Table 1 Comparative strengths and weaknesses of three sampling approaches [18]
Sampling approach | Strengths | Limitations |
|---|---|---|
Convenience | Rapid Low cost | Sampling depends on collectors choice of outlet (risk of bias) Poor documentation—findings hard to replicate Prevalence estimates are not reliable |
Random | Sampling frame is defined to obtain representative sample from all types of outlets and/or brands Results can be replicated | Need to authenticate and update sampling frame increases time and cost of survey |
Mystery clients | Outlets are unaware of survey so less chance of sample bias | Information on sources of poor quality drugs limited to brand, batch and country of manufacture as stated on packaging |
Overt | Additional information is collected at minimal additional cost to mystery approach as provider is aware of the aims of the study | Risk of sampling bias in samples collected, if some outlets refuse to sell or samples are deliberately withheld as poor quality is known by the seller |