Evidence Building
Documenting Research Software as Original Contributions in O-1A
Research software is increasingly the primary output of computational scientists, yet USCIS adjudicators rarely encounter it as original contributions evidence. This article explains what documentation a research software author needs to satisfy the O-1A original contributions criterion, what USCIS discounts, and how to frame borderline adoption data.
Research software and the original contributions criterion
Research software — code written to advance scientific inquiry rather than to power a commercial product — occupies an ambiguous position in O-1A petitions. The original contributions criterion at 8 C.F.R. § 214.2(o)(3)(ii)(A) requires evidence of original scientific, scholarly, artistic, athletic, or business-related contributions of major significance in the field. Software development, even when it is the primary research output of a computational scientist or bioinformatician, does not fit neatly into the evidentiary categories USCIS adjudicators encounter most often. The agency's experience is heaviest with traditional academic outputs — peer-reviewed papers, grants, named awards — and a petition built primarily around software authorship requires proactive contextualization to be evaluated fairly.
The practical challenge is that research software contributions can be transformative — a widely adopted genomics pipeline, a Python library for causal inference, a climate modeling framework used by hundreds of research groups — without generating the kinds of primary evidence that immigration adjudicators recognize readily. A paper may have 500 citations; a software repository may have 50,000 downloads and 2,000 dependent packages without producing an equivalent formal citation record. The petition must bridge that gap by translating software adoption metrics into terms that carry the same inferential weight as citation data, and by securing expert letters from researchers who can explain why the tool represents a genuine original contribution rather than a convenience utility.
The totality-of-evidence standard, as interpreted by the Ninth Circuit in Kazarian v. USCIS and applied by the AAO in subsequent precedent decisions, requires that the original contributions criterion be assessed by asking whether the petitioner's contributions have substantially influenced the field — not merely whether the contributions exist. A research software author whose work has been adopted as a standard tool by independent research groups in multiple countries and cited as a methodological foundation in published papers is in a fundamentally different position from a researcher whose software is used primarily within their own lab. The petition's job is to make that distinction legible to an adjudicator who will not independently investigate the software's adoption history.
What the regulation actually requires
The text of 8 C.F.R. § 214.2(o)(3)(ii)(A) specifies that evidence of original contributions must demonstrate that those contributions are of major significance in the field. This phrase has two operative components: the contribution must be original — not a replication or incremental extension of prior work — and it must be significant within the relevant discipline, not merely within the petitioner's own institution or research group. For research software, originality typically means that the software introduces a novel algorithm, methodology, or capability that was not previously available in open-source or commercial form. Significance means that the tool has been adopted and relied upon by researchers who had no personal connection to its creation, and whose research would have been materially more difficult or less rigorous without it.
USCIS policy guidance requires that original contributions criterion evidence be supported by documentation, not simply by the petitioner's own characterization. A researcher who states in a support letter that their software is widely used and highly significant without supporting exhibits will typically receive an RFE requesting objective evidence. Acceptable documentation includes peer-reviewed papers published by the petitioner describing the software's methodology and validation, citation counts for those associated methods papers in Google Scholar or Web of Science, repository metrics (stars, forks, dependent repositories) from GitHub or equivalent platforms, download statistics from package managers such as PyPI, CRAN, Bioconductor, or Conda-forge, and independently published papers by unaffiliated researchers who cite the software as a methodological tool.
The distinction between software as a research contribution and software as a service is important in petition drafting. A tool that allows other researchers to replicate or extend the petitioner's published findings is closely tied to the scholarly articles criterion and may not independently satisfy the original contributions criterion. A tool that enables new research that the petitioner's own publications do not directly address — because it opens up methodological capabilities that did not previously exist — is more clearly an independent original contribution. Expert letters should explicitly identify what research is now possible because of the tool that was not feasible before, framing the contribution in terms of capability expansion rather than task automation.
Evidence that typically satisfies USCIS
The most persuasive evidence packages for research software contributions typically include four elements presented together: the associated methods paper, third-party citation data for that paper, adoption metrics from the software repository or package manager, and at least two independent expert letters from researchers who use the tool. The methods paper establishes that the contribution has been peer-reviewed and published — satisfying the scholarly articles criterion simultaneously — while citation data demonstrates that peer researchers have formally acknowledged reliance on the methodology. Repository metrics extend that evidence to users who may not have cited a formal paper, and expert letters synthesize both data streams into a qualitative assessment of the tool's significance within the field.
High adoption rates on package managers are particularly useful evidence when the software fills a function that researchers must perform routinely. A bioinformatics tool available through Bioconductor that shows hundreds of thousands of downloads per year, and that appears in dozens of published methods sections as the designated preprocessing pipeline, is documented in ways that are objectively verifiable. Package manager download statistics should be captured at a specific date and included as an exhibit with a cover explanation; some platforms publish annual download statistics or trend charts that provide additional context. Where the petitioner's tool is listed in a curated resource list — a field-specific wiki, an NIH-funded resource center's recommended tools page, or a journal's required software for data submission — that listing is meaningful third-party validation.
Letters from researchers at peer institutions who rely on the software are the strongest expert evidence and should be recruited before the petition is filed. The ideal letter describes the specific research problem the writer was solving, why existing alternatives were insufficient, how the petitioner's software addressed the gap, and what the writer's own published research would have required without it. Letters that engage with specific versions or features of the software, and that reference published papers in which the tool was used, are more credible than letters that praise the tool's ease of use or documentation. The petitioner's own characterization of the tool's significance, however accurate, is presumed self-interested and should not carry the evidentiary load alone.
Evidence USCIS often discounts
Letters from collaborators within the petitioner's own research group, from co-authors on the software's associated paper, or from researchers who contributed to the codebase as direct collaborators carry limited weight for the original contributions criterion. USCIS evaluates independence of recognition, and a letter from a doctoral student the petitioner supervised, a postdoc who contributed modules to the repository, or a co-PI on the grant that funded the software's development is presumed to lack the objectivity that independent recognition requires. Even a well-written letter from a close collaborator may be treated as evidence of collaboration rather than external recognition. The petition should clearly distinguish between letters from independent users and letters from contributors or collaborators.
GitHub star counts, raw download numbers, and social media mentions are regularly questioned in RFEs because they reflect availability and ease of installation as much as scientific significance. A widely-downloaded tool that is popular among students or practitioners — because it is well-documented and easy to use, not because it represents a frontier contribution — may accumulate impressive metrics without demonstrating major significance in the research community. Adjudicators who have encountered prior petitions using gaming or consumer application download counts as evidence have developed skepticism about pure download metrics. The petition must connect adoption metrics to research outputs — showing that the downloads translate into published papers, funded grants, or clinical protocols that would not have been possible without the tool.
Vague descriptions of software significance are regularly cited in RFEs as insufficient. Statements such as the tool has been widely adopted by the scientific community or researchers around the world use this software, without specific documentation, do not satisfy the regulatory standard. Similarly, describing the technical elegance or computational efficiency of the software — its speed advantage over prior implementations, its memory efficiency, its modular architecture — addresses software engineering quality rather than scientific significance. The original contributions criterion is not satisfied by building a technically superior implementation of an existing algorithm. The contribution must be original in the sense of introducing a methodology or capability that was not previously available, not merely in the sense of executing an existing approach more efficiently.
Framing borderline software evidence
When a petitioner's software adoption is real but hard to document — because the tool is used in non-publishing contexts such as industry labs, regulatory settings, or internal government research — the petition can invoke the comparable evidence provision at 8 C.F.R. § 214.2(o)(3)(ii) to present alternative documentation. Industry adoption, if documented through licensing agreements or vendor relationships rather than academic citations, can demonstrate major significance in a commercial or applied research context. A tool adopted by a pharmaceutical company for biomarker discovery, by a government agency for environmental modeling, or by a hospital system for clinical decision support may represent a more significant real-world contribution than one used exclusively in academic publications, but the evidence chain requires more careful construction.
When the petitioner's software is new enough that citation and adoption data have not yet accumulated, the petition can present prospective significance evidence — though it must be used carefully. Expert letters from researchers who have evaluated the software and describe their plan to use it in funded research, along with the grant abstracts for those projects, establish that the tool has been independently reviewed and found to be of sufficient quality and novelty to incorporate into serious research programs. The prospective argument is weakest when presented alone and strongest when combined with at least some adoption data and an established methods paper, even if the citation count is still modest.
Preprint adoption data — citations in arXiv, bioRxiv, or SSRN preprints — is useful supplemental evidence but should not be presented as the primary citation record because preprints have not been peer-reviewed. Some fields publish primarily through preprints, and where the petition can document that the relevant community's practice is preprint-first, the preprint citation record carries more weight. Expert letters from researchers in preprint-heavy communities who can explain the community's publication norms and attest to the petitioner's standing within it help contextualize what would otherwise appear as a thin formal citation record. The petition's task in borderline cases is not to manufacture additional evidence but to frame the available evidence against the field's specific norms.
Building and auditing your evidence file
A complete evidence file for a research software original contributions argument should begin with a master exhibit list that maps each regulatory element — originality, major significance, independent adoption, field recognition — to specific documents. The methods paper and its citation data establish originality and scholarly recognition. Repository metrics and download statistics establish adoption. Expert letters from independent researchers establish that the adoption reflects significant contribution rather than incidental use. If any element in this mapping is thin or absent, the petition is at heightened risk of an RFE on the original contributions criterion. Identifying gaps before filing allows the petitioner to recruit additional expert witnesses or gather supplemental documentation while there is still time.
Before filing, the petitioner or attorney should verify that every expert letter explicitly addresses the major significance standard and does not simply describe the software's features or the petitioner's technical skill. Each letter should answer the question: why does this contribution matter to researchers who are not affiliated with the petitioner? A letter that answers that question specifically — describing a concrete research problem that the software solves, and identifying published work that would not have been possible without it — satisfies the standard. A letter that praises the petitioner's collaboration skills, their availability for technical support, or their generosity in open-sourcing the tool addresses qualities that are valuable but legally irrelevant to the original contributions criterion.
The original contributions criterion for research software authors is typically argued alongside the scholarly articles criterion — because most research software authors have also published methods papers, and those papers contribute to both criteria simultaneously. A petition that presents research software as the central contribution should build the case around originality and adoption first, and then use the scholarly articles record to reinforce the finding that the petitioner's work has been peer-reviewed and recognized by the relevant scientific community. An immigration attorney experienced in O-1A petitions for computational scientists can help the petitioner identify which evidence is strongest, how to sequence arguments within the brief, and whether any elements are likely to draw scrutiny during adjudication.
What we typically gather for this kind of case
| Document | Where to source | Why it matters |
|---|---|---|
| Peer-reviewed publications | Web of Science / Scopus exports | Anchors original-contributions and authorship criteria |
| Citation analysis | Google Scholar profile + ESI top-1% data | Quantifies major significance in the field |
| Salary benchmark | BLS OEWS for SOC code + locality | Documents high-salary criterion at 90th-percentile or above |
| Critical-role letters | Direct supervisor + program director | Establishes role's importance, not just title |
What we see go wrong, again and again
- 01Treating extraordinary ability as a credentials checklist rather than a story of field-wide impact.
- 02Submitting bibliometric data (h-index, citation counts) without explaining what makes those numbers high relative to peers in the same sub-field.
- 03Relying on letters from collaborators or co-authors rather than independent experts who can speak to influence.