Apply the formula: Percent Change = (New - Old) / Old × 100. Estimation often works faster because SAT wrong choices are spaced apart—if a bar rises from about 50 to 62, the percent change is roughly 12/50 or 24%. You can select an answer like 24% immediately if the choices are 8%, 12%, 24%, 40%. Only compute precisely if two answers remain close after bounding.

When an axis starts above zero, use relative comparisons carefully since bars may look dramatically different but represent small actual changes. Check whether the scale is linear, logarithmic, or has a break. For non-linear scales, a visually larger jump between ticks may represent the same numeric increment as a smaller-looking jump.

Use column-first scanning: read headers left to right to name each column, then filter rows to isolate only those matching the question. Perform the minimum math after isolating the exact cells. For tables with multiple schools or years, box the relevant rows first to avoid drowning in numbers.

Calculate the slope for each line using slope = Δy / Δx. The line with the higher slope increased faster per unit of time, regardless of its starting level. For example, if School A rises from 400 to 520 over 4 years (slope = 30 per year) while School B rises from 450 to 560 (slope = 27.5 per year), School A increased faster despite starting lower.

The main traps are: unit shifts (computing dollars instead of thousands of dollars), wrong baseline (using the wrong 'Old' value for percent change), visual bias (choosing the tallest bar when the axis doesn't start at zero), group confusion (using all students when the question specifies left-handed students), and correlation-causation errors (assuming causation from a positive scatterplot trend).