CHAPTER 2 — Descriptive Statistics

Complete Homework Solutions

Elementary Statistics: Picturing the World (7th Edition) — Larson & Farber

Super Bowl Scores — 51 Winning Teams (Page 37)

Raw: 35, 33, 16, 23, 16, 24, 14, 24, 16, 21, 32, 27, 35, 31, 27, 26, 27, 38, 38, 46, 39, 42, 20, 55, 20,
37, 52, 30, 49, 27, 35, 31, 34, 23, 34, 20, 48, 32, 24, 21, 29, 17, 27, 31, 31, 21, 34, 43, 28, 24, 34
Sorted: 14, 16, 16, 16, 17, 20, 20, 20, 21, 21, 21, 23, 23, 24, 24, 24, 24, 26, 27, 27, 27, 27, 27, 28,
29, 30, 31, 31, 31, 31, 32, 32, 33, 34, 34, 34, 34, 35, 35, 35, 37, 38, 38, 39, 42, 43, 46, 48, 49, 52, 55

2.1 Frequency Distributions and Their Graphs

TRY IT YOURSELF 1 (p.40) — Frequency Distribution (6 Classes)
Range = 55 − 14 = 41 | Class width = ⌈41/6⌉ = 7
Class Frequency (f)
14 – 20 8
21 – 27 15
28 – 34 14
35 – 41 7
42 – 48 4
49 – 55 3
Total 51

TRY IT YOURSELF 2 (p.41) — Midpoints, Relative Frequency & Cumulative

Frequency
Class Freq (f) Midpoint Rel. Freq Cum. Freq
14–20 8 17 0.157 8
21–27 15 24 0.294 23
28–34 14 31 0.275 37
35–41 7 38 0.137 44
42–48 4 45 0.078 48
49–55 3 52 0.059 51

★ Pattern: Most scores (≈57%) fall in the 21–34 range. The distribution is skewed right.
TRY IT YOURSELF 3 (p.43) — Frequency Histogram
Draw bars at class boundaries (13.5, 20.5, 27.5, 34.5, 41.5, 48.5, 55.5) on the x-axis with frequency
on the y-axis. The tallest bar is 21–27 with height 15.
★ Pattern: Bell-shaped with a right skew. Most winning teams scored between 21 and 34 points.

TRY IT YOURSELF 4 (p.43) — Frequency Polygon

Connect the midpoint-frequency pairs with line segments:
(10, 0) → (17, 8) → (24, 15) → (31, 14) → (38, 7) → (45, 4) → (52, 3) → (59, 0)
Add anchor points at (10, 0) and (59, 0) to close the polygon.

TRY IT YOURSELF 5 (p.44) — Relative Frequency Histogram

Same shape as TIY 3, but the y-axis shows relative frequencies (proportions) instead of counts.
Class Relative Frequency
14–20 0.157
21–27 0.294
28–34 0.275
35–41 0.137
42–48 0.078
49–55 0.059

TRY IT YOURSELF 6 (p.45) — Ogive (Cumulative Frequency Graph)

Plot cumulative frequencies at upper class boundaries, starting from (13.5, 0):
Upper Boundary Cumulative Frequency
20.5 8
27.5 23
34.5 37
41.5 44
48.5 48
55.5 51

2.2 More Graphs and Displays

TRY IT YOURSELF 1 (p.52) — Stem-and-Leaf Plot
Stem Leaves
1 4 6 6 6 7
2 0 0 0 1 1 1 3 3 4 4 4 4 6 7 7 7 7 7 8 9
3 0 1 1 1 1 2 2 3 4 4 4 4 5 5 5 7 8 8 9
4 2 3 6 8 9
5 2 5

★ Pattern: The data clusters in the 20s and 30s. Most winning teams scored between 20 and 39
points.

TRY IT YOURSELF 2 (p.52) — Two Rows per Stem

Stem Leaves
1L (14–16) 4 6 6 6
1H (17–19) 7
2L (20–24) 0 0 0 1 1 1 3 3 4 4 4 4
2H (25–29) 6 7 7 7 7 7 8 9
3L (30–34) 0 1 1 1 1 2 2 3 4 4 4 4
3H (35–39) 5 5 5 7 8 8 9
4L (40–44) 2 3
4H (45–49) 6 8 9
5L (50–54) 2
5H (55–59) 5

★ Pattern: More detail visible. The 20s–30s still dominate. High scores (50s) are rare outliers.

TRY IT YOURSELF 3 (p.54) — Dot Plot

Place a dot above each score on a number line from 14 to 55. Multiple identical scores stack
vertically.
★ Pattern: Scores cluster between 20–35. Distribution is skewed right, with gaps at the high end.

TRY IT YOURSELF 4 (p.55) — Pie Chart (Earned Degrees 1990)

Total = 455 + 1051 + 330 + 104 = 1,940 thousand degrees
Degree Count (thousands) Percentage Angle
Associate's 455 23.5% 84.6°
Bachelor's 1,051 54.2% 195.1°
Master's 330 17.0% 61.2°
Doctoral 104 5.4% 19.3°

TRY IT YOURSELF 5 (p.56) — Pareto Chart (BBB Complaints)

Arrange bars from greatest to least:
Industry Complaints
Collection agencies 19,277
Auto dealers (used cars) 16,281
Insurance companies 8,384
Travel agencies 6,985
Mortgage brokers 3,634

★ Greatest cause of complaints: Collection agencies with 19,277 complaints.

TRY IT YOURSELF 6 (p.57) — Scatter Plot (Employment vs. Salary)

Years of Employment Salary ($) Years of Employment Salary ($)
2 25,000 7 41,650
3 28,000 8 40,000
4 27,350 9 45,100
4 32,000 10 43,000
5 32,000 6 39,225

★ Trend: Strong positive correlation — more years of employment → higher salary.

TRY IT YOURSELF 7 (p.58) — Time Series Chart (Burglaries 2005–2015)

Plot years on the x-axis and number of burglaries on the y-axis. Connect each data point with a line.
★ Trend: Burglaries generally decreased over the 2005–2015 period.

2.3 Measures of Central Tendency

TRY IT YOURSELF 1 (p.62) — Mean of 51 Scores
x̄ = 1,541 / 51 ≈ 30.2 points

TRY IT YOURSELF 2 (p.63) — Median of 51 Scores

n = 51 → Median = 26th value in the sorted list
Median = 30 points

TRY IT YOURSELF 3 (p.63) — Median of 2001–2016 Super Bowl Scores

Data: 20, 48, 32, 24, 21, 29, 17, 27, 31, 31, 21, 34, 43, 28, 24, 34
Sorted: 17, 20, 21, 21, 24, 24, 27, 28, 29, 31, 31, 32, 34, 34, 43, 48
n = 16 → Median = average of 8th and 9th values = (28 + 29) / 2 = 28.5 points

TRY IT YOURSELF 4 (p.64) — Mode of 51 Scores

The score 27 appears 5 times (most frequent).
Mode = 27 points

TRY IT YOURSELF 5 (p.64) — Mode of Survey Responses

Response Count
A great deal 550
Some 578 ← most frequent
Not too much 274
Not at all 119
No answer 13

Mode = "Some" (578 respondents)

TRY IT YOURSELF 6 (p.65–66) — Effect of Removing Outlier (65)

Original data (n=20): 20,20,20,20,20,20,21,21,21,21,22,22,22,23,23,23,23,24,24,65
Measure With Outlier (65) Without Outlier Change
Mean 475/20 = 23.8 yrs 410/19 ≈ 21.6 yrs Decreased significantly
Median (21+22)/2 = 21.5 yrs 21 yrs (10th value) Decreased slightly
Mode 20 years 20 years No change

★ The mean is most affected by outliers. The median changed only slightly, and the mode was
unaffected.

TRY IT YOURSELF 7 (p.66) — Weighted Mean (Grade Changed to B)

From Example 7 — Original grades: C (3 cr), C (4 cr), D (1 cr), A (3 cr), C (2 cr), B (3 cr)
Change: The 2-credit course grade changes from C (2 pts) to B (3 pts)
Final Grade Credit Hours (w) Points (x) x×w
C 3 2 6
C 4 2 8
D 1 1 1
A 3 4 12
B (was C) 2 3 (was 2) 6 (was 4)
B 3 3 9
Total Σw = 16 Σ(xw) = 42

New Weighted Mean = 42 / 16 = 2.625 ≈ 2.63

★ The weighted mean increased from 2.5 to 2.63 by changing the 2-credit course grade from C to
B.

TRY IT YOURSELF 8 (p.67) — Estimated Mean from Frequency Distribution

Class Midpoint (x) Frequency (f) x·f
14–20 17 8 136
21–27 24 15 360
28–34 31 14 434
35–41 38 7 266
42–48 45 4 180
49–55 52 3 156
Total — 51 1,532

Estimated Mean = 1,532 / 51 ≈ 30.0 points

★ The estimated mean (30.0) is very close to the actual mean (30.2). The frequency distribution
gives a good approximation.

2.4 Measures of Variation

TRY IT YOURSELF 1 (p.74) — Range for Corporation B
Corporation B starting salaries (in $1,000s): 40, 23, 41, 50, 49, 32, 41, 29, 52, 58
Range = Maximum − Minimum = 58 − 23 = 35 → Range = $35,000
Comparison: Corporation A's range was $10,000. Corporation B has much greater variability
($35,000).

TRY IT YOURSELF 2 (p.76) — Population Variance & Standard Deviation for

Corporation B
Corporation B data (in $1,000s): 40, 23, 41, 50, 49, 32, 41, 29, 52, 58
Population Mean: μ = 415 / 10 = 41.5
x x−μ (x − μ)²
40 −1.5 2.25
23 −18.5 342.25
41 −0.5 0.25
50 8.5 72.25
49 7.5 56.25
32 −9.5 90.25
41 −0.5 0.25
29 −12.5 156.25
52 10.5 110.25
58 16.5 272.25
Sum: 1,102.5

Population Variance: σ² = 1,102.5 / 10 = 110.25

Population Standard Deviation: σ = √110.25 = 10.5 → σ = $10,500
 ★ Corporation B's SD ($10,500) is much larger than Corporation A's, confirming greater salary
spread.

TRY IT YOURSELF 3 (p.77) — Sample Variance & Std Dev (Recovery Times)
Group 2 data (days): 43, 57, 18, 45, 47, 33, 49, 24 n=8
Sample Mean: x̄ = 316 / 8 = 39.5 days
x x − x̄ (x − x̄ )²
43 3.5 12.25
57 17.5 306.25
18 −21.5 462.25
45 5.5 30.25
47 7.5 56.25
33 −6.5 42.25
49 9.5 90.25
24 −15.5 240.25
Sum: 1,240.00

Sample Variance: s² = 1,240 / 7 ≈ 177.14

Sample Standard Deviation: s = √177.14 ≈ 13.3 days

TRY IT YOURSELF 4 (p.78) — Mean & Std Dev (Dallas Office Rental Rates)
Dallas rates ($/sq ft/yr), n=24:
18,27,21,14,20,20,24,11,16,7,12,22,10,15,21,34,23,13,38,16,18,30,15,30
Sum = 475 Sample Mean: x̄ = 475 / 24 ≈ $19.79 per sq ft/yr
x (x−19.79)² x (x−19.79)²
18 3.20 34 201.92
27 51.98 23 10.30
21 1.46 13 46.10
14 33.52 38 331.60
20 0.04 16 14.36
20 0.04 18 3.20
24 17.72 30 104.24
11 77.26 15 22.94
16 14.36
7 163.58
12 60.68
22 4.88
10 95.84
15 22.94
21 1.46
 Sum: 1,387.86

Sample Variance: s² = 1,387.86 / 23 ≈ 60.34

Sample Standard Deviation: s = √60.34 ≈ $7.77 per sq ft/yr

TRY IT YOURSELF 5 (p.79) — Create a Data Set (n=10, mean=10, σ≈3)

One valid answer: 7, 7, 7, 7, 10, 10, 10, 13, 13, 16
Check: Sum = 100 → Mean = 10 ✓
Σ(x−10)² = 9+9+9+9+0+0+0+9+9+36 = 90 → σ = √(90/10) = √9 = 3 ✓
★ There are many correct answers. Any 10 values with mean=10 and σ≈3 is acceptable.

TRY IT YOURSELF 6 (p.80) — Empirical Rule (Women's Heights)

From Example 6: μ = 64.2 in, σ = 2.9 in (women ages 20–29)
Range: 64.2 to 67.1 = μ to (μ + σ) → this is half of the ±1σ region
By the Empirical Rule: 68% of data falls within ±1σ, so from μ to μ+σ = 68%/2 = 34%
≈ 34% of women ages 20–29 have heights between 64.2 and 67.1 inches.

TRY IT YOURSELF 7 (p.81) — Chebyshev's Theorem (Iowa, k=2)

Iowa: μ = 39.3 years, σ = 23.5 years (from Example 7)
Chebyshev (k=2): At least (1 − 1/4) × 100% = 75% of data within 2σ
Lower bound: 39.3 − 2(23.5) = −7.7 (effectively 0, since age ≥ 0)
Upper bound: 39.3 + 2(23.5) = 86.3 years
Conclusion: At least 75% of Iowa residents' ages fall between 0 and 86.3 years.
Is age 80 unusual? Since 80 < 86.3, age 80 is within 2σ of the mean → NOT unusual.
★ Age 80 is within the expected range for Iowa residents under Chebyshev's Theorem.

TRY IT YOURSELF 8 (p.82) — Effect of Changing Three 6s to 4s

From Example 8: n=50 children per household. Original: x̄ ≈ 1.82, s ≈ 1.7
Original frequencies: x=0(f=10), x=1(f=19), x=2(f=7), x=3(f=7), x=4(f=2), x=5(f=1), x=6(f=4)
Change: 3 sixes → 4s. New: x=4 (f=5), x=6 (f=1)
New Σxf = 91 − 3(6) + 3(4) = 85 New Mean = 85/50 = 1.70
x f (x−1.70) (x−1.70)² (x−1.70)²·f
0 10 −1.70 2.89 28.90
1 19 −0.70 0.49 9.31
2 7 0.30 0.09 0.63
3 7 1.30 1.69 11.83
4 5 2.30 5.29 26.45
5 1 3.30 10.89 10.89
6 1 4.30 18.49 18.49
Total 50 Sum: 106.50
New s² = 106.50 / 49 ≈ 2.17 New s = √2.17 ≈ 1.47
★ Effect: Mean decreased (1.82→1.70) and SD decreased (1.7→1.47) because extreme 6s were
replaced with 4s, closer to the mean.

TRY IT YOURSELF 9 (p.83) — Effect of Changing Midpoint from 599.5 to 650

Original (Example 9): midpoint for '$500+' = 599.5, mean = $192, s ≈ $160.30
Old xf (last class): 599.5 × 70 = 41,965
New xf (last class): 650 × 70 = 45,500
New Σxf = 192,000 + 3,535 = 195,535
New Sample Mean = 195,535 / 1,000 ≈ $195.54
Class Midpoint (x) f (x−195.54)²·f
0–99 49.5 380 8,104,518
100–199 149.5 230 487,526
200–299 249.5 210 611,453
300–399 349.5 50 1,185,184
400–499 449.5 60 3,869,741
500+ 650 70 14,457,372
Sum: 28,715,794

New s = √(28,715,794 / 999) ≈ $169.54

★ Conclusion: A higher midpoint increases the mean from $192 to $195.54 and the SD from
$160.30 to $169.54.

TRY IT YOURSELF 10 (p.84) — Coefficient of Variation (LA vs. Dallas)

Formula: CV = (s / x̄ ) × 100%
City Mean (x̄ ) Std Dev (s) CV
Los Angeles (Example $36.88 $17.39 47.2%
4)
Dallas (TIY 4) $19.79 $7.77 39.3%

Conclusion: LA has higher relative variability (47.2%) than Dallas (39.3%).

★ Even though LA has a higher standard deviation in dollars, the CV shows its rates are more
variable relative to the mean. CV is useful for comparing data sets with different units or scales.

2.5 Measures of Position

TRY IT YOURSELF 1 (p.91) — Quartiles for 51 Super Bowl Scores
n = 51 → Q2 (Median) = 26th value = 30
Lower half (values 1–25): Q1 = 13th value = 23
Upper half (values 27–51): Q3 = 13th value from position 27 = 35
Quartile Value Interpretation
Q1 (First Quartile) 23 points 25% of winning teams scored ≤
23 pts
Q2 (Median) 30 points 50% of winning teams scored ≤
30 pts
Q3 (Third Quartile) 35 points 75% of winning teams scored ≤
35 pts

★ Half of all winning Super Bowl teams scored between 23 and 35 points.

TRY IT YOURSELF 2 (p.92) — Quartiles for Tuition Costs (25 Universities)

Data: 44,30,38,23,20,29,19,44,29,17,45,39,29,18,43,45,39,24,44,26,34,20,35,30,36
Sorted: 17,18,19,20,20,23,24,26,29,29,29,30,30,34,35,36,38,39,39,43,44,44,44,45,45
n=25 → Q2 = 13th value = $30,000
Lower half (1–12): Q1 = avg of 6th & 7th = (23+24)/2 = $23,500
Upper half (14–25): Q3 = avg of 6th & 7th = (39+43)/2 = $41,000
Quartile Value
Q1 $23,500
Q2 (Median) $30,000
Q3 $41,000

★ The middle 50% of tuitions range from $23,500 to $41,000. Wide spread among universities.

TRY IT YOURSELF 3 (p.93) — IQR & Outliers for 51 Scores

IQR = Q3 − Q1 = 35 − 23 = 12
Lower fence: Q1 − 1.5(IQR) = 23 − 18 = 5
Upper fence: Q3 + 1.5(IQR) = 35 + 18 = 53
Score of 55 > 53 → 55 is an outlier!
★ There is one outlier in the data set: 55 points.

TRY IT YOURSELF 4 (p.94) — Box-and-Whisker Plot for 51 Scores

Five-Number Summary Value
Minimum 14
Q1 23
Q2 (Median) 30
Q3 35
Maximum 55 (outlier)

Draw a box from Q1=23 to Q3=35 with a vertical line at the median=30.
Left whisker extends to min=14. Right whisker extends to upper fence=53.
Plot the outlier (55) as a separate dot beyond the upper whisker.
★ The distribution is slightly right-skewed. The outlier at 55 extends the right side.

TRY IT YOURSELF 5 (p.95) — 10th Percentile from Ogive

10% of 51 = 5.1 → corresponds to approximately the 5th–6th values in sorted data
5th value = 17, 6th value = 20
The 10th percentile ≈ 17 points
★ Interpretation: About 10% of winning Super Bowl teams scored 17 points or fewer.

TRY IT YOURSELF 6 (p.96) — Percentile for $26,000 (Tuition Data)

Sorted data: 17, 18, 19, 20, 20, 23, 24, 26, 29, ...
Number of values below 26 = 7
Percentile = (7 / 25) × 100 = 28th percentile
★ A tuition of $26,000 is at the 28th percentile — lower than about 72% of the universities.

TRY IT YOURSELF 7 (p.97) — Z-Scores for Utility Bills

μ = $70, σ = $8 Formula: z = (x − μ) / σ
Bill (x) Calculation z-score Interpretation
$60 (60−70)/8 z = −1.25 Below average;
moderately low
$71 (71−70)/8 z = 0.13 Very close to mean;
typical
$92 (92−70)/8 z = 2.75 Unusually high (>2σ
above mean)

★ A utility bill of $92 is unusual (z=2.75). Bills of $60 and $71 are within normal range.

TRY IT YOURSELF 8 (p.97) — Z-Scores: 5-Foot Man vs. 5-Foot Woman

Height = 5 feet = 60 inches
Men: μ = 69.9 in, σ = 3.0 in
Women: μ = 64.2 in, σ = 2.9 in
Person Calculation z-score Conclusion
5-foot man (60−69.9)/3.0 z = −3.30 Very unusual for a man
5-foot woman (60−64.2)/2.9 z = −1.45 Somewhat below
average; not unusual

Conclusion: A 5-foot man (z=−3.30) is far more unusual than a 5-foot woman (z=−1.45).
★ Being 5 feet tall is very rare among men, but relatively less uncommon among women.

— End of Chapter 2 Solutions — | Elementary Statistics: Picturing the World, 7th Edition | Larson & Farber