Chapter 4: Sampling

Question 1

What are the four elements of a sampling plan?

Answer 1

1. Who will be surveyed
2. How many people should be surveyed
3. how should the sample be chosen?
4. When will the survey be given?

Question 2

What is the best predictor for reliable results in (survey) research?

Answer 2

1. The quality of the sample

Question 3

What is a census and how is it different from a sample?

Answer 3

1. Census - data collection of every member of the population.

Difference with a sample is that every member of the population is studies in a census

Question 4

What is a sampling frame?

Answer 4

1. the list or resource that has elements of the defined population. It is the group of people that will participate in the study

Question 5

What are often pitfalls of a sampling frame?

Answer 5

1. Impossible to generalize conclusion from web surveys to general populations (unless working with offline recruitment methods)
2. Impossible to draw a sample out of the entire population

Question 6

What is coverage error?

Answer 6

1. A mismatch between the target population and the frame population. The frame does not contain every member of the population

Question 7

What are the different forms of probability samples? (3)

Answer 7

1. simple random sample –> every person has the same chance of being selected
2. Stratified random sample - divide population into groups and draw samples of every group
3. No sampling list available? Cluster sampling - large clusters divided into smaller clusters

Question 8

What are the different types of non-probability samples (5)?

Answer 8

1. Convience sample - choose anyone who meets basic criteria
2. Purposive sample - Choose people by interest, qualifications or typicallity
3. Judgement sample - Choose based on judgement of the researcher
4. Quota sample - Obtain participants in proportion to population size
5. Snowball sample - identify one good participant who will invite the next and so on

Question 9

What is statistical power?

Answer 9

1. The probability that you accept H1 (there is a difference) in case that there is an actual effect or difference.

Question 10

What is Type 1 error in statistics?

Answer 10

1. A false rejection of H0. You say there is an effect when in real life there is not

Question 11

What is a Type 2 error in statistics?

Answer 11

1. Failure to reject null hypothesis. You say there is no difference when in fact there is one!