Get the entire Solution Manual for Interactive Statistics (Classic Version), 3rd Edition, Martha Aliaga in PDF format instantly.
Get the entire Solution Manual for Interactive Statistics (Classic Version), 3rd Edition, Martha Aliaga in PDF format instantly.
com
http://testbankbell.com/product/solution-manual-for-
interactive-statistics-classic-version-3rd-edition-martha-
aliaga/
OR CLICK HERE
DOWLOAD NOW
https://testbankbell.com/product/test-bank-for-interactive-
statistics-3-e-3rd-edition-0131497561/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-thermodynamics-
an-interactive-approach-subrata-bhattacharjee/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-introduction-to-
environmental-engineering-si-version-3rd-edition/
testbankbell.com
https://testbankbell.com/product/principles-of-marketing-
version-3-0-3rd-tanner-solution-manual/
testbankbell.com
Introduction to Psychology Version 3 0 3rd Stangor
Solution Manual
https://testbankbell.com/product/introduction-to-psychology-
version-3-0-3rd-stangor-solution-manual/
testbankbell.com
https://testbankbell.com/product/solution-manual-for-elementary-
statistics-3rd-edition-william-navidi-barry-monk-2/
testbankbell.com
https://testbankbell.com/product/probability-statistics-and-
reliability-for-engineers-and-scientists-3rd-ayyub-solution-manual/
testbankbell.com
https://testbankbell.com/product/test-bank-for-elementary-statistics-
a-brief-version-5th-edition-bluman/
testbankbell.com
2.2
(a) The value of 9% is a parameter – it is a numerical summary of the population.
(b) The value of 12.5% is a statistics – it is a numerical summary of a sample from the population.
2.3
The proportion of register voters in Ann Arbor who would vote “Yes” on Proposal A is an example of (c) a
parameter.
2.4
(a) The population consists of (the planned vote for) the 100 U.S. Senators.
(b) N = 100.
(c) The sample consists of (the planned vote for) the ten selected U.S. Senators.
(d) n = 10.
2.5
For a simple random sample of n = 200 and the number of items that were defective = 5. All we can say is that (c)
$ = 5/200 = 0.025 or 2.5%.
the percent of defective items in the sample is 5/200 = 2.5%. p
2.6
(a) 22.4.
(b) 19. No.
(c) 22.5. No. No.
(d) 18 and 21. The sample mean age is 19.5.
18 and 26. The sample mean age is 22.
20 and 27. The sample mean age is 23.5.
20 and 21. The sample mean age is 20.5.
20 and 26. The sample mean age is 23.
27 and 21. The sample mean age is 24.
27 and 26. The sample mean age is 26.5.
2.7
(a) The value of 34 is a statistic.
(b) Response bias.
2.8
(a) Statistic since the 54% is a sample percentage, not of the population.
(b) Nonresponse bias.
2.9
Response bias.
2.11
Nonresponse bias.
2.13
The best answer is (c).
2.14
Results will vary.
(a) Using the calculator with a seed value of 291, the selected persons are 39 (79 years old), person 3 (75 years old)
and person 24 (70 years old). The average age is 74.67 years.
(b) 22 (36 years old), 34 (89 years old) and 29 (89 years old). The average age is 71.33 years, different from the
mean calculated in part (a).
(c) 69.19 years.
(d) 74.67 and 71.33 are statistics, while the mean in part (c) is a parameter.
2.15
(a) Yes, each sample of size 20 has the same chance as any other sample of size 20 to be selected (assuming all
tags are exactly the same and the box is thoroughly mixed).
(b) It is drawn without replacement.
2.16
(a) 12/100.
(b) With the graphing calculator the selected employees are: 77, 51, 72, 40, 71, 42, 17, 34, 62, 23, 35, and 12.
From the random table the selected employees are: 7, 5, 69, 76, 28, 33, 78, 70, 99, 98, 42, and 80.
2.17
(a) H0: The population proportion of dissatisfied customers equals 0.10.
H1: The population proportion of dissatisfied customers is less than 0.10.
(b) With the calculator the selected customers are: 34318, 15553, 8461, and 614. With the random table the
selected customers are: 15409, 23336, 29490, and 30414.
(c) Type II error.
(d) 0.21.
(e) No, the p-value is > 0.05.
(f) (ii) Statistic.
2.18
Using the TI: Label the sites from 1 to 80. With a seed value of 29 the five labels, and thus sites, selected at random
are as follows: Sites #50, #66, #43, #49, and #74. Using the Random Number Table: Since there are more than 10,
but less than 100, sites, we can use double-digit labels, such as 01, 02, 03, through 80. With Row 10 of the table,
starting at Column 1, the simple random sample of sites consists of sites #47, #53, #68, #57, and #34.
2.19
Stratified random sampling. You are dividing up your population by class rank then selecting 100 students from
within each stratum at random.
2.20
(a) Stratifying by declared major (field of study) would be a good stratification variable. The cost of textbooks is
likely to vary from one field of study to the next, but not vary as much within a particular field of study (many
students have to take the same basic classes for their field of study and hence may be paying similar textbook
costs.) You would also learn about the costs for each field of major as well as costs overall.
(b) Stratifying by gender may not be as useful as a stratification variable. The cost of textbooks may vary a lot for
females and vary a lot for males, and there may not be much variation between males and females. However, if
you wanted to learn about the costs for each gender as well as overall, you might wish to stratify by gender.
(c) Stratifying by class rank may be useful as a stratification variable. The cost of textbooks may vary from one
class-rank to another. However, there may be quite a lot of variation within each class-rank as well. If you
wanted to learn about the costs for each class-rank as well as overall, you might wish to stratify by class-rank.
2.22
(a) 100/200 = 1/2 = 0.50 or 50%.
(b) 100/1000 = 1/10 = 0.10 or 10%.
(c) The chance of being chosen is 0.10. This is NOT a simple random sample. For this stratified random sampling
plan each possible sample would contain exactly 100 males and 20 females. All samples of size 120 are not
equally likely (as it should be for simple random sample). Some samples of size 120 are not even possible, for
example, having 120 males.
2.23
We should use a weighted average: ⎛⎜ 40 ⎞⎟ (70) + ⎛⎜ 60 ⎞⎟ ( 63) = 65.8 inches
⎝ ⎠
100 ⎝ 100⎠
2.24
(a) 20/200 = 0.10.
(b) With the calculator the first five females selected in the sample are: 132, 195, 147, 171, and 85. With the
random number table: To each woman we assigned 5 numbers, for example 1, 201, 401, 601 and 801. The
numbers from the table were 521, 625, 391, 646, and 369. So the first five selected females are: 121, 25, 191,
46, and 169.
(c) The weighted average is: (0.75)8 + (0.25)5 = 7.25.
2.25
(a) Stratified random sampling.
(b) With the calculator the first five selected homes are: 386, 81, 379, 211, and 156. With the random number table
the first five selected homes are: 94, 299, 396, 378, and 363.
(c) (0.60)(2100) + (0.40)(2600) = 2300 square feet.
2.26
(0.20)(16) + (0.50)(43) + (0.30)(71) = 46 years.
2.27
(a) Stratified random sampling.
(b) With the calculator the first six homes selected from the 1000 homes in County I are: 918, 193, 902, 502, 370,
and 5. With the random number table the first six homes selected from the 1000 homes in County I are: 963,
19, 197, 705, 463, and 79 where the homes are labeled 0 to 999.
(c) (0.50)(175) + (0.30)(200) + (0.20)(195) = 186.5 thousands of dollars or $186,500.
2.28
(a) With the calculator, the label of the first student selected is 3. With the random number table, the label of the
first student selected is 3.
(b) The students in the sample are those with ID numbers 3, (3 + 4 =) 7, (7 + 4 =) 11, and (11 + 4=) 15, for a total
of 4 students.
2.29
(a) With the calculator with the first 100 addresses labeled 1 through 100, the sample is: Addresses = #79, #179,
#279, #379, #479. Using the random number table with the first 100 addresses labeled 01, 02, ... , 98, 99, 00,
the sample is: Addresses = #30, #130, #230, #330, #430.
(b) 1/100 = 0.01 or 1% There are 100 possible systematic samples of size 5, each equally likely.
2.31
Yes, the chance is 1/5 or 0.20 or 20%. There are 5 possible systematic samples, each equally likely.
2.32
(a) 1/14 = 0.0714.
(b) Note that there are 555 members so 555/14 Æ 39.64 or 39 groups of 14 and one last group of 9. For the 17th
member to be selected, the starting point must have been a 3, that is, the 3rd member in each group of 14 is
selected. This will result in a sample of 40 members.
2.33
(a) 1/8 = 0.125.
(b) Note that there are 350 members so 350/8 Æ 43.75 or 43 groups of 8 and one last group of 6. For the 300th
member to be selected, the starting point must have been a 4, that is, the 4th member in each group of 8 is
selected. This will result in a sample of 44 members.
2.34
(a) 1/40 = 0.025.
(b) Using the TI calculator with seed of 19, and with the first 40 members labeled 1 through 40, the first five
members are 7, 47, 87, 127, and 167. Using row 15, column 1, and labeling the first 40 members as 01, 02, ... ,
39, 40, the first 5 members are 7, 47, 87, 127, and 167.
(c) Note that there are 2220 members so 2220/40 Æ 55.5 or 55 groups of 40 and one last group of 20. Since the
starting point was a 7, that is, the 7th member in each group of 40 was selected. This would result in a complete
sample of 56 members.
2.35
False, the chance depends on the number of clusters.
2.36
(a) With the calculator, the two departments selected were 2 and 4; Chemistry and Mathematics. Using the random
number table, the two department selected were 2 and 3; Chemistry and Biology.
(b) With calculator, the sample size is 80. With random number table, the sample size is 90.
2.37
(a) Cluster sampling. The faculty are grouped into departments which serve as clusters. Six of the clusters are
selected at random. All of the units in the cluster are in the sample.
(b) Label the list of the 60 departments 1, 2, 3, ..., 60. The numbers generated and thus the departments
selected using the TI calculator with seed = 79 are: 54, 37, 49, 5, 15, and 43. If you are using the random number
table, you might label the list of the 60 departments 01, 02, 03, ... , 60. The numbers generated and thus the
departments selected starting at row 60, column 1 are: 23, 22, 47, 40, 25, and 37.
(c) Yes we can determine the chance that any specific professor will be selected. The chance of selecting a
professor is the same chance that his/her department or cluster will be selected. The reason being if a cluster is
selected, every element in the cluster is selected. Therefore the probability that a cluster is selected is 6/60 =
0.10, the chance that any specific professor will be selected.
2.39
(a) The type of sampling performed in each dorm is cluster sampling, with the rooms forming the clusters and 3
clusters were selected at random from each of the four dorms.
(b) If each cluster selected has one student, we would have 3 students from each of the four dorms for a minimum
sample size of 12 students.
(c) If each cluster selected has three students, we would have 9 students from each of the four dorms for a
maximum sample size of 36 students.
(d) We do not know, it will depend on how many clusters selected are rooms with women. The number of women
could be as low as 0 and as high as 36.
(e) No, there will be anywhere from 3 to 9 freshmen students sampled from the freshmen dorm, as well as from 3
to 9 sophomores, from 3 and 9 juniors, and from 3 to 9 seniors.
2.40
(a) Cluster sampling.
(b) No, if the number of adults per city block is unknown. Yes, if you have the seed value and you know how
many adults are in each block.
2.41
(a) With the calculator or the random number table, the selected region is 3 = Southwest.
(b) Stratified random sampling.
(c) (i) 1-in-10 systematic sampling.
(ii) 0.10.
(iii) With the calculator, the first five selected cans are 7, 17, 27, 37, and 47. With the random number table we
might label the first can 1, the second can 2, …, and the 10th can 0. Then the first five selected cans are 7,
17, 27, 37, and 47.
(iv) Note that 125/10 = 12.5 Æ 12 or 13 cans. However, there will not be a 7th can to select in last group. Thus
the total number of cans in the sample will be 12.
(d) (i) No.
(ii) Two possible values are 0.12 and 0.15.
(iii) Yes, a Type II error.
2.42
(a) (i) Convenience sampling.
(ii) Yes, a selection bias.
(iii) The calculated average is expected to be higher than the true average, as all of the books in the sample
have already been checked out at least once, and may include some of the more popular books.
(b) Cluster sampling.
(c) (i) Stratified random sampling.
(ii) Overall estimate: (40/1200)(20) + (200/1200)(15) + (600/1200)(10) = 14.2 times checked out.
(d) For each of the three categories of books the following stages are followed.
Stage 1: Divide the books into clusters according to the last digit of the call number (0 through 9). Take a
simple random sample of 3 digits from the list of 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9.
The clusters of books (in that category) with call numbers ending with those selected digits are selected.
Stage 2: Within each of the selected clusters of books from Stage 2, select a simple random sample of 7 books.
Note that with this multistage sampling plan, we will have a total of 3 categories x 3 clusters x 7 books = 63
books.
2.44
Answer is (c) statistic is to a sample.
2.45
The proportion 153/200 is a parameter since the instructor polled her entire class and that was the group that she
was interested in learning about.
2.46
(a) Population: Adults U.S. residents. The sample size n = 1500.
(b) Population: Today’s shipment of 1-gallon milk cartons. The sample size n = 5.
(c) Population: The 740 members of the local women’s business association. The sample size n = 100.
2.47
Nonresponse bias.
2.48
(a) Selection bias, where the sampling frame is either incomplete (such as sampling only one dealership) or
incorrect.
(b) Nonresponse bias, where persons who do not respond to a survey (such as the lazy car owners) may have different
opinions to those who do.
(c) Response bias, where respondents may have a tendency to lie (such as car dealerships that believe in lower
miles per gallon) or refuse to answer.
2.49
This survey may be subject to nonresponse bias. Only those alumni who respond and report their income will be
included. Alumni who perhaps are currently unemployed or in a low paying position may elect not to respond.
Therefore, the reported average income based on such a survey may be biased upwards -- the average may be larger
than the actual average for all alumni.
2.50
(a) The sampling design is a simple random sample of size 227 taken from the cocaine users who called.
(b) The population this sample is drawn from is cocaine users who called the National help line between February
and March.
(c) The sample is not from the population of workers, but only from those who called the hotline, so the survey results
do not generalize to the population of workers. Also, "more people" implies a comparison for which no data were
given.
2.51
(a) With the calculator the selected ID numbers are: 179, 2274, and 3327. With the random number table the
selected ID numbers are: 2398, 2258, and 3540.
(b) A statistic.
2.52
(a) True. If we fail to reject the null hypothesis, them the population consists of all males.
(b) False.
(c) True.
2.54
You first need to label the 4000 signatures. Hopefully no one signed the petition more than one time. If you use a
calculator or computer, the labels can simply be 1 to 4000. If you use the random number table, you could label the
signatures from 0001 to 4000. Using a seed value of say 29 with the calculator or row 120, column 11 of the
random number table, you can proceed to take a simple random sample of 400 signatures.
2.55
Stratified random sampling. You are dividing up your population by gender then selecting your sample within the
strata at random.
2.56
(a) Stratified random sampling.
(b) 75/416.
(c) (0.20)(15/25) + (0.80)(60/75) = 0.76.
(d) Statistic.
2.57
(a) All former university graduate students.
(b) Stratified random sampling.
(c) False.
2.58
(a) Stratified random sampling.
(b) Selection bias.
2.59
(a) Stratified random sampling
(b) The chance is 0, only 1 of the two fiction books (A, B) will be selected.
(c) Two books that could be selected are Book A and Book C.
(d) Another pair of books that could be selected is Book A and Book D.
(e) The chance that the total number of pages exceeds 800 is the same as the chance that the two selected books are
Book B and Book D. There are 4 possible pairs of books that could be selected of which the (B, D) pair is 1, so
the chance is 1/4 = 0.25.
2.60
(a) Stratified random sampling.
(b) High: 20 clients, Moderate: 125 clients, Low: 45 clients.
(c) With the calculator the selected clients are: 163, 2196, 214, 2462, 740.
With the random number table the selected clients are: 1887, 1209, 2294, 954, and 1869.
2.62
(a) Systematic sampling (1-in-30). Once you select the sample you can determine the number of freshmen you
selected. Also knowing the population list you can calculate the number of freshmen you would select for each
of the 30 possible samples (depending on the starting point.)
(b) Stratified random sampling. Yes, you would have 25 freshmen.
2.63
(a) (i) 0.02(500)+0.03(1200)+0.05(18000) = 10+36+900 = 946.
(ii) Stratified Random Sampling.
(iii) With the calculator the selected labels are 432, 232, 304, 412, and 372. With the random number table, we
might assign the first High category driver the labels 001 and 501. We would assign the second High
category driver 002 and 502. This assignment pattern would continue until the 500th High category driver
who would be assigned the labels 000 and 500. Reading off labels from row 60, column 1, we have: 789,
191, 947, 423, and 632. This would correspond to selecting the 289th, 191st, 447th, 423rd, and 132nd High
category drivers in the list of 500 High category drivers.
(b) (i) With the calculator or the random number table, the first selected label is 15. Thus the selected labels will
be 15, 35, 55, 75, 95, and so on.
(ii) Since the 500 High category drivers divide evenly into groups of 20 (500/20 = 25), there will be a total of
25 High category drivers in the systematic 1-in-20 sample.
2.64
(a) 907.
(b) Cluster sample.
(c) (iii) Selection bias.
2.65
These results are based on a study of 125 aerobic classes in five health clubs, not all aerobic classes in all health
clubs. Thus, the 60% figure is a statistic and the sample size is n = 125.
2.66
(a) It is a systematic 1-in-45 sampling resulting in 50 students (1 from each of the 50 sections, the 33rd in each of
the 50 lists of 45 students).
(b) It is a cluster sample and you cannot know the sample size (number of students selected) because we do not
know how many students are in the various major clusters.
2.67
(a) Cluster sampling.
(b) 1/5.
(c) Response bias.
2.68
(a) False.
(b) False.
(c) True.
2.69
(a) A 1-in-40 systematic sample.
(b) 1/40 is the chance that any specific address is chosen since one of the first 40 addresses is picked at random.
The other addresses are directly linked to that first random pick.
2.71
(a) Multistage, with the first stage being a cluster sample of 3 lab sections and the second stage being a simple
random sample of 25% of the students in the 3 selected labs.
(b) 0.0625.
(c) The 78% is a statistic since it was based on the sample of students surveyed.
2.72
Nonresponse bias is the distortion that can arise because a large number of units selected for the sample do not
respond or refuse to respond, and these nonresponders have a tendency to be different from the responders. So
nonresponse bias has to do with who responds. Response bias is the distortion that can arise because the wording of
a question and the behavior of the interviewer can affect the responses received. So response bias has to do with
how the responders answer.
2.73
(a) Since 15 patients were selected with a 1-in-9 systematic sample, there were at least 15 groups of 9 patients each
or 135 patients in all.
(b) (ii) Statistic.
2.74
(a) Since there is no prior information, we will sample in proportion to the size of the stratum relative to the size of
the population. So the sample size from Stratum I is 3 (since there are 10/40 large facilities in all, 25% of the
sample could be large facilities, for 25% of 12 = 3), and the sample size from Stratum II is 12 – 3 or 9.
(b) Using the calculator, the stratum I selected facilities and responses are: 7 (Yes), 10 (Yes), 8 (No), and the
stratum II selected facilities and responses are: 7 (No), 17 (No), 9 (Yes), 24 (No), 3 (No), 21 (No), 8 (Yes),
12 (No), and 1 (Yes),
Using the random number table we will label the stratum I facilities as: 1 has label 1, Facility 2 has label 2, ... ,
Facility 10 has label 0. Then the selected facilities and responses are: 2 (Yes), 6 (Yes), 4 (Yes). We will label
the stratum II facilities as: 1 has label 01; Facility 2 has label 02; up to Facility 30 has label 30. Then the selected
facilities and responses are: 05 (No), 36 (skip), 60 (skip), 42 (skip), 13 (Yes), 25 (Yes), 66 (skip), 92 (skip), 64
(skip), 22 (Yes), ..., 04 (Yes), 06 (No), 20 (No), 12 (No), 18 (Yes).
10 2 30 3
(c) Using the calculator we have: p̂ = estimate = ⎜ ⎟⎜ ⎟ + ⎜ ⎟⎜ ⎟ = 0.416 .
⎛ ⎞⎛ ⎞ ⎛ ⎞⎛ ⎞
⎝ 40 ⎠⎝ 3 ⎠ ⎝ 40 ⎠⎝ 9 ⎠
10 3 30 5
Using the table we have: p̂ = estimate = ⎜ ⎟⎜ ⎟ + ⎜ ⎟⎜ ⎟ = 0.667 .
⎛ ⎞⎛ ⎞ ⎛ ⎞⎛ ⎞
⎝ 40 ⎠⎝ 3⎠ ⎝ 40 ⎠⎝ 9 ⎠
(d) The true population proportion is p = 20/40 = 0.50.
(e) In general, estimates may not be exactly equal to p. The calculator estimate was slightly too small, while the
random number table estimate was too large.
In this chapter, the students learn about the various sampling methods by actually doing them -- sampling their
fellow classmates. A few of the Let’s Do It! exercises rely on group work. If in Chapter 1 you have made an
effort to have students work together at least in groups of 2 or 3, then forming larger groups for these sampling
exercises should come fairly easy. Some of the LDI exercises can take 15 to 20 minutes. Thus it is important to
watch the time. For our larger classrooms it is not easy to start a group project and have the exact same groups be
able to continue on it at the start of the next class. And if you do not allow for a few minutes to wrap up and recap,
the main ideas or features of the LDI exercise can be forgotten. If you spend the time on simple random sampling
so students have a good understanding of it, the other sampling methods follow more easily and quickly.
This is the first chapter which can make use of a calculator or computer. If you are using a TI-84 graphing
calculator, details on how to generate random integers with the TI are presented on page 102 and in the TI Quick Steps
appendix which follows the exercises at the end of this chapter (page 144). The steps are fairly straight forward
and no data entry is required, so students are not overwhelmed with TI details and options. They become familiar
and comfortable with the calculator gradually throughout this text. If you do just one sampling example using the
random number table with multiple labels and some labels unused, and then do the same example using the
calculator, students see the benefit of using a calculator or computer. The random number generating steps for other
TI graphing calculators are similar, and if the same seed value is used, the output should be the same.
29 Chapter 2
Producing Data 29
Let’s Do It (LDI) Solutions
Let’s Do It! 2.1 Parameter or Statistic?
According to the Campus Housing Fact Sheet at a Big-Ten University, 60% of the students living in campus
housing are in-state residents. In a sample of 200 students living in campus housing, 56.5% were found to be in-
state residents. Circle your answer.
(a) In this particular situation, the value of 60% is a (parameter, statistic).
(b) In this particular situation, the value of 56.6% is a (parameter, statistic).
LDI 2.1
How long? 2 minutes
How might it be done? Ask students to read through the scenario (or you read it aloud with the class), complete
the choices, and compare with a neighbor.
How important? We recommend you do this exercise and/or bring in other examples from recent news to
share with the class.
LDI 2.2
How long? 2-3 minutes
How might it be done? Ask students to read through the scenario and discuss possible answers with a neighbor.
How important? This is a nice, short exercise that reinforces the idea of a voluntary sample. Such call-in
polls are common enough, that you may even be on the look out for an actual example to
share with the class.
LDI 2.3
How long? 4-5 minutes
How might it be done? Ask students to read through the scenario and discuss possible answers with a neighbor.
How important? This exercise reinforces the ideas of population, parameter, and sampling the wrong unit.
30 Chapter 2
Producing Data 30
This is also an example of length-biased sampling.
31 Chapter 2
Producing Data 31
Let’s Do It! 2.4 A Simple Random Sample of Companies
An investment magazine publishes data on sales, profits, assets, dividends, shares, and earnings per share for the
nation’s 500 most valuable companies. You are to select a simple random sample of 10 companies from the list of
500 companies. Explain how you would label the companies and then use your calculator (with a seed value of
53) or the random number table (Row 26, Column 1, reading from left to right) to identify the labels of the 10
companies that would be selected from the list of 500 companies.
With the Table: Give each of the 500 companies 2 labels as follows:
Company 1: 001
Company 2: 002
etc ...
Company 499: 499
Company 500: 500
Starting at row 26, column 1, the labels of the 10 selected companies are: 815 (skip), 257, 229, 504 (skip),
839 (skip), 964 (skip), 232, 487, 882 (skip), 651 (skip), 665 (skip), 661 (skip), 477, 876 (skip), 797 (skip),
147, 801 (skip), 330, 087, 074, 796 (skip), 669 (skip), 572 (skip), 529 (skip), 676 (skip), 205.
LDI 2.4
How long? 5 minutes
How might it be done? Recap the scenario with the whole class. Most of the time we work through this
exercise together as the whole class and then have them work own their own in groups
for the next LDI 2.5 exercise. Alternatively if, you spent time going over the material
from pages 86 to 89, you might have students work with their neighboring classmate on
this exercise to practice taking a simple random sample.
How important? Not completely necessary to do in class, but it is a good basic exercise on taking a simple
random sample, to reinforce the basic steps. This exercise could be assigned as a
homework problem.
Steps:
1. In the space provided below, write the names of the people in your group.
Labels if using the random number table Labels if using a calculator
0 Susan 5 Matt 1 Susan 6 Matt
1 Martha 6 Linda 2 Martha 7 Linda
2 Peter 7 Kathy 3 Peter 8 Kathy
3 Brenda 8 Karl 4 Brenda 9 Karl
4 John 9 Albert 5 John 10 Albert
32 Chapter 2
Producing Data 32
3. Select your sample by selecting labels at random.
If you will be using a calculator, use a seed of 21 and your population size N = 10.
If you willisbethe
What using
firstthe random
label number
selected? = table, start at Row 13, Column
TI: 1.7, Table: 0
Who is the first person selected from your group? TI: Linda, Table: Susan
What is the second label selected? = TI: 10, Table: 9
Who is the second person selected from your group? TI: Albert, Table: Albert
What is the third label selected? = TI: 8, Table: 4
Who is the third person selected from your group? TI: Kathy, Table: John
Suppose we wish to learn about the proportion of women in your population, denoted by p .
Count the number of women in your population, COUNT = 5
Count the number of people in your population, N = 10
Compute the proportion of women in your population p = COUNT = 5/10 N
Next, let’s look at the results for your simple random sample of size n = 3. In many cases, the corresponding
symbol computed for the sample is the same as that for the population, but a hat “ ^” is written over the top, like this
p$ (read p-hat)
Count the number of people in your sample: n = 3
Count the number of women in your sample, count = TI: 2, Table: 1
Compute the proportion of women in your sample, p$ = count n = TI: 2/3, Table: 1/3
LDI 2.5
How long? 15-20 minutes
How might it be done? There are many ways to approach this exercise so it can be adaptable to fit your needs.
Since we have large lecture sections and this is the first sampling exercise, we often take
about 5 minutes to briefly go through a various steps with our own mock population of
10 people. You could prepare the transparency before class, having already filled in
names (step 1). With the whole class, discuss how you might label the units (step 2 -- if
you are only using the table, perhaps ask what if we had 12 in our population), select the
first person at random (step 3, using a DIFFERENT seed so they will have to do this
themselves in their groups later), and compute your population proportion. Then explain
that it is their turn -- to form groups of about 10 (this number may have to be different for
some groups) and complete the full exercise. If they have questions, they should first
ask their neighboring classmate, and then raise their hand. After most to all groups
have finished, gather the class back together to review the last part of step 3. You might
finish out your mock example to wrap this up, then go right into a discussion of the think
about it questions following LDI 2.5.
How important? Fairly important. Since simple random sampling is used within the remaining sampling
methods (random selection within a stratum, of a systematic starting point, of a cluster),
it is important that students understand how to do it. By having them do it themselves,
with their own population, it reinforces the basic steps.
33 Chapter 2
Producing Data 33
Let’s Do It! 2.6 Accounting Practices
Accountants often use stratified random sampling during audits to verify a company's records of such things as
accounts receivable. The stratification is based on the dollar amount of the item, and often includes 100%
sampling of the largest items. One company reports 5000 accounts receivable. Of these, 200 are in amounts over
$100,000, another 1000 are in amounts between $10,000 and $100,000, and the remaining 3800 are in amounts
under $10,000. Using these groups as strata, you decide to verify all of the largest accounts (over $100,000) and
to take a simple random sample of 5% of the midsize accounts ($10,000 to $100,000) and 1% of the small accounts
(under $10,000).
(a) Based on this sampling design, how many accounts will be sampled?
# of large accounts to be sampled = 100% of 200 200
# of midsize accounts to be sampled = 5% of 1000 50
# of small accounts to be sampled = 1% of 3800 38
------------------------------------------------------------------------------------
total # of accounts to be sampled 288
(b) Describe how you will label the accounts in the small stratum to select some small accounts to be included in
the sample. Use your calculator (seed value = 25) or the random number table (Row 18, Column 1) to select
only the first 5 small accounts to be verified. Think carefully about how many small accounts there are to
label.
TI: Label the 3800 small accounts from 1 to 3800. Using a seed value of 25 and N = 3800,
the first 5 selected small accounts have the following labels: 2429, 1511, 2410, 2777, 636.
Table: Label the 3800 small accounts from 0001 to 3800. Starting at row 18, column 1,
the first 5 selected small accounts have the following labels:
0101, 1540, 9233 (skip), 3629, 4904 (skip), 3127, 3041
LDI 2.6
How long? 5-7 minutes
How might it be done? Recap the scenario with the whole class. Have students work with their neighboring
classmate to practice setting up and starting to take a stratified random sample.
Sometimes we work through this exercise together with the whole class.
How important? Not completely necessary to do in class, but it is a good basic exercise on stratified
random sample, to reinforce the basic steps. Many students rush into it and make the
mistake of taking a sample of 5 small accounts from some list of 200 small accounts, instead
of sampling from a list of N = 3800 small accounts. This exercise could be assigned
as a homework problem.
The question of interest: How many times per year do you get a haircut?
We want to learn about the average number of haircuts per year for your population.
(Note: you may come up with a different question of interest.)
In the space provided below, write the names of the people in your group.
Name # haircuts per year Name # haircuts per year
Monica 3 Sarah 4
Jon 3 Emily 4
David 2 Mike 5
Mary 1 Steve 2
34 Chapter 2
Producing Data 34
Ask the question of interest to each member of your population and record their responses next to their name.
Compute the average response for your population. Add up all of the responses and then divide by the number of
students in your population, N.
You are able to take a sample of size n = 4. Take a simple random sample of size n = 4.
Steps:
2. Select a place to start in your random number table (row 10, column 22) and read off labels until 2 different
labels have been selected or use your calculator (seed value = 270) to select your sample of size 4 from your
population of N. Who did you select from your group and what are their responses?
TI: (seed of 270): => the selected sample is label = 3 David with response 2
label = 1 Monica with response 3
label = 7 Mike with response 5
label = 6 Emily with response 4
3. Compute the average response for your simple random sample of size n = 4 --
add up the above 4 responses and divide by 4.
sum sum
Table: Average = n = 12/4 = 3 TI: Average = n = 14/4 = 3.5
This number is a ( parameter , statistic )
Now you are able to take a sample of size 4, but you want to have 2 females and 2 males in your sample. How?
STRATIFY!
Steps:
1. In the space provided below, write a list of all the males and all of the females in your group, that is,
form the strata. Also include their response next to their name in parentheses, for example, Mary (2).
Label FEMALES (Stratum 1) Label MALES (Stratum 2)
1 Monica (3) 1 Jon (3)
2 Mary (1) 2 David (2)
3 Sarah (4) 3 Mike (5)
4 Emily (4) 4 Steve (2)
35 Chapter 2
Producing Data 35
2. Assign a label to each unit in each stratum. Note that you can start with the same label for each stratum.
For example, if there were four females and four males in your group, the females could be labeled 1 through 4
and the males could be labeled 1 through 4.
3. Select a simple random sample of size n = 2 females (start at row 14, column 1 or use your calculator with a
seed value = 24) and a simple random sample of size n = 2 males (start at row 23, column 20 or use your
calculator with a seed value = 35). Record the selected responses below.
Table: Monica and Sarah are selected for ... responses: 3, 4
Jon and Mike are selected for ... responses: 3, 5
TI: Emily and Sarah are selected for ... responses: 4, 4
Jon and Steve are selected for ... responses: 3, 2
⎞ ⎛⎞ ⎛4⎞ ⎛4⎞
⎜⎛ 4 4
Table: ⎟( 3.5) + ⎜ ⎟( 4 ) = 3.75 TI: ⎜ ⎟( 4) + ⎜ ⎟( 2.5) = 3.25
⎝8⎠ ⎝8⎠ ⎝8⎠ ⎝8⎠
LDI 2.7
How long? 15-20 minutes
How might it be done? As with LDI 2.5, there are many ways to approach this exercise. You could take a few
minutes to briefly go through a various steps with our own mock population of 8 (or actually
any number you wish) people. If you decide to increase the population size, you might
also increase the sample size from within each stratum. You could discuss some or all
of the steps with your mock population. We have found that if you have spent enough
time teaching the details about taking a simple random sample, students do not have
much difficulty performing the remaining sampling methods. After a brief introduction to
the idea of stratified random sampling and doing LDI 2.6, we have gone right into asking
students to form groups and complete the full exercise. You might need to spend some of
36 Chapter 2
Producing Data 36
the time wrapping up this class exercise explaining the weighted average concept (also
presented in Example 2.16 which precedes LDI 2.7).
How important? Fairly important. By having them do it themselves, with their own population, it reinforces
the basic steps. They see that the labeling is assigned separately within each stratum.
They experience computing a weighted average.
37 Chapter 2
Producing Data 37
Let’s Do It! 2.8 Faculty Salaries
A study is being conducted of the faculty salaries for a public university. Of the 1200 tenure-track faculty, 480 are full
professors, 336 are associate professors and the remaining are assistant professors. A simple random sample of 100 full
professors, 50 associate professors and 50 assistant professors will be taken and information about salary will be
obtained.
(a) What type of sampling method is used to obtain the 200 selected faculty members? Stratified random sampling.
(b) Use your calculator (with a seed value of 18) or the random number table (Row 14, Column 1) to give the labels for
the first 5 full professors to be selected.
TI: 202, 195, 28, 215, and 174.
Table: 103, 112, 267, 339, and 401.
(c) The table summarizes the average salary for the sampled faculty members by rank.
Give the overall estimate of the average salary for all faculty members based on this sample information.
Show all work and include your units.
(480/1200)($95,000) + (336/1200)($70,000) + (384/1200)($55,000) = $75,200.
LDI 2.8
How long? 5 - 7 minutes
How might it be done? Recap the scenario with the whole class. Go through part (a) as a group. It should be
somewhat obvious since the exercise is in the Stratified Random Sampling section.
You might have students work on parts (b) and (c) in groups of size 2 or 3. When
polling the class for their answers to part (c), see if they included the units.
How important? Not completely necessary to do in class, but it is a good basic exercise on stratified
random sample, to reinforce the basic steps. It gives students another opportunity to
apply the idea of a weighted average to obtain an overall estimate.
In the map, the ID numbers start at 101, which is along the right side, and end at 514 in the upper left corner. Notice
that some of the numbers in between are skipped. For example, the numbers jump from 414, located in the middle
of the bottom edge, to 501 for Sinai Hospital. Interviewers can be assigned to a set of blocks, such as the blocks
with ID numbers in the 100s. The ID numbers are also assigned in a serpentine pattern so that the blocks that form a
set are close to one another.
38 Chapter 2
Producing Data 38
Your task is as follows: Take a 1-in-10 systematic sample of the blocks in this tract.
(a) Following the ID numbers for ordering the blocks in this population, use your pen to continue to trace out this
order on the above map (that is, connect the blocks with a line in ascending order of ID number).
(b) For a 1-in-10 systematic sample the first ten blocks in the map form your first group. Label these 10 blocks
and
use your calculator (seed value = 39 and N = 10) or the random number table (row 16, column 6) to randomly
select your starting block, which is the first block in your sample.
What is the ID number of the first block selected?
TI: Labels are #101=1, #102=2, #103=3, #104=4, #105=5,
#106=6, #107=7, #108=8, #109=9, #111=10
Using a seed value of 39 and N = 10, the first selected label is 1, which is block #101.
Table: Labels are #101=1, #102=2, #103=3, #104=4, #105=5,
#106 =6, #107=7, #108=8, #109=9, #111=0
Using a row 16, column 6, the first selected label is 1, which is block #101.
(c) Now starting at the first block selected, count off every tenth block, in order, to be included in your sample.
List the block ID numbers which form your sample: #101, #202, #212, #310, #407, #503,
#513
LDI 2.9
How long? 8-10 minutes
How might it be done? Recap the scenario with the whole class. Have students work in groups of size 2 or 3.
You might do the tracing to form the list up front, so that everyone starts off with the
correct list.
How important? This is a nice, real-data exercise to do in class. Since the ID numbers are in order but
skip some numbers, the students do have to ‘count’ out every tenth block to be included
in the sample. This exercise could be assigned as a homework problem.
39 Chapter 2
Producing Data 39
The meeting organizers would like to survey the programmers to learn of their impressions regarding the annual
meeting. It was decided to take a 1-in-50 systematic sample of the programmers using the ID number.
40 Chapter 2
Producing Data 40
(a) With this type of sampling plan, is each presentation guaranteed to be represented? That is, will the sample
include at least one programmer from each presentation?
Circle one: Yes No
Explain: Some presentations had less than 50 registered.
If yes, use your calculator with seed value = 18 (or Row 33, Column 1) and give the ID numbers for the
programmers that will be selected. If no, give the maximum value for k such that the 1-in-k systematic sample
will always include at least one programmer from each presentation.
The smallest number of programmers registered is 9 for presentation #13, so the maximum value of k is
9.
(b) With the correct k from part (a), use your calculator with seed value = 18 (or Row 33, Column 1) and give the
first 10 ID numbers for the programmers that will be selected.
Using a k = 9 and the TI, we have 4, 13, 22, 31, 40, 49, 58, 67, 76, 85.
Using a k = 9 and the Table, we have 6, 15, 24,33, 42, 51, 60, 69, 78, 87.
(c) How many programmers will be included in your systematic sample? We have that 1268/9 = 140 with a
remainder of 8. With either the TI or the Table we will sample one more from the last 8 attendees, so
the total sample size will be 141. If the first selected digit were a 9, we would have just 140.
(d) Suppose the organizers would like to survey exactly two programmers from each presentation. Suggest a
sampling plan to accomplish this.
They should do a stratified random sample with the 20 presentations representing 20 strata.
LDI 2.10
How long? 10-15 minutes
How might it be done? Talk through the scenario with the whole class. Have students work in groups of size 2
or 3 to discuss part (a). You might gather the class back together to discuss and arrive
at
k =9, so that everyone can do the remaining parts with the same value for k.
How important? Very Important. Students have to really “think” to complete this exercise. It should
generate some good discussion. You could extend part (d) as a homework exercise by
having students carry out their suggested plan, showing all details. We have also
assigned this as a homework question and given them some time in their weekly lab to
discuss it.
The blocks in the tract have been grouped to form clusters of blocks, corresponding to the Census Bureau
identifying numbers.
41 Chapter 2
Producing Data 41
Exploring the Variety of Random
Documents with Different Content
details of a like kind, all helping to a horrible wildness of appearance
to one who viewed, as I did, from an eminence, the crazy, fire-
blackened, dismasted old basket, that wallowed as though every
head of swell that rolled at her must overwhelm and drown her
hollow interior.
I again sent my eyes in another passionate search, then descended.
As I sprang from the shrouds on to the deck, my eye was taken by
the brig’s bell, that dangled from a frame close against the foremast.
Dreading lest some increase in the swell should start it off into
ringing in some dismal hour of gloom and heighten Miss Temple’s
misery and terror, I unhooked the tongue of it, and threw it down,
and rejoined my companion, whose white face put the piteous
question of her heart to me in silence.
‘No,’ said I, swaying in front of her as I held on to the door; ‘there is
nothing to be seen.’
‘Oh it is hard! it is hard!’ she cried. ‘If one could only recall a few
hours—be able to go back to yesterday! I do not fear death: but to
die thus—to drown in that dreadful sea—no one to be able to tell
how I perished.’ She sobbed, but with dry eyes.
There was no reasoning with such a fit of despair as this, nor was it
possible for me to say anything out of which she might extract a
grain of comfort, seeing that I could but speak conjecturally, and
with no other perception than was to be shaped by the faint light of
my own hopes. My heart was deeply moved by her misery. Her
beauty showed wan, and was inexpressibly appealing with its air of
misery. The effects of the long and fearful vigils of the night that was
gone were cruelly visible in her. There was a violet shadow under
her eyes, her lips were pale, her lids drooped, her hair hung in some
little disorder about her brow and ears; her very dress seemed
significant of shipwreck, mocking the eye with what the grim usage
of the sea had already transformed into mere ironical finery. Yet
there was too much of the nature she had familiarised me to on
board the Indiaman still expressed in the natural haughty set of her
lips, even charged as they were with the anguish that worked in her,
to win me to any attempt of tender reassurance. I watched her
dumbly, though my soul was melted into pity. Presently she looked
at me.
‘I suppose there is nothing to be done, Mr. Dugdale?’
‘Indeed, then,’ said I, ‘there is a deal to be done. First of all, you
must cheer up your heart, which you will find easy if you can credit
me when I tell you that this hull is perfectly buoyant; that though
the weather is thick and gloomy, the sun, as he gains power, is
certain to open out the ocean to us; that there are two ships close at
hand searching for us; that there are provisions enough below to
enable us to support life for days and perhaps weeks; and that, even
if the Indiaman or the corvette fail to fall in with us, we are sure to
be sighted by one of the numerous vessels which are daily traversing
this great ocean highway. What, then, are we to do but compose our
minds, exert our patience, keep a bright lookout, be provided with
means for signalling our distress, and meanwhile not to suffer our
unfortunate condition to starve us? And that reminds me to overhaul
the pantry for something better than biscuit to break our fast with.’
A softness I should have thought impossible to the spirited fires of
her eyes when all was well with her entered her gaze for a moment
as it rested upon me, and a faint smile flickered upon and vanished
off her lips; but she did not speak, and I dropped through the hatch
to ascertain if the pantry could yield us something more nourishing
than ship’s bread.
The sullenness of the day without lay in gloom below. I was forced
to return for a candle, with which I entered the little cabin that I had
visited on the previous day; but when I came to make a search I
could find nothing more to eat than cheese, biscuit, and marmalade.
There was a number of raw hams, but the galley was gone, and
there was no means to cook them. There were two casks of flour, a
sack of some kind of dried beans, and a small barrel of moist sugar.
These matters had probably been overlooked when the crew
hurriedly removed themselves from the brig. No doubt, at the time
of jettisoning such commodities as the hold might have stored, they
had broken out as much food and water as they could take with
them. There was more than a bottle of wine in the deck-house;
down here, stowed away in straw and secured by a batten, were
some three or four scores of full bottles, all, I supposed, holding the
same generous liquor contained in the first of them we had tasted.
But there was no fresh water. I sought with diligence, but to no
purpose. Possibly the people might have left some casks of it in the
hold; but that was a search I would not at present undertake.
I took some cheese and marmalade and another handful of biscuits,
along with a knife and a couple of tin dishes. As I passed through
the cabin, the light of the candle I held glanced upon a stand of
small-arms fixed just abaft the short flight of the hatch-ladder. There
were some thirty to forty muskets of an old-fashioned make, even
for those days, and on either hand of them, swinging in tiers or rows
from nails or hooks in the bulkhead, were a quantity of cutlasses,
half-pikes, tomahawks, and other items of the grim machinery of
murder. I placed the food upon the deck-house table.
‘A shabby repast, Miss Temple,’ said I, ‘but we may easily support life
on such fare until we are rescued.’
She ate some biscuit and marmalade, and drank a little wine; but
she incessantly sent her gaze through the windows or the open door,
and sighed frequently in tremulous respirations, and sometimes
there would enter a singular look of bewilderment into the
expression of her eyes, as though her mind at such moments failed
her, and did but imperfectly understand our situation. I would then
fear that the horror which possessed her might end in breaking
down her spirits, and even dement her, indeed. Already her eyes
were languid with grief and want of rest, and such strength and life
as they still possessed seemed weakened yet by the shadowing of
the long fringes. I endeavoured to win her away from her thoughts
by talking to her.
I possessed a pocket-book, which supplied me with pencil and paper,
and I drew a diagram of the two ships’ and the wreck’s position, as I
was best able to conceive it, and made arrows to figure the direction
of the wind, and marked distances in figures, and enlarged freely
and heartily upon our prospects, pointing with my pencil to the
paper whilst I talked. This interested her. She came round to the
locker on which I sat, and placed herself beside me, and leaned her
face near to mine, supporting her head by her elbow whilst she
gazed with eyes riveted to the paper, listening thirstily. I had never
had her so close to me before saving that day when we swung
together on to the hencoop, but then it was a constrained situation,
and she had let me suspect that it was very distasteful to her. It was
far otherwise now. She was near me of her own will; I felt her warm
breath on my cheek; the subtle fragrance of her presence was in the
air I respired. I talked eagerly to conceal the emotions she excited,
and I felt the blood hot in my face when I had made an end with my
diagram, and drew a little away to restore the book to my pocket.
She now seemed able and willing to converse, but she did not offer
to leave my side.
‘Suppose the ships are unable to find us, Mr. Dugdale?’
‘Some other vessel is certain to fall in with us.’
‘But she may be bound to a part of the world very remote from India
or England.’
‘True,’ said I; ‘but as she jogs along she may encounter a vessel
proceeding to England, into which we shall be easily able to tranship
ourselves.’
‘How tedious! We may have to wander for months about the ocean!’
‘It is always step by step, Miss Temple, in this life. Let us begin at
the beginning, and quit this wreck, at any rate.’
‘All my luggage is in the Indiaman. How I am to manage I cannot
conceive,’ said she, running her eyes over her dress, and lifting her
hand to her hat.
‘Pray let no such consideration as dress trouble you. The experience
will gain in romance from our necessities, and we shall be able to
read “Robinson Crusoe” with new enjoyment.’
She faintly smiled, with just a hint of peevishness in the curl of her
lip.
‘If this be romance, Mr. Dugdale, may my days henceforth, if God be
merciful enough to preserve us, be steeped in the dullest prose.’
‘I wonder where Colledge and the cutter’s crew are?’ said I.
‘I do not think,’ she exclaimed, ‘if Mr. Colledge were in your place he
would show your spirit.’
‘He was a great favourite of yours, Miss Temple.’
‘Not great. I rather liked him. I knew some of his connections. He
was an amiable person. I did not know that he was engaged to be
married.’
I was astonished that she should have said this, but I was eager to
encourage her to talk, and in our state of misery it would signify but
little what topic we lighted upon.
‘Did he inform you he was engaged?’ said I.
‘No. I perceived it in his looks when his cousin asked him the
question. Did he ever tell you who the young lady was?’ she added
listlessly, and though she spoke of the thing it was easy to see that
she was without interest in it.
I could not tell a lie, and silence would have been injurious to my
wishes for her. Besides, she had guessed the truth by no help from
me, and then, again, our situation rendered the subject exquisitely
trifling and insignificant.
‘Yes,’ I replied; ‘we were cabin fellows, and intimate. He showed me
the girl’s portrait—a plump, pretty little woman. Her name is Fanny
Crawley, daughter of one of the numberless Sir Johns or Sir
Thomases of this age.’
She was looking through the cabin door at the sea, and scarcely
seemed to hear or to heed me. Am I strictly honourable in this?
thought I. Pshaw! it was no moment to consider the rights and
wrongs of such a thing. Her discovery had freed me from all
obligation of secrecy, and what I had supplied she would have easily
been able to ascertain for herself on her return home, if, indeed,
home was ever to be viewed again by either of us.
‘What horrible weather!’ she exclaimed, bringing her eyes to my
face; ‘there is no wind, and the sea rolls like liquid lead. When you
were at sea, were you ever in a situation of danger such as this?’
‘This is an uneasy time,’ said I; ‘but do not call it a situation of
danger yet. I am going shortly to overhaul the wreck. I must keep
her afloat until we are taken off her.’
‘How long were you at sea, Mr. Dugdale?’
‘Two years.’
‘Is your father a sailor?’
‘No; my father is dead. He was captain in the 38th Regiment of Foot,
and was killed at Burmah.’
There was a kind of dawning of interest in her eyes, an expression I
had not noticed when she talked of Colledge and his engagement.
‘My father was in the army, too,’ said she; ‘but he saw very little
service. Is your mother living?’
‘She is.’
She sighed bitterly, and hid her face whilst she exclaimed:
‘Oh, my poor mother! my poor mother! How little she knows! And
she was so reluctant to let me leave her.’ She sighed again deeply,
and let her hands fall, and then sank into silence.
I quitted the deck-house to take another look round. Just then rain
began to fall, and the sea became shrouded with the discharge. So
oil smooth now was the swell that each drop as it fell pitted the
lead-coloured rounds with a black point, and the water alongside
looked to be spotted with ink. As I had met with no fresh water in
the little room that I call the pantry, and as there might be none in
the hold, or none that with my single pair of hands I should be able
to come at, I resolved to take advantage of the wet that was
pouring down, and dived into the cabin to search for any vessel that
would catch and hold it. The flour and sugar casks in the pantry
would not do. I peered into the other berths, but could see nothing
to answer the purpose. It was of the first consequence, however, to
us that we should possess a store of drinking water to mix with our
wine, for we were in the tropics; the atmosphere was heavy with
heat, even under a shrouded heaven; it was easy to figure what the
temperature would rise to when the sun should shine forth; and the
mere fancy of days of stagnation and of vertical suns, of this hull
roasting; under the central broiling eye, of the breathless sea,
stretching in feverish breathings into the dim, blue distance,
unbroken by any tip of sail, and no fresh water to drink, was horribly
oppressive, and rendered me half crazy to find some contrivance to
catch the rain, which might at any moment cease. The thought of
the lockers in the deck-house occurred to me. I mounted the ladder
and searched them, and to my unspeakable joy, found in the locker
upon which Miss Temple had been seated during the night, four
canvas buckets, apparently brand new, as I might judge, from the
cloth and from the rope handles. The rain fell heavily, and the water
gushed in streams from the roof of the deck-house at many points of
it. In a very short time the buckets were filled, but they were of a
permeable substance, and it was necessary to decant them as soon
as possible. There was no difficulty in doing this, for there were
several empty bottles in the shelves below along with a couple of
large jars, some tin pannikins, and so forth. These I brought up,
washed them in the rain, and then filled them, and in this manner
contrived to store away a good number of gallons, not to mention
the contents of the buckets, which I left hanging outside to fill up
afresh, meaning to use them first, and taking my chance of loss
through the water soaking through them.
All this, that is to be described in a few lines of writing, signified a
lengthy occupation, that broke well into the day. Miss Temple
watched my labours with interest, and begged to be of service; but
she could be of little use to me, nor would I suffer her to expose
herself to the wet.
‘Will not this rain fill the hull,’ she exclaimed, ‘and sink her?’
‘It would need to keep on raining for a long while to do that,’ said I,
laughing. ‘I am going below to inspect the forepart of her, and to
ascertain, if possible, what her hold contains. Will you accompany
me? The hull rolls steadily; you will not find walking inconvenient,
and it is very necessary that you should occupy your mind.’
‘I should like to do so,’ she answered; ‘but ought not one of us to
stay here in case the sea should clear and show us the ships?’
‘Alas!’ said I, ‘there is no wind, and the ships probably lie as
motionless as we. This weather will not speedily clear, I believe. We
shall not be long below, and any sort of exertion is better than
sitting here in loneliness and musing upon the inevitable, and adding
the misery of thought to the distress of our situation.’
‘Yes, you are right,’ she exclaimed, rising. ‘You give me some heart,
Mr. Dugdale, yet I do not know why. There is nothing that you can
say to encourage me to hope.’
To this I made no reply, but took her hand, and assisted her to
descend the ladder. She came to a stand at the foot of it, as though
terrified by the gloom.
‘It is dreadful,’ she exclaimed in a low voice, ‘to think that only a few
short hours ago the poor lieutenant whose heart was beating high
with thoughts of returning home, should have been laughing and
joking—here! I can hear his voice still; I can hear Mr. Colledge’s
laughter. Hark! What noises are those?’
‘Rats!’ I exclaimed.
The squeaking was shrill and fierce and close to. I lighted a candle,
she meanwhile coming to my side, her elbow rubbing mine, as
though she would have my hand within an instant’s reach of her
own. The squeaking continued. It sounded as though there were
some score of rats worrying something, or fighting among
themselves.
‘Hold this candle for a moment,’ said I, and I advanced to the
bulkhead and grasped a cutlass, and then peeped into the little
passage that divided the after cabins. The rats were somewhere
along it, but it was too dark to see; so laying the cutlass aside, I
took down a musket and sent the heavy weapon javelin-fashion
sheer into the thick of the hideous noise. A huge rat as big as a
kitten rushed over my feet; Miss Temple uttered a shriek, and let fall
the candle.
‘Do not be alarmed!’ I shouted; ‘the beasts know their way below;’
and seeing the pallid outline of the candle upon the deck I picked it
up and relighted it.
‘Oh, Mr. Dugdale,’ she cried, in a voice that trembled with disgust
and fear, ‘what am I to do? I dare not be here, and I dare not be
above, alone. What is more shocking and terrifying than a rat?’
I told her that rats were much more afraid of us than we could
possibly be of them; but, commiserating her alarm, I offered to
escort her to the deck-house.
‘But you will not leave me there,’ she exclaimed.
‘It is very necessary,’ said I, ‘that I should examine the state of the
hull.’
‘Then I will stay with you,’ said she. ‘I cannot endure to be alone.’
She gathered up her dress, holding the folds of it with one hand,
whilst she passed the other through my arm. I could feel her
shuddering as she clung to me. Her eyes were large with fright and
aversion, and they sparkled to the candle-flame as she rolled them
over the deck. At the extremity of the passage that separated the
foremost berths from the pantry stood what I believed a bulkhead;
but on bringing the candle to it I discovered that it was a door of
very heavy scantling that slided in grooves with a stout iron handle
for pulling it by. It travelled very easily, as something that had been
repeatedly used. The moment it was open there was plenty of
daylight, for the open square of the main hatch yawned close by
overhead, of dimensions considerable enough to illuminate every
part of this interior. I stood viewing with wonder a scene of
extraordinary confusion. There were no hammocks, but all about the
decks, in higgledly-piggledly heaps and clusters, were mats of some
sort of West Indian reeds, rugs and blankets, bolster-shaped bags, a
few sea-chests, most of them capsized, with their lids open, and a
surprising intermixture of hook-pots, tin-dishes, sea-boots, oilskins,
empty broken cases, staves of casks, tackles, and a raffle of gear
and other things of which my mind does not preserve the
recollection. Several large rats, on my swinging the door along its
grooves, darted from out of the various heaps and shot with
incredible velocity down through the large hatch that conducted into
the hold, and that lay on a line with the hatch above.
‘By all that’s—— Well, well! here’s been excitement, surely,’ said I.
‘Was ever panical terror more incomparably suggested? But this brig
was full of men, and there was manifestly a tremendous scramble at
the last. Would not anyone think that there had been a fierce fight
down here?’
‘Do you think there are any dead bodies under those things?’
exclaimed Miss Temple in a hollow whisper.
‘See!’ cried I; ‘lest there should be more rats about, suppose I
contrive some advantage for you over the beasts;’ and so saying I
dragged one of the largest of the sea-chests to the bulkhead and
helped her to get upon it.
This seemed to make her easier. Filled as my mind was with
conflicting emotions excited by the extraordinary scene of hurry and
disorder which I surveyed, I could yet find leisure to glance at and
deeply admire her fine, commanding figure, as she stood with
inimitable, unconscious grace, swaying upon the chest to the regular
rolling of the hull. It was a picture of a sort to live as long as the
memory lasted. There she stood, draped in the elegancies of her
white apparel, her full, dark eyes large and vital again in the shadow
of her rich hat, under which her face showed colourless and faultless
in lineament as some incomparable achievement of the sculptor’s
art: her beauty and dignity heightened in a manner not to be
expressed or explained by the character of the scene round about—
the uncovered square of hatch through which the rain was falling,
the wild disorder of the deck, the rude beams and coarse sides of
the interior.
I approached the edge of the hatchway and looked down. Little
more was to be seen than ballast, on the top of which lay a couple
of dismounted guns, apparently twelve-pounders. A short distance
forward in the gloom were the outlines of some casks and cases.
The hull was dry, as the lieutenant had said. Water there
undoubtedly must have been, washing to and fro under the ballast
and down in the run, but too inconsiderable in quantity to give me
the least uneasiness. One glance below sufficed to assure me that
the fabric of the wreck was tight.
I considered a little whether it might not be possible to so protect
the yawning hatches as to provide against any violent inroads of
water should this dirty shadow of weather that overhung the wreck
in wet end in wind; but there were no tarpaulins to be seen, no
spare planks or anything of a like kind which could be converted into
a cover, nothing but mats and rugs, which were not to be put to any
sort of use in the direction I had in my mind.
I left Miss Temple standing on the chest, darting alarmed glances at
the huddled heaps which littered the decks, and walked forward to a
doorway in a stout partition that bulkheaded off a short space of
forecastle from these ’tweendecks. There was an open forescuttle
here that made plenty of light. This was the interior that had been
burnt out, as the lieutenant had told me, to the condition of a
charred shell. The deck and sides were as black as a hat, and the
place showed as if it had been constructed of charcoal. A strong
smell as of fire still lingered. Whatever had been here in the shape
of sea-furniture was burnt, or removed by the people. I picked up a
small handspike, and entering the cindery apartment, beat here and
there against the semi-calcined planks, almost expecting to find the
handspike shoot through; but black as the timber looked it yielded a
hearty echo to my thumps, and I returned to Miss Temple satisfied
that the hull was still very staunch, and, but for her uncovered
hatches, as seaworthy as ever she had been at any time since her
launch.
Whilst turning over some of the mats and wearing apparel on the
deck with my foot I spied a large cube of something yellow, and,
picking it up and examining it, I was very happy to discover that it
was tobacco. I made more of this than had I found a purse of a
hundred guineas, for, though I had my pipe in my pocket, I was
without anything to smoke, and I cannot express how hungrily
during the night I had yearned for the exceeding solace of a few
whiffs, and with what melancholy I had viewed the prospect of
having to wait until we were rescued before I should obtain a cigar
or a pipe of tobacco.
‘What have you there, Mr. Dugdale?’ cried Miss Temple.
‘A little matter that, coming on top of the discovery that this hull is
as good as a cork under our feet, helps very greatly towards
reestablishing my peace of mind—a lump of very beautiful tobacco,’
and I smelt it fondly again.
‘Oh, Mr. Dugdale, I thought it was a dead rat,’ she exclaimed. ‘What
are all those mats?’
‘The privateersmen used them to sleep on, I expect. The quantity of
them tells us how heavily manned this old waggon went.’
‘There is no wind, Mr. Dugdale. The rain falls in perfectly straight
lines. Let us return to the deck-house.’
I took her hand and helped her to dismount. She gathered her dress
about her as before, and passed with trepidation through the
darksome cabin, holding tightly by my arm, and then, with a wearied
despairful air, seated herself upon a locker and leaned her chin in
her hand, biting her under lip whilst she gazed vacantly through the
little window at the sullen raining gloom of the sky.
CHAPTER XXI
WE SIGHT A SAIL