SAT-54: Scatterplots — Lines of Best Fit and Trends
Read scatterplots, describe correlation, and use a line of best fit to estimate and predict values.
SAT-54: Scatterplots — Lines of Best Fit and Trends
Description: A scatterplot shows many (x, y) points to reveal a relationship between two quantities. A line of best fit is a straight line drawn through the cloud of points to summarize the trend. The SAT asks you to read the line, find its slope, and make predictions.
Correlation — what the cloud is telling you
- Positive correlation: as x goes up, y goes up (points rise left-to-right).
- Negative correlation: as x goes up, y goes down (points fall).
- No correlation: no clear pattern.
(Oʻzbekcha: musbat bogʻlanish — x oshganda y ham oshadi; manfiy — x oshganda y kamayadi.)
The line of best fit
It is the straight line that comes closest to all the points (roughly equal points above and below). You read it just like any line: it has a slope and a y-intercept, and its equation is y = (slope)·x + (intercept). (Oʻzbekcha: eng mos chiziq nuqtalarga eng yaqin oʻtadigan toʻgʻri chiziq.)
What slope and intercept mean here
- The slope = how much y changes for each +1 in x (the rate, with units).
- The y-intercept = the predicted y when x = 0 (the starting value).
Worked Example 1 — describe the relationship
A scatterplot of "hours studied" vs "test score" rises from lower-left to upper-right. Describe it.
- As hours increase, score increases → positive correlation. More study tends to mean a higher score.
Worked Example 2 — interpret the slope
A best-fit line is score = 5·(hours) + 60. What does the 5 mean, and the 60?
- Slope 5: each extra hour of study predicts about 5 more points.
- Intercept 60: a student who studies 0 hours is predicted to score 60.
(Oʻzbekcha: nishab (5) — har bir qoʻshimcha soat uchun ball oʻzgarishi; kesma (60) — boshlangʻich qiymat.)
Worked Example 3 — predict a value
Using score = 5·(hours) + 60, predict the score for 4 hours.
- score = 5(4) + 60 = 20 + 60 = 80.
Interpolation (predicting inside the data range) is reliable; extrapolation (far outside the range) is risky — the trend may not continue. (Oʻzbekcha: maʼlumot oraligʻidan tashqarida bashorat qilish xavfli.)
Correlation is not causation
This is one of the SAT's favorite ideas. A strong correlation means two quantities move together — it does not prove that one causes the other. Ice-cream sales and drowning numbers rise together, but ice cream doesn't cause drowning; hot weather drives both. So if a question asks what you can conclude from a scatterplot, "there is an association/relationship" is safe, while "X causes Y" is usually a trap unless the data came from a controlled experiment (see SAT-62). (Oʻzbekcha: bogʻlanish (korrelyatsiya) sababiy bogʻlanishni isbotlamaydi — birga oʻzgarishi sabab degani emas.)
Reading residuals (how good is the fit?)
The vertical gap between a real data point and the best-fit line is called the residual. Small residuals everywhere mean the line fits well; large, patterned residuals mean a straight line may be the wrong model. You don't usually compute them, but knowing the word helps you read SAT answer choices.
Practice 1
A best-fit line is cost = 0.5·(miles) + 3 for a taxi. What does the 0.5 mean?
Show answer
Each additional mile adds about $0.50 to the cost (the per-mile rate). The $3 is the base fare at 0 miles.
Practice 2
With cost = 0.5·(miles) + 3, what is the predicted cost of a 10-mile ride?
Show answer
cost = 0.5(10) + 3 = 5 + 3 = $8.
Key words — Kalit soʻzlar
- Scatterplot — sochma diagramma
- Line of best fit — eng mos chiziq
- Correlation — bogʻlanish (korrelyatsiya)
- Positive / Negative — musbat / manfiy
- Slope — nishab
- y-intercept — y-kesishma
- Predict / Estimate — bashorat qilish / baholash
- Trend — yoʻnalish (tendensiya)
- Extrapolation — chegaradan tashqari bashorat
Summary
- Scatterplots show correlation: positive (up), negative (down), or none.
- The line of best fit summarizes the trend; slope = rate, intercept = value at x = 0.
- Plug an x into the line's equation to predict y; be cautious extrapolating far outside the data.