Nice article (sent on by Margot Black from Lewis & Clark College) on the persuasiveness of charts in people’s belief in scientific claims. The study titled, Blinded with Science, showed text to some participants explaining the efficacy of a new drug, while others saw the text and this graph:
Of those who saw the text and graph, 97% believed the drug worked, versus 68% of those who saw only text. This makes sense for those of us who struggle with reading comprehension in our QR classes. It is interesting in reading the actual text used in the study to find the statement: “a drop of forty percent (from 87% to 47%)”.
This should be stated as a 40 percentage point drop (or 46.0% decline). The authors acknowledged the incorrect usage of percent change but stated (in a very kind email response): “It’s a question of pragmatics (language use), and so we figured being 40% below 87% is how most people would take it.” This of course is unfortunately true, and probably does not skew the results of the study. The problem occurs when a town raises the tax rate from 6% to 9% citing a small 3% increase (as opposed to the true 50%). If scholarly articles published in journals like, Public Understanding of Science, don’t hold citizens accountable for quantitative illiteracy then who will?