Question 1

All of the following are components of a basic multiple regression equation except: A) Intercept constant B) Betas C) b-weights D) X₁ to X_k (k = number of predictors)

Accepted Answer

B

Question 2

In a standardized multiple regression equation, which of the following is absent?&#10;A) Intercept constant&#10;B) b-weights&#10;C) Raw scores on the predictors&#10;D) All of the above

Accepted Answer

In a standardized multiple regression equation, the intercept constant, b-weights, and raw scores on the predictors are all absent because the equation uses standardized scores (z-scores) for both predictors and outcome variables, which inherently adjusts for the mean and standard deviation of those variables, making the intercept and raw scores unnecessary.

Question 3

In a standardized multiple regression equation, a beta weight is associated with a:&#10;A) z score for the corresponding predictor variable&#10;B) z score for the dependent variable&#10;C) Raw score for the corresponding predictor variable&#10;D) Raw score for the dependent variable

Accepted Answer

Beta weights are standardized regression coefficients, which means they are based on z-scores of the predictor variables. Therefore, the beta weight is associated with a z score for the corresponding predictor variable.

Question 4

Which of the following statements about R²is not true? A) It is sometimes called the coefficient of determination. B) Its values can range from -1.00 to +1.00. C) Its value indicates the percentage of variation of Y explained by all predictors as a set. D) It is a measure of magnitude but not direction of relationships.

Accepted Answer

The correct term, likely referring to R-squared (R²), ranges from 0 to 1, not -1 to +1. R² is the coefficient of determination (A), indicating the proportion of the variance in the dependent variable that is predictable from the independent variable(s) (C), and it measures the strength of the relationship between variables, but not the direction (D).

Question 5

The zero-order correlation between Y and three predictors is as follows: r_y₁= .44, r_y2=
)51, and r_y3= .29. Which of the following is true?

A) Rwould be a value between .29 and .50.
B) Rwould be a value between .08 and .26.
C) Rwould be a value of .51 or greater.
D) Rwould be a value greater than .51.

Accepted Answer

The zero-order correlations represent the individual correlation of each predictor with the outcome variable Y. The combined predictive power of all predictors together (R) is typically greater than any individual predictor's correlation with Y, assuming the predictors add unique information. Since .51 is the highest individual correlation, the combined predictive power would be a value of .51 or greater, but not necessarily greater than .51 without additional information about the predictors' intercorrelations and unique contributions.

Question 6

Which of the following statements about adjustments to R²is not true? A) The formula for adjusting R²involves sample size. B) The formula for adjusting R²takes number of predictors into account. C) Adjustments to R²can either increase or decrease the original R²value. D) Adjustments to R² are more substantial when N = 50 than when N = 250.

Accepted Answer

Adjustments to R² always decrease or keep the original R² value the same; they do not increase it.

Question 7

A regression coefficient is:&#10;A) Equal to zero when the r between the predictor and dependent variable is .00&#10;B) Less interpretable as an index of a predictor's importance when it is standardized&#10;C) Always a positive value when the r between its associated predictor and the dependent variable is positive&#10;D) The weight associated with a given predictor when the effects of other predictors are taken into account (held constant)

Accepted Answer

A regression coefficient is the weight or slope associated with a given predictor variable when the effects of other predictors in the model are held constant. Option A is incorrect because a regression coefficient can be non-zero even when the correlation between the predictor and dependent variable is zero. Option B is only partially true - a standardized regression coefficient may be difficult to interpret in its original units, but it still represents the change in the dependent variable associated with a change of one standard deviation in the predictor. Option C is also incorrect because a regression coefficient can be negative even when the correlation between the predictor and dependent variable is positive.

Question 8

The statistic used to test whether the population multiple correlation coefficient is zero is: A) F B) t C) R² D) R

Accepted Answer

The statistic used to test whether the population multiple correlation coefficient is zero is F, which follows the F-distribution. The null hypothesis for this test is that there is no linear relationship between the variables, which is equivalent to the population multiple correlation coefficient being zero.

Question 9

The statistic used to test whether individual regression coefficients are different from zero in the population is: A) F B) t C) R² D) b

Accepted Answer

The t-test is used to test whether individual regression coefficients are different from zero in the population. This is because it evaluates the ratio of the sample estimate to its standard error, which allows us to determine how likely it is that the observed coefficient is different from zero in the population. The other options (A, C, D) are not appropriate for testing individual coefficients. The F-test is used to test the overall significance of a model, R² is not a valid statistical test, and b refers to the slope coefficient itself, not a statistical test.

Question 10

Predictor variables are entered into a regression equation in blocks controlled by the researcher in which of the following?&#10;A) Simultaneous regression&#10;B) Standard regression&#10;C) Hierarchical regression&#10;D) Stepwise regression

Accepted Answer

Hierarchical regression is the only option in which the researcher can control the order in which predictor variables are entered into the equation. In this method, the researcher enters variables in blocks based on theoretical or practical considerations, rather than relying on statistical selection approaches (as in stepwise regression) or entering all predictors at once (as in simultaneous regression). Standard regression is not a named method and does not refer to a specific procedure.

Question 11

Predictor variables are entered into a regression equation based on statistical criteria in which of the following?&#10;A) Simultaneous regression&#10;B) Standard regression&#10;C) Hierarchical regression&#10;D) Stepwise regression

Accepted Answer

The answer of Predictor variables are entered into a regression...

Question 12

Cross validation of a regression equation using a different sample or a random subset of a large sample is especially valuable for which of the following?&#10;A) Simultaneous regression&#10;B) Standard regression&#10;C) Hierarchical regression&#10;D) Stepwise regression

Accepted Answer

The answer of Cross validation of a regression equation using...

Question 13

Which of the following statements about variables in a regression analysis is not true?&#10;A) The dependent variable can be on an interval scale.&#10;B) The dependent variable can be on a nominal scale.&#10;C) The independent variables can be on an interval scale.&#10;D) The independent variables can be on a nominal scale.

Accepted Answer

The answer of Which of the following statements about variables...

Question 14

If one nominal level variable had four categories, how many variables would used to represent the variable in a regression analysis?&#10;A) 1&#10;B) 2&#10;C) 3&#10;D) 4

Accepted Answer

The answer of If one nominal level variable had four...

Question 15

For which of the following would a code of &#189; (.5) be a legitimate value?&#10;A) Dummy coding&#10;B) Reference group coding&#10;C) Effect coding&#10;D) Orthogonal coding

Accepted Answer

The answer of For which of the following would a...

Question 16

For which of the following would a code of -1 not be a legitimate value?&#10;A) Dummy coding&#10;B) Effect coding&#10;C) Orthogonal coding&#10;D) All of the above

Accepted Answer

The answer of For which of the following would a...

Question 17

The term &#34;reference group&#34; in the context of multiple regression refers to:&#10;A) The omitted group when coding nominal-level variables&#10;B) The group representing the most important category in the analysis&#10;C) The group coded 1, as opposed to 0, when coding nominal-level variables&#10;D) The group of interest in planned (a priori) comparisons

Accepted Answer

The answer of The term &#34;reference group&#34; in the context...

Question 18

If two dummy-coded variables were used to create an interaction term for multiple regression, which of the following would be true?&#10;A) The two original dummy-coded variables could be omitted from the analysis&#10;B) Codes for the interaction term would be either 1 or -1&#10;C) Codes for the interaction term would be either 1 or 0&#10;D) There would be a minimum of four predictors in the analysis

Accepted Answer

The answer of If two dummy-coded variables were used to...

Question 19

Although no solution is ideal, the relative importance of predictors can best be evaluated by examining: A) b weights B) Beta weights C) Changes to R² D) Squared semipartial correlation coefficients

Accepted Answer

The answer of Although no solution is ideal, the relative...

Question 20

In multiple regression, the term "suppression" refers to: A) The noninclusion of a predictor in a stepwise regression because of insufficient explanatory power B) The obscuring or alteration of relationships between a predictor and an outcome due to overlapping variability among predictors C) The successively smaller increments to the value of R² as additional predictors are added to the equation D) The instability to the regression equation that can occur when predictor variables are too highly correlated

Accepted Answer

The answer of In multiple regression, the term &#34;suppression&#34; refers...

Question 21

In multiple regression equation, the term "multicollinearity" refers to: A) The noninclusion of a predictor in a stepwise regression because of insufficient explanatory power B) The obscuring or alteration of relationships between a predictor and an outcome due to overlapping variability among predictors C) The successively smaller increments to the value of R² as additional predictors are added to the equation D) The instability to the regression equation that can occur when predictor variables are too highly correlated

Accepted Answer

The answer of In multiple regression equation, the term &#34;multicollinearity&#34;...

Question 22

It would be concluded that multicollinearity was not a problem if: A) Tolerance was .90 or higher B) Tolerance was .001 or lower C) R² was significantly greater than zero D) Singularity did not occur

Accepted Answer

The answer of It would be concluded that multicollinearity was...

Question 23

To examine the viability of the assumption of multivariate normality, which of the following would be examined?&#10;A) Scatterplots that plot residuals against errors of prediction&#10;B) Scatterplots that plot the dependent variable against each independent variable&#10;C) Scatterplots that plot errors of prediction against predicted values of Y&#10;D) Scatterplots that plot beta weights against residuals

Accepted Answer

The answer of To examine the viability of the assumption...

Question 24

A multivariate outlier would be indicated if:&#10;A) A standardized residual value of 3.0 or higher was observed&#10;B) Tolerance for the variable was .90 or higher&#10;C) Tolerance for the variable was .01 or lower&#10;D) A beta weight was negative, when the r between the predictor and the dependent variable was positive

Accepted Answer

The answer of A multivariate outlier would be indicated if:&#10;A)...

Question 25

In multiple regression, all of the following circumstances suggest the need for larger sample sizes except:&#10;A) Stepwise regression is being performed&#10;B) The estimated effect size is small&#10;C) The number of predictors is large&#10;D) Multicollinearity is suspected

Accepted Answer

The answer of In multiple regression, all of the following...

Question 26

Questions pertain to the following table (Table 10), which presents results for a fictitious multiple regression analysis predicting women's positive health practices in pregnancy on the basis of seven independent variables.

-Refer to Table 10. Which predictor variable had the strongest bivariate correlation with the women's health practices during pregnancy?

A) Age
B) Highest grade completed
C) Maternal-fetal attachment
D) It cannot be determined.

Accepted Answer