Conducting interviews and surveys

Once you decide the goal of your research, the next step is to select whether you’ll collect qualitative or quantitative data. When possible, collecting both together can be very valuable. If you can’t do both due to time or financial restraints, think about whether the information you are trying to gather would be best told through stories or charts and graphs. For example, are you trying to gather data about whether an event participant would come again next year and about which activity at the event was their favourite? A survey would make the most sense for this kind of information. Or are you trying to learn about what was most inspirational and meaningful to the participant at the event? In that case, listening to the participant’s words by interviewing them would be most beneficial.

After choosing the method you’ll be using, it’s time to create a learning agenda and discussion guide. A learning agenda is an outline of what you’re seeking to learn in your research. It should break down the overall topic into a few sections that will serve to keep your research organised and from becoming too expansive. There is so much we’d like to ask our audience, but we want to ensure we’re asking about the most important information so that we don’t overwhelm them. The learning agenda is the precursor to the discussion guide, which assigns questions to each section of the learning agenda. The discussion guide becomes a script for interviews and the questions for surveys.

Once your discussion guide is ready, you’ll need to decide who you’re asking the questions to and how. If you’ll be conducting interviews, start with a random sample of 10-12 individuals within your target audience.

If you’ll be sending out a survey, calculating how many people you’ll need to send the survey to is a bit more tedious. First, it depends on the margin of error you’re willing to accept. Because this data will be used to influence your organisation’s identity and communications and not to conduct super scientific research, a 5-10% margin of error is fine. Using the chart below, identify how many total people are within your population, then see how many respondents you’ll need based on the margin of error you choose.

For example, if you have 3,000 contacts in your email database that fit the description of your target audience, and if you’re satisfied with a 10% margin of error, then you’ll need 100 survey responses. However, take into account that not everyone to whom you send the survey will respond. If your organisation typically experiences a 20% open rate of emails, then assume you’ll need to send the survey to 500 individuals to receive 100 responses.

Once you know who will be participating in your research, all that’s left to do is conduct the interviews or send out the surveys and run the analyses. Quantitative data is fairly straightforward to analyse, and many survey tools can create useful graphs and charts for you. Analysing qualitative data will involve listening back to recordings of the interviews and gathering themes. What was said by several participants? It’s also helpful to identify if anything was inconsistent in the data. For example, did each participant answer differently when asked what your organisation’s mission is? Perhaps that’s a clue that they are unfamiliar with it and that you should communicate it more frequently.

When analysing both qualitative and quantitative data, be sure to return to your learning agenda and ensure your questions have been answered. Then, apply your findings to continue improving your organisation’s identity and communications to best serve your audience and grow your impact.