Minimum number of clusters

[donaldgloverParticipant]

Can I still use the data collected if I could not reach the 25 minimum number of clusters recommended by SMART due to unforeseen circumstances that occurred after cluster selection? For example, for our survey we could only collect data from 24 clusters due to security issues that arose during data collection?

[SMARTKeymaster]

Hi,

If the data collected was high quality it can be considered valid. There are two potential issues at hand; precision and design effect.

Precision: The confidence intervals will be wider as the sample size (children) was not fully obtained. SMART recommends that if at least 80% of the sample size is obtained, the data collected can still be used.

Design Effect (DEFF): The design effect of the survey will be affected depending on the number of households per cluster. If there was a low number of HH per cluster the DEFF will increase; if there was a high number of HH per cluster then the DEFF will decrease.

As a mitigation strategy, all efforts should be made to include all of the reserve clusters (RC) randomly selected in ENA for SMART (at the same time the other clusters were selected) in order to increase the sample size and improve precision.

In the report it is very important to outline which specific clusters from the sampling frame were not accessed during data collection due to unforeseen circumstances (such as violence), as well as to identify the RC added (the ones that can be accessed).

• This reply was modified 11 years, 2 months ago by  SMART.

[Author]

Posts