Singapore UX portfolio critique: keep a small study’s claim within its evidence

The problem

A Singapore UX researcher has a small portfolio study and needs help judging whether the conclusion is narrower than the recommendation they wrote.

Write the intended claim before debating the sample size

Copy the conclusion your portfolio currently asks a reader to accept. Identify the people, situation and behaviour it describes. “Visitors understand booking status” reaches further than an observation about a few people in one exercise. Put the research question beside the claim and ask whether they have the same scope. A small study can expose a useful difficulty, but the number of participants alone does not tell a reviewer who they represent or what happened. Begin with what your evidence can establish rather than searching for a minimum number that would make any conclusion defensible. This guide supplies a review method, not a statistical sampling rule or a population estimate.

Show the recruitment route and the conditions of the task

Record intended audience, people actually involved, how they were invited, prior familiarity, task, starting material and assistance supplied. DesignSingapore’s customer-research explanation discusses deciding whom to interview and what to learn. That narrow context supports examining those choices; it does not prescribe a sample size or make colleagues representative of customers. Keep relevant differences visible. Someone familiar with the existing service may interpret a term a first-time visitor has never seen. A guided exercise and an unaided attempt can supply different observations. If the record omits a condition, mark it unknown rather than assuming everyone received the same context or completed the task independently.

• DesignSingapore: understanding customers through design ↗ · Source checked 20 October 2026

Work through one fully fictional four-person case

This is an invented study, not research conducted with Singapore residents. A researcher wants to understand how first-time visitors interpret a room-booking status, but invites four colleagues who already use the imagined service. Their task is to identify whether a selected room can be booked from the status information on a supplied screen. In this fictional screen, “hold” means unavailable while an earlier request is checked; completion means correctly identifying the room as unavailable in this exercise. Two colleagues ask what “hold” means and receive an explanation before finishing. A third rereads the screen and finishes without help; the fourth finishes without asking. All four finish, but only two do so without the recorded explanation. The example defines these observations for discussion. It does not show that the explanation caused completion, that the two asking could never succeed alone or that first-time visitors would behave the same way.

Compare a supported sentence with a stronger one

For that fictional record, a bounded sentence is: “Two experienced colleagues asked for the meaning of hold during the supplied task; an explanation was given before they finished.” Another accurate note is that all four completed under the described conditions, including assistance to two. “Everyone could complete unaided” removes that assistance. “First-time residents understand every booking status” changes both the studied people and the task. Keep these candidate sentences beside the actual record so a reviewer can identify which step adds unsupported meaning. Do not turn a fraction from the invented group into an estimate of all users. Narrowing the sentence preserves the useful observation instead of pretending the study established nothing.

Identify the missing comparison that matters to the decision

Make three columns: what was observed; what the proposed claim adds; and evidence still absent. In the fictional case, first-time visitors and their unaided interpretation were not observed. That gap matters if the claim is about their understanding. It does not erase what the four colleagues asked. Ask the reviewer which missing condition is most relevant to the actual research question, rather than promising to recruit an arbitrary larger group. A larger convenience sample with the same familiarity and assistance may leave the same scope problem. Plan further work through the appropriate research and permission process. A public critique can improve the question; its attendees do not automatically become the people the study needs.

Keep the proposed design action separate from the finding

A team may choose to investigate a clearer status explanation because the recorded questions matter to its work. Record that as a proposal or provisional design decision, together with its rationale and unresolved evidence. It is different from claiming the revised interface has improved bookings. In the invented case, preparing a possible explanation for later review is a bounded next action. Its effectiveness still needs the appropriate task evidence; the original observations cannot establish a result for a screen nobody tested. If you later examine another version, identify which version, people, task and assistance produced its observations. Do not silently combine the two exercises into one more impressive result.

Bring a claim-and-support card to the right critique

Prepare six fields: proposed claim; intended people and task; actual recruitment; observed outcomes with assistance; condition missing; and design decision under consideration. Add the origin and sharing status of the extract. Ask a focused question: “Does this sentence stay within the people and conditions shown, and what evidence would the broader version still need?” Confirm that the Singapore session accepts a methods question and the material you can show. A visual portfolio review may help organise the case without examining recruitment. Use fictional practice when real records cannot be shared, labelling it clearly; do not invent actual quotations or make the polished card appear to be a completed study.

Preserve the limits when revising the portfolio

After critique, keep the earlier claim, the revised sentence, the reason for narrowing it, the unchanged evidence and the next question. In the fictional case, the revised note retains the two requests for explanation and the familiar-colleague context, while the first-time visitor question stays open. That is a clearer account, not new research. If feedback concerns an inferred motivation, use the journey-map guide; if it concerns completing a prototype versus wanting the service, use the usability-or-demand guide. Use Boardzip’s free design search for a suitable review and check its actual participation terms. Search access does not supply a methods reviewer or a representative sample. Keep the evidence gap visible when no verified session can examine it.