Many of us have heard that proficiency in literacy is an issue in the US. When I read graphs in the newspaper, I sometimes notice issues that denote science illiteracy.
Read the following graph from The New York Times. This is also good practice for students. Remember, you should look for axes names, labels, units; you should observe the number of points. What do you notice?
Although the above graph has three data points labeled (i.e., “75 headlines, 50, 25”) it’s missing the y-axis. Without the y-axis label and units, we don’t know if the y-axis starts at zero or has any scaling, which makes the data points difficult to read and not directly comparable. Additionally, because the gray bars are not demarcated by year, not only are they difficult to read (i.e., how many data points in total are here?), but the NYT seems to have mis-plotted the number of bars or years between certain time intervals. The period of 1917-1921 should have five bars (1917, 1918, 1919, 1920, 1921), but has four. The period of 1950-53 is a period of four years, but contains five bars.
Now try again to apply science literacy to the graph below:
The graph is missing the x-y coordinate plane, and so it has no “anchor.” For example, if we look at the “Parks” data point, we have no idea how much less than “4” it really is. It’s essentially floating in space. Neither axis has units. In the linked article, the authors explain that they “asked people to rate, on a scale of 1 to 10, how often they interacted with people or touched shared surfaces at various businesses, as well as how much activity in different sectors occurs indoors.” We do not know how these variables were matched to specific numbers — did people count that they touched an object “5” times at a gas station and had their calves massaged for greater than “6” minutes at a nail salon? The y-axis is “how much activity at each location occurs indoors.” Malls are about “7;” does this mean “7 shops visited?” Parks are less than “4;” does this mean less than “4 steps taken in a public restroom?” The size of the data points is related to a third variable - “the product of a business sector’s average weekly visits and its total number of establishments” - however, by adding the additional sizing of the bubbles, this further confounds the objective numbers given for each axis. While this graph tells a nice story, it is not clear how it displays the data that has been collected, which is ultimately the point of a graph.
Here’s another graph to apply science literacy:
The graph has no units at all. It’s also unclear how the dual labels of the x and y axis relate WITHIN those axes to produce a continuum that for example, involves both “communication and critical thinking” and “physical work” at opposite ends. At which point on the x-axis does critical thinking become zero, and physical work become whatever unit is the opposite of zero here? Therefore, it’s unclear how to objectively plot these points (different professions) in space so that for example, “Clergy” is in an extreme opposite quadrant labeled “More emphasis on communication and critical thinking” opposed to “Roofers” or “Rail Car Repairers,” who we should infer from the figure must have little to no communication and critical thinking involved in their work. The data also shows professions in green or purple, which relate to whether they are “less or more automable.” The data in this graph come from a well cited study which has also been criticized for having major problems due to its subjective nature. Like the previous graph, this graph tells a nice story, but it’s not clear how it displays actual data that has been objectively collected.
The last three "bad" graphs from the NYT contain errors that students should be able to notice: missing axes, missing the entire coordinate plane, missing labels and units. Not only is it "bad" science that these items are missing, but it muddles the meaning and impact of the articles, which involve important topics in public health, the workforce, and the economy.
As a teacher, the issues we find with student performance are often foundational. In the case of science, I find that those foundational issues are gaps in knowledge of the scientific method and science literacy, described above. By addressing these gaps, in a short period of time, my students have had major growth in their scores, and more confidence with scientific material. I advocate for students investing time on the correct way to read graphs.
Comments