You can do better, WSJ

   This morning I was perusing my new online subscription to the Wall Street Journal and I came across an article entitled, "Women Own This Summer. The Economy Proves It."  The article outlines the impact of the combination of Barbie, Taylor and Beyoncé on the economy of the country.  I'll leave you to read that at your leisure.  

    What bothered me about this article was not the premise or the writing, but was something we see frequently in much less reputable print, video and other online media: the use of misleading graphs to misrepresent data.  After a picture of several women in their pink Barbie movie attire, a graph is presented with the long title "Labor force participation rate for women with children under six, annual average" (see image at right).  The graph shows that the percentage of such women was roughly 64.5% in 2012 and has increased to almost 68% in 2022.  My complaint about this graph is that since the vertical axis starts at 62 and rises to only 69, the first data point is almost 3/7 of the way up the graph, and the last data point is 6/7 of the maximum value, giving the visual impression that the percentage of women in the workforce has significantly increased in the last 10 years.  It appears visually to have almost doubled, until you carefully read the vertical axis.  

    I estimated the values from the graph for each year and used a spreadsheet to make my own version of the graph, this time with the vertical axis ranging from 0 - 100%.   This graph is shown at left.  This is the same data, plotted on a scale that more accurately demonstrates the fact that a rise from almost 65% to 68% is an increase, but is not really that significant.  Think about it: if your salary rose that much over the same amount of time you would would be very upset: slightly more than 3% in ten years.  This graph, which includes the entire possible range of values, shows the "trend" much more clearly.   The WSJ graph, which shows less than 10% of the range of possible values, misrepresents the trend leading the reader to believe there has been a significant increase.  I'm not sure why the authors chose to represent the data this way (especially since they never refer to the graph in the article).  Perhaps they are just following the advice of their science teachers, urging them to "use the full page;" or perhaps they just accepted the ranges for the axes that the software gave them, given the data they chose. I don't mean to imply malicious intent on their part.  But nonetheless, their graphic misrepresents the trend, showing the values as steeply increasing when, in fact, the trend is much flatter.  However, I must admit that the data as presented in the graph I created doesn't fit the narrative of "women are on the rise" and thus stands very little chance of being included in their article.

    One the most important skills science instructors should work to develop in their students is graphical analysis.  In this case, it is an important life skill for a person to be able to recognize when data is being presented in a way that is misleading or misrepresents the trend in the data being plotted.  I expect bloggers and journalists in some corners of the Internet to represent the data only in a way that best supports their point, even if it is misleading.  I expect better from the Wall Street Journal.