A PM asks you to improve onboarding before planning in 10 days, but has not named a user problem. How do you turn that request into a research question?

I would frame the study around where new users fail to reach first value and what causes those breakdowns. - I would ask the PM which onboarding decision is still reversible, such as removing a step or changing guidance, and record it in a one-page research brief. - I would define the audience as people who first used the product in the last 30 days and examine one recent onboarding attempt rather than opinions about an imagined redesign. - I would have the research lead review the scope because a 10-day study can explain specific barriers, not solve the broad goal of improving onboarding.

A designer asks whether users like a new billing dashboard, and the team must decide in one week whether to revise it. What would you research instead?

I would replace liking with observable comprehension and task performance tied to the revision decision. - I would ask whether six target users can identify an overdue invoice, explain the balance, and start the intended follow-up without assistance. - I would define success and critical errors before sessions, then use neutral follow-up questions to learn why labels or totals were misunderstood. - I would report usability evidence about this dashboard version, not claim that six participants represent overall customer preference.

The team has two weeks to decide whether to explore a budgeting feature, but it knows little about how customers currently plan monthly spending. Which method would you choose?

I would run a small generative interview study about recent budgeting behavior before testing a feature concept. - I would recruit about eight relevant customers and ask each to reconstruct the last time they planned or adjusted a monthly budget using their real tools. - I would collect workflows, triggers, workarounds, and unmet needs in a discussion guide and synthesize them into opportunity themes. - Eight interviews can reveal patterns and vocabulary for the next decision, but they cannot estimate how prevalent each need is.

A PM must decide whether to prioritize bulk export and asks how many of 20,000 active customers export data weekly. Would you use qualitative or quantitative research?

I would use quantitative behavioral data first because the decision depends on prevalence at scale. - I would work with an analyst to define an export event, the active-customer denominator, and a weekly observation window before querying usage. - I would segment the rate by relevant account type and check event quality so failed exports are not counted as successful use. - If the team later needs to know why customers export or avoid it, I would add interviews, but qualitative sessions alone cannot produce the weekly proportion.

A team can recruit 12 participants before choosing between two meal-planning concepts, but it is unsure whether the underlying problem matters. How do generative and evaluative research differ here?

I would use generative research to understand the problem first and reserve evaluative research for concepts that address a supported need. - Early interviews would examine recent meal-planning situations, constraints, and workarounds without showing either concept at the start. - If a clear need emerges, later sessions would give participants realistic tasks with the two concepts and compare comprehension and breakdowns. - The 12 participants provide qualitative evidence for direction, not a statistically significant winner or a market preference percentage.

You have six participants available before release and product metrics due 30 days after release. How would you separate formative and summative research on a new checkout flow?

I would use the pre-release sessions formatively to find fixable problems and the post-release measures summatively to assess performance against defined targets. - Before release, I would test realistic checkout tasks, document breakdowns, and feed prioritized findings into the next design iteration. - After release, I would compare predefined measures such as completion and abandonment with a baseline using an adequate sample and analytics support. - The six formative sessions can expose failure modes, but they cannot establish the release-wide completion rate or statistical significance.

Analytics shows an 18 percent drop at identity verification, and the team has 12 days to decide whether to change instructions or the flow. How would you combine qualitative and quantitative evidence?

I would use the funnel to locate and size the drop, then moderated sessions to investigate why it happens. - I would verify the event sequence and segment the 18 percent drop by device, region, and error code with an analyst. - I would recruit six to eight affected users for a realistic verification task and probe their interpretation of instructions, trust concerns, and technical failures. - I would triangulate both sources in the brief, keeping observed explanations separate from their unknown prevalence unless the product data measures it.

You have five business days and a clickable prototype to decide whether a new refund flow is understandable before engineering starts. What bounded study would you propose?

I would propose a rapid formative moderated usability test with about six target users. - I would define two realistic refund tasks, observable success criteria, and critical errors with the researcher and PM before recruiting. - I would pilot the prototype on day one, run sessions over two days, and deliver a concise findings brief with clips and severity by day five. - The study can identify comprehension and interaction problems in the prototype, but it cannot predict the production refund rate.

Four of five interview participants preferred concept A, and a stakeholder says this proves 80 percent of customers want it. How do you respond before tomorrow’s review?

I would correct the claim and present the result as qualitative evidence from four of five participants, not an 80 percent population estimate. - I would show what participants understood, valued, or struggled with in each concept and attach the relevant notes or clips. - I would state how the five people were recruited and which customer groups were not represented. - If prevalence will determine the decision, I would recommend a properly sampled quantitative follow-up rather than attaching statistical significance to five sessions.

In the eighth generative interview, a new cancellation reason appears, but the team says the study has reached saturation and must stop today. What would you do?

I would not claim guaranteed saturation because the new theme shows that relevant variation may remain. - I would check whether the eighth participant differs on a planned dimension such as tenure, plan, or cancellation history and update the theme matrix. - I would tell the lead what is stable, what is still changing, and whether one or two targeted sessions fit the decision deadline. - If recruitment must stop, I would document the contextual sample and report the new reason as an emerging finding, not a complete taxonomy of cancellation.

You need to interview eight first-time sellers in two weeks so the team can decide which listing setup problem to address. How would you build the interview guide?

I would organize the guide around a recent listing attempt and the decision the team needs to make. - I would start with context, then ask for a step-by-step reconstruction of the last listing, including triggers, tools, interruptions, and workarounds. - I would place broad behavior questions before targeted probes about known setup issues so the team’s assumptions do not frame every answer. - I would create a 45-minute guide with timing, neutral probes, and a closing question, then have the research lead review and pilot it.

During a 45-minute interview, a participant says the reporting tool was confusing, and you have only five minutes left. What neutral probes do you use?

I would ask for the specific episode and let the participant define what confusing meant. - I would say, “Tell me about the last time that happened,” then ask what they were trying to do and what they expected next. - I would follow observable details with “What did you do then?” and “What made you choose that?” rather than offering possible causes. - I would capture any unanswered branch in the notes instead of rushing through leading yes-or-no questions in the final minutes.

A guide for six interviews asks, “How helpful was the simple automatic reminder?” What would you change before sessions start tomorrow?

I would remove the positive and simplifying assumptions and ask about the participant’s experience with the reminder. - I would rewrite it as, “What, if anything, did you notice about reminders?” followed by “What happened after you saw one?” - If the reminder was not noticed, I would record that outcome rather than explain the feature and force an opinion. - I would scan the full guide for words such as easy, useful, and better, then ask the research lead to approve the revised version.

The team will interview 10 people who recently switched expense tools and must decide which migration problem to address. How would you use a Jobs-to-Be-Done lens?

I would reconstruct the circumstances and progress behind the recent switch rather than ask for desired features. - I would probe the trigger, prior workaround, alternatives considered, anxieties, and outcome that made each person choose a new tool. - I would map those episodes into job statements based on situation, motivation, and expected progress, supported by participant evidence. - I would treat the map as a qualitative framing artifact for the migration decision, not as proof of how common each job is among all customers.

Three stakeholders want to observe each of six moderated interviews, and participants must feel comfortable discussing financial mistakes. How would you set observer roles?

I would keep observers silent and out of the participant’s direct view while making their responsibilities explicit. - I would assign one person to capture behavior, one to capture exact quotes, and one to log follow-up questions in a private channel for the moderator. - I would brief observers not to message the moderator constantly, react on camera, or contact participants after the session. - I would disclose observation during consent and let participants continue without observers if that is the approved option.

After two of eight interviews support the PM’s hypothesis, you notice you are probing that theme more than alternatives. What do you do before interview three?

I would reset the guide and make the emerging hypothesis one probe among several rather than the center of later sessions. - I would compare the first two transcripts with the guide, mark where my follow-ups became uneven, and tell the research lead about the risk. - I would use the same core questions and balanced probes for the remaining participants so coverage is comparable. - I would keep a reflexivity note describing the expectation and separate participant evidence from my interpretation during synthesis.

You are moderating six remote interviews alone, automatic transcripts arrive the next day, and the readout is due in four days. How do you take notes without losing rapport?

I would take light, structured notes during sessions and use recordings and transcripts for detailed evidence afterward. - I would prepare a note template with timestamps, observed behavior, key quotes, and follow-up flags aligned to the research questions. - During the interview I would write short markers and maintain eye contact instead of trying to transcribe every sentence. - I would verify important quotes against the recording, correct transcript errors, and avoid treating an automatic transcript as fully accurate.

A 45-minute guide for eight interviews contains 28 main questions, and the first session is tomorrow. How would you use a pilot?

I would pilot immediately and cut the guide to the questions needed for the product decision. - I would run the guide with one representative or near-representative person, timing each section and noting confusing wording or missing probes. - I would prioritize recent behavior and critical topics, moving optional questions to a clearly marked reserve section. - I would document pilot changes and confirm with the research lead whether the pilot data is comparable enough to include in analysis.

In interview four of eight, an observer asks you to show a new concept and ask whether the participant would use it, but the approved study is generative. What do you do?

I would keep the session within the approved generative scope and not introduce the concept mid-study. - I would tell the observer privately that showing it now would change later sessions and invite hypothetical feature feedback instead of comparable behavior evidence. - I would log the concept question as a follow-up study request and clarify the product decision, method, and consent needs after the session. - If the research lead approves a guide change, I would document which participants saw it and analyze that evidence separately.

Six interviews are split: three users rely on templates and three deliberately avoid them, with a roadmap review in two days. How would you report the result?

I would report two observed behavior patterns and the conditions behind them rather than force one average finding. - I would build a small matrix comparing experience, task frequency, trigger, and reason for using or avoiding templates across all six participants. - I would attach representative evidence and explain which roadmap option each pattern supports or puts at risk. - I would state that the split does not estimate a 50/50 population share and recommend quantitative follow-up only if prevalence changes the decision.

UX Researcher interview questions

100 real questions with model answers and explanations for Junior candidates.

See a UX Researcher resume example →

Practice with flashcards

Spaced repetition · Hunter Pass

Questions

research-questionsonboarding

I would frame the study around where new users fail to reach first value and what causes those breakdowns.

I would ask the PM which onboarding decision is still reversible, such as removing a step or changing guidance, and record it in a one-page research brief.
I would define the audience as people who first used the product in the last 30 days and examine one recent onboarding attempt rather than opinions about an imagined redesign.
I would have the research lead review the scope because a 10-day study can explain specific barriers, not solve the broad goal of improving onboarding.

Why interviewers ask this: The interviewer is checking whether you can convert a vague product request into a bounded, decision-linked research question.

design

I would replace liking with observable comprehension and task performance tied to the revision decision.

I would ask whether six target users can identify an overdue invoice, explain the balance, and start the intended follow-up without assistance.
I would define success and critical errors before sessions, then use neutral follow-up questions to learn why labels or totals were misunderstood.
I would report usability evidence about this dashboard version, not claim that six participants represent overall customer preference.

Why interviewers ask this: A strong answer turns a subjective request into evaluative questions with observable criteria and explicit limits.

I would run a small generative interview study about recent budgeting behavior before testing a feature concept.

I would recruit about eight relevant customers and ask each to reconstruct the last time they planned or adjusted a monthly budget using their real tools.
I would collect workflows, triggers, workarounds, and unmet needs in a discussion guide and synthesize them into opportunity themes.
Eight interviews can reveal patterns and vocabulary for the next decision, but they cannot estimate how prevalent each need is.

Why interviewers ask this: The interviewer wants method choice to follow the team’s knowledge gap rather than the availability of a prototype.

qualitative-researchquantitative-researchprioritization

I would use quantitative behavioral data first because the decision depends on prevalence at scale.

I would work with an analyst to define an export event, the active-customer denominator, and a weekly observation window before querying usage.
I would segment the rate by relevant account type and check event quality so failed exports are not counted as successful use.
If the team later needs to know why customers export or avoid it, I would add interviews, but qualitative sessions alone cannot produce the weekly proportion.

Why interviewers ask this: The answer should distinguish a quantitative frequency question from a qualitative explanation question.

evaluative-researchresearch-participants

I would use generative research to understand the problem first and reserve evaluative research for concepts that address a supported need.

Early interviews would examine recent meal-planning situations, constraints, and workarounds without showing either concept at the start.
If a clear need emerges, later sessions would give participants realistic tasks with the two concepts and compare comprehension and breakdowns.
The 12 participants provide qualitative evidence for direction, not a statistically significant winner or a market preference percentage.

Why interviewers ask this: The interviewer is checking whether you know that generative work explores problems while evaluative work examines proposed solutions.

formative-researchsummative-researchresearch-participants

I would use the pre-release sessions formatively to find fixable problems and the post-release measures summatively to assess performance against defined targets.

Before release, I would test realistic checkout tasks, document breakdowns, and feed prioritized findings into the next design iteration.
After release, I would compare predefined measures such as completion and abandonment with a baseline using an adequate sample and analytics support.
The six formative sessions can expose failure modes, but they cannot establish the release-wide completion rate or statistical significance.

Why interviewers ask this: A strong answer links formative work to iteration and summative work to judging an outcome against criteria.

qualitative-research

I would use the funnel to locate and size the drop, then moderated sessions to investigate why it happens.

I would verify the event sequence and segment the 18 percent drop by device, region, and error code with an analyst.
I would recruit six to eight affected users for a realistic verification task and probe their interpretation of instructions, trust concerns, and technical failures.
I would triangulate both sources in the brief, keeping observed explanations separate from their unknown prevalence unless the product data measures it.

Why interviewers ask this: The interviewer is evaluating whether you can combine scale from analytics with explanatory depth from qualitative research.

prototypes

I would propose a rapid formative moderated usability test with about six target users.

I would define two realistic refund tasks, observable success criteria, and critical errors with the researcher and PM before recruiting.
I would pilot the prototype on day one, run sessions over two days, and deliver a concise findings brief with clips and severity by day five.
The study can identify comprehension and interaction problems in the prototype, but it cannot predict the production refund rate.

Why interviewers ask this: The answer should fit the method, sample, artifacts, and claims to a short pre-build decision window.

research-participantscommunicationstakeholder-management

I would correct the claim and present the result as qualitative evidence from four of five participants, not an 80 percent population estimate.

I would show what participants understood, valued, or struggled with in each concept and attach the relevant notes or clips.
I would state how the five people were recruited and which customer groups were not represented.
If prevalence will determine the decision, I would recommend a properly sampled quantitative follow-up rather than attaching statistical significance to five sessions.

Why interviewers ask this: The interviewer is checking whether you resist false precision and preserve the useful qualitative evidence.

generative-interviewssaturationresilience

I would not claim guaranteed saturation because the new theme shows that relevant variation may remain.

I would check whether the eighth participant differs on a planned dimension such as tenure, plan, or cancellation history and update the theme matrix.
I would tell the lead what is stable, what is still changing, and whether one or two targeted sessions fit the decision deadline.
If recruitment must stop, I would document the contextual sample and report the new reason as an emerging finding, not a complete taxonomy of cancellation.

Why interviewers ask this: A strong answer treats saturation as contextual evidence, not an automatic threshold reached at a fixed sample size.

interview-guides

I would organize the guide around a recent listing attempt and the decision the team needs to make.

I would start with context, then ask for a step-by-step reconstruction of the last listing, including triggers, tools, interruptions, and workarounds.
I would place broad behavior questions before targeted probes about known setup issues so the team’s assumptions do not frame every answer.
I would create a 45-minute guide with timing, neutral probes, and a closing question, then have the research lead review and pilot it.

Why interviewers ask this: The interviewer wants to see a guide grounded in recent behavior, neutral sequencing, and a bounded product decision.

interview-probingresearch-participants

I would ask for the specific episode and let the participant define what confusing meant.

I would say, “Tell me about the last time that happened,” then ask what they were trying to do and what they expected next.
I would follow observable details with “What did you do then?” and “What made you choose that?” rather than offering possible causes.
I would capture any unanswered branch in the notes instead of rushing through leading yes-or-no questions in the final minutes.

Why interviewers ask this: A strong junior moderator obtains concrete behavior without putting an explanation into the participant’s mouth.

sessions

I would remove the positive and simplifying assumptions and ask about the participant’s experience with the reminder.

I would rewrite it as, “What, if anything, did you notice about reminders?” followed by “What happened after you saw one?”
If the reminder was not noticed, I would record that outcome rather than explain the feature and force an opinion.
I would scan the full guide for words such as easy, useful, and better, then ask the research lead to approve the revised version.

Why interviewers ask this: The interviewer is checking whether you can identify leading language and replace it with neutral, behavior-based prompts.

discoverymigrations

I would reconstruct the circumstances and progress behind the recent switch rather than ask for desired features.

I would probe the trigger, prior workaround, alternatives considered, anxieties, and outcome that made each person choose a new tool.
I would map those episodes into job statements based on situation, motivation, and expected progress, supported by participant evidence.
I would treat the map as a qualitative framing artifact for the migration decision, not as proof of how common each job is among all customers.

Why interviewers ask this: A strong answer applies Jobs-to-Be-Done to real switching behavior while keeping qualitative claims bounded.

moderated-researchresearch-participantsownership

I would keep observers silent and out of the participant’s direct view while making their responsibilities explicit.

I would assign one person to capture behavior, one to capture exact quotes, and one to log follow-up questions in a private channel for the moderator.
I would brief observers not to message the moderator constantly, react on camera, or contact participants after the session.
I would disclose observation during consent and let participants continue without observers if that is the approved option.

Why interviewers ask this: The interviewer is evaluating whether you can gain stakeholder exposure without compromising moderation or participant comfort.

hypothesis-testing

I would reset the guide and make the emerging hypothesis one probe among several rather than the center of later sessions.

I would compare the first two transcripts with the guide, mark where my follow-ups became uneven, and tell the research lead about the risk.
I would use the same core questions and balanced probes for the remaining participants so coverage is comparable.
I would keep a reflexivity note describing the expectation and separate participant evidence from my interpretation during synthesis.

Why interviewers ask this: A strong answer recognizes moderator bias early and changes the procedure rather than pretending neutrality is automatic.

I would take light, structured notes during sessions and use recordings and transcripts for detailed evidence afterward.

I would prepare a note template with timestamps, observed behavior, key quotes, and follow-up flags aligned to the research questions.
During the interview I would write short markers and maintain eye contact instead of trying to transcribe every sentence.
I would verify important quotes against the recording, correct transcript errors, and avoid treating an automatic transcript as fully accurate.

Why interviewers ask this: The interviewer is checking whether your note-taking supports both good moderation and traceable synthesis under time pressure.

sessions

I would pilot immediately and cut the guide to the questions needed for the product decision.

I would run the guide with one representative or near-representative person, timing each section and noting confusing wording or missing probes.
I would prioritize recent behavior and critical topics, moving optional questions to a clearly marked reserve section.
I would document pilot changes and confirm with the research lead whether the pilot data is comparable enough to include in analysis.

Why interviewers ask this: A strong answer uses a pilot to test timing and clarity rather than treating the first recruited participant as an unplanned rehearsal.

research-participants

I would keep the session within the approved generative scope and not introduce the concept mid-study.

I would tell the observer privately that showing it now would change later sessions and invite hypothetical feature feedback instead of comparable behavior evidence.
I would log the concept question as a follow-up study request and clarify the product decision, method, and consent needs after the session.
If the research lead approves a guide change, I would document which participants saw it and analyze that evidence separately.

Why interviewers ask this: The interviewer is checking whether you protect study consistency while handling a stakeholder request constructively.

roadmap

I would report two observed behavior patterns and the conditions behind them rather than force one average finding.

I would build a small matrix comparing experience, task frequency, trigger, and reason for using or avoiding templates across all six participants.
I would attach representative evidence and explain which roadmap option each pattern supports or puts at risk.
I would state that the split does not estimate a 50/50 population share and recommend quantitative follow-up only if prevalence changes the decision.

Why interviewers ask this: A strong answer preserves meaningful variation and avoids turning a small qualitative split into a market estimate.

Locked questions

21
You have a fragile prototype, six participants, and one week to learn why people fail a multi-step insurance claim. Would you run moderated or unmoderated testing?
moderated-researchunmoderated-researchresearch-participants
22
The team needs results from 40 participants in three days to compare whether two label sets help users find account settings. When is unmoderated testing appropriate?
unmoderated-researchresearch-participantstesting
23
For six usability sessions tomorrow, the task says, “Click Billing and select Download PDF.” How would you rewrite it to test navigation?
usability-testingsessions
24
The team wants accurate time on task for checkout with eight participants, but also asks everyone to think aloud continuously. How do you handle the conflict?
research-participantssoft-skills
25
A pilot participant cannot start task two because the prototype link opens the wrong state, and five paid sessions begin in two hours. What do you do?
research-participantsprototypessessions
26
Before testing password recovery with eight participants, the team says success means “the user seemed to finish.” How would you define task success?
research-participantspasswordstesting
27
In an unmoderated test with 30 participants, mobile users take twice as long as desktop users to find a setting. What do you check before calling mobile worse?
unmoderated-researchresearch-participants
28
During six checkout tests, participants make several detours, and the team asks for an error count by tomorrow. How would you define an error?
research-participantstesting
29
Twelve participants give a prototype a mean SUS score of 68, and a stakeholder calls that 68 percent satisfaction. How do you interpret it?
research-participantsprototypescommunication
30
Four of six participants fail to change their delivery address before payment, and release is in five days. What do you report?
research-participants
31
You need eight active expense-app administrators in 10 days to evaluate approval rules. What would you put in the recruitment screener?
participant-recruitmentrecruitment-screenersdecision-making
32
A study needs six caregivers who booked a medical appointment for someone else in the last 90 days. How do you set inclusion and exclusion criteria?
eligibility-criteria
33
Only about 3 percent of panel members are payroll administrators, but you need six of them within two weeks. How would you recruit this low-incidence audience?
34
For 12 usability sessions, the PM wants both novice and expert inventory managers represented before choosing a default workflow. How would you set quotas?
usability-testingsample-quotassessions
35
You need eight cybersecurity administrators for 60-minute interviews, but the standard $30 incentive produces no qualified sign-ups after five days. What would you change?
participant-incentives
36
Six remote interview participants will be recorded, and two stakeholders will observe. What must happen before the first session tomorrow?
research-participantscommunicationsessions
37
A moderated study includes four screen-reader users next week, and the prototype is only partly accessible. How do you prepare without excluding them?
moderated-researchprototypes
38
A diary study with 10 participants asks for screenshots of medication reminders over seven days. What basic privacy steps would you take?
research-participantsdiary-studies
39
For an eight-person study, 25 screener responses contain identical open-text answers and claim implausible product use. What do you do?
recruitment-screeners
40
Two of eight participants cancel the morning of a three-day usability study, and the readout date cannot move. How do you protect the sample?
usability-testingresearch-participants
41
The team wants a 200-response survey in one week to decide which notification problem to investigate next. How would you define the survey before writing questions?
surveys
42
A survey item for 300 customers asks, “How easy and trustworthy was our improved checkout?” What would you change before launch tomorrow?
surveys
43
A 500-response survey asks age using ranges 18-25, 25-35, 35-45, and 45+, and respondents must answer. What is wrong and how do you fix it?
surveys
44
You have room for one final item in a 250-response cancellation survey, and the team needs reasons it may have missed. Would you use an open or closed question?
surveysresilience
45
The team has 20 participants and no agreed categories for 40 help-center topics. It must propose an information structure in two weeks. Would you use an open card sort?
research-participantscard-sorting
46
A team already has six proposed support categories and can recruit 30 customers before choosing whether topics fit them. Would you run an open or closed card sort?
card-sorting
47
Fifty participants can complete a remote tree test before Friday, and the team must choose between “Plans” and “Billing” as a navigation label. What would you measure?
research-participantstree-testing
48
The team wants 10 commuters to log trip-planning breakdowns for seven days before deciding what to explore next. How would you structure the diary study?
diary-studies
49
By day four of a 14-day diary study, only 6 of 15 participants still submit the required five-minute morning and evening entries. What would you change?
research-participantsdiary-studies
50
A seven-day diary study ends with complete entries from 7 of 12 participants, partial entries from three, and none after onboarding from two. How do you analyze and report compliance?
research-participantsdiary-studiesonboarding
51
You have 8 recorded interviews and automated transcripts that misheard product names in about 1 of every 10 sampled lines; what artifact would you deliver before coding, and what is your exit criterion?
artifacts
52
Two observers took different notes during 6 onboarding interviews, and synthesis starts tomorrow; how would you produce one structured note set with a measurable completion rule?
onboarding
53
You must code 7 interviews about account cancellation, and your first pass produced 34 overlapping labels; what codebook would you create, and when would you stop refining it?
qualitative-codingresilience
54
You and another junior researcher coded 2 of 10 transcripts and agreed on only 61% of excerpts; what artifact and threshold would you use before coding the remaining 8?
artifacts
55
An affinity-mapping session starts with 126 notes from 8 interviews and 45 minutes on the calendar; what board would you create, and what signals that the session is done?
affinity-mappingsessions
56
In 5 usability sessions, your notes say users were confused 17 times but contain no recorded behavior; how would you repair the analysis artifact, and what is the exit criterion?
usability-testingsessionsartifacts
57
You have 9 interviews, 4 tentative clusters, and a request for final themes by Friday; what theme artifact would you produce, and what evidence threshold would you apply?
artifacts
58
Seven of 8 interviewees describe the same renewal problem, while participant P06 reports the opposite; how would you document the negative case, and when is the theme ready?
research-participants
59
A draft from 6 interviews says 70% of users want weekly reports because 4 participants mentioned them; how would you rewrite the finding and audit the 2-page brief?
research-participants
60
From 10 support interviews, one participant reports a rare lockout that blocks payroll for 200 employees; how would you represent frequency and importance in the synthesis, and what closes the review?
research-participants
61
You need to compare 12 claims from 8 interviews before a synthesis review; what evidence table would you build, and what is the minimum completion criterion?
62
Three researchers bring 180 coded excerpts from 9 interviews to a 60-minute synthesis session; how would you create a shared output and decide when to stop?
sessions
63
A synthesis board from 7 interviews contains the statement users hate verification; how would you turn it into one defensible insight statement and define its acceptance rule?
64
A PM challenges 5 findings from 11 interviews and asks to see the source within 10 minutes; what traceability artifact would you prepare, and what makes it complete?
artifacts
65
Eight interviews support 3 findings, and the team asks for recommendations before Tuesday planning; what recommendation artifact would you write, and what is its exit criterion?
artifacts
66
Your 6-participant study has 2 strong patterns, 1 unresolved contradiction, and no mobile participants; how would you report confidence and limitations in a 1-page artifact?
research-participantsartifacts
67
You have 24 hours to turn 9 interviews into a 2-page insight brief for a PM and designer; what sections would you include, and when is the brief ready?
design
68
You have 15 minutes to present findings from 10 interviews to 12 stakeholders; what readout artifact would you prepare, and what exit criterion would you use?
stakeholder-managementcommunicationartifacts
69
A study with 8 participants produces 14 possible findings, but the brief can contain only 5; what prioritization artifact would you use, and when would selection end?
research-participantsprioritizationartifacts
70
After a readout on 7 interviews, Support provides 4 tickets that challenge insight 2; how would you update the research record, and what closes the revision?
71
You are importing 10 interviews and 240 highlights into Dovetail for a 3-person research team; what workspace artifact would you set up, and what is the exit criterion?
artifacts
72
Six moderated sessions are stored in Lookback, and 2 observers need to review task 3 by noon; what artifact would you prepare, and when is it ready?
moderated-researchsessionsartifacts
73
A UserTesting study returns 20 videos, but 3 participants skipped the key task and 2 recordings have no audio; what review artifact and exit rule would you use?
research-participantsartifacts
74
Maze reports results from 32 participants across 4 tasks, and task 2 shows a 41% misclick rate; what evidence artifact would you create before calling it a usability problem?
usability-testingresearch-participantsartifacts
75
An Optimal Workshop card sort has 24 participants, 36 cards, and 5 cards left ungrouped by at least 8 people; what analysis artifact would you deliver, and when is it complete?
research-participantscard-sortingartifacts
76
A repository contains 480 findings from 22 studies and 73 near-duplicate tags; what cleanup artifact would you create for a 2-week junior assignment, and what is the exit criterion?
dedupartifacts
77
Four researchers named clips with labels such as interesting and problem across 9 sessions; what naming artifact would you introduce, and how would you know it works?
sessionsartifacts
78
A Dovetail project for 12 health interviews is visible to 38 employees, although only 6 are approved for raw data; what permissions artifact would you prepare, and when is access fixed?
artifacts
79
From 8 interviews, 3 participants allowed internal clips, 4 allowed notes only, and 1 withdrew consent; what clip register would you create, and what is the release criterion?
research-participantsinformed-consent
80
A PM wants to reuse a 14-month-old finding from 6 interviews for a new checkout market; what reuse assessment would you create, and when may the finding be cited?
81
Only 3 of 6 participants join today's interview block, and the readout is due in 4 days; what recovery artifact would you create, and what exit criterion would you set?
research-participantsjoinsartifacts
82
After 2 of 8 interviews, you learn both participants have never used the required invoicing feature; what incident artifact would you produce, and when would sessions resume?
research-participantsincidentssessions
83
A review of your first 3 of 9 interviews finds 7 questions that suggest the expected answer; what guide revision artifact would you make, and what is the restart threshold?
artifacts
84
During session 4 of 7, a stakeholder interrupts 5 times to explain the product; what session record would you create, and when can the remaining interviews continue?
stakeholder-managementcommunicationsessions
85
The prototype link fails 12 minutes into session 2 of 6, leaving 3 tasks untested; what failure artifact would you produce, and what is the exit rule for resuming?
sessionsprototypesartifacts
86
Recording fails for 1 of 8 interviews, but you have structured notes from 2 observers; what evidence artifact would you keep, and when may the session enter synthesis?
sessionsartifacts
87
A bilingual study has 4 Spanish and 4 English interviews, but synthesis must happen in English within 3 days; what note artifact would you create, and what is the quality threshold?
artifacts
88
You have 5 working days to recruit, run, and summarize 5 interviews before a product checkpoint; what minimum research package would you commit to, and when is it done?
89
A 1-person pilot reveals that task 2 can be completed only because the prototype starts on the answer screen; what defect artifact would you create, and what must pass before 7 scheduled sessions begin?
sessionsprototypesdefects
90
You have 6 English interviews and 4 Spanish interviews, one bilingual teammate, and a readout in 3 days. How would you prepare the synthesis without flattening meaning?
91
A PM says 6 interviews are just anecdotes and rejects 3 findings; what evidence artifact would you bring to a 30-minute review, and what outcome closes it?
artifacts
92
Sales cites 5 customer calls supporting feature A, while Support cites 18 tickets opposing it; what comparison artifact would you create before the 2 teams meet?
artifacts
93
In 8 interviews, 5 participants say they never use saved searches, while Amplitude shows 42% of 1,200 active users triggered save_search last month; what triangulation artifact would you produce?
research-participantsartifacts
94
Six of 9 interviewees report abandoning setup at verification, but Mixpanel shows a 12% drop between verification_started and verification_completed across 4,500 users; what analysis record would you make?
95
A researcher gives you an events table with 30,000 rows and asks how many of 240 users returned within 7 days after onboarding; what SQL artifact would you provide, and how would you validate it?
onboardingsqlvalidation
96
A team discusses 4 findings from 7 interviews but makes 2 product decisions without recording why; what decision-log artifact would you add, and when is the handoff complete?
artifacts
97
Research with 10 participants leads the team to change an error message, and the PM asks how to follow up over 4 weeks; what measurement artifact would you propose?
research-participantsartifacts
98
After 8 interviews, checkout completion rises from 64% to 71% following 3 product changes; how would you write the 1-paragraph impact note without claiming causality?
causal
99
Five interviews suggest notifications are distracting, but an Amplitude cohort of 2,800 users shows recipients return 1.4 times more often; what joint evidence brief would you create?
cohorts
100
Thirty days after 9 interviews led to 3 actions, one shipped, one changed, and one was cancelled. How would you assess research influence without claiming it caused the product metrics?
monitoring

Questions

1A PM asks you to improve onboarding before planning in 10 days, but has not named a user problem. How do you turn that request into a research question?

2A designer asks whether users like a new billing dashboard, and the team must decide in one week whether to revise it. What would you research instead?

3The team has two weeks to decide whether to explore a budgeting feature, but it knows little about how customers currently plan monthly spending. Which method would you choose?

4A PM must decide whether to prioritize bulk export and asks how many of 20,000 active customers export data weekly. Would you use qualitative or quantitative research?

5A team can recruit 12 participants before choosing between two meal-planning concepts, but it is unsure whether the underlying problem matters. How do generative and evaluative research differ here?

6You have six participants available before release and product metrics due 30 days after release. How would you separate formative and summative research on a new checkout flow?

7Analytics shows an 18 percent drop at identity verification, and the team has 12 days to decide whether to change instructions or the flow. How would you combine qualitative and quantitative evidence?

8You have five business days and a clickable prototype to decide whether a new refund flow is understandable before engineering starts. What bounded study would you propose?

9Four of five interview participants preferred concept A, and a stakeholder says this proves 80 percent of customers want it. How do you respond before tomorrow’s review?

10In the eighth generative interview, a new cancellation reason appears, but the team says the study has reached saturation and must stop today. What would you do?

11You need to interview eight first-time sellers in two weeks so the team can decide which listing setup problem to address. How would you build the interview guide?

12During a 45-minute interview, a participant says the reporting tool was confusing, and you have only five minutes left. What neutral probes do you use?

13A guide for six interviews asks, “How helpful was the simple automatic reminder?” What would you change before sessions start tomorrow?

14The team will interview 10 people who recently switched expense tools and must decide which migration problem to address. How would you use a Jobs-to-Be-Done lens?

15Three stakeholders want to observe each of six moderated interviews, and participants must feel comfortable discussing financial mistakes. How would you set observer roles?

16After two of eight interviews support the PM’s hypothesis, you notice you are probing that theme more than alternatives. What do you do before interview three?

17You are moderating six remote interviews alone, automatic transcripts arrive the next day, and the readout is due in four days. How do you take notes without losing rapport?

18A 45-minute guide for eight interviews contains 28 main questions, and the first session is tomorrow. How would you use a pilot?

19In interview four of eight, an observer asks you to show a new concept and ask whether the participant would use it, but the approved study is generative. What do you do?

20Six interviews are split: three users rely on templates and three deliberately avoid them, with a roadmap review in two days. How would you report the result?

Locked questions