In this post
Selecting samples
When we decide who to ask questions to in a survey there are several ways for us to select the participants. Instead of asking questions to the entire population we look to choose a sample. This saves us a lot of time and will hopefully give a fair representation of the entire population that we can draw some conclusions from.
One of the best ways to get a sample is to randomly select an amount that we need from the entire population. This is obviously done by using some way to find a completely random amount that has no bias to any factor at all.When selecting things randomly we do not always get the result that we would expect. So if we randomly selected people for a survey we may end up with a selection that is say 75% male and 25% female. Therefore, we may wish to take a sample that is random for some factors but to make sure that others are kept exactly equal (we could make sure we pick the same number of random females as males).
The main idea of sampling is to find a number of participants that will save us time but also give a fair representation of the entire population. This will then mean that the data we find and conclusions that we draw will be very close to the truth and will not be biased in any way.
Opportunity sampling
This type of sampling is where we ask the people to participate in a survey who are available at the specific time we need them. For example, if you wish to find out people’s views in a certain town you could stand in the town centre and ask people passing by. The only reason that you are asking these people is because they are close by and available to answer questions. Obviously, there are drawbacks to this as the people that live in a certain area may be from a specific background and you will therefore only gain information from people that may have very similar ideas instead of a wide range.
Quota sampling
We have already discussed quota sampling on the previous page in the example of needing the same number of females to ask as males. The name simply comes from the fact that we must fill a quota that is set out so that we will gain a view from different people. This means a population will be split and equal numbers used in a sample for different variables. These can include things like job, income or area of the world that they live in. By asking only people from London about a certain issue we are clearly neglecting the rest of the UK who may have very different opinions.
Systematic sample
A systematic sample is used to create a sample with a specific number used. For example, from a list of 100 people we may wish to interview 10. So we could simply ask every tenth person on the list. This is called a systematic sample as we have a clear system when choosing the participants.
There are many other types of methods that can be used for getting a sample and a mixture of a few that we have discussed can also be used. The main thing to avoid is bias and to make sure that the sample taken is the best possible representation of the entire population.
Sample size
There is no specific answer to the question: how big should a sample be? A sample should be a good representation of the entire population. Obviously, we do not wish to have a sample that is too large as we are looking to save time. However, a sample that is too small would not give a fair representation of the actual population and the data that we collect would then not be a fair insight to the truth of the population.
