Test bank

Chapter 15: Surveys and sampling

1. Even when the distribution of the variable in the population is not normal, the sampling distribution becomes more and more similar to the normal distribution as:

a. the sample size increases.

b. the sampling technique is changed.

c. the sample size decreases.

d. the standard deviation increases.

2. The idea behind sampling is to:

a. obtain a sample of manageable size.

b. make inferences from a sample to a population.

c. understand the properties of specific sample drawn.

d. All of these.

3. Sampling distribution tells us:

a. the probability of obtaining a particular value of a statistic in one sample.

b. the distribution of values on a particular variable in a particular sample.

c. the area distribution of the sample in a particular city.

d. None of these.

4. The difference in the precision of an estimate produced by a complex design and that by a simple random sample is referred to as the design effect.

a. True

b. False

5. A researcher finds that his data are normally distributed around a mean of 168. Is he right in assuming that 95% of his sample cases will be within the range 148.4 to 187.6?

a. Yes

b. No

c. Depends on the nature of the experiment

d. Can’t say. Need more information

6. The precision of an estimate is affected by three factors:

a. population variance, sampling technique, and sample standard deviation.

b. sample size, population variance, and location of participants.

c. population variance, sampling technique, and sample size.

d. population mean, population standard deviation, and population variance.

7. A deft value of 3 indicates:

a. that the standard error is three times as large as it would have been had the design had been a simple random sample.

b. that the standard error is 1.5 as large as it would have been had the design been a simple random sample.

c. that the standard error is one-third times as it would have been had the design been a simple random sample.

d. None of these.

8. Quota sampling divides the sample up into separate sub-groups and then selects random samples from within each group.

a. True

b. False

9. Unit non-response is when:

a. only one of the respondents failed to complete the survey.

b. no information was obtained from one or more of the sample units.

c. one of the items in the survey was not completed.

d. All of these.

10. Which of the following statements is true?

a. Standard deviation is a measure of dispersion from the mean.

b. Design effects are unaffected by the cluster homogeneity.

c. CATI surveys are longer than personal interview surveys.

d. None of these.

11. A researcher wishes to study the correlation of divorce and IQ levels of emerging adults. She randomly samples students from Oxford University where she teaches. What could be a possible problem with her sample?

a. Oxford University students may not have the time to complete the survey.

b. The sample is not representative of the population of emerging adults in terms of IQ, learning ability, social status, etc.

c. Her sample may be too small since the parents of Oxford University students do not get divorced.

d. There is no problem with her sample.

12. Structured questionnaires are the most common form of data collection for surveys.

a. True

b. False

13. The response rate is calculated by dividing the number of achieved interviews by the eligible issued sample size.

a. True

b. False

14. A researcher should be particularly worried about non-response bias when:

a. there is non-response to the survey.

b. the sample is very large.

c. the sample is very small.

d. the design effect is large.

15. In a Quota sampling of 100 individuals, if we know that there are 80% computer-users in our population, we would set quotas of:

a. 80 computer-users and 20 non-users.

b. 20 computer-users and 80 non-users.

c. 50 computer-users and 50 non-users.

d. None of these.

16. Web-surveys may not always be the best option because:

a. there is still a significant minority of the population who do not have access to the Internet.

b. there is no list of Internet users that can be used to draw random samples.

c. the quality of data obtained can be lower in the absence of an interviewer.

d. All of these.

17. If the standard deviation of a sample of 1000 individuals is 10, then the standard error of the mean is:

a. 10.

b. .01.

c. 100.

d. None of these.

18. A simple random sample (SRS) gives every unit in the population an equal probability of selection.

a. True

b. False

19. If the sample mean of a continuous variable is 120 and the standard deviation is 5, where would 99% of the data fall? (Assume normal distribution)

a. 110.2 – 129.8

b. 107.1 – 132.9

c. 148.4 – 187.6

d. Cannot say.

20. If rho= .10 and n= 120. Calculate the design effect (deff).

a. 11.9

b. 12.9

c. 13.9

d. 14.9