Hoofdstuk 12: sampling

Question 1

Explain the sampling plan

Answer 1

1. Target pop.: generalize to
2. Accesible pop: who can we reach ?
3. Sampling frame: through what medium can we reach ?
4. Sample: Wo participates ?

Question 2

Eligibility criteria ?

Answer 2

Inclusion and exclusion (+costs, practical constraints,…)

Question 3

Sampling bias ?

Answer 3

=systematic over or underepresentation of population segment

more homogenity, less bias

Question 4

Explain non-probability sampling (4)

Answer 4

=non random sampling

1. Convenience: easy to reach pop
2. Quota:based on identified strata and size of each group –> proportion is goal –> lowers sampling bias
3. Consecutive: all people accesible on specific location over period of time
4. Purposive: almost never representative

Question 5

Explain probability sampling (4)

Answer 5

=each element has equal chance of being selected

1. Simple random: random selection from list
2. Stratified random: divide in strata and sample at random
3. Multistage cluster: select groups instead of individuals (eg: 10 classes each 20 people-> N=10), is less accurate
4. Systematic: every k’th from list = simple random but more convenient

Question 6

sum up some sample size basics

Answer 6

• power analysis
• large sample = more representative
• large non-probability sample > small non-prob sample

Question 7

What factors affect sample size

Answer 7

1. Effect size (expected relationship between var.)
   - stong relation –> small sample is okay (analog for weak relation)
2. Homogenity: more = smaller sample
3. Cooperation and attrition: percentage of people likely to cooperate and loss over time
4. Sub group analysis: samples must be large enough to support analysis within subgroups