How to Select a Sample Group

There are numerous ways to select a sample. You should choose the strongest possible method that is practical within your resources and the expertise of the evaluation team.

Random samples – as explained in the first table - are the strongest but they usually require the most resource and evaluation expertise. Non-random samples – as explained in the second table – are not as strong but may be more practical.

How Many People Should Take Part?

Ideally, get help from a statistician to determine the number of people to take part in your evaluation. If this is not possible then the following figures can be used as a rough guide.

For questionnaires or surveys (mainly closed questions) try to have at least 50 people take part. If different groups are being compared you will need at 50 people per group e.g. 50 males and 50 females.

In focus groups (mainly open questions) you should have somewhere between 4 and 12 people per focus group. You may do one or more focus groups. The total number of participants, over all groups, is less strict than questionnaires because focus groups are used to obtain rich detail rather than aim to be representative of the population (pupils in a school or year group, for example).

Ways to Select a Sample

Random Sample Techniques
Simple Random

Every member of the population you are interested in has an equal chance of being selected. To use this technique you will need a complete list of everyone in the population. This list is called a sampling frame.

For example, a theatre in education presentation was delivered to 3,000 students. To obtain a simple random sample you will need the names of all 3,000 students who attended. With the names in no particular order (make sure they are not alphabetically listed), make a random selection until you have reached the number of people you want for your sample.

You can randomly select by using a random number generator such as, or by using a random number table, or by pulling out of a hat. The random number generator will produce a list of random numbers. You then select every student who is that number on the list, i.e. the 331st student and the 2982nd if these are the numbers generated, and so on.

Stratified Random (proportionate)

When your population of interest contains different characteristics it may be more helpful to use a stratified sample.

For example, you delivered a presentation to 800 learner motorcyclists, 200 of whom were female and 600 were male. Using a simple random sample it is possible that all 200 female learners would be selected to be in the evaluation sample. Therefore, if your sample size is 300 your sample will not be proportionate to the number of males and females in the population of 800 learners. So the views you collect may not accurately represent the views of the whole group.

Using a proportionate stratified sample you divide the learners into two groups: male and female. Then, using simple random selection, and for a sample size of 300, you select 75 learners from the list of females, and 225 learners from the list of males.


If you do not have a list of every member of the population you are interested in, it is still possible to choose a random sample by using cluster sampling. Cluster sampling divides your population into groups and a simple random selection of those groups is made. You then survey everybody within the selected groups.

For example, a talk by RSOs was given to all Year 11 pupils in the County. You do not have a list of all the pupils who received the talk but you do have a list of all the schools in the County. Therefore, you randomly select a school (your cluster) and survey all Year 11 pupils within that one school. This can save time and money as all your participants are in the same location.

Non-random Sampling Techniques

Only people with specified characteristics are selected. These characteristics can represent the range of characteristics in your population of interest, or you can focus on some characteristics in particular.

For example, if you are evaluating the attitudes of drivers towards speeding, you may want to only sample those who have got penalty points for speeding. Or, you could sample those with extreme characteristics by only selecting drivers who have been disqualified as a result of a more serious speeding offence or multiple offences.


This is another method commonly used in market research. With quota sampling you divide the population (e.g. older drivers) into distinct parts (strata). You then decide how many of each stratum you want to have in the total sample. This forms your quota and you convenience sample (see below) until the quota is reached.

For example, you could stand outside an out-of-town supermarket in the day time. You have decided that you want to survey 50 male drivers aged over 65. Every time you see a male older driver enter the store you ask if he is aged over 65 and you continue until you reach your quota of 50 completed surveys.

This is the non-random form of stratified sampling.

Weakest Sampling Methods:

In a snowball sample you ask your participants if they can suggest someone else, with the relevant characteristics, who they think may be willing to take part in the evaluation. That someone else then nominates another, and so on.

This makes it easier to contact people you might not otherwise have any way of getting in touch with. However, it does introduce a lot of bias into your results because those in your sample are all likely to know each other and to have similar opinions.


With this method you survey whoever you happen to have access to. It is also called 'opportunity sampling'. This is commonly used in market research.

It is a very easy method to use but you do not know if the people you stop are members of the population you are interested in (drivers who have been prosecuted for speeding, for instance).