All are true about cluster sampling except?
|A||Sample size is same as that of simple random sampling|
|B||Less efficient method than simple random sampling|
|C||Results can be generalized to the population|
|D||Clusters are naturally occurring groups|
Ref: Basic & Clinical Biostatistics; Dawson – Saunders & Trapp.
- The sample size is never the same as that of simple random sampling in cluster sampling. The very reason for carrying out cluster sampling is for administrative & operational feasibility.
- In this method, instead of selecting individuals, a subset of clusters or groups are randomly selected and then all respondents or a pre-determined number of respondents in each cluster are included in the study.
- The result is considered to be statistically valid within the error limit of 10%. However the method is definitely less efficient to simple random sampling & requires a larger sample size, but in some situations it is the method of choice, such as in multicentric trials, in order to obtain adequate number of patients.
- Cluster sampling is most frequently used in immunization coverage evaluation surveys.