous–they distort the wedges and make comparison impossible
  • Pie charts are not the same as gauges–gauges are calibrated and measure only one quantity

Comments are closed.

Last updated on Monday, June 27, 2016

Funded by the National Library of Medicine under Contract No. UG4LM012343 with the University of Washington.