Skip to content
AP Statistics · Probability, Random Variables, and Probability Distributions

Linear transformation

Linear transformation appears in 13 reviewed AP Statistics questions and sits in Probability, Random Variables, and Probability Distributions. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Linear transformation is tested

Example 1Medium
Random variable X has mean μ_X and standard deviation σ_X. A constant c is added to X to create Y = X + c. Which of the following is true about the mean and variance of Y?
  1. Aμ_Y = μ_X; Var(Y) = Var(X) + c.
  2. Bμ_Y = μ_X + c; Var(Y) = Var(X) + c^2.
  3. Cμ_Y = μ_X * c; Var(Y) = c^2 * Var(X).
  4. Dμ_Y = μ_X + c; Var(Y) = Var(X).Correct

Why D is right

Adding a constant c to every value shifts the mean by c: μ_Y = μ_X + c. However, adding a constant does not change the spread — Var(Y) = Var(X) because variance measures spread relative to the mean, and all values and their mean shift by the same amount c.
Why the other 3 choices miss
A
Incorrect because adding c to X shifts the mean by c: μ_Y = μ_X + c, not μ_Y = μ_X. Also, the variance does not change by adding a constant (Var(Y) = Var(X), not Var(X) + c).
B
Incorrect because adding a constant to X adds c to the mean but does NOT add c^2 to the variance. Variance measures spread, not location. Adding a constant does not change spread.
C
Incorrect because Y = X + c is an additive operation, not multiplicative. The formulas for multiplying by a constant (Y = c*X: μ_Y = c*μ_X, Var(Y) = c^2*Var(X)) do not apply here.
Example 2Harder
A dataset shows a positive, moderately strong linear association between x and y with r = 0.7. A researcher adds 10 to every value of y. How does this affect r and the slope of the LSRL?
  1. Ar remains the same and the slope remains the same; only the intercept changes.Correct
  2. BBoth r and the slope remain the same but the intercept decreases by 10.
  3. Cr remains the same but the slope increases.
  4. DBoth r and the slope increase by a proportional amount.

Why A is right

Adding a constant to every value of y is a linear transformation that shifts the y-distribution without changing its variability or the linear relationship with x. The correlation r is unchanged (it depends on the relative positions, not the scale or shift). The slope b1 = r*(Sy/Sx) is unchanged because adding a constant does not change Sy. Only the intercept b0 = y-bar - b1*x-bar changes, since y-bar increases by 10.
Why the other 3 choices miss
B
Incorrect about the direction of the intercept change. If y-bar increases by 10, then b0 = y-bar - b1*x-bar increases by 10, not decreases.
C
Incorrect because the slope b1 = r*(Sy/Sx) depends on the standard deviations, which are unchanged when a constant is added to y. The slope stays the same.
D
Incorrect because adding a constant to all y-values does not alter the strength of the linear relationship or the ratio Sy/Sx. Neither r nor the slope is affected by a shift in y.
Example 3Harder
A scatterplot shows a strong positive linear relationship with r = 0.95. The slope of the LSRL is b1 = 2.3. If the units of y are changed from meters to centimeters (multiply all y-values by 100), which of the following best describes the effect on r and b1?
  1. Ar remains 0.95; b1 becomes 230.Correct
  2. Br becomes 95 (multiplied by 100); b1 becomes 230.
  3. Cr remains 0.95; b1 remains 2.3.
  4. Dr becomes 0.0095; b1 becomes 0.023.

Why A is right

Correlation r is unitless and invariant to changes in scale (multiplying by a constant). So r remains 0.95. The slope b1 = r*(Sy/Sx): multiplying all y-values by 100 multiplies Sy by 100, so b1 becomes r*(100*Sy/Sx) = 100*(r*Sy/Sx) = 100*2.3 = 230. The intercept also changes, but the question asks about r and b1.
Why the other 3 choices miss
B
Incorrect because r is a dimensionless number between -1 and 1. Multiplying y by 100 does not change r; it only changes the slope and intercept. r = 95 is impossible.
C
Incorrect because multiplying all y-values by 100 does change the slope. The slope measures the change in y per unit change in x; if y is measured in cm instead of m, the slope must increase by a factor of 100.
D
Incorrect because the directions of the changes are reversed. Multiplying y by 100 increases both Sy and the slope by factor 100, not divides them by 100.

Studied alongside

Other concepts from Probability, Random Variables, and Probability Distributions.