Not just in Lebanon
I was reminded of Mary Ann's post about inaccurate bar graphs ("Visualized Data") when I saw a article in a pharmaceutical industry trade magazine I get. Here are two examples of the bar graphs in the article, which are the results of a survey of pharma industry employees:

Come on now! 42.7% is not almost half of 57.3%. And 24.4% is not almost as much as 75.6%! And to three significant figures, no less. It is hard to believe that with all of the graphing software available, someone would choose to apparently "eye-ball" these graphics. Such sloppiness from a scientific audience publication was not what I would expect. Is it only carelessness, or were they trying to emphasize one difference and down play the other? Probably not intentially from reading the article. Should I e-mail the editor?

2 Comments:
that's my dad! I'd write the letter. Tell them how disapointing this is.
Dear Mr. Wittwer:
Thank you so much for writing in. We had used an automated survey program and , for the first time, had imported the figures directly from that program (but left in the actual percentages). Please be assured that this will be the last time that we take this approach.
Sincerely,
Agnes M. Shanley, Editorial Director
-------------------------------
To: AShaley@putman.net
Subject: Inaccurate graphics
Dear Ms. Shanley,
I am a new recipient of Pharmaceutical Manufacturing, and am not very familiar with it. I was amazed, however, to see the graphics in your article, "A Year of Uncertainty: Results of Our 2007 Job and Salary Survey". An example (Figure 4b) is attached. 24% is not almost 76%, as the bars in the graph might suggest. There are similar errors in almost all the graphs. My only conclusion is that someone must have "eye-balled" the bars on these graphs rather than use any of the graphing options that are so readily available in commonly used software like Excel or even PowerPoint. What is the point of expressing the results to 3 significant figures if the visual representation is so far off? It is a shame to have the reader of an otherwise fine article distracted by such things. We are a technical audience!
Best wishes,
Art Wittwer
Post a Comment
<< Home