How to Solve: Line of Best Fit (SAT) – Complete Guide

How to Solve: Line of Best Fit (SAT) – Complete Guide

Score Impact: This question type appears 2-4 times per SAT Math section—mastering it can boost your Math score by 30-50 points by eliminating careless errors and saving time.

WHAT THIS QUESTION TYPE IS ACTUALLY TESTING

The SAT isn’t testing your ability to draw a perfect line—it’s testing: ✅ Data interpretation – Can you extract trends from scatterplots? ✅ Prediction logic – Do you understand how a line of best fit approximates data, not matches it? ✅ Trap avoidance – Can you resist overcomplicating (e.g., calculating exact slope) when estimation is enough?

ANATOMY OF THE QUESTION

Structure Breakdown

1. Stem: A scatterplot with a line of best fit drawn (or described).
2. Conditions: A question asking for:
3. A predicted value (e.g., "What is the predicted y-value when x = 5?")
4. The meaning of the slope/intercept (e.g., "What does the slope represent?")
5. A comparison (e.g., "Which point is farthest from the line?")
6. Answer Choices: 4 options, often with:
7. One correct answer (based on the line, not the raw data).
8. Two distractors (misreading the line or overcomplicating).
9. One outlier (ignoring the line entirely).

Representative Example

A scatterplot shows the relationship between hours studied (x) and test scores (y). A line of best fit is drawn. The line passes through (2, 60) and (6, 80).

Question: Based on the line of best fit, what is the predicted test score for a student who studies 4 hours? A) 65 B) 70 C) 75 D) 80

(Ignore: Exact data points, outliers, or the "perfect" line—only the given line matters.)

THE DECISION FRAMEWORK (Step-by-Step)

Run this every time:

1. Identify the line’s two points (given or visible on the graph).
2. If not given, pick two clear points where the line crosses gridlines.
3. Calculate slope (m) using:
   [
   m = \frac{y_2 - y_1}{x_2 - x_1}
   ]
4. Round to the nearest 0.5 if decimals are messy.
5. Write the equation in point-slope form:
   [
   y - y_1 = m(x - x_1)
   ]
6. Or use slope-intercept (y = mx + b) if the y-intercept is obvious.
7. Plug in the x-value from the question to predict y.
8. If the question asks for x, rearrange the equation.
9. Match to the closest answer choice.
10. Eliminate options that are ±5+ units away from your prediction.

Worked Examples

Example 1 – Straightforward

Scatterplot: Line of best fit passes through (1, 3) and (3, 7). Question: What is the predicted y-value when x = 4? A) 8 B) 9 C) 10 D) 11

Solution: 1. Points: (1, 3) and (3, 7). 2. Slope:
[
m = \frac{7 - 3}{3 - 1} = \frac{4}{2} = 2
] 3. Equation (point-slope):
[
y - 3 = 2(x - 1) \implies y = 2x + 1
] 4. Plug in x = 4:
[
y = 2(4) + 1 = 9
] 5. Answer: B) 9

Example 2 – Common Trap Version

Scatterplot: Line of best fit passes through (0, 2) and (4, 10). Question: What is the predicted y-value when x = 2? A) 5 B) 6 C) 7 D) 8

Trap: Students calculate slope as 2, then use y = 2x + 2 → y = 6 (B). But the line’s y-intercept is not 2—it’s the point (0, 2)!

Solution: 1. Points: (0, 2) and (4, 10). 2. Slope:
[
m = \frac{10 - 2}{4 - 0} = 2
] 3. Equation (slope-intercept):
[
y = 2x + 2 \quad (\text{since y-intercept is 2})
] 4. Plug in x = 2:
[
y = 2(2) + 2 = 6
] 5. Answer: B) 6 (But the trap is thinking the y-intercept is wrong—always verify!)

Example 3 – Hard Variant

Scatterplot: Line of best fit passes through (2, 5) and (6, 13). Question: Which of the following points is farthest from the line of best fit? A) (3, 6) B) (4, 9) C) (5, 12) D) (7, 15)

Solution: 1. Points: (2, 5) and (6, 13). 2. Slope:
[
m = \frac{13 - 5}{6 - 2} = 2
] 3. Equation:
[
y - 5 = 2(x - 2) \implies y = 2x + 1
] 4. Calculate vertical distance for each point:
- A) (3, 6): Predicted y = 2(3) + 1 = 7 → Distance = |6 - 7| = 1
- B) (4, 9): Predicted y = 9 → Distance = 0
- C) (5, 12): Predicted y = 11 → Distance = 1
- D) (7, 15): Predicted y = 15 → Distance = 0 5. Farthest point: A) (3, 6) (But wait—B and D are on the line! Recheck: A’s distance is 1, but is there a bigger outlier?)
- Recalculate for (4, 9): Predicted y = 9 → Distance = 0 (on the line).
- Recalculate for (5, 12): Predicted y = 11 → Distance = 1.
- Recalculate for (7, 15): Predicted y = 15 → Distance = 0.
- The question asks for farthest—so A and C tie at 1. But A is the only option with a non-zero distance in the choices. 6. Answer: A) (3, 6) (But this is a trick—B and D are on the line, so A is the farthest.)

WRONG ANSWER PATTERNS

Common Mistakes

TIME STRATEGY

• Target time: 45-60 seconds per question.
• Skip if:
• The line’s points aren’t clear (come back after easier questions).
• The slope calculation is messy (estimate instead).
• Minimum work:
• Find two points → calculate slope → plug in x → match answer.

BACKSOLVING AND SHORTCUTS

1. Estimate slope visually:
2. If the line goes up ~2 for every 1 right, slope ≈ 2.
3. Use the line’s equation for elimination:
4. If the line passes through (2, 5), plug in x = 2 → y must be 5. Eliminate options where y ≠ 5.
5. Check answer choices first:
6. If options are 65, 70, 75, 80, and your prediction is 72, pick the closest (70).

1-Minute Recap

"Here’s the deal: The SAT gives you a scatterplot with a line of best fit. Your job? Use that line—not the raw data—to predict values. Here’s how: 1. Find two points on the line (even if they’re not data points). 2. Calculate slope (rise over run). Round if needed. 3. Write the equation (y = mx + b or point-slope). 4. Plug in the x-value and match the closest answer. That’s it. No regression formulas, no overthinking. The line is your best friend—use it, and move on. See you in the next question!

Final Tip:

Always draw the line if it’s not given. Even a rough sketch helps avoid misreading the slope. Under time pressure, estimation beats exact calculation.