How to... use an izwe Survey

Main considerations

Preparation

• Define your information needs and the objectives of the survey; do you want to find out about local opinion and priorities generally, do you want help in defining local options and preferences, or do you want input into specific decisions?
• Plan the individual questions you want to ask: see below
• Design the questionnaire, paying attention to question design to get precise answers. See University of Leeds: Guide to the Design of Questionnaires
• Run a pilot survey to test and review the questions

Question creation

• Do people need more information beforehand? Remember: you can embed videos and documents in the text before the survey or in individual slides
• Do you really need to ask this question? Are you asking it in as simple a way possible
• Make sure you're asking the right kind of questions! See our quick survey guide to types of questions below.

See external resource on survey questions here

Where to host your survey?

You can include your questions:

• As part of a regular residents' survey
• As a standalone survey in its own 'conversation'
• As part of a themed conversation

Acheiving representativeness with private surveys and sampling

When it is important to ensure that the responses are representative of the population as a whole, it is best to use sampling unless you are confident of sufficient numbers of participants. So for surveys of:

• limited public interest
• highly important to be representative


• Set up your survey in a private environment (inside an new or exisiting private conversation)
• Identify your target population and sample (including minimum sample size to ensure representativeness)
• A pre-existing resident's panel might be ideal for this. If you don't run one already it is likely to be worth it
• Otherwise you'll need to recuit participants, and potentially incentivise
• 'Invite' those recruited to the survey, explaining the purpose of the survey and how you intend to use the results

Acheiving representativeness through large-scale participation, targetted publicity, and weighting

Often you will want a survey to be open to those interested to take part. However you may still have an interest in acheiving an overall representative picture

• See tips for publicising surveys below
• Think about the demographic profile of the population, if appropriate make an effort to target publicity to different networks to reflect this as closely as possible to ensure diversity of views
• Add a question to capture any demographic information you need for profiling and weighting
• Sort and weight survey results in your favourite data analysis package/excel

It is very difficult to achieve participation that accurately represents the demographic profile of the local population. Although they enable wide reach, online surveys have a high tendency for self-selecting participants. Even after the demographic weighing of survey results, it is important to be aware of these limitations. Attention should be paid to the potential for combining online and offline data gathering methods to ensure that the most marginalised groups within the population are properly represented.

Tips when publicising Surveys

• Identify the benefits for participants of taking part
• Publicise the survey on any easy digital or regular channels (inluding any consultation database you maintain), and also invite particular relevant groups to take part
• Use existing stakeholder networks - send out an email or bulletin to be forwarded, and ask organisations who might post links on their websites or e-newsletters
• If it is important to get a wide range of views, it might be beneficial to incentivise participation with a prize draw or similar
• If it is high priority or of wide interest, you might want also traditional publicity, including in the local press

Analysis

• izwe has basic data visualisation - see 'Analysing Results' in the technical section
• In addition to the top-level quantitative results from the survey, analysing the data in terms of demographic sub-groups, areas etc will enhance your understanding of the diversity of local opinion and offer potential insights into the underlying factors. Allowing for qualitative comment within the survey will also provide some opportunities to delve deeper into local views and concerns.

Feedback and follow-up

• Post a blog post, and convey top-line results with some suitable communication (perhaps you could give regular updates to the group)
• Explain how you'll act on their contribution/the results
• If you can, give participants access to the whole results by uploading them to a blog, or adding them to your council's open data store. Find out about how you can easily create 'linked data' here. With the right community involvement, you can get rich, free, community-contributed analysis
• Update participants (and your online community) on any actions taken as a result of the survey, even if these are not immediate. This will help you to build and maintain an ongoing relationship with the users.
• Think about using other tools to further explore certain aspects of the topic, to gather additional information or to invite users' suggestions for addressing the issues identified through the survey (such as a call for ideas; eg "90% of respondents felt that services for young people most need improving in the local area. Which services are most in need of improving? What are your ideas for other/different services?")