Critiquing data presentation means checking a graph before accepting its conclusion: do the axes start at zero and use an even scale, is the full time range shown, is the source credible, and is the sample size stated. Learn the checklist that applies to any bar, line, or pie chart.
Reading a graph critically
Critiquing data presentation means checking a graph or chart before accepting the conclusion it seems to suggest. A graph can be accurate, misleading, or somewhere in between, and the same five questions apply whether it is a bar graph, line graph, or pie chart.
Run through this checklist before trusting what a graph seems to show.
Check the axes first
Confirm both axes are labeled with units, and check where the y-axis starts. If it does not start at zero, the graph may be a misleading graph that exaggerates a small difference. Also check that the scale is even — equal-looking steps on the axis should represent equal amounts.
Check the range and the source
Ask whether the graph shows the full relevant time period, or just a favorable slice of it — a trend can look very different if the last few data points are cut off. Then check where the data comes from: a credible, named source matters more than a polished-looking chart.
Check the sample size
A survey of 20 people and a survey of 20,000 people can produce the same-looking bar graph, but they do not carry the same weight. Look for the sample size before trusting a conclusion drawn from it.