Under what circumstance, unlikely as it might be, would the standard error of estimate be zero?
a. The correlation coefficient is either 1.00 or -1.00.
b. The proportionate reduction in error is also zero.
c. The correlation coefficient is also zero.
d. The standard error of estimate can never be zero.
Every year it seems as though last season’s baseball rookie of the year fails to live up to expectations for his sophomore season. What might explain this phenomenon?
a. regression to the mean
b. proportionate reduction in error
c. overestimation of effect size
d. standard error of the estimation
With regression we are concerned about variability around the ________, rather than variability around the ________ which would be the case in t tests or ANOVAs.
a. outliers; line of best fit
b. median; tails of the distribution
c. mean; outliers
d. line of best fit; mean
In a study designed to predict blood cholesterol levels from amount of daily saturated fat in grams (X1) and number of hours of daily exercise (X2), we determine that the slope of X1 is 5, the slope of X2 is -4, and the y intercept is 130. Which of the following formulas is the regression equation for these data?
a. ? = 130 + 5(X1) -4(X2)
b. ? = 130 + 5(X1) + 4(X2)
c. ? = 130 + 1(X)
d. ? = 130 -5(X1) -4(X2)
We can examine a graph to get a sense of how much error there is in a regression equation. Which of the following describes a graph that reveals there will be a high amount of error when using our regression equation?
a. Data points cluster very close to the line with several outlier exceptions.
b. The data points consistently cluster far away from the line of best fit.
c. Data points cluster close around the line of best fit.
d. Data points fall directly on the line.
In the equation for a regression line, the slope is the:
a. z score of the amount that Y is predicted to increase as X increases.
b. value for X when Y is equal to 0.
c. predicted value for Y when X is equal to 0.
d. amount that Y is predicted to increase for a one-unit increase in X.
The statistical analysis that allows us to use one scale variable to predict outcome on a second scale variable is called:
a. linear analysis.
The standard error of the estimate indicates:
a. the typical distance between the regression line and each of the observed data points.
b. how much error there is in any single prediction we make from a given regression equation.
c. how far, on average, the regression line is from the mean.
d. how far two regression lines are from each other.
An independent variable that makes a unique contribution to the prediction of a dependent variable is a(n) ________ variable.
A researcher wants to be able to predict first-semester grade point average with as much accuracy as possible, so she would like to use both high school grade point average and SAT score as predictor variables. Which of the following techniques would be most appropriate to make this prediction?
a. standardized regression coefficient
b. multiple regression
c. simple linear regression
d. proportionate reduction in error
To determine the Y intercept, we determine the value of Y when X is:
a. at the minimum of the data set.
b. at the mean.
c. at the maximum of the data set.
In regression, the discovery of additional predictor variables that are separate and distinct serves to:
a. over-complicate the analysis.
b. help researchers assess error of prediction.
c. explain more of the variability in our outcome variable.
d. create redundancy in our equation.
If two variables, independently, can help us predict the outcome of a third variable, we say that they are:
The predicted z score for the dependent variable will always be ________ the individual’s z score for the independent variable.
a. less than
b. the same as
c. two times
d. more than
As the standard error of estimate becomes larger, predictions become:
a. less accurate.
c. more accurate.
The tendency for very tall parents to have children who are somewhat short illustrates the phenomenon:
b. regression to the mean.
c. central limit theorem.
d. central tendency.
A small standard error of the estimate means that:
a. your two variables are poorly correlated.
b. variability is high in your Y variable.
c. confounding variables may be present.
d. you are making predictions with great accuracy.
For a simple linear regression, the standardized regression coefficient is:
a. the square root of the slope.
b. the square of the r statistic.
c. equal to the Pearson correlation coefficient.
d. unrelated to the correlation value.
The regression line is also called the:
a. line of best fit.
b. line of central limit.
c. prediction estimate.
d. error of estimate.
Once we have an equation for a straight line through our data, we can:
a. use an independent-samples t test to compare means.
b. speculate about the causal relationship.
c. compute post hoc tests.
d. look at each value on the x-axis and predict its corresponding value on the y-axis.