SAT Linear Regression and Best-Fit Lines: Interpreting Slope, Intercept, and Predictions

A line of best fit passes through a scatter plot to minimize the total distance between each data point and the line. On the SAT, the line is always given in graph or equation form; you never need to calculate it. Your task is to read the slope and intercept, use the equation to predict values, and interpret what the slope and intercept mean in context.

The slope tells you how much the dependent variable changes for each one-unit increase in the independent variable. If the regression line for hours studied vs. test score has slope 8, each additional hour studied is associated with an average 8-point increase in score. Always attach units to slope and intercept interpretations; a slope of 8 means nothing without specifying it is 8 points per hour rather than 8 hours per point.

If the best-fit equation is y=12x+40, predicting y at x=5 means substituting: y=12(5)+40=100. The y-intercept (40) gives the predicted y when x=0, which may or may not have a practical meaning depending on the context. These substitution questions are the most common regression question type and are designed to be fast if you recognize the pattern immediately.

Three micro-examples: (1) y=2.5x+30 predicts y=80 when x=20. (2) The y-intercept of y=5x+15 means the predicted value is 15 when x=0. (3) If x is years since 2010 and y is sales in thousands, slope 3 means sales grow by 3,000 per year. Choosing an answer that attaches incorrect units or misidentifies which axis is the independent variable is the most common prediction question error on the SAT.

A regression line shows association, not causation. Even a perfectly fitting line does not mean one variable causes changes in the other. The SAT tests whether you can identify that regression describes a relationship without proving that one variable causes another. Correct interpretation uses language like "is associated with" or "predicts," not "causes" or "determines."

Practice prompt: given y=3.2x+10 where x=daily TV hours and y=calorie intake, which interpretation of slope is correct? Option A: "TV causes people to eat 3.2 more calories per hour." Option B: "Each additional hour of TV is associated with 3.2 more calories consumed per day on average." Option B is correct because it uses association language. Any SAT answer that uses causal language (causes, leads to, results in) for a regression relationship is almost certainly a trap answer.

Predicting within the data range (interpolation) is generally reliable. Predicting far outside the data range (extrapolation) is unreliable and often flagged explicitly by SAT questions that ask whether a prediction is reasonable. If your data spans x=0 to x=10 and the question asks for x=100, the SAT is testing whether you recognize this as unreliable extrapolation rather than valid prediction.

Apply this two-step check before trusting any regression prediction: (1) is the x value inside the data range shown in the graph or described in the problem? (2) if not, is the question specifically asking you to evaluate the reliability of the prediction? If the problem states a data range and the x value is outside it, the correct answer will identify the prediction as unreliable regardless of what value the equation produces at that x.