Last month, the EXPO-ENGAGE team took to the streets of Caerphilly, South Wales, and Camborne and Redruth, Cornwall, to ask local people for their thoughts on air quality. To do this, we designed a short survey, then set up our EXPO-ENGAGE stall and invited passers-by to participate in our research at several busy locations, including a weekly market and shopping centre.
While we did lots of planning before our fieldwork, designing the survey, undergoing ethical review, and preparing posters for the stall, there were still practical challenges on the day. Once we had packed up and returned home, we had a think about this. We noted that while there are many resources for designing research and doing data analysis, there are fewer that offer guidance on the practical aspects of conducting surveys, such as how to make the survey as easy as possible for participants to complete, and ensure technology works correctly. To support other researchers who may be planning similar fieldwork, we share three challenges we faced during our fieldwork and potential solutions.
1. Making QR codes easy to scan
This problem could be solved by printing a stack of business cards containing the QR code and a short description of the project. These could also be used to share information sheets with participants, instead of handing out pages of A4. Online information would be more portable, allowing people to easily check the project’s goals, and reach out to the lead researcher if necessary.
Lesson 1: QR codes alone are not enough to make your survey simple to access. It’s important to carefully consider where these signs are printed and whether this is convenient for participants.
2. Preparing survey questions for verbal participation
QR codes aside, many people preferred not to use their smartphones or mess around with unfamiliar devices. This meant a member of the EXPO-ENGAGE team had to read the questions aloud. There were some questions that were difficult to communicate in this way, such as this question where people were asked to select up to three factors:
By limiting the number of responses to three, we designed this question to understand which factors were most important to people. However, because of this participants had to wait until all possible answers had been read out before selecting their responses. They often found it difficult to remember all the options, potentially biasing responses. In the future, we would not limit the number of possible responses to three. Instead, we would run through the possible answers in order so that respondents could indicate if a response was relevant immediately after it was read out. This would still provide insight into the issues that people see as relevant, while making the survey easier to complete.
Lesson 2: if you’re going to make a survey available in multiple formats, it’s important to make sure your questions are easy to understand, whether in writing or when spoken aloud.
3. Recording response rate with a handheld tally
A handheld tally like the one pictured above would have helped us to calculate the number of people who declined our survey. It’s relatively easy to track this number for online or telephone surveys, but keeping a record of our survey’s response rate in the street while juggling a tablet and information sheets proved challenging. With a handheld tally, we could have simply clicked for each non-response. This data would have been useful for understanding how many residents were not interested in engaging with an academic study about air quality.
Lesson 3: when planning an in-person survey, especially when researchers will be recruiting passers-by, it’s important to consider how you will easily and effectively record any non-responses.
We had lots of interesting conversations about measuring air quality in Cornwall and South Wales. We’re excited to begin the next phase of our project, where we’ll work with local people and organisations to develop plans to measure air pollution in Caerphilly and Camborne, Pool, and Redruth. As we move forward, we will draw on the above learnings to develop our research practice, while we also hope they will help other researchers doing similar studies.