202 CHAPTER 3 D e s c r i b i n g R e l at i o n s h i p s

Chapter 3 Chapter Review Exercises

These exercises are designed to help you review the important R3.3 Stats teachers’ cars A random sample of AP® Sta-
ideas and methods of the chapter. tistics teachers was asked to report the age (in years)
and mileage of their primary vehicles. A scatterplot
R3.1 Born to be old? Is there a relationship between the of the data, a least-squares regression printout, and
gestational period (time from conception to birth) a residual plot are provided below.
of an animal and its average life span? The figure Predictor Coef SE Coef T P
shows a scatterplot of the gestational period and av- Constant 3704 8268 0.45 0.662
erage life span for 43 species of animals.30 Age 12188 1492 8.17 0.000

S = 20870.5 R-Sq = 83.7% R-Sq(adj) = 82.4%

40 A B
160,000
140,000
Life span (years)

30 120,000
100,000

Mileage
80,000
20
60,000
40,000
10 20,000
0

0 2 4 6 8 10 12
0
0 100 200 300 400 500 600 700 Age
Gestation (days) 60,000
50,000
(a) Describe the association shown in the scatterplot. 40,000
30,000
(b) Point A is the hippopotamus. What effect does this
Residual

20,000
point have on the correlation, the equation of the 10,000
least-squares regression line, and the standard de- 0
viation of the residuals? 10,000
20,000
(c) Point B is the Asian elephant. What effect does this 30,000
point have on the correlation, the equation of the 0 2 4 6 8 10 12
least-squares regression line, and the standard de- Age
viation of the residuals?
(a) Give the equation of the least-squares regression
R3.2 Penguins diving A study of king penguins looked line for these data. Identify any variables you use.
for a relationship between how deep the penguins (b) One teacher reported that her 6-year-old car had
dive to seek food and how long they stay under 65,000 miles on it. Find and interpret its residual.
water.31 For all but the shallowest dives, there is a (c) What’s the correlation between car age and mile-
linear relationship that is different for different pen- age? Interpret this value in context.
guins. The study gives a scatterplot for one penguin (d) Is a linear model appropriate for these data? Ex-
titled “The Relation of Dive Duration (y) to Depth plain how you know.
(x).” Duration y is measured in minutes and depth
(e) Interpret the values of s and r2.
x is in meters. The report then says, “The regression
equation for this bird is: y^ = 2.69 + 0.0138x.” R3.4 Late bloomers? Japanese cherry trees tend to blos-
(a) What is the slope of the regression line? Interpret som early when spring weather is warm and later
this value. when spring weather is cool. Here are some data
on the average March temperature (in °C) and the
(b) Does the y intercept of the regression line make any day in April when the first cherry blossom appeared
sense? If so, interpret it. If not, explain why not. over a 24-year period:32
(c) According to the regression line, how long does a
Temperature (°C): 4.0 5.4 3.2 2.6 4.2 4.7 4.9 4.0 4.9 3.8 4.0 5.1
typical dive to a depth of 200 meters last? Days in April
(d) Suppose that the researchers reversed the variables, to first bloom: 14 8 11 19 14 14 14 21 9 14 13 11
using x = dive duration and y = depth. What effect Temperature (°C): 4.3 1.5 3.7 3.8 4.5 4.1 6.1 6.2 5.1 5.0 4.6 4.0
will this have on the correlation? On the equation Days in April
of the least-squares regression line? to first bloom: 13 28 17 19 10 17 3 3 11 6 9 11
 AP® Statistics Practice Test 203

(a) Make a well-labeled scatterplot that’s suitable for pre- lost Julie’s final exam but knows that her total before
dicting when the cherry trees will bloom from the the exam was 300. He decides to predict her final-
temperature. Which variable did you choose as the exam score from her pre-exam total.
explanatory variable? Explain. (a) Find the equation for the appropriate least-squares
(b) Use technology to calculate the correlation and the regression line for Professor Friedman’s prediction.
equation of the least-squares regression line. Inter- (b) Use the least-squares regression line to predict ­Julie’s
pret the correlation, slope, and y intercept of the line final-exam score.
in this setting. (c) Explain the meaning of the phrase “least squares” in
(c) Suppose that the average March temperature this year the context of this question.
was 8.2°C. Would you be willing to use the equation (d) Julie doesn’t think this method accurately predicts
in part (b) to predict the date of first bloom? Explain. how well she did on the final exam. Determine r2.
(d) Calculate and interpret the residual for the year when the Use this result to argue that her actual score could
average March temperature was 4.5°C. Show your work. have been much higher (or much lower) than the
(e) Use technology to help construct a residual plot. De- predicted value.
scribe what you see. R3.6 Calculating achievement The principal of a high
R3.5 What’s my grade? In Professor Friedman’s econom- school read a study that reported a high correlation
ics course, the correlation between the students’ total between the number of calculators owned by high
scores prior to the final examination and their final- school students and their math achievement. Based
examination scores is r = 0.6. The pre-exam totals for on this study, he decides to buy each student at his
all students in the course have mean 280 and stan- school two calculators, hoping to improve their
dard deviation 30. The final-exam scores have mean math achievement. Explain the flaw in the princi-
75 and standard deviation 8. Professor Friedman has pal’s reasoning.

Chapter 3 AP® Statistics Practice Test

Section I: Multiple Choice Select the best answer for each question.
T3.1 A school guidance counselor examines the number alcoholic beverages for each of 11 regions in Great
of extracurricular activities that students do and their Britain was recorded. A scatterplot of spending on
grade point average. The guidance counselor says, alcohol versus spending on tobacco is shown below.
“The evidence indicates that the correlation between Which of the following statements is true?
the number of extracurricular activities a student par- 6.5
ticipates in and his or her grade point average is close
to zero.” A correct interpretation of this statement 6.0
would be that
5.5
Alcohol

(a) active students tend to be students with poor grades,

and vice versa. 5.0
(b) students with good grades tend to be students who
4.5
are not involved in many extracurricular activities,
and vice versa.
(c) students involved in many extracurricular activities are 3.0 3.5 4.0 4.5
just as likely to get good grades as bad grades; the same is Tobacco
true for students involved in few extracurricular activities.
(a) The observation (4.5, 6.0) is an outlier.
(d) there is no linear relationship between number of activ-
(b) There is clear evidence of a negative association be-
ities and grade point average for students at this school.
tween spending on alcohol and tobacco.
(e) involvement in many extracurricular activities and (c) The equation of the least-squares line for this plot
good grades go hand in hand. would be approximately y^ = 10 − 2x.
T3.2 The British government conducts regular surveys (d) The correlation for these data is r = 0.99.
of household spending. The average weekly house- (e) The observation in the lower-right corner of the plot is
hold spending (in pounds) on tobacco products and influential for the least-squares line.


204 CHAPTER 3 D e s c r i b i n g R e l at i o n s h i p s

T3.3 The fraction of the variation in the values of y that is  T3.5 What was the activity level rating for the fish at a
explained by the least-squares regression of y on x is temperature of 20°C?
(a) the correlation. (a) 87  (b) 84  (c) 81  (d) 66  (e) 3
(b) the slope of the least-squares regression line.  T3.6 Which of the following gives a correct interpreta-
(c) the square of the correlation coefficient. tion of s in this setting?
(d) the intercept of the least-squares regression line.
(a) For every 1°C increase in temperature, fish activity
(e) the residual.
is predicted to increase by 4.785 units.
T3.4 An AP® Statistics student designs an experiment to see (b) The typical distance of the temperature readings
whether today’s high school students are becoming too from their mean is about 4.785°C.
­calculator-dependent. She prepares two quizzes, both (c) The typical distance of the activity level ratings
of which contain 40 questions that are best done using from the least-squares line is about 4.785 units.
paper-and-pencil methods. A random sample of 30 stu- (d) The typical distance of the activity level readings
dents participates in the experiment. Each student takes from their mean is about 4.785.
both quizzes—one with a calculator and one without—
(e) At a temperature of 0°C, this model predicts an ac-
in a random order. To analyze the data, the student con-
tivity level of 4.785.
structs a scatterplot that displays the number of correct
answers with and without a calculator for each of the 30  T3.7 Which of the following statements is not true of
students. A least-squares regression yields the equation the correlation r between the lengths in inches and
weights in pounds of a sample of brook trout?
Calculator = −1.2 + 0.865(Pencil) ​ ​
r = 0.79 (a) r must take a value between −1 and 1.
Which of the following statements is/are true? (b) r is measured in inches.
I. If the student had used Calculator as the explanatory (c) If longer trout tend to also be heavier, then r > 0.
variable, the correlation would remain the same. (d) r would not change if we measured the lengths of
II. If the student had used Calculator as the explanato- the trout in centimeters instead of inches.
ry variable, the slope of the least-squares line would
(e) r would not change if we measured the weights of
remain the same.
the trout in kilograms instead of pounds.
III. The standard deviation of the number of correct an-
swers on the paper-and-pencil quizzes was larger than  T3.8 ​When we standardize the values of a variable, the
the standard deviation on the calculator quizzes. distribution of standardized values has mean 0 and
(a) ​I only (c) ​III only (e) ​I, II, and III standard d ­ eviation 1. Suppose we measure two
(b) ​II only (d) ​I and III only variables X and Y on each of several subjects. We
Questions T3.5 and T3.6 refer to the following setting. Scien- standardize both variables and then compute the
tists examined the activity level of 7 fish at different tempera- least-squares regression line. Suppose the slope of
tures. Fish activity was rated on a scale of 0 (no activity) to 100 the least-squares regression line is −0.44. We may
(maximal activity). The temperature was measured in degrees conclude that
Celsius. A computer regression printout and a residual plot (a) the intercept will also be −0.44.
are given below. Notice that the horizontal axis on the residual (b) the intercept will be 1.0.
plot is labeled “Fitted value.” (c) the correlation will be 1/−0.44.
Predictor Coef SE Coef T P
(d) the correlation will be 1.0.
Constant 148.62 10.71 13.88 0.000
Temperature -3.2167 0.4533 -7.10 0.001 (e) the correlation will also be −0.44.

S = 4.78505 R-Sq = 91.0% R-Sq(adj) = 89.2%  T3.9 There is a linear relationship between the number of
chirps made by the striped ground cricket and the air
7.5
temperature. A least-squares fit of some data collect-
ed by a biologist gives the model y^ = 25.2 + 3.3x,
5.0 where x is the number of chirps per minute and y^
2.5 is the estimated temperature in degrees Fahrenheit.
Residual

What is the predicted increase in temperature for an

0.0 increase of 5 chirps per minute?
2.5 (a) 3.3°F (c) 25.2°F (e) 41.7°F
5.0
(b) 16.5°F (d) 28.5°F
T3.10 A data set included the number of people per ­television
55 60 65 70 75 80 85 90 95 set and the number of people per physician for
Fitted value 40 countries. The Fathom screen shot below displays

Starnes-Yates5e_c03_140-205hr3.indd 204 11/13/13 1:21 PM



AP® Statistics Practice Test 205

a scatterplot of the data with the least-squares regres-

sion line added. In Ethiopia, there were 503 people
per TV and 36,660 people per doctor. What effect
would removing this point have on the regression line?
(a) Slope would increase; y intercept would increase.
(b) Slope would increase; y intercept would decrease.
(c) Slope would decrease; y intercept would increase.
(d) Slope would decrease; y intercept would decrease.
(e) Slope and y intercept would stay the same.

Section II: Free Response ​Show all your work. Indicate clearly the methods you use, because you will be graded on
the correctness of your methods as well as on the accuracy and completeness of your results and explanations.
T3.11 Sarah’s parents are concerned that she seems short grass more heavily, so there are fewer fires and more
for her age. Their doctor has the following record of trees grow. Lions feed more successfully when there are
Sarah’s height: more trees, so the lion population increases. Research-
ers collected data on one part of this cycle, wildebeest
Age (months): 36 48 51 54 57 60
abundance (in thousands of animals) and the percent
Height (cm): 86 90 91 93 94 95 of the grass area burned in the same year. The results of
(a) Make a scatterplot of these data. a least-squares regression on the data are shown here.33
(b) Using your calculator, find the equation of the least-
squares regression line of height on age.
(c) Use your regression line to predict Sarah’s height at
age 40 years (480 months). Convert your prediction
to inches (2.54 cm = 1 inch).
(d) The prediction is impossibly large. Explain why this
­happened.
T3.12 Drilling down beneath a lake in Alaska yields chemi-
cal evidence of past changes in climate. Biological
silicon, left by the skeletons of single-celled creatures
called diatoms, is a measure of the abundance of life
in the lake. A rather complex variable based on the
ratio of certain isotopes relative to ocean water gives 40

an indirect measure of moisture, mostly from snow. 30

As we drill down, we look further into the past. Here 20

is a scatterplot of data from 2300 to 12,000 years ago:

Residual

10

-10

-20

-30
500 750 1000 1250 1500
Wildebeest (1000s)

Predictor Coef SE Coef T P

Constant 92.29 10.06 9.17 0.000
Wildebeest (1000s) −0.05762 0.01035 −5.56 0.000

S = 15.9880  R-Sq = 64.6%  R-Sq(adj) = 62.5%

(a) Identify the unusual point in the scatterplot. Explain
what’s unusual about this point. (a) Give the equation of the least-squares regression
(b) If this point was removed, describe the effect on line. Be sure to define any variables you use.
i. the correlation. (b) Explain what the slope of the regression line means
ii. the slope and y intercept of the least-squares line. in this setting.
(c) Find the correlation. Interpret this value in context.
iii. the standard deviation of the residuals.
(d) Is a linear model appropriate for describing the rela-
T3.13 Long-term records from the Serengeti National Park tionship between wildebeest abundance and percent
in Tanzania show interesting ecological relationships. of grass area burned? Support your answer with appro-
When wildebeest are more abundant, they graze the priate evidence.

Starnes-Yates5e_c03_140-205hr3.indd 205 11/13/13 1:21 PM