Q: Suppose the regression equation is $$\hat{y} = 0.5x + 10. $$Predict the length of a song released in 2001.

$$\hat{y} = 0.5 \times 2001 + 10 = 1010.5$$

Q: A song released in 2008 has a duration of 242 seconds. If the regression equation is $$\hat{y} = 0.5x + 10$$, what is the residual for this song?

Residual $$= 242 - (0.5 \times 2008 + 10) = 242 - 1014 = -772$$

Question 1

A linear regression model is fit to a dataset that includes an influential outlier at (1900, 105). What is the most likely effect of including this outlier on the regression line?

Accepted Answer

It will distort the regression line, pulling it toward the outlier and affecting predictions.

Question 2

After removing a high-leverage outlier from a regression analysis, which of the following changes is most likely to occur?

Accepted Answer

The regression line will better represent the general trend of the data.

Question 3

Given a regression equation $$\hat{y} = 0.5x + 10$$, what does the slope of 0.5 represent in the context of predicting song length from year of release?

Accepted Answer

For each additional year, the predicted song length increases by 0.5 units.

Question 4

If the $$R^2$$ value for a regression model is $$0.036748$$, what does this indicate about the model?

Accepted Answer

About 3.7% of the variation in the response variable is explained by the model.

Question 5

Given $$R^2 = 0.036748$, what is the value of the correlation coefficient r$?

Accepted Answer

$$r = 0.1917$$

Question 6

Suppose the regression equation is $$\hat{y} = 0.5x + 10. $$Predict the length of a song released in 2001.

Accepted Answer

$$\hat{y} = 0.5 	imes 2001 + 10 = 1010.5$$

Question 7

A song released in 2008 has a duration of 242 seconds. If the regression equation is $$\hat{y} = 0.5x + 10$$, what is the residual for this song?

Accepted Answer

Residual $$= 242 - (0.5 	imes 2008 + 10) = 242 - 1014 = -772$$

Question 8

Which of the following $$X$$ values would be considered extrapolation if the data range is from 1950 to 2020?

Accepted Answer

$$X = 2050$$

Question 9

Why is the y-intercept in a regression model predicting song length from year of release often nonsensical?

Accepted Answer

Because it predicts the song length for year zero, which is outside the range of the data.

Question 10

Given the regression output, the error sum of squares (SSE) is 5000 and the total sum of squares (SST) is 5200. What is the $$R^2$$ value?

Accepted Answer

$$R^2 = 1$$ - \frac{5000}{5200} = 0.0385$$

Question 11

If the error sum of squares (SSE) is 5000 and there are 100 data points, what is the estimate of the error standard deviation $$s$$?

Accepted Answer

$$s = \sqrt{\frac{5000}{100 - 2}} = \sqrt{\frac{5000}{98}} \approx 7.12$$

Question 12

Which value represents the total squared distance from the prediction line to $$y in a $$regression analysis?

Accepted Answer

The error sum of squares (SSE): $$\sum (y_i - \hat{y}_i)^2$

Question 13

If a regression model has a very low $$R^2$$ value, what does this suggest about the model's predictive power?

Accepted Answer

The model explains very little of the variation in the response variable.

Question 14

Which of the following best describes the effect of an influential outlier on regression analysis?

Accepted Answer

It can significantly change the slope and intercept of the regression line.

Question 15

A regression model is used to predict song length from year of release. Which of the following is an example of extrapolation?

Accepted Answer

Predicting the length of a song released in 2050 when the data only includes years up to 2020.

Question 16

If the correlation coefficient $$r is $$negative, what does this indicate about the relationship between the two variables?

Accepted Answer

As one variable increases, the other tends to decrease.

Question 17

Which of the following is the correct formula for the residual of a data point in regression analysis?

Accepted Answer

$$e_i = y_i - \hat{y}_i$$

Question 18

In regression analysis, what does the error sum of squares (SSE) measure?

Accepted Answer

The total squared difference between observed and predicted values.

Question 19

If a regression model's $$R^2$$ value increases after removing an outlier, what does this suggest?

Accepted Answer

The outlier was reducing the model's explanatory power.

Regression Analysis: Outliers, Model Interpretation, and Error Measures

Study Guide - Practice Questions

Study Guide - Flashcards