S3 June 2016 Q1
1.
(a) State two reasons why stratified sampling might be a more suitable sampling method than simple random sampling. (2)
(b) State two reasons why stratified sampling might be a more suitable sampling method than quota sampling. (2)
| Scheme | Marks |
|---|---|
| e.g. Analyse / find estimates for a particular subgroup of the population. Stratified guarantees representation of all groups, srs does not. Observe relationships between subgroups – srs does not guarantee equal or proportionate representation. Rare or extreme cases as part of a small subgroups can be represented proportionately in stratified i.e. stratified represents the structure of the population– srs does not allow this. Stratified typically require large sample size compared to srs due to lower variability within subgroups compared to entire population. Any 2 distinct reasons | B1B1 |
| (2) |
| Scheme | Marks |
|---|---|
| e.g. It (a stratified sample) is not biased as the members are chosen randomly. You can estimate the sampling errors (for a stratified sample) It (a stratrified sample) gives more accurate estimates as it is a random process. A quota sample may be (interviewer / process) biased. It’s not possible to estimate/find the sampling errors for a quota sample (whereas you can for a stratified sample) Any 2 distinct reasons | B1B1 |
| (2) | |
| (4 marks) |
Notes
Award B1B1 two correct, B1B0 one correct.
Allow ‘it’ for ‘stratified’.
Do not award marks for vague responses such as ‘cheap’, ‘easy’ ‘quick’ ‘random’ etc.
Mentioning ‘sampling frame’ alone is not sufficient for a mark.
Mentioning ‘non-response are not recorded’ alone is not sufficient for a mark.