Survey

Generally, a set of simple questions that enables the creation of statistics. These statistics are used to generalise about how a larger groups feels, acts and thinks regarding a topic or issue. Surveys are a quick and relatively cheap method of consultation, with the biggest cost being sourcing an appropriate sample of the population in order to make the data meaningful.

There are four modes of survey data collection that are commonly used, with selection generally based on the reach and perceived efficacy of the population being surveyed:

• in-person surveys – the surveyor completes the form for the participant
• telephone surveys – useful for dispersed populations with mixed digital capability; the surveyor completes the form for the participant
• self-administered paper-and-pencil surveys – useful for feedback in a physical location, such as ‘How did we do?’, or as a mail-out to dispersed populations with mixed technical capability
• self-administered computer surveys – the fastest and cheapest method, but unlikely to produce representative information due to the need for digital capability and equipment.

Surveys are most useful when you only have a short amount of time and need insights that can be shared as statistics. They can help you better understand an issue or learn more about participant priorities.

As survey outputs are easy to convert into statistics, they are particularly helpful if gathering evidence to provide a rationale for a decision or set of priorities.

Immediate

• Statistically relevant results.
• Directly comparable responses.
• Fast evidence-gathering and input for decision-making.
• Better access to time-poor, geographically distant participants.
• Anonymity may improve the honesty of answers.
• Groupthink or power dynamics will not skew the results.

 

Long term

• Ability to track changes in sentiment over time.
• Ability to measure outcomes through sentiment changes.

• The questions used in the survey will have the biggest impact on the results. Leading or poorly worded questions will result in outcomes that do not represent actual opinions.
• Statistical relevance will be relative to the size of the sample group. Be cautious regarding how conclusions are made, and be transparent about the limitations of the data.
• Surveys seldom provide ‘new’ information, as the questions asked rely on the knowledge of the person creating the survey. Surveys are best combined with focus groups both before and after, to improve your understanding of the questions to ask and also uncover what’s behind the results.

Beforehand

• Carefully design and test your survey. It’s surprising how often people misunderstand seemingly obvious questions.
• Be very specific about the structure of the sample you need and the number of people you’ll be able to reach.
• Choose a delivery method that is appropriate for your participants’ needs, your ability to reach them, and the size of the sample.
• Have a plan for how you will use the survey question answers to address your research challenge. Code responses in advance to save time once you receive the results – there are software packages that can assist with this.
• Have a plan for recruiting participants, as advertising might not be enough. You may need to find an agency to source participants; however, there is usually a per-result cost for this.

During the process

• Watch for trends regarding when participants are completing the survey, and extend the time available to contribute responses if need be.
• Make sure you get the right number of results from each demographic group before ending the survey.

Afterwards

• Identify themes in the qualitative data in order to codify and make this feedback more useful to you.
• Allow enough time to analyse the results. The more qualitative questions you have, the longer you’ll need.