If you’re a decision maker, then you need to engage those affected by your decisions. The problem is – this can be a lot of people! The same problem can exist if you’re a social researcher - when there’s so many people that you could talk to, how do you choose who you will talk to?
We face this question every day here at Traverse, as talking to people is kind of what we do! Because we do it every day, we feel we’ve learned a bit about it, and this blog is designed to step you through the main parts of what you need to do. Specifically, in this blog, we take you through how to create a sample and how to reach them. This is frequently essential in our research and engagement work.
Step 1: Identify your population
Your population is the group of people whose opinions are essential for your purposes, usually because they are the ones who will be affected by the decision(s) you are consulting or engaging on. You need to know who exactly this is. Examples of populations include,
- For an infrastructure project, it will likely include all those living close enough to the site(s) to be affected by construction and operation, as well as any others who are likely to be affected;
- For a business plan, it is likely to be all of your customers;
- For a policy decision, it will be all those affected by your change of policy (which may be the entire UK population, or may be just those who use a particular service);
- For a regional utility issue; it will likely be all those who live in the region and are connected to the particular utility;
- For a local issue, it will be everyone living in the local area – this may be everyone in a village, suburb, town or city;
- For a national issue, it will likely be the entire UK population.
Step 2: Find out about your population
You then need to find some things out about your population. You will want to gather some basic statistics about them, such as the breakdown of your population by:
- Socio-economic group
- Household size and composition (e.g. number of people in the household and whether this includes dependent children)
You can often find this information on government websites, such as the Office for National Statistics ‘Population and Migration’ and 2011 census pages.
However, don’t rely just on standard population statistics - you should consider all aspects of your population which may affect a person’s opinion on the issues. The examples above are included because, often, people in different age groups usually have different opinions on issues, so age must be one of the statistics you gather. However, you should think about what other factors will influence opinion. If it’s a political issue, you may also need to gather information on how people vote. Otherwise there’s a risk that your sample (see below), whilst perfectly balanced for things like age and socio-economic group, does not represent a balance of political views. For an energy issue, your sampling may need to take account of data on public attitudes (such as that produced by BEIS), to ensure you build a sample which is representative of the energy attitudes of the population.
Step 3: Create your sample
You will need to use the information you have gathered about your population to create a sample. A sample is a smaller group of people who, between them, are statistically representative of your population. When your sample is statistically representative, it means that you can be confident, to within a specific margin of error (usually 5%), that the people and opinions in the sample are reflective of the people and opinions in the population. There are two aspects to this:
The size of your sample needs to be large enough. Calculating this depends on a number of factors, including which engagement method(s) you plan to use, which in turn can depend on budget. You can find out more about how to calculate sample size online. A full guide to how to calculate a sample size is well beyond the scope of this blog, but if you search online you will find a number of websites which will calculate the sample for you, and online guides for those with more complex needs.
Your sample needs to contain the same mixture of important characteristics (see Step 2) as your population. For example, if 20% of your populations is aged 18-25, then 20% of your sample must be too.
Optional step: Targeting specific groups
Sometimes you don’t want your sample to be fully representative, because you want to find extra people in particular groups to ensure their voices are very clearly heard. It is good to consider who will be most affected by your decision, or most vulnerable to its effects, and consider whether you want boost this area of your sample. Examples of groups in your sample where you may wish to boost the numbers include those on low incomes; those with health problems; and those who live closest to the affected area (if there is one).
Alternatively, or additionally, you may want to target groups of people who don’t usually engage much in the sort of research you are doing, and therefore whose opinions are often under-represented. This sometimes includes young people, those with English as a second language or those in lower socioeconomic groups.
Step 4: Reaching your sample
Finally, having worked out exactly who to talk to, you need to actually talk to them! This bit can be tricky, depending on the characteristics and size of your sample. If you’re lucky you’ll be able to reach most of your sample online, such as through posting an online survey or using a recruitment company to reach an online panel (see below). However, some groups of people are very under-represented online, so you’ll need to use other methods to reach them in sufficient numbers. You will probably need to use a mix of the following methods to reach your sample:
- Use a market research /fieldwork/ recruitment company to reach all of your sample (or some of your sample if you can reach the rest yourself). A recruitment agency can recruit people to
- Fill in a survey (online, face to face, or both)
- Attend an event where you engage them directly
- A combination of the above.
- Go out and talk to your sample directly – door to door or on a busy shopping street; and
- If you already have contact details for your sample (such as customers or parents of school children), and permission to use these details in this way, you can contact all those in your sample directly.
The title of this blog is ‘how to talk to a crowd’, and we hope we’ve shown you how you can do this – by identifying your population, finding out about them, creating a sample and then talking to them. So once you’ve spoken to them, what do you then do with the data? You may already know exactly what you want to do with the data, but if you’re not sure, there are many types of quantitative and qualitative analysis that can be done. We’ll be digging into some of these in our next blog posts.