I will recommend the use of a probability sampling method which involves random selection, so as to make make strong statistical inferences about the whole group (health workers). In view of this, I would suggest stratified sampling to ensure that every subgroup (Nurses, doctors, medical lab technicians and other health workers) is properly represented in the sample.

Thank you.

Best of luck.

In addition to the above I would recommend that you look at the attached screenshot. This book tells you how to do stratified in full detail. It's available for download from the z-library. Btw most statistics computer dont do the stratified calculations. Without adding a sub package. I have attached a second screen shot about doing this in R. Replace R in the query by any other package you might prefer. Best wishes David Booth

Recalling that stratified sampling aims to mininmise sampling error, note that a sampling frame is required for each stratum and sampling “weight” should be taken into account for the

disproportionate sampling of

strata.

In addition to that, if you want to use stratified random sampling, sample sizes within each stratum and cost of sampling needs to be put into consideration.

In the same vein, variability in the sample needs to be considered as well

@Adejumo Taiwo Joel thank you