Chapter 5 No Because The Sampled Population Consists The

Chapter 5

5.1 In an observational study, there is no attempt to control factors that might influence the

5.2a The study is observational. The statistics practitioner did not randomly assign stores to buy

cans or bottles.

b Randomly assign some stores to receive only cans and others to receive only bottles.

5.3 Randomly sample smokers and nonsmokers and compute the proportion of each group that has

5.4a A survey can be conducted by means of a personal interview, a telephone interview, or a self–

administered questionnaire.

b A personal interview has a high response rate relative to other survey methods, but is expensive

5.5 Five important points to consider when designing a questionnaire are as follows:

(1) The questionnaire should be short.

5.6a The sampled population will exclude those who avoid large department stores in favor or

smaller shops, as well as those who consider their time too valuable to spend participating in a

5.7a The Literary Digest was a popular magazine in the 1920s and 1930s which had correctly

predicted the outcome of many presidential elections. To help predict the outcome of the 1936

presidential election, the Literary Digest mailed sample ballots to 10 million prospective voters.

5.8a A self-selected sample is a sample formed primarily on the basis of voluntary inclusion, with

little control by the designer of the survey.

5.9 We should ignore the results because this is an example of a self–selected sample.

5.11 We used Excel to generate 40 three-digit random numbers. Because we will ignore all

randomly generated numbers over 800, we can expect to ignore about 20% (or about 8 to 10) of

the randomly generated numbers. We will also ignore any duplication. We therefore chose to

those to be ignored.

5.12 We used Excel to generate 30 six–digit random numbers. Because we will ignore any

duplicate numbers generated, we generated 30 six-digit random numbers and will use the first 20

unique

5.13 Stratified random sampling is recommended. The strata are the school of business, the faculty

5.14 A stratified random sampling plan accomplishes the president’s goals. The strata are the four

5.15 The operations manager can select stratified random samples where the strata are the four

departments. Simple random sampling can be conducted in each department.

5.16 Use cluster sampling, letting each city block represent a cluster.

5.17a Sampling error refers to an inaccuracy in a statement about a population that arises because

the statement is based only on sample data. We expect this type of error to occur because we are

5.18 Three types of nonsampling errors:

(1) Error due to incorrect responses

5.19 Yes. A census will likely contain significantly more nonsampling errors than a carefully

conducted sample survey.