Developing research questions/frameworks

Research should start with questions, followed by decisions on which methods can best answer these questions. Research questions are refined through discussions with fellow researchers, reading up on the existing literature and considering what might be practical and feasible with your current resources.

Research questions can either be open-ended (exploratory) or assumption testing (confirmatory). You’re aiming for a clear, concise question by the end of your refining exercise so that it is clear what the study is aiming to achieve.

Types of questions

Open-ended research questions: Open-ended research questions are suited to research areas that are new, where we know little about the subject, and exploring complex human experiences, systems or processes.

In our experience, open research questions have provided a good platform to freely explore people’s lived experience of a particular health condition. For example, in our insomnia research our open-ended research questions were:

• What do people’s everyday experiences with insomnia look like?
• What factors influence the way people treat their insomnia?

These questions gave us (as researchers) a frame from which we could broadly explore participants’ daily experiences of insomnia and their treatment.

Assumption-testing research questions: Assumption-testing research questions are suited to research where we already have a specific focus, e.g. medication routines and testing certain features in a product development cycle.

The advantages of assumption testing questions are that they give a focus to the research project. It lets you know immediately whether what you are studying meets or does not meet your predictions.

For example, in a user research study where designers are testing a messaging feature in an app, researchers have made the assumption that people who are depressed would like to receive motivational messages every day to remind them to take their medication. This assumption might have originated from the research literature or previous observational studies.

Our assumption testing questions for this were:

• Do patients with depression find motivational messages helpful in remembering to take their medication?
• Do participants in a clinical trial find the first month of using an app more challenging compared to the sixth month?

Research methodologies

Now that you know what you’re looking to explore or find out in your research, you need to choose which method will be best placed to give you these answers.

These are some of the methods we use:

• Interview: a key methodology that allows you to dig around a topic area. Asking the right questions in the right way is crucial. We look at this in more detail below under “Framing Questions”.
• Observation: what people say doesn’t always reflect what they actually do. They may be unaware of some of their behaviours or feel bad about them (when asked about weekly alcohol consumption, for example, people tend to underestimate). Therefore, observing behaviour can be very revealing. This can involve spending time with an individual to see how they go about daily tasks (with a physical impairment, for instance) or observing a group in a particular situation (using public transport, perhaps).
• Digital ethnography: this is a way of gathering real-time live insight into people’s lives, capturing their experiences in the moment as text, images, audio or video, for example, before they might forget. You can use tools or apps specifically designed for this purpose (we use dscout), or harness the fact that many people already use social platforms or channels such as Whatsapp or Facebook, and work with those.
• Stimuli: it can be hard for people to respond to abstract concepts in an interview, or to think about the way they do things when they haven’t thought about it much before. It can be easier to get a stronger, more focussed response when there is something more concrete to respond to, a mock up of a tool, for example. We’ll look at this in more detail below under “Bringing stimulus and prototypes into the discussion”.
• Surveys: if you need to reach a large number of people at a distance, or on a particular platform, you could use a survey. These are not good for generating rich qualitative data and tend to be skewed to the most active and engaged users (offering an appealing incentive helps with this) but can be useful for quantitative data, particularly at a preliminary stage.
• Passive data monitoring: Establishing a back-end platform aggregating data from a digital prototype being tested in order to use interaction metrics as research data (e.g. amount of time spent on app, time of day of interactions, etc.).
• Co-creating: Offering participants the chance to apply their ideas to prototypes as they are being developed, and contribute to possibilities for future designs.
• Participatory decision-making: A process by which participants are involved in key points of a project to contribute to decisions about next steps. This can take the form of workshops, which need to be carefully facilitated to ensure that roles, expectations and outcomes are clearly managed.
• Situated devices: Placing a physical device, which could be a button, tablet or other interface in a relevant location to the research group, and which invites people who pass by to answer simple survey questions through interactive features.

Framing questions

If you decide to carry out interviews as part of your research, you’ll need to think about framing interview questions. The way you ask questions is important, and needs to be well planned. These questions should be driven by the primary and secondary research question that you’ve developed, but they should not be the same questions. For example, if your research question is: “Could technology improve the quality of life for dementia patients?” it wouldn’t be helpful to ask this question directly. Instead you might ask questions that aim to find out more about the frustrations that dementia patients and their carers experience.

Bringing stimulus and prototypes into the discussion

Giving users something concrete and visual to respond to can make a big difference in the quality of responses that you get. In deciding what stimulus to use its useful to disconnect this from what the final product or service needs to be. We sometimes refer to prototypes that are just used for research as being ‘sacrificial prototypes’ created for the sole purpose of being used to stimulate research.

Prototypes can take many forms. Here are some examples of prototypes we’ve used in research:

• Rich picture maps - showing how processes and services fit together and allowing participants to scribble and draw idealised journeys and flows.
• Click-through prototypes - helping participants have a basic interaction with key functions. It helps to let users handle technology and try and use it to complete tasks.

Principles for recording data

Field notes:

• Taking field notes by hand rather than digitally is important during face-to-face interviews in order to build rapport. These then need to be digitalised for analysis.
• To facilitate note taking in the field, a template is useful. This should be organised according to the interview structure for easy note taking and also provide space to record timestamps that correspond with a recording of the interview. The timestamps allow listening and clarifying the notes later on.
• For remote interviews digital notes are time-efficient.

Interview transcripts:

• Full interview transcripts can be valuable for explorative projects but are very resource intensive.
• If the researcher transcribes themselves it can be a useful exercise to go through the research activity again and familiarise themselves with the data before analysis.