Recently I engaged in a debate concerning whether to use nonstandard linear graphs (NLG) or Standard Celeration Charts (SCC) for time series behavior. I took the side of the SCC. A question arose asking which took longer to explain to teachers, the NLG or the SCC.
I responded that generally it takes me about three hours to get people familiar with the SCC so they understand and can use its basic capabilities. The other person noted it takes very little time to familiarize people with nonstandard linear graphs, therefore, we should use the NLG instead of the SCC.
The logic train seems straightforward:
1. Nonstandard linear graphs appear easier to understand than the Standard Celeration Chart.
2. Teachers do not have a lot of time to learn a new graph or chart.
3. We should use the NLG because teachers can pick it up quickly and don’t have the time to learn something more involved.
I don’t know what people thought in the audience but I saw a few head nods at my fellow debater’s assertion. Due to the format of the discussion I could not rebut the proposition. But I can now :-)
Ease of understanding versus complexity
My debating colleague had a point. Nonstandard linear graphs take less time to explain to teachers probably because they already have extensive experience with the visual display. We see linear graphs everywhere! Newspaper stories, popular magazine articles, scientific journals, even in sarcastic political commentaries.
Figure 1. The ubiquity of linear graphs, even in political cartoons.
Regardless of the “familiarity argument” (i.e., more people have used the linear graph and know it better), if we dissect the logic from above we run into a serious problem. Why should familiarity or how long it takes to explain the any graph guide our decision for adoption? Or for that matter, how does the complexity of a graph signal a superior choice?
Let’s insert some different targets in the above logic and see how we appraise my colleague’s point.
1. Calculators take less time to understand than computer driven statistical programs (e.g., Excel).
2. Teachers do not have a lot of time to learn a new statistical program.
3. We should use the calculators to solve advanced statistical problems because teachers can learn them quickly and don’t have the time to learn something more complicated.
Perhaps you agree the complexity of a chart or how long it takes to learn it should drive the decision to adopt a visual display. If not, what should guide our decision making? What the chart can actually do!
The SCC provides different information from the linear graph. Many sources explain in great detail why we should prefer a ratio (or semilogarithmic) chart to a linear graph (e.g., advantages in finance , education, science).
And then we have the whole matter of why a standard view provides a 10x or better advantage over nonstandard graphs.
The next time someone tells you we should select the nonstandard linear graph over the Standard Celeration Chart because teachers (or parents, students, or whomever) understand it better, remember the problem with such logic. A magnifying glass takes way less time to learn compared to a microscope, but which resolving power would you rather have to see the world with clarity?