Deck 9: Correlation and Simple Regression

Question 1

The basic null hypothesis in the context of Pearson's r is:

A) H₀: ρ > .00
B) H₀: ρ = .00
C) H₀: ρ ≠ .00
D) H₀: ρ_A ≠ ρ_B

Answer 1

B

Question 2

When ρ = .00, which of the following is true?

A) The sampling distribution of ρ is centered on 1.0.
B) The sampling distribution is positively skewed.
C) The sampling distribution of ρ cannot be constructed.
D) None of the above

Answer 2

D

Question 3

Which of the following is not an assumption or requirement for Pearson's r?

A) One variable is conceptualized as a cause, the other as the effect.
B) Measurement of both variables is on an interval or ratio scale.
C) The distribution of the two variables is bivariate normal.
D) Values on the two variables are homoscedastic.

Answer 3

A

Question 4

For Pearson's r, degrees of freedom is equal to:

A) N
B) N - 1
C) N - 2
D) N - Number of groups

Question 5

Which of the following is not true about the r-to-z transformation?

A) It is needed to test differences between two correlations.
B) It is needed with this null hypothesis: H₀: ρ_A ≠ ρ_B.
C) It can be used to calculate confidence limits.
D) It is used in power analyses.

Question 6

Magnitude of effects in the context of correlations is usually expressed as:

A) r
B) r²
C) eta²
D) d

Question 7

Which of the following is a coefficient of determination?

A) r
B) r²
C) τ
D) ρ

Question 8

Which of the following equations is correct?

A) r = SS_Explained ÷ SS_Total
B) r = eta²
C) r² = SS_Explained ÷ SS_Total
D) r² = SS_Residual ÷ SS_Total

Question 9

Which of the following statements is true?

A) In a power analysis, effect size = r²
B) The calculation of confidence limits around r involves calculating the SE for r
C) If the 95% CI around r is .01 to .22, the null hypothesis is true
D) In the context of a correlation, .50 is considered a large effect

Question 10

The correlation between height and weight would likely be greatest in which of the following populations?

A) 15-year-old adolescents
B) College basketball players
C) 1-year-old infants
D) People with growth hormone deficiency

Question 11

Which of the following statements is not true?

A) Low reliability of measures can reduce the magnitude of r.
B) There are formulas for correcting r when low reliability results in attenuation.
C) Outliers should routinely be dropped from correlation analyses to avoid problems.
D) A sample comprised only of people who are extremely underweight and extremely overweight could inflate correlations between weight and height.

Question 12

Which of the following is a correlation coefficient that uses ranks?

A) Kendall's tau (τ)
B) Pearson's r
C) Point-biserial correlation (r_pb)
D) Spearman's rho (r_S)

Question 13

What is an intercept?

A) The angle of a straight line
B) The point at which a straight line crosses the Y axis
C) The "X" part of the equation for a straight line
D) The "b" part of the equation for a straight line

Question 14

The basic equation for a straight line is:

A) Y' = b + aX
B) X' = bY
C) Y = a + bX
D) Y = bX

Question 15

In a regression equation, the "b" is called the:

A) Regression coefficient
B) Intercept
C) Residuals
D) Error of prediction

Question 16

Regression is used to:

A) Corroborate results from a correlation analysis
B) Make predictions about values for a variable based on known values for another
C) Determine the magnitude of effects of correlations among variables
D) Estimate sample size needs when planning a study involving correlation

Question 17

Simple regression is sometimes called ordinary least-squares regression because the equation:

A) Requires information about r squared
B) Minimizes the value of r squared
C) Is equal to the sum of squared prediction errors
D) Minimizes the sum of squares of prediction errors

Question 18

The term residuals in regression refers to:

A) Standard errors of estimate
B) Predicted values
C) Errors of prediction
D) The term Σ(Y - Y')²

Question 19

Which of the following statements is true?

A) The larger the value of r, the more accurate the prediction of Y'.
B) The larger the SE_Estimate, the larger the value of r.
C) The larger the SE_Estimate, the more accurate the prediction of Y'.
D) The larger the value of r, the larger the value of e².

Question 20

Correlation coefficients can be used in all of the following applications except:

A) Assessing test-retest reliability
B) Testing hypotheses in assessments of construct validity
C) Evaluating interrater reliability
D) Assessing internal consistency reliability via coefficient alpha