What: Observation is the ‘mother’ of all Human Centred Design techniques. It is a method based on observing users or stakeholders while they use a product or service or carry out an action. There are many variants, similar and yet different.
When: You practice observation when you want to know ‘how’ people use products/services.
Advantages: Observation is one of the most straightforward ways to gather insights. Even though anyone can ‘look’, observation is not simple, as you must also want and be able to ‘see’.
There are many observation techniques. Your choice will depend on the type of information you wish to obtain from the user. Generally, there are four major variables:
1. The purpose of the observation: Why are you observing? Are you aiming to get an answer to a specific question, or do you want to obtain general insights? The more concrete the description of the purpose of the observation, the more you have to prepare and execute it in a structured way.
2. The context of the observation: In many cases, the context has a major impact on a person’s behaviour. It is not always easy to observe people in their private life. Solutions may involve allowing the subjects to record the observations themselves; alternatively, you can place people in a context (test set-up) that you have devised. Both approaches influence behaviour and therefore result, but this can be applied perfectly in more concrete research questions.
3. The researcher’s mode of participation: Is the researcher observing without intervening actively, or is he/she approaching, directing and questioning the subjects actively? Did the researcher draft a scenario and briefing for the observation beforehand?
4. The method of involvement of the ‘subjects’: There are techniques requiring a certain effort from the subject as well: to formulate his/her ideas and feelings during the observation, for example; or by helping to evaluate and interpret the results of the observation afterwards.
Some common observation techniques
• Fly on the wall observation –This is the purest form of observation: the study of a user in his/her natural context, with no interference from the researcher.
• Shadowing – ‘Tailing’ of a user in his/her normal environment. The fact that you are visibly present will have an unintentional effect on the behaviour.
• Lab test – Observation in an unnatural context, test set-up. The context influences behaviour.
• Think Aloud – The subject is asked to comment on his/her actions. What he/she does, thinks and feels is expressed out loud and recorded.
• Participatory observation – The researcher approaches the subject actively, asks questions, and may even direct the subject.
What the term also involves: the techniques used should always boil down to observation or participatory observation.
- Think carefully about the purpose of the observation and put it on paper so that everyone involved also has a good understanding. What do you want to know and who are the most suitable stakeholders to achieve this? Generally speaking, what is the context of the problem?
- Define an approach for the test and the technique you wish to apply.
- Think carefully about the legal restrictions concerning privacy. It is not legal to start filming complete strangers in private contexts.
- Also think about the way you want to document the observation (notes, photographs, videos?).
- Draft an observation protocol including the scenario for carrying out the observation, as well as the agreed behaviour for the persons conducting the observation. Create forms to fill out, checklists, etc.
- Define the practicalities regarding location, test subjects, logistics, any remuneration for participants, etc. An NDA (non disclosure agreement) can be useful in strategic projects.
During a pure observation, you study a person in his/her own enviornment. You also avoid any kind of interaction between you and your subject so as not to affect the results. The idea is that the subject should get used to you to the point when he/she begins to ignore you. This requires trust, and you can speed things up by explaining yourself and your intentions before the observation begins. Keep in mind that your behaviour as a researcher also has an impact on results. Be a ‘fly on the wall’, also with your technical tools.
Immediately upon concluding the observation – while details are still fresh in your memory –, take down all the data that comes to your mind. Fill in the sections left blank between the already noted details, entering as much information as possible.
Important: at this stage, be curious, objective and delay your judgment: note EVERYTHING, even if it does not seem important. Things that may seem trivial may suddenly offer you breakthrough insights during the processing (analysis) of the results.
Overall, the approach followed in participatory observation is similar, except that the researcher has a more active and inquiring role in the script to make the observation as complete as possible.
Participatory observation requires even more preparation and a more extensive list of what you might be able to query than a standard observation. Questionnaires drawn up beforehand can be useful, but before you know it you may find yourself interviewing instead of observing.
- Prepare a discreet checklist and procedure.
- Introduce yourself to the subject and explain why you are observing him/her.
- Give the subject space to carry out the tasks quietly at his/her own pace.
- Ask the subject why he/she does something in a certain way, ask until you find out the real reason (the 5x Why technique can be useful here: keep asking ‘why’ until you get a deeper understanding).
- Anticipate what you see, be spontaneous and curious and only fall back on a possible list if the conversation threatens to die out.
Analysis of the results
Once concluded, an observation leaves you with all kinds of data in various forms. Start by reducing the information to a common denominator, e.g. in an affinity diagram.
Use your best audio and video clips to report and communicate on the research and its results. Design is communication.