🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
scale_manipulation
Scale manipulation involves using uneven intervals, non-linear scales, dual axes, or inconsistent unit sizes to distort the visual impression of data. Unlike simple axis truncation, scale manipulation may use logarithmic scales without disclosure, change interval sizes partway through an axis, or employ dual y-axes with different scales to make two unrelated trends appear correlated. These techniques can make flat trends look dramatic or dramatic changes look gradual.
A graph shows tax revenue and crime rate on dual y-axes. The left axis for tax revenue runs from $0 to $100 billion; the right axis for crime runs from 400 to 410 incidents per 100,000. The two lines appear to track each other perfectly, implying causation, but the crime axis covers a trivial range while the revenue axis spans the full range.
An energy drink advertisement plots its product's caffeine content alongside competitors using a bar chart where the y-axis jumps from 0 to 140mg and then has an unannounced break before continuing to 160mg. The brand's bar, sitting just above the break, appears dramatically taller than rivals with nearly identical caffeine levels.
A climate skeptic blog displays a temperature anomaly graph with the y-axis spanning –10°C to +10°C, compressing a consistent 1.2°C warming trend over a century into a nearly flat, visually imperceptible line, making the change appear trivial.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Are the intervals on the axes evenly spaced?
Type: binaryIs the scale type (linear, logarithmic) clearly labeled?
Type: binaryDoes switching to an even scale change the visual impression of the trend?
Type: binaryAre different parts of the graph using different scales?
Type: binaryScale manipulation involves using uneven intervals, non-linear scales, dual axes, or inconsistent unit sizes to distort the visual impression of data. Unlike simple axis truncation, scale manipulation may use logarithmic scales without disclosure, change interval sizes partway through an axis, or employ dual y-axes with different scales to make two unrelated trends appear correlated. These techniques can make flat trends look dramatic or dramatic changes look gradual.
Viewers instinctively compare visual positions of lines and bars without carefully reading axis labels and intervals. Dual-axis charts are especially deceptive because two unrelated series can always be made to appear correlated by adjusting the scales.
Check whether axes start at zero, whether intervals are even, and whether dual axes are being used. Re-plot the data with consistent scales to see if the visual impression holds.
Scale manipulation is common in climate change denial graphics, pharmaceutical advertising, financial market commentary, and political data visualization.
Use these tools to detect, analyze, or train this aspect.