7: Sampling Methods

Question 1

What are the reasons for sampling?

Answer 1

Cost, time, possibility of generalisation.

Question 2

What is the population?

Answer 2

The group we want to draw conclusions about. Population must be clearly defined.

Question 3

What is the sample?

Answer 3

Sub-group of the population.

Question 4

What is a representative sample?

Answer 4

Sample where every major attribute of the larger population is present in roughly in proportion or frequency with which those attributes occur in the larger population.

Question 5

What is probabality sampling?

Answer 5

All potential cases must have the same probability of entering the sample. Allows for generalisability.

Question 6

What are the types of probability sampling?

Answer 6

Simple random samples, systematic random sampling, cluster sampling, stratified random sample.

Question 7

What is simple random sampling?

Answer 7

Sample chosen - every individual or case in the population has an equal opportunity to be selected for analysis. e.g. random number generator.

Question 8

Why is systematic random sampling used?

Answer 8

Used for large populations.

Question 9

What is systematic random sampling?

Answer 9

Used for large populations. Selection from lists - telephone books, student directories, etc. Count no, of total cases - divide by no. desired in the sample (select every kth case). Randomised Ist draw.

Question 10

What are the problems with systematic random sampling?

Answer 10

Risk of systematic bias, sometimes pratical problems.

Question 11

What is cluster sampling?

Answer 11

Multistage random area sampling. Identify members of the sample as residents of particular housing units.

Question 12

Why is stratified random sampling used?

Answer 12

Used if a sub-group of the population is too small to permit detailed analysis therefore. e.g. e/m, etc.

Question 13

What is stratified random sampling?

Answer 13

Sample is divided into 2 samples. Sub-groups need to be known in advance. Employed as a second-order method - not a replacement for random sampling methods (bias).

Question 14

What is non-probability sampling?

Answer 14

Not selected randomly. Greater chance for bias and distortion. Still used as random techniques are often infeasible. Awareness of potential shortcomings.

Question 15

What are the types of non-probability sampling?

Answer 15

Convenience sampling, volunteer samples, purposive samples, snowball samples, quota samples.

Question 16

What is convenience sampling?

Answer 16

Use participants that are readily avaliable. Used for basic psychological or physiological attributes.

Question 17

What is volunteer sampling?

Answer 17

Participation due to material incentive. Self-selection makes these studies tricky with regards to representative conclusions.

Question 18

What is purposive sampling?

Answer 18

Selects participants which are a common experience - e.g. shopping in a shopping centre. Representative for a very specific target population only.

Question 19

What is snowball sampling?

Answer 19

For some studies if might be hard to identify subjects. e.g. homosexulatiy of sub-Saharan Africa. Find someone who is gay - knows someone else, increases no. of subjects = bias - restrained.

Question 20

What is quota sampling?

Answer 20

If samples need to be balanced by a certain characteristic. e.g. age, sex, race, etc. Individuals displaying the specific traits are selected in proportion to their share in the general population. Ensures that all combinations are present in their sample.