Difficulty: Medium
Which formula links the Y-on-X regression coefficient to r?
A rσX/σY
B σXσY/r
C rσY/σX
D r²σX
Answer: C. rσY/σX The Y-on-X coefficient is r multiplied by the ratio of Y’s standard deviation to X’s.
Which formula links the X-on-Y regression coefficient to r?
A rσY/σX
B σXσY/r
C r²σY
D rσX/σY
Answer: D. rσX/σY The X-on-Y coefficient is r multiplied by the ratio of X’s standard deviation to Y’s.
If b_yx=0.8 and b_xy=0.45, what is r?
A 0.6
B 0.36
C 0.8
D 0.45
Answer: A. 0.6 The product is 0.36=r²; both coefficients are positive, so r=+0.6.
If b_yx=−0.5 and b_xy=−0.72, what is r?
A 0.6
B −0.6
C −0.36
D −0.5
Answer: B. −0.6 The product is 0.36=r²; the common negative sign makes r=−0.6.
If b_yx=2 and b_xy=0.125, what is r?
A 0.25
B 2
C 0.5
D 1
Answer: C. 0.5 Their product is 0.25=r²; the positive sign gives r=0.5.
If b_yx=1.5 and b_xy=0.4, what is their product?
A 1.9
B 0.375
C 0.4
D 0.6
Answer: D. 0.6 Multiply the two regression coefficients, 1.5×0.4=0.6.
If b_yx=0.6 and b_xy=0.6, what is r?
A 0.6
B 0.36
C −0.6
D 1
Answer: A. 0.6 The product is 0.36=r², and positive coefficients give r=+0.6.
What sign do two nonzero regression coefficients share?
A Opposite signs
B The same sign
C No fixed relation
D Always positive
Answer: B. The same sign Both coefficients have the sign of the correlation coefficient.
If one regression coefficient is zero, what is the correlation?
A One
B Minus one
C Zero
D Undefined always
Answer: C. Zero Since the product of the regression coefficients equals r², a zero coefficient implies r=0.
The product of the two regression coefficients equals what?
A r divided by two
B The covariance
C The variance ratio
D r squared
Answer: D. r squared The product b_yx b_xy equals the square of the correlation coefficient.
What is the largest possible product of the two regression coefficients?
A One
B Two
C The sample size
D The standard deviation
Answer: A. One Because the product equals r² and |r|≤1, it cannot exceed one.
Can a regression coefficient exceed one in magnitude?
A No, never
B Yes
C Only if r exceeds one
D Only when r is zero
Answer: B. Yes A regression coefficient includes a standard-deviation ratio and can exceed one.
If r=0.5, σY=4, and σX=2, find b_yx.
A 0.5
B 2
C 1
D 4
Answer: C. 1 Use b_yx=rσY/σX=0.5×4/2=1.
If r=−0.6, σY=5, and σX=3, find b_yx.
A −0.6
B 1
C −1.8
D −1
Answer: D. −1 Using the standard-deviation ratio gives −0.6×5/3=−1.
If covariance is 24 and variance of X is 12, find b_yx.
A 2
B 0.5
C 12
D 36
Answer: A. 2 The Y-on-X coefficient equals covariance divided by variance of X, 24/12=2.
If covariance is 24 and variance of Y is 48, find b_xy.
A 2
B 0.5
C 24
D 72
Answer: B. 0.5 The X-on-Y coefficient equals covariance divided by variance of Y, 24/48=0.5.
If covariance is negative, what are both regression coefficients?
A Positive
B Opposite signs
C Negative
D Both zero always
Answer: C. Negative Each regression coefficient divides the covariance by a positive variance, preserving its sign.
If both regression coefficients are positive, what is the sign of r?
A Negative
B Zero
C Undefined
D Positive
Answer: D. Positive The correlation has the same sign as both regression coefficients.
If both regression coefficients are negative, what is the sign of r?
A Negative
B Positive
C Zero
D Undefined
Answer: A. Negative The shared negative sign of the regression coefficients makes r negative.
Can two nonzero regression coefficients have opposite signs?
A Yes, always
B No
C Only when r is one
D Only with equal means
Answer: B. No Both coefficients share the sign of r when both variables have positive variance.
Changing the origins of X and Y affects regression slopes how?
A They reverse signs
B They become zero
C They stay unchanged
D They are squared
Answer: C. They stay unchanged Shifting either variable by a constant leaves deviations and regression slopes unchanged.
Changing measurement scales can affect regression coefficients how?
A They always stay fixed
B They become unit-free
C They become correlations
D Their sizes may change
Answer: D. Their sizes may change A regression slope depends on the chosen units for X and Y.
When r=0 and both variables vary, how do the regression lines meet?
A At right angles
B They coincide
C They are parallel
D They never meet
Answer: A. At right angles Zero correlation makes both regression slopes zero, giving horizontal and vertical lines through the means.
When |r|=1, how are the regression lines related?
A They are perpendicular
B They coincide
C They are parallel
D They are unrelated
Answer: B. They coincide Perfect correlation means the data lie on one exact line, so both regression equations coincide.
If σX equals σY, how does b_yx relate to r?
A b_yx is r squared
B b_yx is negative r
C They are equal
D No relation exists
Answer: C. They are equal When the standard deviations match, b_yx=rσY/σX simplifies to r.