Interview guide for research roles: a loop for research associates, scientists and study coordinators

Research hiring has an unusual problem: the evidence on a resume is collective. Papers have many authors, posters list whole labs and a project described as "my work" may have been one person's assay inside someone else's design. The best research hires also share traits that are hard to see in a conversation about results, such as careful records, honest reporting of failed experiments and the judgment to question their own data. A good loop separates what the candidate did from what the lab did, tests scientific reasoning on fresh material, and checks the integrity and regulatory points specific to research employers.

This guide covers research assistants, research associates and staff scientists in academic, nonprofit and industry R&D, plus clinical research coordinators at study sites. Clinical labs that report patient results are covered in the interview guide for lab roles, because CLIA personnel rules apply there. For computational research roles, the interview guide for data scientists has a modeling exercise, and the first call is in data scientist phone screen questions.

The loop at a glance

For senior scientists, add a 30-minute research talk before stage 2 and a conversation about funding and direction with the department or R&D head.

Stage 1: the screen (20 minutes)

• "This position is funded through [example: a grant ending in two years]. How does that fit your plans?" Be honest about term-limited funding at the start.
• "Which of the techniques in the posting have you run independently, start to finish?" Ask for the last time they ran each one.
• "Does the work schedule work for you, including time-sensitive experiments or animal care checks?" Describe it concretely.

Stage 2: the methods deep dive (45 minutes)

Ask the candidate to choose one project from their resume. Spend the whole stage on it. Follow the same sequence with every candidate:

1. "What question was the project answering, and why did it matter to the lab?" Listen for: a clear question in plain language.
2. "Which parts did you design, which did you run and which were done by others?" Listen for: precise boundaries. Strong candidates name collaborators without being asked.
3. "Tell me about an experiment in this project that failed. How did you find out why?" Listen for: controls, hypotheses about failure and what they changed. This question separates hands-on work from proximity.
4. "If you started over with the same budget, what would you do differently?" Listen for: honest critique of the design, not just "more time".
5. "How did you record what you did? Could someone repeat it from your notes?" Listen for: dated, complete records, versioned protocols and raw data kept alongside results.

Do not ask for confidential data from a current employer. If the work is unpublished or proprietary, ask the candidate to describe the approach at the level they are allowed to share, and score the reasoning rather than the result.

Stage 3: the figure or data critique (25 minutes)

Build a one-page result in your field with a planted problem. Examples: a bar chart with no error bars and an n that does not match the methods text; a western blot without a loading control; a survey result reported as significant with a very small sample and many comparisons; a clinical lab value trend where one site's data were collected differently. Ask: "What would you want to check before you believed this?"

What to score

• Detection: finds the planted problem, or a real one you did not plant.
• Reasoning: explains why it matters to the conclusion, not just that it looks wrong.
• Next step: proposes a specific check or control.
• Tone: critiques the work, not the person who made it, which predicts how they will handle lab meetings.

Stage 4: integrity and documentation interview (30 minutes)

1. "Tell me about a time results did not support what your supervisor expected. What did you do?" Listen for: reporting the result as it was, and discussing it openly.
2. "Have you ever found an error in data that had already been shared or published? What happened?" Listen for: telling the PI and correcting the record.
3. "How do you decide whether to exclude a data point?" Listen for: criteria set in advance and documented, never exclusion because a point is inconvenient.
4. "Which compliance trainings have you completed, and which approvals did your last project need?" Listen for: familiarity with the approvals relevant to your work, such as human subjects, animal use, biosafety or export control review, without expecting them to know your institution's forms.

For clinical research coordinators, replace the third question with a protocol deviation scenario: a participant came outside the visit window and a required lab draw was missed. Score whether they report the deviation, document it and tell the investigator, rather than adjust dates in the record. Add a short informed consent role-play in which the "participant" asks whether they can leave the study later.

Stage 5: the team conversation (30 minutes)

Two lab members ask about shared equipment, authorship disagreements and covering each other's experiments. Give them two questions from the scorecard so the conversation adds evidence rather than impressions. A useful one: "Tell me about a disagreement over credit or authorship. How was it resolved?"

Scorecard competencies and weights

Example weights for a research associate. Lock yours before the first candidate.

For staff scientists, add research direction and mentoring. For coordinators, replace technical contribution with protocol compliance and participant communication.

References and verification

Ask the candidate's former PI or manager the same three questions: what the candidate personally did on the named project, how they handled a failed experiment, and whether you could rely on their records without checking them. Confirm degrees with the issuing institution, and confirm dates and titles as described in how to verify employment history. The HHS Office of Research Integrity publishes case summaries of research misconduct findings in research it oversees; your research integrity office can advise whether and how to consult them.

Legal and regulatory points to check

Checked against the linked primary sources as of October 2026. Not legal advice; your export control officer, sponsor and counsel decide.

• Export controls without citizenship screens. Under 15 CFR 734.13, releasing controlled technology or source code to a foreign person in the United States is a "deemed export". That can require a license for some roles, but the Justice Department's citizenship status discrimination fact sheet (updated April 2024) says the ITAR and EAR do not contain hiring requirements and do not require employers to limit jobs to U.S. citizens. Ask whether candidates are authorized to work and need sponsorship; route export licensing questions to your export control office, separately from the Form I-9.
• FDA debarment in drug research. Under 21 U.S.C. 335a(k), drug product applications must certify that the applicant did not and will not use in any capacity the services of a person debarred under that section. If you hire for clinical trials or regulated drug development, check the FDA debarment list and your sponsor's requirements.
• Exclusions. Study sites that bill federal health programs should also screen against the HHS OIG exclusion list.
• Background checks. When a consumer reporting agency runs the check, 15 U.S.C. 1681b(b) requires a stand-alone disclosure, written authorization and pre-adverse action notice. See the FCRA background check process.

The decision rule

1. Floor: at least 3 on hands-on contribution; nothing below 2 on integrity.
2. Weighted total: in this example, 2.8 or higher on a 1–4 scale is an offer.
3. Contingencies: references from a supervisor on the named project, degree verification, and any export licensing or debarment checks the role needs.

Mistakes that weaken research loops

• Counting publications. Authorship order and lab culture vary too much to compare.
• Letting the research talk replace the interview. Polished talks reward rehearsal, not contribution.
• Skipping failure questions. Most research is failed experiments; ask how they handle them.
• Adding citizenship requirements by habit. Confirm a legal requirement first.