Stats 240 Mid-Semester Test PART A 2020

Hello, dear friend, you can consult us at any time if you have any ques tions, add WeChat: daixieit

Stats 240 Mid-Semester Test

PART A: Survey Design and Conduct (Andrew Sporle)

2020

A1 Senior management at the local Auckland hospital have teamed up with

researchers at the University of Auckland to do a survey on what patients think about possible future data linkage of medical records to other official statistics. Will they be required to get ethics approval for this survey? List the reasons for your answer (brief bullet points only)

A2 The hospital mentioned in A1 serves a large and diverse community. The

researchers want to collect demographic information from survey participants – what is the best source of demographic questions for their survey?

A3 The researchers wanted to survey a random sample of patients. However, data

processing delays meant that there were likely to be patients missing from the list  of admitted patients, which was the intended sampling frame for this survey. What are the three (3) main reasons why errors in sampling frames are seldom                corrected?

A4 The researchers planning the survey decide to use a multicultural team of well-

trained interviewers to collect the data in order to minimise what type of bias?


A5 List the four (4) reasons why a good response rate is important.



A6 What type of missing data cannot be corrected by imputation using other data?


A7 What are the four (4) steps the researchers should take in evaluating the effect of  bias in a study?

A8 High quality design of a survey instrument will minimise the potential for coverage bias by affecting what two (2) quality measures of a survey process?

A9 Which one of the following statements is false?

i. Every step of the survey process is a potential source of error

ii. Piloting a survey process can identify sources of both selection and information bias

iii. One reason for using a survey instrument (e.g. a questionnaire) from an official statistics survey is that it shouldn’t need pre-testing

iv. When using survey data from a survey conducted by others, the technical information about how the survey was conducted can provide useful        information about the possible sources and extent of non-sampling error.

v. Increasing the sample size is an effective way of resolving non-response bias.

vi. Postal surveys are used as they are cheap and have a good response rate.

vii. Re-using data from an existing survey dataset doesn’t remove the risk of information bias.

viii. The total error in a survey is seldom reported as it is difficult to accurately estimate the effect of the sources of non-sampling bias.

2

(Please highlight one)      (i)      (ii)      (iii)      (iv)      (v)      (vi)      (vii)      (viii)

A10 Which one of the following statements is false?

i. Area based measures of Socio-Economic Position (SEP) are often used as collecting information is relatively simple.

ii. Ethnic categories are not internationally consistent

iii. Using occupation-based measures of SEP for comparisons across time   requires consideration of possible changes in employment rates between time periods.

iv. Outputs from using official administrative data can have information bias.

v. A sampling frame should not include units which are not part of the target population.

vi. A survey can have a low response rate but a high completion rate.

vii. Socio-economic measures are all consistent across a person’s lifecourse.

viii. The purpose of piloting a survey is to test the survey instrument.

2

(Please highlight one)      (i)      (ii)      (iii)      (iv)      (v)      (vi)      (vii)      (viii)





2022-08-24