Samples and Sampling

Question 1

what makes a good sample?

Answer 1

that it is representative of the population you aare studying

Question 2

what determines external validity?

Answer 2

population definition

size of sample

type of sample

Question 3

what is the difference between a probability and non-probability sample?

Answer 3

probability sample = drawn in such a way that each member of the population has a defined probability of being included

non-probability sample = sample drawn of participants whoa re most likely to be informative

Question 4

when do you use probability vs. non-prabbility samples?

Answer 4

probability = used when data is being analyzed statistically - allow results to generalize the population

non-probability samples= cannot relate to overall population, only subgroups selected for sample

Question 5

what is cluster sampling?

Answer 5

identifying naturally - occuring clusters in the population

ex) select schools at random and sample random children within those schools

Question 6

what is stratified sampling?

Answer 6

to make comparisons between groups which have different prevalences we use stratified sampling

ex) if you were to do a study on male and female nurses, there are more female nurses so you would have to account for that

Question 7

What is systematic sampling?

Answer 7

selecting “every 10th employee on payroll” = it’s methodical in selection, gets random results but not quite a random process

Question 8

can you draw a random sample without identifying members of the population?

Answer 8

no, but you could still draw a representative sample

Question 9

what is a consecutive series sample?

Answer 9

ex) study phone calls for one entire month

or study all patients over a period of months

there is argument, however that this is not representative

Question 10

what is a grab sample?

Answer 10

done by taking whatever is available -

often called convenience samples

Question 11

what are quota samples?

Answer 11

grab samples with specified quotas of people

ex) interview 10 male and 10 female colleagues

make comparisons easier but do not remove bias

Question 12

what are ‘snowball samples’?

Answer 12

used with hard-to-access groups like drug abusers and sex workers

researcher asks each person to recruit more people - pass survey along