It is important to make sure that when you undertake any targeted or ad hoc survey work, you are clear about why you are conducting this survey and understand which group of people you need to speak to.
Some council departments already conduct targeted surveys, for example in adult social care. If you are interested in conducting an ad hoc, or targeted survey, you may want to consider speaking with colleagues in departments who already conduct this type of survey to benefit from their experience.
Above all, it is important to make sure that when you undertake any targeted or ad hoc survey work that you are clear about why you are conducting this survey and understand which group of people you need to speak to. It is also important that you make your survey as representative of your target group as possible and present your results clearly.
Ways to make sure your survey is as representative as possible
If your survey is not representative, then the findings it generates will not give you a fully accurate picture of people’s views. You will have the views of only some, not a representation of all of the population, which will be less useful to you. The first step is to identify your population, and ensure that you understand their characteristics. This means that, at the end of the survey, you will be able to check the extent to which your achieved sample reflects the real make-up of that population.
For geographical areas, the Census is a good option, and data can be obtained here. The characteristics that it will be useful to look at will vary depending on the subject matter but age, gender, working status and ethnicity can be a good starting point. If you want to target your survey at particular service users you should speak directly to colleagues in that service area as they may already hold data and information that can be used for this purpose.
Importantly, in the survey itself, you will need to ask respondents to provide a small amount of information about themselves (their age, gender etc) to enable you to make this judgement about representativeness.
The second step is to think about the most appropriate way to contact your population. An important factor in this is to think about what sections of the population may be excluded by your chosen method and take steps to address this. For example:
- Online methods may exclude some people. Community centres or community groups could be targeted to include a wider sample.
- A face-to-face town centre survey conducted on a weekday may exclude younger residents and those in work, so one step to mitigate this would be for data collection to take place on evenings and weekends as well.
- A survey on views of a community facility that is conducted at that facility will give a picture of the views of service users but won’t give any insight as to the reasons why people don’t use that service (perhaps it is not accessible enough for them or they are unaware of its existence). Service user profile statistics could be used to see which groups are underrepresented, and their views could be sought separately.
Where you have been able to identify contact details for the population you want to survey you will also need to decide whether to go for a census approach, where you invite everyone in the given population to participate, or to select a sample to invite. The census approach is often appropriate where the given population is small, and it is important to get as many responses as possible. If the population is large, a randomly selected sample will be more cost effective. A sample should be selected randomly as a simple and free option, and Excel has functions available to help with this. This blog post outlines one way of doing this.
A note on using personal details
When using email addresses and other contact details to survey residents, it is important to make sure that people gave consent for their contact details to be used for this purpose when they provided the information to the council.
A survey can be greatly affected by the quality of the contact list. You should make sure that your contact list is up to date.
When you collect survey data, you must make it very clear to respondents how their data will be used and who will have access to it. Providing respondents with a confidentiality statement alongside your survey may assist. You will also need to make sure that any data collected it is stored securely and in line with the General Data Protection Regulation. Further details about data protection can be found here.
When it is appropriate to conduct ad hoc surveys that may not be fully representative
There are two common situations where ad hoc surveys could be appropriate. The first is where the population is not easily defined or contact details do not exist.
The second is where you need a rapid and low cost indication of the views or priorities of a particular group. This is often used when you need to get a broad steer, rather than make high profile statements or take important policy or spending decisions based on the results.
How the results should be presented
It is really important that you provide the results to everyone who took part. Results should always be presented with a full, plain-English description of the method that was used to collect the data. You should also make sure that your results clearly explain that the findings are not statistically representative of your entire population, and are instead a sample. The results will reflect the views of the people who were surveyed rather than the wider population.
To avoid any confusion and to make sure that your findings are not misinterpreted you should consider using phrases such as “50 per cent of respondents strongly agreed with the proposal” rather than, for example, “50 per cent of service users strongly agreed with the proposal”.
Further guidance and resources
- The Local Area Research and Intelligence Association (LARIA) runs a Knowledge Hub group which is a useful place to seek advice from fellow researchers on methodological challenges. Knowledge Hub is a social networking site for local government and it is free and straightforward to register.
- The Social Research Association also runs regular training courses on various aspects of research practice, including running a survey.
- The Market Research Society has produced a range of useful guidance, including ten tips for DIY surveys. (See link annex)
- The Government Social Research Service has produced guidance on undertaking evaluations, which includes some pointers on conducting surveys.
- Ipsos MORI has produced a guide to conducting surveys of tenants of social housing, and the tips will also be of use in a range of other scenarios.
- For advice on how to evaluate and measure the effectiveness of your communications activity visit the International Association for the Measurement and Evaluation of Communication.