SAMPLE SIZE CALCULATION MRCPsych Paper B

MRCPsych Paper B

SAMPLE SIZE CALCULATION

Critical Review

Question

You are checking the frequency of “the Perceived stress scale among doctors serving at various Government and private hospitals in Faisalabad. (JPMA-2020-02-232)”

If you aimed for a specific category of doctors, then that will be your target population. For example juniors, seniors, permanent visiting, etc.

Introduction

Once we plan a survey or a research experiment, then we have a target population in mind. For example, some surveys cover the entire population of the country like census and political elections. If we take an example for census first, then what sample size do you think would be the correct representative in that case? The obvious answer is 100% of the population of the country, not a single percent less.

Examples

Now take the example of the political elections, let’s say 50% was the turnout on polls. Would that be an excellent election turn out? Probably not, because half the population of the country didn’t take Part in it. If the turnout was 80% then it is stronger because a good proportion took part and we can conclude that the results represent the voters in our country.

The above examples highlight that sample size determination will depend upon what population we aim to get the data from and what is the purpose of the data collection.

You are checking the frequency of “the Perceived stress scale among doctors serving at various Government and private hospitals in Faisalabad. (JPMA-2020-02-232)”

What you need to ask yourself is that what is the total population of such doctors in this hospital? If you aimed for a specific category of doctors, then that will be your target population. For example juniors, seniors, permanent visiting, etc.

Let's assume that you know the population of such doctors in question. And for sake of a hypothetical exercise, let’s say that 100 doctors are your target population. If you are checking the frequency then how much sample size do you think would represent these doctors?

Follow this link on the sample size calculator website and put the values Sample Size Calculator. Once you put a population size of 100, then you find that to be 95% confident that this sample truly represents the target population you need to examine at least 80 people out of 100. Now 5% is the margin of error. If you increase the margin of error to 20%, you will need a sample size of 20 only. Now put your population size into this calculator and choose your margin of error, and you will find the sample size that you must have aimed for.