Using surveys to gather patient insight

Picking the right questions

Think about what type of questions will help you get the insights you need.

Think about your respondents

When writing a question, ask yourself whether the respondents:

• can understand the question
• will be able to answer the question
• will be willing to answer the question
• will be led to answer the question in a certain way

Make it relevant

Make sure respondents are only asked questions that are relevant to them. Ask yourself whether the respondents have the right knowledge or experience to answer the question. If not, you may need to use filter questions to direct relevant questions to the relevant respondents. For example, if you want to ask about someone’s experience with a certain treatment, you may need to ask if the respondent has had that treatment before asking specific questions. You can also provide additional explanations and information within the survey if required.

Be clear

Make the questions easy to understand and be explicit with what you mean. There shouldn’t be room for interpretation, remember:

• not all respondents will understand the question in the same way
• don't be afraid to use definitions or give examples (e.g. a biopsy is a medical procedure where a small sample of body tissue is removed so it can be examined under a microscope)
• to follow the guidelines for writing for a lay audience

One at a time

Avoid double barreled questions by making sure you only ask one question at a time. ‘Is this toolkit interesting and useful?’ is an example of a double-barreled question. Some people may think this toolkit is interesting but not useful, or useful but not interesting. This question doesn’t allow for accurate responses.

Question order matters

Question order is important to engage respondents and minimise bias. Questions should be ordered in a logical sequence and start with broad questions before moving on to the more specific or sensitive questions. Think about whether the content of a question could influence how respondents answer subsequent questions.

Don’t allow your data to be skewed

Ensure people can answer your survey accurately by:

• including ‘none of these’ or ‘other’ for multiple choice questions
• using ‘don’t know’ as an answer option for closed questions. Without this as an option, respondents may guess or pick any option, which leads to unreliable data. 2%-3% usually choose this option.
• ‘don’t know’ is different to ‘not applicable’. ‘Not applicable’ should be used when the question is not relevant to the respondent
• there are different ways to phrase this, such as ‘not sure’, ‘can’t recall’ etc.
• include ‘prefer not to say’ when necessary (e.g. sensitive questions)

Think about the time reference

Don’t ask questions which will overstretch the respondent’s memory – this will encourage inaccuracy, generalisations and guessing. Define the time period so that each respondent understands the question in the same way, ensuring answers are consistent across respondents. For example, rather than saying ‘have you recently…’ say ‘in the last month have you…’

Introducing your survey

A good introduction can help increase the completion rate, so it’s important to clearly introduce your survey. Your introduction should help establish credibility, build trust and ensure only relevant respondents complete it. The introduction should include:

• a brief summary of your research
• the aim of the survey, how you’ll use their answers and why their insight is important to your research
• the length of the survey
• if responses will be anonymous and confidential
• a thank you message

If you plan on asking sensitive or personal questions, it’s important to mention this at the start and explain how responses will help you. This will help people prepare to share or opt out if they don’t want to.

Keep it short and simple

Longer surveys can result in lower completion rates and lower quality data, so keep respondents engaged by:

• taking out any unnecessary questions and any ‘nice to have’ questions
• keeping your survey to a maximum of 15 questions
• using open ended questions wisely, as they will take respondents longer to respond to and take you longer to analyse

Presentation matters

Having a cluttered and text heavy survey may overwhelm respondents, so they might not complete your survey. Make your survey inviting by using colour, white space and, if appropriate, pictures to break up text and questions. It’s also useful to add a progress bar, so respondents can see how many questions are left.

Check grammar and spelling

Make sure to check spelling, grammar, punctuation and readability. This will reassure respondents that the survey is credible and that the data will be used for professionally.

Set your goals

Be clear about your timeframes and how much data you need so you know when to end your survey. For example, you may want to end the survey on 1 January or when you get 25 responses – whichever comes first. We usually recommend keeping the survey open for at least a month.

Anonymity and confidentiality

Consider whether respondents will have the option to remain anonymous. Anonymity can help people feel more comfortable sharing their views and opinions. If you’d like to capture respondents' contact details, you should ask for them at the end of your survey. It’s particularly useful to capture contact details if you’d like to involve respondents in other stages of your research.

You’ll also need to ensure you’re compliant with General Data Protection Regulation (GDPR) [EU] in the collection, storing, sharing and disposal of information collected.

Test your survey

Don’t forget, you can also involve patients to help you write your survey. This is a great technique if you’re aiming to reach a wider group of patients. This will help ensure your survey is easy to understand, impactful and that you’re asking questions patients can answer and that are important to them.

You can also test your survey with staff or friends not in your research team to make sure it’s easy to understand and complete. They can also help test how long the survey takes to complete.

Be objective

When you are talking with patients they shouldn’t be able to tell your opinion. Make sure questions are balanced, neutral, and don’t lead respondents to answer in a certain way. You should watch out for any non-verbal cues that might make the patients feel like you approve or disapprove of certain answers.

Be aware of groupthink

Groupthink is where individuals agree with the groups’ or individuals’ opinions, often without voicing their own. This is sometimes an effort to maintain harmony or a belief that their idea isn’t as good. There are many techniques to help you avoid groupthink. One is asking people to write their ideas on post it notes individually before sharing with the wider group for discussion.

Be sensitive

The people you involve may be very close to their cancer experience. Sensitivity is important. Try not to ask questions that have risk of offending or upsetting your audience. It can be difficult discussing cancer experiences so it’s important all attendees and staff are being sensitive and respectful to each other’s personal experiences.