210 answered questions · Clinical Data Management
Prepare the skills this role needs. Start with your resume, practise technical questions, then work through scenarios and salary conversations.
Know every claim on your resume. Be ready to explain your internships, certificates, projects and tools: what you personally did, why, the result and what you learned. Distinguish observation from hands-on experience.
Choose a practice topic
210 questions · Tap to read answers
No matching questions. Try a shorter term or clear your search.
Resume, motivation & HRQuestions 1–25
001Tell me about yourself.
- Give a 60–90 second introduction: your qualification, one relevant experience, a skill you can demonstrate and why this role fits.
- A useful structure is: 'I studied …, worked on …, personally handled …, and learned …'.
- Replace each example with truthful details; avoid reciting your entire resume.
002Walk me through the parts of your resume most relevant to this job.
- Select two or three relevant experiences and explain the link to the job description.
- For each, state the setting, your responsibility and a concrete output.
- Clearly identify classroom, simulated and workplace experience.
- Expect follow-up questions about every technical term you mention.
003What exactly did you do during your internship?
- Explain the department, duration, supervisor and tasks you actually performed.
- Separate observations from hands-on work.
- Describe one activity from input to output, including checks and escalation.
- If you only observed a process, say so and explain what you understood rather than claiming independent responsibility.
004Which internship activity could you explain from beginning to end?
- Choose one real activity.
- Explain its purpose, required input, sequence, safety or quality checks, documentation and the person who approved the result.
- Be ready to explain why each step existed and what would happen if it was missed.
- Do not disclose confidential patient or company details.
005What did your certificate teach you, and what did you practise?
- Name the course, provider, completion date and main modules.
- Then distinguish lectures, assessed exercises and practical tasks.
- For example: 'I completed the theory module and practised with a training dataset; I have not used the production system.' Explain one concept well instead of listing certificates without understanding them.
006Explain the objective of a project on your resume.
- State the problem, why it mattered, the objective and your own contribution.
- Explain the setting and limitations before describing results.
- If the work was simulated or academic, say so.
- Do not imply that a classroom exercise produced real patient, manufacturing or commercial outcomes.
007How did you choose your project's method?
- Explain why the method addressed the objective, what alternatives you considered and which constraints mattered.
- Identify the protocol, reference or supervisor input used.
- Discuss one limitation honestly.
- A strong answer shows reasoning; naming a technique without explaining its purpose does not demonstrate competence.
008How did you check the accuracy of your project results?
- Describe the real checks you performed: source comparison, calculations, controls, peer review or supervisor verification as applicable.
- Explain how you handled missing information and discrepancies.
- Avoid saying that the result was 100% accurate unless you can define and substantiate that claim.
009Your resume mentions an improvement percentage. How was it calculated?
- State the baseline, final value, denominator and time period.
- Explain whether the result is an absolute difference or a relative change, and identify your contribution.
- If you cannot reconstruct a number from authorised evidence, correct the claim rather than invent a calculation during the interview.
010Which instrument or software function have you personally used?
- Name the specific task and your level of access or supervision.
- Walk through the input, action, output and checks.
- Separate familiarity with terminology from hands-on competence.
- Be prepared for a practical demonstration using nonconfidential training material and do not claim certification or administrator access you do not have.
011Which academic subject is most relevant to this role?
- Choose a subject you understand and connect one concept to a real role activity.
- Explain the concept in simple language, then give an example and limitation.
- Revise any subjects or clinical-trial phases you highlight on your resume; mentioning a topic invites deeper questions about it.
012Why do you want to work in this department?
- Connect your interest to actual responsibilities in the job description.
- Support it with a truthful course, project or work example.
- Explain what you want to learn and contribute.
- Avoid claims that the department is easy, guarantees high pay or provides a shortcut around the technical learning required.
013Why are you interested in this employer?
- Use two current facts from the employer's own website or job description and explain their relevance to the role.
- Then connect your demonstrated skills to the team's work.
- Do not invent company rankings, employee counts, client relationships or promises about training that you have not verified.
014What is a strength you can demonstrate?
- Choose one relevant strength and prove it with a brief example.
- 'I use a source-check checklist; in my project it helped me catch a date mismatch before review' is stronger than 'I never make mistakes'.
- State what you personally did and what evidence supports the result.
015What is a weakness you are working on?
- Describe a genuine, manageable development area and the steps you are taking.
- For example: 'I used to give overly long explanations, so I practise a one-minute summary before the detail.' Explain progress without disguising a boast as a weakness or suggesting that essential safety requirements are optional.
016How do you explain a career gap or change of field?
- Give a brief factual explanation at a level you are comfortable sharing, then discuss relevant learning and readiness for the role.
- You do not need to invent employment, training or personal circumstances.
- Connect transferable skills to the job while acknowledging the department-specific training still needed.
017Give an example of teamwork.
- Use Situation, Task, Action and Result.
- Describe the shared objective, your own contribution and how you handled communication or a handoff.
- An academic project is a valid example when labelled accurately.
- Do not take credit for the whole team's work or reveal confidential information.
018How do you work under pressure?
- Explain how you identify the highest-impact tasks, confirm priorities, work through a checklist and communicate delays early.
- Support the answer with a real example.
- Speed should not mean skipping required checks, hiding errors or working outside your training and authority.
019Can you work shifts, relocate or meet the joining date?
- Answer honestly about your availability and constraints, then ask for the actual schedule and role expectations.
- Do not automatically say yes to conditions you cannot meet.
- You can be professional and interested while clarifying notice periods, transport arrangements or practical limitations.
020What would you do if you realised your resume overstated a skill?
- Correct the claim promptly and explain your actual level.
- Offer an accurate example of what you can demonstrate and a realistic learning plan.
- Honesty protects both you and the employer; rehearsing an invented answer is especially risky in work involving safety, regulated records or confidential information.
021Why should we hire you for this role?
- Choose two requirements from the job description and support each with something you have actually done.
- For example: "Your role needs accurate records and reliable follow-up. In my academic project, I compared our results with the source records, corrected a mismatch with my supervisor and completed the revised report on time."
- Use this example only if it reflects your own experience; otherwise replace it with a truthful internship, project or work example.
- Explain how those habits will help the team and acknowledge the role-specific training you still need.
022How will you contribute as an individual within our team?
- Describe a deliverable you can own, the checks you will perform and the handoffs you will communicate.
- For example: 'I will keep my assigned work traceable, raise blockers early and provide a clear status update so colleagues can depend on my output.' Add a truthful supporting example.
023What would your first 30, 60 and 90 days look like?
- First learn the team's procedures and complete training; then perform supervised tasks and act on feedback; finally take ownership of work for which you are qualified.
- Agree actual milestones with your manager.
- Do not promise an arbitrary productivity target before understanding the process and task complexity.
024Why are you leaving your current job or changing employers?
- Give a brief, truthful reason and connect it to the responsibilities you seek.
- 'I want more experience in [specific activity] and this role includes it.' Avoid attacking colleagues or inventing a promotion story.
- Explain how your notice period and handover will be managed.
025Tell us about a failure and what you changed afterward.
- Choose a real, manageable example.
- Explain the intended result, your decision, what went wrong and how you corrected it.
- End with a specific change and evidence of learning.
- 'I am a perfectionist' or blaming an entire team does not demonstrate accountability.
Technical foundations & workflowQuestions 26–65
026How would you explain source data?
- Source data are original recorded information or authorised source records supporting trial observations.
- Their location and use must be defined for the study.
027How would you explain an electronic case report form (eCRF)?
- An eCRF collects protocol-required information in an electronic system.
- It is not automatically the original source for every data item in every study.
028How would you explain an EDC system?
- Electronic Data Capture systems support collection and management of clinical-trial data.
- Access controls, validation and audit trails are important alongside the visible entry forms.
029How would you explain a CTMS?
- A Clinical Trial Management System supports operational trial activities such as sites, milestones and monitoring.
- Its purpose differs from collecting participant-level eCRF data.
030How would you explain a Data Management Plan?
- A DMP describes study-specific data-management activities, responsibilities and conventions.
- It should align with the protocol and applicable controlled procedures.
031How would you explain a data-review plan?
- A review plan defines which data are reviewed, how discrepancies are identified and who resolves them.
- It translates broad requirements into operational review activities.
032How would you explain an automated query?
- An automated query is raised by configured logic.
- Correct configuration does not guarantee clinical relevance; testing and review are needed to avoid misleading or excessive queries.
033How would you explain a manual query?
- A manual query is initiated after authorised review identifies a discrepancy needing clarification.
- It should be clear, specific and nonleading.
034How would you explain a range check?
- A range check flags values outside a defined range.
- An unusual value may be genuine; do not replace it with a normal value without source-supported correction.
035How would you explain a cross-form check?
- A cross-form check compares related information across forms, such as treatment and event dates.
- Study conventions and clinically plausible exceptions matter.
036How would you explain data correction?
- Correction changes an erroneous recorded value through an authorised, traceable process.
- The person performing it must have the appropriate responsibility and supporting evidence.
037How would you explain data clarification?
- Clarification resolves ambiguity or explains a value that may already be correct.
- Not every resolved query should lead to a changed value.
038How would you explain database freeze?
- A freeze restricts specified changes at a defined milestone.
- Its exact meaning depends on the study process; it must not be assumed identical to lock.
039How would you explain database lock?
- Lock follows defined readiness checks and authorisations.
- Unresolved items, external-data reconciliation and required review status must be assessed under the study's process.
040How would you explain authentication?
- Authentication establishes that an account user is who they claim to be.
- Shared credentials undermine accountability and are not an acceptable shortcut.
041How would you explain authorisation?
- Authorisation determines which actions an authenticated user may perform.
- Study, site and role scope should follow approved access arrangements.
042How would you explain SDTM?
- The Study Data Tabulation Model organises trial data for consistent tabulation.
- It is not the same thing as the operational EDC database or an analysis-ready dataset.
043How would you explain ADaM?
- The Analysis Data Model structures analysis datasets and metadata to support reproducibility and traceability.
- Its derivations should connect back to defined source information.
044How would you explain an adverse event dataset?
- This dataset represents collected event information under study conventions.
- Compare relevant fields with safety records without assuming two systems have identical purposes.
045How would you explain SAE reconciliation?
- SAE reconciliation identifies relevant discrepancies between clinical and safety data.
- Owners must resolve mismatches without delaying necessary safety handling or inventing missing facts.
046How would you explain CDASH?
- CDASH supports consistent clinical data collection and traceability toward tabulation standards.
- A form still needs to meet the actual protocol and system requirements.
047How would you explain controlled terminology?
- Controlled terminology standardises permitted values and meanings.
- Use the agreed version and mapping rules rather than changing categories casually during review.
048How would you explain an edit-check specification?
- It defines the logic, inputs, conditions and expected behaviour of a check.
- Clear requirements allow meaningful positive, negative and boundary testing.
049How would you explain User Acceptance Testing?
- UAT checks whether a configured system meets approved intended-use requirements.
- Record inputs, expected results, actual results and defects before approving readiness.
050How would you explain an audit trail?
- An audit trail supports reconstruction of relevant changes, including who changed what and when.
- Review must consider context rather than treating a trail as an automatic guarantee of integrity.
051How would you explain an external-data transfer?
- A transfer brings data from a source such as a laboratory or vendor under agreed specifications.
- Verify identifiers, units, dates and completeness before integration.
052How would you explain query ageing?
- Query ageing measures how long queries remain in defined states.
- It supports follow-up but should be interpreted with complexity, dependencies and safety relevance.
053How would you explain missing-data review?
- Review distinguishes an expected but missing observation from data not required or not applicable.
- Use study conventions and do not populate unobserved values to satisfy completeness metrics.
054How would you explain an outlier?
- An outlier is an observation unusual relative to an expected pattern.
- Investigate units, timing and source before deciding it is an error.
055How would you explain duplicate records?
- Apparent duplicates need comparison of identifiers, visit context and source.
- Similar values can be valid repeated observations; document the resolution rather than delete automatically.
056How would you explain medical coding in CDM?
- Coding maps reported clinical terms and medicines to study dictionaries.
- Preserve verbatim information and use qualified review for ambiguity.
057How would you explain access reconciliation?
- Compare active system accounts against approved study and site personnel.
- Resolve inappropriate scope or departed users through authorised access controls.
058How would you explain data lineage?
- Lineage shows where a data item came from and how it was transformed.
- It helps reviewers reproduce outputs and trace discrepancies to their source.
059How would you explain privacy protection?
- Use approved identifiers, access controls and secure transfer methods.
- Data available for a task should not be copied into personal email or unauthorised public tools.
060How would you explain an error after lock?
- A post-lock error needs documented impact assessment and the approved change or unlock process.
- Do not quietly change a final dataset outside governance.
061Describe an end-to-end Clinical Data Management workflow.
- A useful answer is: Confirm protocol and review requirements; check authorised data; identify discrepancies; raise clear nonleading queries; track clarification; reconcile external and safety data; complete documented readiness checks.
- Explain who owns each decision and distinguish your own performed tasks from those you have only studied.
- The actual employer's procedures and role scope take priority over this preparation example.
062What would you verify before starting an assigned technical task?
- I would confirm the purpose, my training and authority, and the current approved requirements.
- In this role, relevant inputs include protocol, DMP, approved specifications and authorised datasets.
- I would verify identity, version, status and completeness, then clarify any material gap before relying on the information.
063What outputs would you check before considering the work complete?
- Relevant outputs include reviewed data, query records and documented reconciliation.
- I would compare the result with its requirements and source evidence, resolve discrepancies through the approved process and confirm required review.
- I would not mark a task complete just because a file exists or a tracker demands closure.
064Which performance measures matter in Clinical Data Management?
- Examples include query ageing, unresolved critical discrepancies and review quality.
- I would define the numerator, denominator, time window and data source before interpreting a metric.
- Volume and speed need to be assessed alongside complexity, quality and risk; I would not improve a score by hiding errors or excluding difficult work.
065Which tools should you be prepared to discuss?
- Relevant examples include EDC, approved listings, spreadsheets and authorised data-review tools.
- I would explain the specific functions I have actually used and my supervision or training level.
- For unfamiliar systems, I would describe transferable understanding and how I would learn through approved training, rather than pretend to have production experience.
Study setup and form designQuestions 66–80
066How do you turn a protocol requirement into an eCRF field?
- Identify the required observation, visit and collection method.
- Define type, units, permitted values and conditional logic.
- Trace the field to the protocol and test its intended use.
067What belongs in eCRF completion guidelines?
- Give field-specific entry instructions.
- Explain dates, units, unknown values, corrections and repeat events.
- Include examples that agree with the protocol and validated database configuration.
068How do you avoid collecting unnecessary patient data?
- Map every field to a justified study purpose.
- Remove duplicate or unused identifiers and assessments.
- Review privacy, participant burden and analysis needs before approving collection.
069When should a form use repeating rows?
- Use repeating rows for multiple occurrences, such as concomitant medications.
- Define the unique row identifier and start/stop rules.
- Test additions, corrections and deleted or inactivated entries.
070How do you design a screening-failure pathway?
- Collect the approved screening information and failure reason.
- Prevent inappropriate treatment or follow-up forms from appearing automatically.
- Preserve consent, eligibility and disposition information required by the protocol.
071How should unscheduled visits be represented?
- Use the study's defined unscheduled-visit structure.
- Retain the actual date and assessment identity.
- Avoid forcing observations into a scheduled visit merely to remove a missing-form flag.
072How do you distinguish a visit window from a visit date?
- The date is the actual occurrence; the window is the protocol-permitted interval.
- Calculate against the correct reference event.
- Review exceptions without altering the actual date.
073What makes an edit-check requirement testable?
- Specify input fields, triggering conditions and expected action.
- Define missing-value and boundary behaviour.
- Include examples where the query should and should not fire.
074What are positive, negative and boundary tests for a date check?
- Positive tests demonstrate the intended trigger.
- Negative tests confirm valid chronology is accepted.
- Boundary tests cover equal dates, partial dates, leap days and the allowed window edges.
075How do you test conditional form visibility?
- Exercise every triggering answer and subsequent correction.
- Check that required forms appear and inappropriate forms disappear according to specification.
- Confirm existing records remain traceable when the trigger changes.
076Why do derived fields need specifications?
- The derivation must be reproducible.
- Define formula, input timing, units, rounding and missing-data rules.
- Test expected outputs before relying on derived values in review or analysis.
077How do you control a mid-study database amendment?
- Assess affected forms, participants and downstream transfers.
- Approve requirements, test changes and plan deployment.
- Reconcile migrated data and document training and release evidence.
078What is a database build discrepancy log?
- It tracks differences between requirements and configuration.
- Record severity, owner, expected result and retest evidence.
- Do not equate closing a ticket with verified correction.
079How should pick-list changes be assessed?
- Assess meaning and existing records, not only display text.
- Version the terminology and map affected values explicitly.
- Test exports, edit checks and analysis consequences before release.
080What does role-based UAT cover?
- Test permitted and prohibited actions for each role.
- Include site scope, review, signing, correction and export permissions.
- A successful administrator test does not demonstrate appropriate site-user access.
External data and integrationsQuestions 81–95
081What belongs in a laboratory data-transfer agreement?
- Define identifiers, analytes, units, ranges, formats and frequency.
- Specify corrected files, reconciliation and acceptance checks.
- Name the owners of transfer failures and clinical discrepancies.
082How do you validate a vendor file before loading it?
- Check delivery identity, expected structure and record counts.
- Validate keys, formats, duplicates, units and controlled values.
- Quarantine failed transfers and retain acceptance evidence.
083How do you reconcile a central laboratory record without an EDC visit?
- Compare subject, specimen, collection time and visit mapping.
- Check unscheduled collections and transfer timing.
- Resolve the ownership gap rather than discard an otherwise valid laboratory result.
084Why must laboratory units accompany results?
- A number without its unit can be clinically misleading.
- Preserve reported units and reference ranges.
- Apply only specified, validated conversions and retain the original value.
085How do you handle changing laboratory reference ranges?
- Link the result to the applicable laboratory range and date.
- Consider age, sex and method where specified.
- Do not apply the latest range retrospectively without an approved rule.
086What is a full transfer versus a delta transfer?
- A full transfer contains the defined complete extract; a delta contains changes since a specified point.
- Confirm replacement and deletion semantics.
- Reconcile against the agreed baseline and transfer sequence.
087How do you prevent duplicate loads?
- Use validated file and record identifiers with load tracking.
- Distinguish retries from corrected deliveries.
- Reconcile counts and rejected records rather than rely only on the file name.
088How should electronic patient-reported outcome data be reviewed?
- Check participant mapping, assessment schedule and completion timestamps.
- Distinguish unavailable devices from unanswered items.
- Preserve the participant's actual response and approved correction restrictions.
089Why are time zones important in external data?
- Events near midnight can map to different dates.
- Retain the source time zone and conversion convention.
- Test daylight-saving changes and chronology around dosing or safety events.
090What are specimen identifiers used for?
- They link collection, shipment, laboratory receipt and results.
- Check uniqueness and participant association.
- Investigate relabelling or mismatches through the defined specimen process.
091How do you assess external-data completeness?
- Compare expected participants, visits and specimens with accepted deliveries.
- Account for not-collected and cancelled assessments separately.
- A matching total count can conceal missing and duplicated records.
092What should happen to records rejected during import?
- Retain the rejection reason and source identity.
- Assign correction ownership and track resubmission.
- Reconcile all rejects before declaring the transfer complete.
093How do you reconcile randomisation and treatment data?
- Compare authorised identifiers, dates and treatment-accountability information.
- Respect blinding and restricted access.
- Escalate discrepancies through the designated unblinded or clinical process where required.
094What makes an integration mapping specification useful?
- It defines source and target fields, keys, transformations and exceptions.
- Version terminology and formats.
- Test representative records, not only an empty file or successful connection.
095How do you handle a vendor's corrected result?
- Identify the affected record and correction reason.
- Preserve prior versions and load the authorised correction.
- Recheck queries, reviews and downstream outputs impacted by the change.
Clinical review and query writingQuestions 96–110
096What makes a query nonleading?
- It identifies the discrepancy without suggesting an unsupported answer.
- State the conflicting fields and request verification.
- Avoid telling the site to enter a preferred clinical result.
097When should an unusual value not generate a query?
- When it is plausible, adequately documented and consistent with study rules.
- Review units, timing and context first.
- Querying every extreme value can create unnecessary site burden.
098How do you review an incomplete medication stop date?
- Check ongoing status and available precision.
- Apply the collection convention for partial dates.
- Do not convert an unknown day into an invented exact stop date.
099How do you assess overlapping medication records?
- Compare product, dose, route, indication and dates.
- Overlap may reflect dose adjustment or concurrent treatment.
- Clarify only the unresolved inconsistency rather than automatically merge rows.
100How do you review missing assessments after early withdrawal?
- Compare the protocol's discontinuation and follow-up requirements.
- Distinguish required-but-missing assessments from no-longer-required visits.
- Preserve the reason for noncollection.
101What should a discrepancy review meeting produce?
- Decisions with evidence, owners and due dates.
- Separate clinical judgement from data-entry correction.
- Record unresolved issues and follow-up rather than circulate only attendance minutes.
102How do you avoid repeated queries for the same issue?
- Review query history and previous site responses.
- Consolidate related questions where appropriate.
- Fix a defective rule or unclear instruction instead of repeatedly reopening resolved records.
103What is query closure evidence?
- It shows the response or correction resolves the actual discrepancy.
- Review supporting context and audit history.
- A response of 'done' is insufficient if conflicting values remain.
104How should clinical significance be handled in data review?
- Qualified clinical staff assess clinical significance.
- Data managers identify discrepancies and route them appropriately.
- An automated flag is not a medical diagnosis or treatment instruction.
105How do you review treatment-emergent chronology operationally?
- Check event onset, treatment exposure and relevant dates.
- Preserve partial dates and uncertainty.
- Analysis classification follows the prespecified rules; do not change collected dates to force a category.
106How do you compare event dates across forms?
- Use the same participant and event identity.
- Review precision, time zones and definitions of onset versus awareness.
- Resolve meaningful differences without assuming every date must match.
107How do you review an impossible age calculation?
- Check birth-date precision, reference date and calculation logic.
- Consider masking conventions.
- Correct the source-supported error rather than overwrite a derived value manually.
108What is a clinically important data-review finding?
- A finding that may affect participant safety or reliable trial conclusions.
- Assess context, severity and affected records.
- Escalate through the study's risk and clinical review process.
109How do you calculate query ageing consistently?
- Define the starting state, elapsed-time rule and current state.
- Separate site response time from sponsor review time if required.
- Report denominators and excluded categories transparently.
110How should a data-review listing be validated?
- Check its specification, source extract and transformation logic.
- Verify representative records and edge cases.
- Reconcile counts and confirm the listing contains the intended study population.
Coding, safety and oversightQuestions 111–125
111How do you prepare verbatim terms for dictionary coding?
- Preserve the original reported text.
- Check context and obvious transcription issues through authorised clarification.
- Do not replace a symptom with an inferred diagnosis before coding.
112How do you manage ambiguous medication names?
- Use available formulation, country, strength and indication details.
- Check the authorised drug dictionary and coding convention.
- Request clarification when the available evidence cannot distinguish products.
113What is a dictionary upgrade impact assessment?
- It identifies records affected by a new dictionary version.
- Review changed mappings, coding decisions and downstream outputs.
- Retain version history and approved migration evidence.
114Why should safety reconciliation not wait until lock?
- Early reconciliation exposes missing cases and inconsistent critical fields.
- Schedule checks throughout conduct.
- Necessary safety reporting follows its own obligations and must not wait for database-cleaning completion.
115Which fields commonly need SAE reconciliation?
- Study identifiers, event terms, seriousness information and relevant dates.
- Product exposure and outcome may also need comparison.
- Use the reconciliation plan because clinical and safety systems have different purposes.
116How do you preserve blinding during data review?
- Apply approved access and masked displays.
- Route unblinded questions to designated personnel.
- Avoid exports or query wording that reveal treatment unnecessarily.
117What belongs in a data-issue escalation?
- A clear discrepancy, affected records and potential impact.
- Include evidence, urgency, owner and decision needed.
- Exclude unnecessary personal identifiers and unsupported blame.
118How does risk-based review affect sampling?
- Review intensity follows the data's importance and identified risks.
- Justify sampling rules and escalation triggers.
- Low volume alone does not make a safety-critical field suitable for reduced review.
119How do you review audit trails efficiently?
- Focus on relevant changes, timing, users and patterns.
- Investigate unusual bulk changes or unsupported corrections.
- Document the scope and findings rather than merely confirm that an audit trail exists.
120Why must withdrawn staff access be reviewed?
- Departed or reassigned users may retain inappropriate permissions.
- Reconcile accounts with approved personnel records.
- Remove access through authorised administration without losing attribution of historical work.
121What is a data-management deviation?
- A departure from an approved study data process.
- Assess impact on data quality, privacy and milestones.
- Document correction and any required preventive action under the applicable procedure.
122How should sensitive free-text fields be checked?
- Look for unnecessary direct identifiers and inappropriate disclosures.
- Use approved redaction or correction processes.
- Preserve required source traceability and report privacy concerns promptly.
123What is oversight of an outsourced data-management activity?
- The responsible organisation verifies delegated work and quality.
- Review deliverables, risks and performance against the agreement.
- Outsourcing a task does not mean ignoring its outcomes.
124How should reconciliation decisions be recorded?
- Identify the records, discrepancy, evidence and agreed resolution.
- Name the responsible owner and review status.
- Preserve why a difference was accepted when identical values were not appropriate.
125How do you distinguish a data query from a safety notification?
- A query clarifies collected data; a safety notification communicates a potential reportable concern.
- Route each through its required process.
- Opening an EDC query alone may not satisfy safety communication obligations.
Lock, delivery and analysis readinessQuestions 126–140
126What should a database-lock checklist cover?
- Required data, query disposition, coding and reconciliation.
- Confirm clinical review, approvals and external-data status.
- Document accepted outstanding items instead of concealing them to achieve a milestone.
127What is a soft lock in practice?
- It is a study-defined restricted-change milestone.
- Specify which data and users are affected.
- Never assume the phrase has identical operational meaning across employers or systems.
128Who authorises a database unlock?
- The study's approved governance determines the authorisers.
- Document the reason, scope and downstream impact.
- Reconcile changes and obtain required review before relocking and redelivering data.
129How do you demonstrate that a data extract is reproducible?
- Record the source snapshot, extraction logic and versions.
- Retain execution evidence and record counts.
- Use controlled reruns rather than reconstructing final data from untracked spreadsheets.
130What should a transfer manifest contain?
- Identify the study, files, versions and extraction date.
- Include counts, formats, integrity checks and delivery status where specified.
- Ensure the recipient can recognise a complete authorised delivery.
131How do you reconcile operational EDC data with SDTM outputs?
- Trace mapped fields to collected data and documented transformations.
- Check identifiers, timing and controlled terminology.
- A structural difference can be legitimate; investigate unsupported data changes.
132Why are annotated CRFs useful?
- They connect collected fields to intended tabulation variables.
- They support traceability and review.
- They do not replace the mapping specification or prove every transformation is correct.
133What belongs in a data-management closeout report?
- Summarise scope, deliveries, quality issues and resolutions.
- Record deviations, accepted limitations and archival locations.
- Make outstanding obligations visible after the operational team closes.
134How do you archive data-management records?
- Follow the agreed retention and access plan.
- Preserve data, metadata, essential specifications and review evidence.
- Test retrievability rather than treat a copied folder as a complete archive.
135How should missing data be described to analysis teams?
- Identify what is missing, why if known and which records are affected.
- Preserve collection reasons separately from analysis imputations.
- Do not fill missing observations merely to improve completeness statistics.
136How do you calculate a database-cleanliness metric?
- Define the unit, denominator, time point and issue severity.
- Separate unresolved critical discrepancies from routine queries.
- A single percentage should not hide unsafe or analysis-critical problems.
137What does data provenance add to a final delivery?
- It explains where data originated and how it changed.
- Preserve source identifiers and transformation versions.
- This allows a reviewer to reconstruct important conclusions and investigate discrepancies.
138How do you manage a late vendor transfer after final delivery?
- Assess affected participants, endpoints and released outputs.
- Route the change through approved governance.
- Issue a controlled revised delivery if authorised and notify recipients of the superseded version.
139How should protocol-deviation data be reconciled?
- Check definitions, participant identity and event timing.
- Compare relevant clinical and operational records.
- Preserve the qualified study decision; data managers should not invent a deviation classification.
140What makes a data-management handover usable?
- Provide current milestones, risks, open issues and source locations.
- Identify owners, expected deliveries and review status.
- Confirm the receiving person can locate evidence and act on urgent items.
Practical data checksQuestions 141–155
141How would you identify duplicate subject-visit rows in a spreadsheet?
- Define the expected unique key first.
- Group or count by subject, visit and assessment identity.
- Investigate repeats before removing them because repeat assessments may be valid.
142When is a left join useful in data reconciliation?
- Use it to retain every record from the expected-data table.
- Match accepted external records on validated keys.
- Review unmatched and multiply matched records separately.
143Why can a join unexpectedly increase row counts?
- One or both join keys may have multiple matches.
- Check key uniqueness and the intended relationship.
- Do not suppress duplicated outputs before understanding the underlying many-to-many join.
144How do you check date-format ambiguity in a transfer?
- Confirm the specified format and sample values.
- Distinguish day/month from month/day using source conventions.
- Reject ambiguous imports rather than guess based on the reviewer's locale.
145How should zero and missing be distinguished?
- Zero is a recorded value; missing indicates absent information.
- Use the defined representation for unknown and not applicable.
- Do not convert blanks to zero during numeric cleaning.
146What is a denominator check in a review report?
- It verifies which participants or forms were eligible for the measure.
- Reconcile exclusions and reference dates.
- A correct numerator can still produce a misleading percentage with the wrong denominator.
147How do you verify a calculated BMI field?
- Confirm weight and height units and the approved formula.
- Test typical, missing and implausible inputs.
- Check rounding and timing conventions without replacing source measurements.
148How do you review an apparent negative treatment duration?
- Check start/stop dates, partial-date handling and time-zone conversion.
- Review the derivation and source chronology.
- Correct the supported cause instead of forcing duration to zero.
149Why should review code be version controlled?
- It preserves the exact logic used for a result.
- Review changes and retain tested versions.
- Link significant outputs to the input snapshot and approved code release.
150How do you explain a technical check during a practical interview?
- State its purpose, required inputs and expected outcome.
- Demonstrate with nonconfidential sample data and an edge case.
- Explain assumptions and what needs clinical or lead review.
151How do you test a calculated treatment-compliance percentage?
- Define dispensed, returned and expected quantities under the protocol.
- Check dates, dose changes and missing accountability.
- Report values above the plausible range for investigation instead of capping them silently.
152What is the difference between missing a form and missing an item?
- A missing form means an expected assessment record is absent.
- Item missingness concerns unanswered fields within an existing record.
- Use the correct denominator and follow-up route for each.
153How should data-entry lag be measured?
- Compare the actual assessment date with the defined initial-entry timestamp.
- State whether weekends and corrected records affect the calculation.
- Separate entry delay from query-response delay.
154What should a study data cut specify?
- Define the purpose, cutoff date and included records.
- State whether dates concern events, entry or receipt.
- Keep later corrections traceable without silently changing an already released snapshot.
155How do you assess a clinical-data migration?
- Map old to new fields and preserve meaning.
- Reconcile counts, values, audit history and access requirements.
- Test transformations and obtain approval before retiring the old environment.
Situational & workplace practiceQuestions 156–185
156You find a medication end date before its start date. What would you do?
- I would verify the discrepancy against the authorised source, preserve the original information and assess its immediate impact.
- I would pause affected work where necessary and involve the data-management lead and relevant clinical or safety team.
- I would document the clarification or correction and check whether other records or decisions are affected.
157Required information is missing from an eCRF record and its authorised source. What next?
- I would identify exactly what is missing and whether it prevents a safe or valid next step.
- I would seek the information from the authorised owner and document follow-up.
- I would not invent a value, and I would escalate any deadline or risk implication through the data-management lead and relevant clinical or safety team.
158You notice a discrepancy potentially affecting participant safety. What is your priority?
- I would follow the relevant immediate safety, quality or incident process and promptly involve the data-management lead and relevant clinical or safety team.
- I would preserve facts, restrict affected activity where required and avoid acting beyond my competence.
- Routine productivity targets would not justify delaying an urgent concern or bypassing a protective control.
159A deadline is close, but work on an eCRF record and its authorised source remains incomplete. What would you do?
- I would separate critical issues from lower-impact work, identify dependencies and communicate the realistic status early.
- With the data-management lead and relevant clinical or safety team, I would agree priorities and an authorised plan.
- I would not conceal incomplete checks or certify an unsupported result merely to meet the original deadline.
160A colleague asks you to skip a required check because the team is busy. How would you respond?
- I would explain the purpose of the check and look for an authorised way to manage the workload.
- If a genuine alternative exists, I would seek appropriate approval; otherwise I would complete the requirement or escalate.
- The relevant escalation route here includes the data-management lead and relevant clinical or safety team.
161You discover the wrong version was used for an eCRF record and its authorised source. What next?
- I would identify the affected work, compare it with the correct baseline and inform the owner.
- I would assess whether any output was already shared or used, then follow the controlled correction process with the data-management lead and relevant clinical or safety team.
- I would preserve version history rather than overwrite evidence silently.
162You made an error in a record that has already been reviewed. How do you handle it?
- I would acknowledge the error, establish its impact and notify the responsible reviewer promptly.
- I would correct through the approved process, keeping the original history and rationale.
- I would consider downstream outputs and inform the data-management lead and relevant clinical or safety team if the error affects safety, quality or an external decision.
163You accidentally send confidential information to the wrong recipient. What would you do?
- I would stop further sharing and immediately follow the organisation's privacy or security incident process.
- I would notify the authorised incident team, preserve facts and follow their instructions on containment and communication.
- Deleting my sent message or privately asking for deletion would not replace incident reporting.
164The same problem, a medication end date before its start date, keeps recurring. How would you address it?
- I would review examples to identify the common failure point, assess the impact and involve the process owner.
- I would consider instructions, training, system design and handoffs rather than assume individual carelessness.
- Corrective actions should have owners and an effectiveness check, not only a promise to be more careful.
165The approved system becomes unavailable during urgent work. What would you do?
- I would preserve available work and follow the approved outage or continuity process.
- I would contact the support route and assess the deadline or safety impact with the data-management lead and relevant clinical or safety team.
- I would not move confidential information into a personal account or create an uncontrolled workaround to appear productive.
166Someone asks to use your login or grants you access beyond your role. How do you respond?
- I would not share credentials or use unnecessary privileges.
- I would explain the accountability and confidentiality risks and request properly authorised access for the right person and task.
- Unexpected excessive access should be reported through the access-control process, with the facts recorded.
167Your manager gives you two assignments with overlapping deadlines. How would you prioritise?
- I would compare risk, external commitments and dependencies, then explain the conflict with realistic time estimates.
- For this role I would consider query ageing, unresolved critical discrepancies and review quality alongside the immediate consequences.
- I would agree priorities and ownership rather than silently choose which commitment to miss.
168A reviewer disagrees with your technical conclusion. What would you do?
- I would clarify the exact point, compare the evidence and applicable requirements, and explain my reasoning respectfully.
- If the issue remains unresolved, I would use the data-management lead and relevant clinical or safety team and document the decision appropriately.
- I would not change source facts to make the disagreement disappear.
169A source record contradicts a summary or dashboard. Which do you trust?
- I would investigate the origin, version, definitions and transformations rather than blindly choose either.
- I would trace the summary back to authorised evidence and determine whether the mismatch affects other outputs.
- I would raise the issue through the data-management lead and relevant clinical or safety team and preserve the correction trail.
170How would you hand over unfinished work to another colleague?
- I would record the current status, completed checks, source locations, unresolved issues and time-sensitive risks.
- I would identify the next owner and confirm understanding, using approved systems.
- In Clinical Data Management, the handover should make the status of an eCRF record and its authorised source unambiguous.
171A productivity metric improves while significant errors increase. What would you say?
- I would check the metric definitions and investigate the quality trend before calling the change a success.
- Relevant measures include query ageing, unresolved critical discrepancies and review quality.
- I would present both speed and quality evidence, propose proportionate improvements and avoid treating hidden rework or safety risk as acceptable throughput.
172You are asked to perform an unfamiliar technical task without adequate training. What would you do?
- I would explain my current competence and ask for the required training, supervision and authorisation.
- I could assist within an agreed safe scope while learning, but would not claim independent capability.
- I would involve the data-management lead and relevant clinical or safety team if the request could create a material risk.
173A stakeholder wants a more favourable answer than the evidence supports. How do you respond?
- I would explain what the evidence does and does not show, identify uncertainty and offer an accurate alternative statement or further work.
- I would not alter results, remove inconvenient facts or invent justification.
- If pressure persists, I would use the appropriate quality, ethical or compliance escalation route.
174How would you explain a problem with an eCRF record and its authorised source to a nontechnical colleague?
- I would describe what is wrong, why it matters, what activity is affected and the decision needed in plain language.
- I would separate known facts from uncertainty and provide a clear owner and next step.
- I would avoid excessive jargon and unsupported accusations about who caused it.
175How would you respond to repeated critical feedback during training?
- I would ask for concrete examples, compare them with the required standard and identify the main gap.
- I would agree a practice and review plan, use a focused checklist where appropriate and ask for reassessment.
- I would show progress with evidence rather than become defensive or hide difficult tasks.
176A laboratory transfer uses mg/dL but the database expects mmol/L. What do you do?
- Stop the affected import.
- Verify analyte-specific conversion and approved mapping with the vendor and lead.
- Retain the reported units and test the corrected transfer before acceptance.
177An edit check raises 800 queries after an amendment. How do you respond?
- Check whether the new logic matches the approved requirement.
- Sample valid and invalid triggers before sending more queries.
- Correct faulty logic under change control and document the affected records.
178A site answers Correct as entered, but an impossible date remains. What next?
- Review the exact contradiction and prior response.
- Ask a focused follow-up through the authorised channel.
- Escalate persistent uncertainty; do not silently change the site's date.
179An external file is missing one subject but its total count matches. What do you investigate?
- Check record-level keys and duplicate subjects.
- Compare expected visits and specimen identifiers.
- Resolve the missing and extra records separately instead of accepting the aggregate count.
180An unblinded treatment field appears in your review export. What is your action?
- Restrict further distribution and follow the blinding-incident process.
- Notify the authorised lead and preserve evidence.
- Use a corrected masked export for subsequent work.
181A safety case exists but the corresponding EDC event is absent. What should happen?
- Confirm participant and event identity with authorised teams.
- Follow the reconciliation plan and request site clarification.
- Ensure required safety handling continues independently of EDC correction.
182A vendor resends yesterday's file with changed results and the same name. What do you check?
- Compare content and delivery history.
- Obtain correction scope and reason from the vendor.
- Load only through the authorised corrected-file process and reassess downstream outputs.
183An early-withdrawal subject has many missing-visit flags. How do you decide which matter?
- Check withdrawal date, disposition and protocol follow-up requirements.
- Distinguish expected assessments from no-longer-applicable visits.
- Resolve genuine gaps without manufacturing records for absent visits.
184The statistician requests manual values to complete a final dataset. How do you respond?
- Clarify whether these are approved analysis derivations or unobserved source data.
- Preserve the collected dataset and document authorised derivations separately.
- Do not enter invented observations into EDC.
185After lock you discover a wrongly mapped laboratory unit affecting an endpoint. What next?
- Notify the data lead and analysis owner immediately.
- Assess affected records and conclusions through approved governance.
- Follow authorised unlock, correction, verification and controlled redelivery steps if required.
Salary negotiation & offersQuestions 186–195
186How should you answer What are your salary expectations?
- Ask about the approved range and clarify location, responsibilities and experience level.
- Give a researched range you would genuinely consider: 'For this scope I am targeting, depending on fixed pay and the complete package.' Avoid inventing competing offers or treating an online average as a guaranteed entitlement.
187How do you negotiate salary after receiving an offer?
- Thank the recruiter, confirm your interest and make one clear, evidence-based request: 'Based on [relevant experience or skill] and the responsibilities discussed, could the fixed component be revised to?' Ask politely, then listen.
- Negotiate against the complete written terms, not an assumed verbal package.
188What is the difference between CTC, fixed pay and take-home pay?
- CTC can include employer costs and benefits as well as salary.
- Fixed pay is the non-variable compensation component as defined in the offer; take-home reflects deductions and applicable taxes.
- Request an itemised breakdown.
- Do not estimate your monthly bank credit simply by dividing total CTC by twelve.
189How should a fresher negotiate without work experience?
- Use relevant project work, assessed practical skills, certifications with demonstrable learning and the role's responsibilities.
- Ask whether the fresher band is flexible and whether a defined review is possible.
- Remain honest about competence; a certificate alone does not establish production experience or justify a fabricated market rate.
190How can you discuss compensation when your current salary is low?
- Bring the discussion back to the new role's scope and the skills you can demonstrate.
- 'My present package reflects a different role; for these responsibilities I am looking for.' Answer factual questions truthfully and avoid altering payslips or claiming a nonexistent offer.
191What if the recruiter says the salary budget is fixed?
- Ask whether any genuine flexibility exists in joining support, shift allowance, location, training or a documented review date.
- Evaluate these separately from recurring fixed pay.
- If the package does not meet your needs, decline professionally rather than accept terms you already know you cannot sustain.
192How should you evaluate variable pay, incentives and bonuses?
- Ask how each component is calculated, when it pays, whether it is guaranteed, what conditions apply and what happens during probation or departure.
- Distinguish recurring salary from a one-time joining bonus.
- Ask for the terms in writing instead of counting the maximum possible payout as certain income.
193How do you compare two offers with different packages?
- Compare role content, fixed pay, realistic variable pay, hours, location costs, training, benefits, probation and notice terms.
- Use the same annual period for both.
- Identify your own priorities; the larger CTC may not produce better monthly income or the experience needed for your career direction.
194How do you ask about a bond, recovery clause or notice period?
- Request the complete written terms and clarify amounts, triggers, duration and repayment conditions before accepting.
- Say: 'Could you explain when this recovery applies and provide the policy?' Do not rely on a recruiter's informal assurance if the written offer differs; obtain appropriate advice where needed.
195How should you respond if pressured to accept an offer immediately?
- Express interest and request a reasonable review deadline: 'Thank you.
- I would like to review the complete written offer and confirm by.' Clarify outstanding questions first.
- Never resign solely on an unclear verbal promise, and do not manufacture an emergency or competing offer to gain leverage.
When you do not know an answerQuestions 196–200
196What should you say when you do not know an answer?
- Say calmly: 'I do not know that detail confidently, and I do not want to guess.
- I understand …; I would verify the remaining point in the approved reference or ask the appropriate expert.' Explain any sound knowledge you do have.
- Do not turn uncertainty into a fabricated rule.
197What if you understand the concept but forget the exact term or number?
- Explain the concept accurately and separate it from the detail you cannot recall.
- For example: 'I understand the sequence, but I would check the current specification before quoting a limit.' If invited to reason aloud, state your assumptions clearly rather than presenting an estimate as an approved value.
198How should you respond to an unclear or unfamiliar question?
- Ask a focused clarification: 'Do you mean the clinical-trial process or the postmarketing process?' or 'Which product and market should I assume?' Then answer within that scope.
- Clarification is useful professional judgement, not an excuse to avoid a question you understand.
199What if the interviewer points out that your answer is wrong?
- Listen, acknowledge the correction and restate the corrected understanding briefly.
- If your earlier answer depended on a different setting, explain that respectfully without arguing.
- A useful response is: 'Thank you; I mixed those concepts.
- In this setting the distinction is …'.
- Show that you can learn and correct errors.
200Would you guess when asked an unfamiliar safety-critical question?
- No.
- Explain what action you would pause, which authorised person or reference you would consult and how you would preserve the record.
- In an interview, describe a safe decision process.
- Do not imply that you would independently change a patient treatment, release a batch or override a control to avoid admitting uncertainty.
Questions to ask & interview follow-upQuestions 201–210
201What should you say when asked Do you have any questions for us?
- Prepare two or three relevant questions rather than asking all of them mechanically.
- A useful opening is: 'Yes, thank you.
- What would you most like the person in this role to achieve during the first three months?' Listen to the answer and ask one focused follow-up.
202How can you ask about training and supervision?
- Ask: 'How is initial training organised, and how do you assess readiness for independent work?' This shows that you want to build competence responsibly.
- Avoid implying that you expect to work without required training or that a course certificate should exempt you from the employer's qualification process.
203How can you ask about the team's actual work?
- Ask which responsibilities, systems, products or projects this position would support and how work is divided across the team.
- Use the department guide to make the question specific.
- Do not ask for confidential client names, patient information or unpublished company plans.
204What is a good question about performance expectations?
- Ask: 'How do you balance quality, timeliness and workload when assessing this role?' Follow up on the review and feedback process.
- This gives a more useful picture than asking only how many items you must finish each day without considering their complexity or quality.
205When can you ask about shifts, compensation or practical conditions?
- Ask the appropriate recruiter or interviewer at a suitable point, especially if a condition affects your availability.
- Be polite and specific about what you need to understand.
- You do not need to promise acceptance before receiving the terms, and there is no universal rule that candidates must never discuss pay.
206How should you close the interview?
- Thank the interviewers, briefly reaffirm your interest and ask about the next stage and expected communication process.
- For example: 'Thank you for explaining the role.
- I am interested in the opportunity; could you share the next steps?' Avoid demanding an immediate decision or claiming qualifications you did not demonstrate.
207How can you ask about progression without sounding focused only on promotion?
- Ask which skills distinguish a strong performer at the next level and how feedback and development are organised.
- This connects progression to competence.
- Avoid demanding a guaranteed promotion date or assuming every associate follows the same path regardless of business needs and performance.
208What should you ask about the role's current challenges?
- Ask: 'What is the main challenge the team would like this hire to help solve?' Listen for specific work and dependencies, then connect one relevant skill honestly.
- Avoid promising to solve a complex quality or delivery problem before understanding the evidence and authority required.
209How do you clarify whether the role is contract, permanent or client-deployed?
- Ask who the legal employer is, the employment type, contract duration if applicable, work location and reporting structure.
- Direct employment and client assignment are different arrangements.
- Obtain the relevant written terms rather than assume that a client brand in the job advertisement is your payroll employer.
210How should you follow up after the interview?
- Use the agreed recruiter channel and timeline.
- Send a short note thanking them, naming the role and asking about the next step if the stated date has passed.
- Avoid repeated messages to several interviewers or sending confidential work samples to demonstrate enthusiasm.
Your final interview check
- Explain one real task from input to result.
- Revise the technical topics in your job description.
- Prepare two truthful examples using Situation, Task, Action and Result.
- Know your salary priorities and prepare two questions for the interviewer.