Research into the extent of COVID-19 would only be 100% accurate if we tested the entire population. The same would apply if we wanted to test the effectiveness of a particular vaccine. However, this would be a slow process and extremely expensive. So instead scientists use random sampling in the search for certainty.
Principles of Random Sampling in the Search for Certainty
The accuracy of random sampling depends on choosing subjects objectively, and without bias. First, researchers statistically determine their minimum sample size. Then they ensure every member of their chosen population has an equal chance of being included. Finally, they analyze their randomly-chosen sample in terms of their criteria of interest.
Following the principles of random sampling in the search for certainty means researchers can infer their findings to the population. However, there is always a margin of uncertainly, which is why the word ‘probability’ pops up in their reports.
World Health Organization and American administrations including CDC and FDA approve of vaccines trialed scientifically. But the accuracy of the results depends on a representative sample size. This is why vaccine testing takes a long time to the frustration of those who want fast results.
Bending the Random Rules to Get the Job Done Sooner
Randomly testing a vaccine on a sample of the entire world population would be a massive task, we should probably never get done. Researchers therefore have to ‘bend the rules’ within statistical limits. This typically involves randomly selecting participants within a subset of the entire population of interest in one of four ways:
1… Convenience sampling of the most accessible individuals
2… Volunteer sampling by making an open invitation to participate
3… Purposeful sampling involving thoughtfully selecting participants
4… Snowball sampling whereby participants recommend family and friends
Scientists can then randomly select contributors from these lists, and complete their research in a relatively short time. However, while the results may increase broader understanding, their chances of representing the entire population 100% accurately are statistically unlikely.
Related
Double Blind Random Experimental Testing
WHO Queries Benefits of Four COVID Drugs
Preview Image: Complex, Multistage, Probability Sampling
