Unit 2 · Lesson 7 of 20
Separate the symptom from its causes
Compare plausible explanations and choose evidence that separates them.
01 / Learn
Learn the method
Fictional case
Mira asks reception whether her check-in worked after the kiosk displays Done.
Terms for this lesson
- Symptom
- The visible difficulty a team wants to explain.
- Hypothesis
- A possible explanation that predicts an observable signal.
- Discriminating check
- Evidence that helps choose among competing explanations.
A symptom is the event the team wants to explain. Mira's question at reception is a symptom. Several causes could contribute: the Done message may leave its meaning unclear, the queue may update after she reaches reception, or a nearby sign may suggest that staff confirmation is required. Each explanation points toward a different change.
Turn the explanations into hypotheses. A hypothesis states a possible cause and predicts an observable signal. If wording causes uncertainty, Mira's explanation may focus on the message. If the queue delay causes it, her question may follow a staff screen that still shows an earlier state. A sign-related cause can be checked by observing where she looks and asking about her expectation.
Choose a discriminating check for each pair of explanations. Combine a kiosk recording, queue timestamp, and short follow-up question. The aim is to find evidence that changes the team's confidence among the explanations. A cause map keeps the symptom, candidate causes, and checks on one page.
A provisional status can still help while research continues. It can describe the receipt event and direct Mira to the waiting area, while staff use a current lookup when she asks. The final design decision follows the evidence about the main mechanism.
Use the method
- Write the symptom as an observed event.
- List at least two plausible causes and their predicted signals.
- Choose a check that changes confidence among the causes.
A visible symptom can arise from interface language, system timing, or the wider setting.
- GOV.UK Service Manual: Analyse a research session ↗Move from notes to observations, findings, and actions in clear steps.
- Nielsen Norman Group: Triangulation: Get Better Research Results by Using Multiple UX Methods ↗Use different evidence types to test an explanation's limits.
- Nielsen Norman Group: Problem Statements in UX Discovery ↗Frame affected people, the problem, and consequence while leaving unproven causes open.
02 / See the reasoning
Worked example
The team compares a video with logs. Mira reads Done and asks reception within 20 seconds. The service receipt arrived within three seconds, and the queue refreshed 70 seconds later. Mira says she expected reception to call her name on screen. These facts support a delayed staff view and an expectation about confirmation. The team adds a check of the lobby signage before choosing the final message.
Try a decision →03 / Decide
Guided decision
Which next check would best separate an unclear kiosk message from a delayed staff queue?
Feedback on your choice
What could change the answer: A queue timestamp that precedes Mira's question would increase attention on wording, signage, and expectation.
Use this feedback to examine your choice and written reasoning.
Build your case file →04 / Build the case
Your case file
Fictional case update
Another patient reads a wall sign saying 'Tell reception when you arrive' and then approaches the kiosk.
Evidence and constraints
- Fictional observation: Mira asks reception 20 seconds after reading Done.
- Fictional log: her receipt arrives three seconds after submission; the staff queue displays it 70 seconds later.
- Fictional participant report: Mira expected reception to call her name on screen.
- Fictional setting: one entrance sign directs arriving patients to reception.
Add the wall sign as a candidate influence in the cause map. Write an observation and a follow-up question that would clarify its role.
Create: A cause map with symptom, three hypotheses, supporting facts, and discriminating checks.
The cause map keeps observed facts distinct from explanations and selects checks that can shift the decision.
- Each hypothesis predicts an observable signal.
- Every case fact has a clear evidence type.
- The next checks distinguish among the candidate causes.
The cause map makes uncertainty visible. The next lesson frames the supported problem for design.
Lesson complete
You added to your case file.
You studied the method, practiced a decision, and added an artifact to your case file. Carry that reasoning into the next lesson.
05 / Keep working