Author: Daniel Luke

It is possible to manipulate a graphical depiction of data to misrepresent information. The two graphs above, Version 1 and Version 2, both show identical information correctly. However, as is visually evident, the shape of the data is quite different in each of the graphs. The curve has been distorted by changing the y-axis scale. In Version 1, which possesses a vertical axis range of [40, 80], the scale has been calibrated in a manner that is most appropriate for the data set, and does not extend far beyond the numbers imputed into the graph. In Version 2, however, the y-axis scale has been set to [0, 100]. The effect, though not erroneous, is misleading. In Version 1, the curve is quite steep, which implies a drastic a change in values over time. In Version, the curve is flattened, which gives the impression that the change over time has not been notable at all. It is not the data that is being changed, therefore no fault can be made to human error or falsification. However, by varying the y-axis scale endpoints, it is possible to manipulate the curve of the graph and warp the shape of the data.

Both of these graphs depict the same information correctly. However, visually, the shape of the data is quite different. Curve manipulation is possible through the variability of the y-axis range. By changing the vertical axis field limits, the graph itself changes. In Version 1 of the data, the y-axis min/max scale has been calibrated to [40, 80]. The “slope” of the curve is steep, implying a drastic change. However, in Version 2 of the data set, the vertical axis limits have been set to [0, 100], which produces a flattened shape of the curve, which suggests a mellow increase. Here, two graphs depicting identical data sets can be represented in notably different ways, showing how the y-axis scale technique can be used to distort statistical details.