Does Research Actually Matter in the Match? What the 2026 Data Shows
If you have spent any time on residency applicant forums, you have heard some version of: "you need research to match into a competitive specialty." NRMP's 2026 Charting Outcomes data lets us actually test that claim, measure by measure, specialty by specialty — and the honest answer is more specific than the forum wisdom suggests.
One methodology note before the numbers: starting with the 2026 report, NRMP separated what used to be a single combined "scholarly activity" item into three distinct measures: abstracts, presentations, and publications, plus a fourth measure for research experiences generally. That is a meaningfully more precise dataset than prior years, and it's what makes this breakdown possible.
The Counterintuitive Headline
Across all specialties combined (n=14,889 matched, n=1,636 not matched), the aggregate numbers run against the conventional wisdom:
| Measure | Matched (Median) | Not Matched (Median) |
|---|---|---|
| Research experiences | 4.0 | 5.0 |
| Abstracts | 2.0 | 2.0 |
| Presentations | 4.0 | 5.0 |
| Publications | 2.0 | 3.0 |
Taken at face value, that table says not-matched applicants out-published matched ones. That is real data, not a typo — but it is also a textbook example of why aggregate statistics can mislead. The explanation is a composition effect: applicants who do not match are disproportionately concentrated in the handful of ultra-competitive, research-saturated specialties (Dermatology, Neurological Surgery, Plastic Surgery) where everyone in the pool, matched or not, has enormous research output. That pulls the "not matched" aggregate upward even though, within most individual specialties, the opposite relationship holds.
What the Specialty-Level Data Actually Shows
Looking within individual specialties tells a cleaner story. Here is matched vs. not-matched medians across nine specialties spanning primary care to the most research-intensive surgical subspecialties:
| Specialty | Research Exp. (M / NM) | Abstracts (M / NM) | Presentations (M / NM) | Publications (M / NM) |
|---|---|---|---|---|
| Internal Medicine | 4.0 / 3.0 | 2.0 / 1.0 | 4.0 / 3.0 | 2.0 / 2.0 |
| Anesthesiology | 5.0 / 4.0 | 2.0 / 2.0 | 4.0 / 3.0 | 2.0 / 1.0 |
| Psychiatry | 4.0 / 3.0 | 1.0 / 1.0 | 3.0 / 2.0 | 2.0 / 1.0 |
| Emergency Medicine | 3.0 / 3.0 | 1.0 / 1.0 | 3.0 / 2.0 | 1.0 / 1.0 |
| Orthopaedic Surgery | 7.5 / 7.0 | 5.0 / 4.0 | 11.0 / 8.0 | 7.0 / 5.0 |
| Neurological Surgery | 6.0 / 6.0 | 9.0 / 7.5 | 11.0 / 11.0 | 12.5 / 9.0 |
| Plastic Surgery (Integrated) | 6.0 / 7.0 | 6.0 / 6.0 | 20.0 / 14.5 | 10.0 / 9.0 |
| Dermatology | 7.0 / 8.0 | 4.0 / 4.0 | 12.0 / 9.0 | 9.0 / 6.0 |
(M = matched, NM = not matched. Family Medicine is excluded from this table — its not-matched group was only 11 applicants, too small to read any pattern into.)
The Real Pattern: Output Beats Volume
Across these eight specialties, raw "research experiences" is the weakest and least consistent differentiator — it ties or even runs slightly higher for not-matched applicants in three of the eight, including both Dermatology and Plastic Surgery. But presentations and publications — the measures that require you to actually finish something, not just join a lab — favor matched applicants in the large majority of specialties, including every surgical subspecialty in this table.
The practical reading: once you are in a competitive applicant pool, simply accumulating research "experiences" stops differentiating you — everyone around you has plenty of those too. What separates matched from not-matched applicants within that pool is whether those experiences turned into finished output: an abstract that got presented, a manuscript that got published. Starting five projects you never finish is weaker than finishing two.
A Caveat Worth Taking Seriously
None of this is causal. NRMP's data cannot tell you whether strong research output causes a match, or whether the kind of applicant who finishes publications also tends to be stronger across the board — clinical grades, letters, interview performance — in ways this dataset does not capture. NRMP's own introduction to the report makes the same point: these trends "do not convey the full story," and course evaluations, letters of recommendation, and the MSPE all factor into outcomes without appearing anywhere in this data.
What This Means for Your Application
- If you are early in medical school and specialty-undecided: prioritize finishing projects over starting new ones. A completed abstract beats three half-finished chart reviews.
- If you are applying to a research-saturated specialty (Dermatology, Plastic Surgery, Neurological Surgery): assume research output alone will not differentiate you from the rest of the pool. It is necessary to be competitive, not sufficient — your rank list length and Step 2 CK score are doing more of the differentiating work (see our full breakdown of the 2026 Match data).
- If you are applying to a less research-intensive specialty (Internal Medicine, Psychiatry, Emergency Medicine): even modest, completed research output appears to correlate with better outcomes, and the bar to stand out is much lower.
Figuring out where to put research on your actual application once you have it? Our guide on where to put your research on ERAS covers exactly how MyERAS's Scholarly Work section works, entry type by entry type.