Medpace is a scientifically-driven, global clinical research organization (CRO) that helps biotech and pharmaceutical companies bring life-saving therapeutics to market. As a Data Scientist at Medpace, you do not just build predictive models in a vacuum; you work at the critical intersection of clinical trial operations, medical data, and Biostatistics. Your work directly impacts the design, execution, and analysis of clinical trials, ensuring that clinical data is processed accurately, analyzed rigorously, and prepared for regulatory approval.
The data you will work with is highly complex, regulated, and critical to patient safety. You will collaborate closely with Biostatisticians, clinical database programmers, and medical monitors to optimize trial designs, build data pipelines, automate quality control, and extract insights from clinical trial databases. This role is highly visible and intellectually demanding, requiring a balance of advanced statistical knowledge, programming expertise, and an understanding of clinical trial methodology.
For professionals who want their analytical skills to have a direct, positive impact on global healthcare, the Data Scientist role at Medpace offers a unique platform. You will tackle challenges ranging from analyzing survival data to optimizing trial protocol compliance, all while maintaining the highest standards of scientific and regulatory integrity.
Phone Screen
reportedMost candidates lose this call inside the first two minutes, during the walkthrough of their own background. The account runs chronologically, sits at the level of tools and titles, and never arrives at a decision anyone could have disagreed with. Anchor on a problem instead of a timeline: what the team could not answer, what you did about it, what happened next. Ninety seconds is enough, and stopping on time leaves room for the half of the call that belongs to you. What you ask about how work gets prioritised signals your level more reliably than the walkthrough does.
What to demonstrate
- Whether your background summary has a shape (problem, decision, consequence) or is a chronological list of tools and employers
- Whether you can account for gaps, short stints and the reason you are looking, unprompted and without hedging
- The substance of the questions you ask back, which an experienced screener reads as a level signal
How to prepare
- Time your opening walkthrough against a clock. If it runs past two minutes, compress the earliest role into a single clause and spend the recovered time on the most recent one
- Write one honest sentence for every gap or short stint visible on your resume and offer it before being asked about it
- Prepare questions about how work arrives and gets prioritised: who writes the request, how often priorities change, and what happens to an analysis after it is delivered
Panel Interviews
reportedWhere a loop includes a partner from outside the data team, that conversation usually carries the same weight as the technical ones and gets the least preparation. The person opposite you will not follow a derivation and does not need to. They are working out whether having you involved would make their decisions better or slower. The failure mode is not being too technical. It is answering a question about a decision with a description of your method, leaving the translation to them. What they carry into the debrief is the sentence you handed them, not the analysis underneath it.
What to demonstrate
- Whether a statistical result arrives as something the partner could act on, with the one caveat that would change their decision kept and the rest left out
- Whether you can state what you need from their side, in their terms: instrumentation that does not exist yet, a definition they own, or a holdout they have to agree to
- Whether uncertainty is given as a range someone can plan against, rather than as hedging that invites them to ignore the result
- Whether you ask what decision is actually on the table before explaining anything
How to prepare
- Take a result you know well and write the version for someone who stops reading after one sentence, then the three-minute version, and check the short one is not the long one with the qualifications stripped out
- For a past project, list everything you asked a non-technical partner for and how you phrased it, then rewrite each ask so it names what goes unmeasured without it
- Practise saying where a result does not apply, out loud, in one sentence that a partner could repeat accurately to someone else
Technical Presentation
reportedThis round decides whether someone can hand you a schema and a question and trust the number that comes back. Correctness under a clock is the bar, not clever syntax. The habit that separates strong from weak answers is checking the grain: after every join, know how many rows you expect and whether the count moved. Most wrong answers in this format are not wrong logic, they are a fan-out from a key that turned out not to be unique, or a filter applied before an aggregate when it belonged after. Say what you expect before you run it.
What to demonstrate
- Whether your row counts survive each join, and whether you notice on your own when they do not
- Deliberate handling of rows that fail to match, including whether the question needs an inner join or a left join with the non-matches kept and counted
- Whether NULLs are treated on purpose, given that a NULL compares equal to nothing and that COUNT of a column skips it
- Reaching a defensible answer inside the window instead of a refined one after it
How to prepare
- Take a two-table schema, write a join that fans out on purpose, then fix it by collapsing the many-side to one row per key before joining. Repeat until the fix is reflex rather than recall.
- Write a funnel as one query and print the distinct user count at each stage, then confirm each stage is a subset of the one above it rather than assuming it
- Do a few timed runs in a plain text box with no autocomplete and no formatter, since assessment editors often have neither
PracHub editorial advice for the preparation topics above.
Ignoring clustering by provider, facility, or site
Patients within a provider panel, and subjects within a trial site, are correlated on both case mix and practice pattern. Standard errors computed as if observations were independent are too narrow by roughly the design effect, 1 + (m - 1) * ICC, where m is the average cluster size. With 50 patients per provider and an intracluster correlation of only 0.02, variance is understated by about a factor of two, which manufactures significant provider differences out of noise. Cluster-robust errors or a random intercept per provider, and cluster-level randomisation when you design the test, are the corrections.
Reading the most recent months of a claims-based series as real
Claims incur before they are reported and paid, so recent incurred months are systematically undercounted until runout completes. The lag is not uniform: pharmacy adjudicates in days, professional claims in weeks, inpatient facility claims in months. That means recent data is both too low and mix-shifted toward cheap services, which reads as a cost improvement and a utilisation drop at once. The fix is to hold the last three incurred months back or apply completion factors, and to state the paid-through date on every chart.
Reaching for a model before the target metric exists
Before naming an algorithm, write down the label, the prediction time, and the action that changes when the score crosses a threshold. If you cannot say what decision the output drives, any modelling choice is guesswork dressed up as method.
Dropping rows with missing values without naming the mechanism
Say whether the values are missing at random, missing by a known process, or missing in a way that depends on the outcome, and handle them accordingly. Deleting incomplete rows silently redefines the population whenever missingness correlates with what you are measuring.
Choose a category, try a prompt, then open its approach, worked solution or follow-up when you need it.
What is the difference between parametric and non-parametric statistic…
What is the difference between parametric and non-parametric statistical tests, and when would you choose one over the other?
Approach
- Write down the assumption the method needs before you use the method.
- Quantify uncertainty explicitly rather than reporting a point estimate alone.
- Sanity-check the answer against a simple bound or a simulated case.
Follow-up
- What sample size would you need to detect an effect half this size?
- How would you explain this result to someone who does not know statistics?
Explain the concept of survival analysis and how a Cox Proportional Ha…
Explain the concept of survival analysis and how a Cox Proportional Hazards model works.
Approach
- Say how the offline result would be validated online before it is trusted.
- Pick an evaluation metric that matches the cost of each error type, not a default.
- Frame the prediction: the label, the moment of prediction, and the action it triggers.
Follow-up
- Where could label leakage enter this setup?
- What would you monitor after launch to know the model is still valid?
Collapse claim versions before summing allowed amounts by member
You are given medical_claim_line as a pandas DataFrame: claim_line_id, claim_id, claim_version, frequency_code (1 original, 7 replacement, 8 void), member_id, service_start_date, procedure_code, allowed_amount, claim_status. Return total allowed_amount per member for service dates in one stated calendar month, counting each claim once at its surviving version and excluding any claim whose surviving version is a void. Multi-line claims must keep every line of that surviving version. Output columns: member_id, allowed_amount, sorted descending. Do not deduplicate on claim_id alone.
Approach
- Separate the two grains out loud before writing code: versions live at the claim level, dollars live at the line level. Every bug in this exercise comes from mixing them.
- Compute the surviving version per claim with a groupby transform of max over claim_version, then keep the lines whose claim_version equals that value. A transform keeps the frame at line grain, which a sort plus head(1) does not.
- Drop whole claims whose surviving version carries frequency_code 8. A void is not a zero-dollar line, it retracts the claim, so filtering rows rather than claims leaves the prior version's lines behind.
- Filter to claim_status 'paid' and state the assumption: denied, pended and reversed lines carry an allowed_amount that never became a liability, so including them overstates cost.
- Restrict service_start_date to the month boundaries, then groupby member_id and sum. Sort descending and return.
Worked solution 20 min
- Build a max-version lookup per claim_id with groupby('claim_id')['claim_version'].transform('max') and keep rows where claim_version equals it.
- From the surviving rows, collect the claim_ids whose frequency_code is 8 and drop those claim_ids entirely.
- Filter claim_status to 'paid' and service_start_date into the month window.
- Group by member_id, sum allowed_amount, sort descending, reset the index.
Follow-up
- A claim has versions 1, 7, 7 where the second replacement has fewer lines than the first. What does your code return and is that correct?
- How would you handle a claim whose lines span two calendar months at the boundary of the reporting window?
- The same logic has to run over 400 million lines in a warehouse. What changes, and what stays the same?
How would you optimize a slow-running data step or query when working …
How would you optimize a slow-running data step or query when working with millions of clinical patient records?
Approach
- Check whether any join is one-to-many before aggregating, or the sums inflate.
- Say which table is the grain you start from, and join outward from it.
- Handle the rows that do not match: a LEFT JOIN with a NULL check is usually the question.
Follow-up
- What breaks if events arrive late or out of order?
- How would you verify this result without re-running the same query?
Explain the difference between a `MERGE` and a `JOIN` operation, and h…
Explain the difference between a MERGE and a JOIN operation, and how you handle many-to-many relationships in SAS.
Approach
- State the window function and its partition and ordering out loud before writing it.
- Say which table is the grain you start from, and join outward from it.
- Compute rates by summing numerator and denominator separately, never by averaging rates.
Follow-up
- How would you verify this result without re-running the same query?
- What breaks if events arrive late or out of order?
Turn overlapping coverage spans into fractional member-month denominators
member_enrollment holds span_id, member_id, plan_id, product_type, effective_date and termination_date, which is NULL while the span is active. A member can hold several spans in a year, and back-dated plan changes make spans overlap. Return one row per calendar month of 2025 with total member-months, where a member contributes covered days in the month divided by that month's length, and a day covered by two overlapping spans counts once. Both span endpoints are inclusive. Collapse overlaps with window functions before joining a month calendar, and do not assume one span per member.
Approach
- Normalise the open end first. COALESCE(termination_date, DATE '9999-12-31') or clip to the reporting end, because a NULL termination_date makes every BETWEEN comparison unknown and drops exactly the members who are still covered.
- Merge overlapping spans per member before the calendar join. Order by effective_date, carry a running MAX of prior termination_date with a frame ending one row before, and open a new island where effective_date exceeds that running max plus one day, which keeps adjacent spans contiguous rather than splitting them.
- Clip each merged island to the reporting year, then cross join to a twelve-row month calendar and keep pairs that intersect.
- Compute covered days as LEAST(island_end, month_end) minus GREATEST(island_start, month_start) plus one, and divide by the actual length of that month rather than a flat 30, so February and the 31-day months are not distorted.
- Sum the fractions by month. Merging after the calendar join also works but forces a DISTINCT over member-days, which stops scaling at population size.
Worked solution 30 min
- CTE bounded: clip spans to 2025-01-01 through 2025-12-31 with COALESCE on termination_date, discarding spans that do not intersect the year.
- CTE merged: gaps-and-islands per member using LAG and a running MAX over preceding rows to collapse overlapping and adjacent spans.
- CTE months: twelve rows of month_start, month_end and days_in_month.
- Join merged to months on interval intersection and compute covered_days per pair.
- SUM(covered_days / days_in_month::numeric) grouped by month, ordered by month.
Follow-up
- Where does a continuous-enrolment filter with an allowable 45-day gap belong, and how does that change the island boundary rule?
- One human appears under two member_ids after a master person index merge. What does that do to member-months, and how would you detect it?
- How do you treat a span that terminates on the day it starts, and what does that row usually represent?
Evaluate a book-wide benefit change with no untreated members
On 1 January a therapeutic class moves from formulary tier 2 to tier 3, raising member cost share, for the entire commercial group book at once. No untreated subset of that book exists. You have 36 months of pharmacy_claim and member_enrollment covering that book plus the Medicare Advantage and Medicaid managed books, which are unaffected. Outcomes are class fills per 1,000 member-months and the share of members with PDC at or above 0.80. Give the design, the identifying assumption, and how you would test it.
Approach
- Rule out the single-series interrupted time series first and say why. The change lands on 1 January, which is also when deductibles reset, benefit designs change, and members switch plans, so a level shift at the new year is confounded with the entire annual benefit cycle. Any design that cannot separate those two is not usable here.
- Build an event-study difference-in-differences with the unaffected product books as controls, estimating month-by-month coefficients relative to December. The lead coefficients are the test of the identifying assumption, not decoration: they should be indistinguishable from zero, and you should say in advance what magnitude of lead violation would make you abandon the design.
- State what the control books are doing for you and what they are not. They experience the same January reset, which is precisely why they are useful, but they differ in population age, cost-sharing structure, and eligibility churn, so parallel counterfactual trends is a real assumption and residual confounding should be named rather than waved at.
- Add a synthetic control as the robustness check: construct a weighted combination of donor series (other product books crossed with other therapeutic classes) that tracks the treated series through the pre-period, and do inference by permutation, assigning the intervention to each donor and comparing post-to-pre RMSPE ratios. Require that the treated pre-period sits inside the convex hull of the donors, and abandon the approach if pre-period fit is poor.
- Fix the population before estimating anything. If the tier move pushes members out of the book, the post-period population differs from the pre-period one, so restrict to members continuously enrolled across the window with a stated allowable gap, and report sensitivity to that gap. Note that pharmacy claims adjudicate in days, so unlike a medical-cost series this outcome is essentially complete within weeks.
Worked solution 45 min
- Build a monthly panel of fills per 1,000 member-months and PDC share by product book, restricted to members continuously enrolled across the window with a 45-day allowable gap.
- Plot the treated and control series for 24 pre-period months and inspect visually before fitting anything.
- Estimate an event-study DiD with month-relative-to-January coefficients, product fixed effects, and calendar-month fixed effects, clustering standard errors at the product-book level or higher.
- Run the placebo: re-estimate as if the change happened on the prior 1 January, which isolates the annual reset from the tier move.
- Fit a synthetic control on donor product-class series, check pre-period RMSPE, and run the permutation inference across donors.
Follow-up
- The lead coefficients are not flat: the treated book was already drifting down for six months. What do you do without abandoning the question?
- Therapeutic substitution is expected. How do you tell a true reduction in therapy from members moving to a cheaper drug in the same class?
- How would you report the result so a clinical committee can act on it, given that fills per 1,000 member-months is not an outcome anyone treats?
Redesign an adherence metric an outreach team cannot quietly inflate
An outreach team is paid on the share of members reaching proportion of days covered of 0.80 or higher in a therapeutic class, computed from pharmacy_claim (member_id, therapeutic_class_code, fill_date, days_supply, refill_number, reversal_flag, reversed_fill_id). The denominator is members with at least two fills in the class and at least 91 days of follow-up, over a 12-month window. Name three ways the team raises this number without any member taking more medication, then redesign the metric and its guardrails so each lever is closed or visible.
Approach
- Attack the denominator first, because it is the cheapest lever. A member who fills once and abandons never enters the denominator at all, so the worst outcome in the class is invisible, and any effort spent on first-fill abandonment lowers the reported share by moving those members in. The incentive runs backwards.
- Attack the window arithmetic second. With the denominator running from first fill to window end, a member whose first fill lands late in the year has a short denominator that is trivially easy to cover, so steering new starts to the last quarter raises the metric mechanically. Report by index quarter or fix the window and require the index fill in the first quarter.
- Attack the threshold third. Days of supply dispensed late in the window count only up to the window end, but a fill placed 60 days before close, on a member whose last on-hand day has already passed, adds up to 60 covered days, about 0.16 of a 365-day denominator. That alone moves a member from 0.77 to 0.93 with no change in what anyone swallows.
- Close the dispensing-side hole: a fill that is billed and reversed is not exposure. Exclude both rows of a reversed pair via reversal_flag and reversed_fill_id, and collapse overlapping fills by pushing each interval start to the later of fill_date and the prior interval's end, so early refills shift coverage forward rather than stacking.
- Redesign the scorecard as a pair rather than a single number: the threshold share stays as the headline, and mean PDC plus the share at 0.90 or higher travel with it, because a distribution cannot be bunched at a threshold without showing it.
- Add the guardrails that make each lever visible: first-fill abandonment rate (fills with refill_number 0 never followed by a second fill, plus reversed first fills), the share of covered days originating from fills in the final 60 days of the window, and the distribution of PDC in the bands just below and just above 0.80.
Follow-up
- The share at 0.80 or higher rose four points and mean PDC rose 0.4 points. What is your reading, and what do you look at next?
- Would you pay on 0.80 at all, or on the continuous measure? Say what each choice costs you.
- Possession is not ingestion. Name a measurable consequence in another table that would corroborate a real adherence improvement, and state its own weakness.
Protocol deviations fell as new sites came online
Protocol deviations per 100 completed visits across a multi-site study fell from 4.6 to 2.9 over two months while randomisation accelerated. study_subject_visit carries study_id, site_id, subject_id, visit_number, visit_name, planned_visit_date, actual_visit_date, visit_window_days, visit_status, protocol_deviation_flag, deviation_severity, data_entry_ts and site_activation_date. Operations wants to credit a retraining rollout that began in that window. Establish whether deviation behaviour changed, and deliver a version of the series that is safe to review monthly.
Approach
- Treat deviation flags as late-arriving data. Build a lag triangle of data_entry_ts minus actual_visit_date, because deviations are largely identified during monitoring review weeks after the visit, so the two most recent months are structurally undercounted in exactly the way an improvement looks.
- Hold back the months that are not developed, or apply development factors fitted on fully developed months, and restate. Do this before any explanation involving training.
- Attack the denominator. It counts visits with visit_status 'completed', while out_of_window and missed visits are excluded even though out-of-window attendance is itself a deviation in most protocols, so a site that pushes visits out of window lowers the rate twice over.
- Standardise the visit mix. Newly activated sites contribute mostly screening and baseline visits, which carry fewer procedures and fewer opportunities to deviate than later treatment visits, so compute the rate within visit_number strata and reweight to a fixed mix before comparing months.
- Split major from minor deviation_severity and report them separately. A single major deviation can remove a subject from the per-protocol analysis set while ten minor ones may not, so a pooled rate can fall while the consequential series rises.
- Before attributing anything to training, model at the right level: subjects are nested in sites, so use a random intercept per site or cluster-robust standard errors. With average cluster size m and intracluster correlation rho the variance is understated by the design effect 1 + (m - 1) times rho, which at m = 60 and rho = 0.02 is about 2.2, so naive intervals are roughly 1.5 times too narrow.
Follow-up
- Design a defensible evaluation of the retraining rollout, given it was deployed site by site and cannot be withheld.
- Your reweighted series is flat but major deviations rose. What do you escalate and to whom?
- How would you decide the holdback period, and how would you communicate a provisional number without it being quoted as final?
For someone who can already write the query and train the model but stalls when asked what to measure or whether a change is worth making. Metric definition and case structure come first; the technical work is kept as maintenance rather than the centre of the week.
Prepare, practise & reflect
One practical outcome each day. Spend longer where you need it.
0 / 7 done01Metric anatomy
- For three products you use daily, write one primary metric, two input metrics that plausibly move it, and one guardrail that would catch a cheap way of moving the primary at the cost of the product.
- For one of them, specify the metric precisely enough that two analysts would return the same number: numerator, denominator, unit of observation, time window, and how returning and deleted accounts are treated.
- Pick a ratio metric and write what happens to it when the denominator shrinks for reasons unrelated to the numerator, with a concrete example of that happening.
Deliverable: A one-page metric tree for one product, with the primary metric written as an unambiguous spec.
Practice prompt ↗Practice prompt ↗Worked solution ↗02Diagnosing a drop without guessing
- Take the prompt "weekly active users fell 8 percent week over week" and write the segmentation plan before proposing any cause: platform, region, tenure cohort, acquisition channel, and whether the movement sits in the numerator or in a changed denominator.
- List the instrumentation failures that manufacture fake drops (a client release that stopped firing an event, a bot filter change, a shifted date boundary or timezone) and write the query that rules out each one.
- Rehearse stating the boring explanations first, seasonality and day-of-week composition, before reaching for a product cause.
Deliverable: A drop-diagnosis checklist short enough to recite from memory in under a minute.
Practice prompt ↗Practice prompt ↗03Should we build it
- Take a feature idea and write it as a bet: what you believe is true, what would have to be true for it to pay off, the metric that would confirm it, and the effect size that would justify the engineering cost.
- Size the opportunity top-down and bottom-up, then reconcile the two numbers in writing instead of quoting whichever is friendlier.
- Write the counter-metric that would make you kill the feature even if it wins on the primary metric.
Deliverable: A one-page product memo ending in a decision rather than a list of considerations.
Practice prompt ↗Practice prompt ↗04The places aggregate numbers lie
- Construct a Simpson's paradox numerically: two segments where the treatment wins within each segment yet loses overall, and identify the shift in segment weights that causes it.
- Take a heavy right-tailed quantity such as revenue per user and write why the mean is the wrong summary, which percentile you would report instead, and what a moving mean with a stable median tells you.
- Write your definition of a session for the product from day one, then name two real behaviours it misclassifies.
Deliverable: One page holding a worked Simpson's paradox table and a session definition with its two known failure cases.
Practice prompt ↗Practice prompt ↗Worked solution ↗05Technical maintenance, aimed at metrics
- Solve four timed SQL prompts that all end in a ratio metric, so the question of grain stays live in every answer.
- Compute a 95 percent confidence interval for a proportion on a small sample, and state why the normal approximation is unreliable when either np or n(1 minus p) falls below roughly 10, along with which interval you would use instead.
- Take one metric from your day-one tree, write the query that computes it correctly, then write the query that computes it wrong in the most plausible way and explain how you would notice.
Deliverable: Four solved prompts plus a matched correct and plausible-wrong query for one metric.
Practice prompt ↗Practice prompt ↗06Turning engineering work into data science stories
- Write three project stories as situation, decision, trade-off, outcome, each carrying one number and one thing you got wrong.
- For the story you will lead with, prepare an answer to "what would you do differently" that names a decision you made, not a constraint you were handed.
- Practise the sentence that reframes a systems project as a question project: the question the work answered, ahead of the pipeline it shipped.
Deliverable: Three written stories with the lead story delivered aloud and timed under four minutes.
Practice prompt ↗07Mock case and gap list
- Run a 40-minute mock case with someone playing a product manager who pushes back on your metric choice, and record it.
- Listen back and mark every moment you proposed a solution before the success metric existed.
- Rewrite those moments as the question you should have asked, and rehearse the first 90 seconds of the case until scoping comes before solving.
Deliverable: A recorded case plus a rewritten opening 90 seconds.
Practice prompt ↗Worked solution ↗Expand any day for tasks and deliverables. Your progress is saved on this device.
Interviewers here are not checking whether you can describe a project. They want the decision you made, why you made it under the information you had, and what changed afterwards that someone else could measure. A story that ends at 'I built a model' has no ending. Say what the model caused, or what you stopped doing because of it.
How do you handle missing data or dropouts in a longitudinal clinical …
How do you handle missing data or dropouts in a longitudinal clinical trial dataset?
Approach
- Name the disagreement or constraint, and how you resolved it with evidence.
- Close with what you would do differently, concretely.
- Pick a story where you drove the decision, not one where you observed it.
Follow-up
- How did you know the outcome was caused by your change?
- What did you decide not to do, and why?
Disagree with a product manager about an adherence nudge
A product manager proposes shipping a refill-reminder feature to the whole member base, citing an internal analysis: members with 12-month proportion of days covered at or above 0.80 had 31 percent fewer admissions than members below it. The roadmap date is three weeks out and engineering is scoped. You believe the comparison builds survival and continuous coverage into the exposure definition, and is separately confounded by the kind of member who refills on time. Deliverable: what you say in the room, what you send afterwards, and the smallest study you would accept as sufficient to ship.
Approach
- The probe is whether you can be specific and non-territorial at once. A disagreement that arrives as 'the analysis is flawed' costs you the room; one that names a mechanism and offers a cheaper path keeps it.
- Name the two defects separately, because they have different fixes. Classifying a member as adherent over 12 months requires them to stay alive and stay covered for those 12 months, and that guaranteed event-free interval is assigned to the adherent group. Separately, members who refill on schedule differ from those who do not in ways claims never record.
- Show rather than assert. Rerun the same comparison as a landmark analysis: classify adherence over days 1 to 90, start follow-up at day 91 for everyone still enrolled and event-free, and report how much of the 31 percent survives.
- Separate the association claim from the intervention claim out loud. Nothing in the analysis speaks to whether a reminder changes refill behaviour, which is the actual product question and is separately testable.
- Offer the shippable path: a randomised holdout on a subset with proportion of days covered as the primary endpoint, powered on the plausible effect of a reminder rather than on the 31 percent, with a stated read date.
Follow-up
- The landmark rerun still shows a 20 percent difference. Do you ship?
- The product manager says a holdout delays launch. How do you cost that argument?
- What primary endpoint do you choose, and why not admissions?
Defend a null result against a programme sponsor
A care management programme enrolled the top 1 percent of members by prior-year allowed spend. The sponsor's deck shows allowed PMPM for enrollees falling 34 percent from the year before enrollment to the year after, and asks for budget to triple the programme. You rebuild the evaluation with a concurrent comparison group selected by the identical spend rule in the same period. The difference-in-differences estimate is a 3 percent reduction with a confidence interval spanning zero. Deliverable: how you present this, to whom, in what order, and what you propose next.
Approach
- Name the probe: whether you can deliver a finding that costs somebody their programme without softening it into uselessness or creating an adversary who routes around you next time.
- Lead with the mechanism, not the verdict. Show the comparison group's own unadjusted drop, which will be large, because a cohort selected on an extreme of the outcome regresses toward the mean whether or not anyone intervenes. The sponsor's 34 percent is mostly that, and it is a property of the selection rule rather than a criticism of their clinicians.
- Give the sponsor the finding privately before it appears in any deck their leadership sees. Being surprised in a room is what turns a methods disagreement into a political one.
- State the estimate with its interval and say what it rules out as well as what it fails to establish. A 3 percent point estimate whose interval crosses zero is not evidence of no effect; it is insufficient power to separate a modest effect from none, and those are different claims.
- Arrive with a design rather than only an objection. Propose a regression discontinuity at the enrollment threshold if the rule is applied sharply, or a randomised rollout across the next wave of eligible members, and state the sample size needed to detect the effect size the sponsor believes in.
Follow-up
- The sponsor says withholding the programme from a comparison group is unethical. What do you propose instead?
- Leadership wants one number for the board next week. What do you give them?
- What result would change your mind and make you believe the 34 percent?
- 01
How do you handle missing data or dropouts in a longitudinal clinical trial dataset?
- 02
A product manager proposes shipping a refill-reminder feature to the whole member base, citing an internal analysis: members with 12-month proportion of days covered at or above 0.80 had 31 percent fewer admissions than members below it. The roadmap date is three weeks out and engineering is scoped. You believe the comparison builds survival and continuous coverage into the exposure definition, and is separately confounded by the kind of member who refills on time. Deliverable: what you say in the room, what you send afterwards, and the smallest study you would accept as sufficient to ship.
- 03
A care management programme enrolled the top 1 percent of members by prior-year allowed spend. The sponsor's deck shows allowed PMPM for enrollees falling 34 percent from the year before enrollment to the year after, and asks for budget to triple the programme. You rebuild the evaluation with a concurrent comparison group selected by the identical spend rule in the same period. The difference-in-differences estimate is a 3 percent reduction with a confidence interval spanning zero. Deliverable: how you present this, to whom, in what order, and what you propose next.
Is this an official Medpace interview guide?
No. It is PracHub's own research and practice material for the Data Scientist role at Medpace. Rounds and questions reflect what candidates have reported, not a process Medpace has published, and they change over time. Confirm the current format and scope with your recruiter.
PracHub interview research ↗How much SAS experience do I really need if I am highly proficient in Python and R?
While modern data science relies heavily on Python and R, SAS remains a critical tool within clinical research and regulatory reporting. You do not need to be a certified SAS master, but you must show a willingness to learn and a solid understanding of SAS data steps, SQL procedures, and basic macro programming.
PracHub interview research ↗What should I focus on for the final presentation?
Focus on a project where you had ownership of the data analysis. Ensure you can clearly explain the scientific or business problem, why you chose your specific statistical methodology, and how you validated your results. Be prepared to defend your analytical decisions during the Q&A.
PracHub interview research ↗Who will be interviewing me during the panel rounds?
You will meet with a diverse group of professionals from the Biostatistics and data science departments. This typically includes junior biostatisticians, mid-senior level biostatisticians, and senior directors. This allows the team to evaluate how you communicate across different levels of seniority.
PracHub interview research ↗Is the working environment remote, hybrid, or on-site?
Medpace typically operates with a strong preference for on-site collaboration at their corporate campuses to maintain high security and close teamwork, though specific hybrid arrangements may vary by location and team needs. It is best to clarify current expectations with your HR recruiter during the initial screen.
PracHub interview research ↗Sources & methodology 3 sources ↗
No official company page is cited. Rounds and questions come from candidate reports and PracHub editorial material; each source shows the date it was read.
- 01PracHub interview research ↗
PracHub editorial research into this company and role, maintained with this guide. Candidate-reported, not an employer publication.
PracHub page · Accessed 2026-09-22 - 02PracHub Data Scientist practice ↗
Cross-company practice questions for this role.
PracHub page · Accessed 2026-09-22 - 03PracHub interview preparation framework ↗
The framework the preparation plan follows.
PracHub page · Accessed 2026-09-22
Explore related engineering environments and compare the preparation focus.
Veracyte Data Scientist Interview Questions & Guide 2026
Causal-inference exercises for Veracyte data scientist prep: difference-in-differences, matching, instrumental variables, and knowing when observational data simply cannot answer the question asked.
Zoetis Data Scientist Interview Questions & Guide 2026
Metric-diagnosis practice for Zoetis data scientist candidates: a number moves week over week and you work out why, splitting by segment, ruling out a logging or release change, and separating a mix shift from a real change in user behaviou
Vizient Data Scientist Interview Questions & Guide 2026
Communication practice for Vizient data scientist candidates: turn an analysis into a recommendation, name the uncertainty honestly, and answer the follow-up a skeptical stakeholder will raise.
Vibrant Emotional Health Data Scientist Interview Questions & Guide 2026
Case practice for the Vibrant Emotional Health data scientist interview, where the prompt is deliberately underspecified and your first move is scoping it into something measurable.