Pathovio
Back to the blog
EB-2 NIW

EB-2 NIW for Researchers and Postdocs: Documenting the Impact of Your Work

Learn how to translate citation counts, peer review invitations, and collaborations into a documented evidence record a researcher can assemble and organize for an EB-2 NIW petition.

9 min read TR ZH PT ES
Share this article

What 'impact' means to an adjudicator, not just a bibliometrician

An officer reviewing your Form I-140 is not scoring you as a bibliometrician would. A high h-index or citation count tells the officer that other people wrote the number down somewhere — it does not, by itself, show what any of those citations did. The evidence that holds up is evidence that connects a specific piece of your work to a documented effect occurring outside your own lab, department, or institution: an independent group citing your method and using it in their own experiments, a dataset you built appearing in someone else's published analysis, a clinical guideline or regulatory document referencing your findings, a tool you released being adopted in another organization's pipeline.

Metrics versus narrative

Treat raw metrics and narrative evidence as two different deliverables, not one. A citation count answers "how many." A narrative answers "who used it, for what, and with what result." Both are useful, but only the second demonstrates reach beyond your own circle. An adjudicator who sees a citation count with no accompanying explanation has no way to tell whether those citations come from your own co-authors, your own prior papers, or genuinely independent researchers.

The task ahead

The sections below walk through producing both layers: a clean, sourced quantitative record (citation counts, by database, with dates) and a documented set of independent downstream uses that gives the numbers meaning. Neither one substitutes for the other.

Pulling and organizing your citation record

Before writing anything about impact, pull the raw numbers from each major database. Counts will not match across sources — record which database and date each figure came from, every time.

Google Scholar

  1. Open your Scholar profile (or build one if you don't have it).
  2. Note your total citation count and h-index as displayed.
  3. For each paper, click "Cited by" and export or copy the citing-paper list.
  4. Screenshot the profile page showing the date, since Scholar has no formal export function.

Web of Science

  1. Search your name or ORCID under "Researcher Profile" or run an author search.
  2. Use "Create Citation Report" to get per-paper and total citation counts.
  3. Export the report as PDF or Excel; it includes a generation timestamp.

Scopus

  1. Find your Scopus Author ID.
  2. Open "Citation Overview" for your publication list.
  3. Export the citation overview to PDF or CSV; this includes the run date.

Building the tracker

Create one spreadsheet with a row per paper and these columns: paper title, year, venue, your role (first author, corresponding, co-author), total citations, citations by independent authors, database source, date pulled. Leave the independent-citation column blank for now — filling it in is the next step. Keep raw exports (PDF/CSV) alongside the spreadsheet so every number in it can be traced back to a source document.

Identifying and documenting independent citation

A raw citation total includes citations from you, your co-authors, and sometimes your own later papers. An adjudicator wants to know how many citations came from people with no connection to you — that is the number worth isolating and documenting.

Step 1: Screen out self-citation and co-author citation

For each paper in your tracker, open the full list of citing works in Scopus or Web of Science (both let you view "citing documents" from the record page). Check each citing paper's author list against your own publication history and your co-author list. Remove any citing paper where you, or any of your past co-authors, appear as an author. Scopus's Author Identifier and Web of Science's Researcher Profile pages help here — they disambiguate common names and show each author's institutional affiliation at the time of publication.

Step 2: Confirm independence of the remaining set

For a sample of the remaining citing papers, open each one and check the author affiliations listed on the paper itself. Note the institution, department, and country. This is what shows the citing authors worked somewhere you did not.

Step 3: Build the annotated list

From this independent set, select 5–10 representative citing works spanning different institutions or countries if possible. For each, record: citing paper title and year, citing authors' institution, and one or two sentences describing specifically how they used or built on your work — not just that they cited it. Save each citing paper as a PDF with the citation itself visible and highlighted.

Documenting peer review and editorial service as evidence of recognized standing

Reviewer and editorial activity needs its own paper trail, separate from citation evidence. Collect every artifact that shows a journal, publisher, conference, or funding agency asked for your expertise specifically.

What to collect

  • Reviewer invitation emails, including the original request and any thank-you or completion confirmation.
  • Journal reviewer-acknowledgment pages (many journals publish an annual list of reviewers; save the PDF or page with the journal name and year visible).
  • Editorial board listings from a journal's masthead or website, dated or archived.
  • Conference program committee rosters, pulled from the conference website or proceedings front matter.
  • Grant review panel invitations or appointment letters from funding agencies.

Reconstructing a record if emails are gone

If invitation emails were deleted, contact the journal's editorial office directly and request a letter or printout confirming your review history — number of manuscripts reviewed, approximate dates, and your role. Most editorial management systems (Editorial Manager, ScholarOne) can generate this on request; ask the handling editor or journal administrator by name if you know one.

Logging each item

Build a dated table:

Role Organization Date range Source document
Peer reviewer Journal name 2021–present Editor confirmation letter, dated 2026-01-10
Editorial board member Journal name 2023–present Masthead screenshot, accessed 2026-01-12
Program committee member Conference name 2024 Conference website archive

Each row should map to one exhibit later in the petition.

Documenting downstream use of your work by others

Citation counts show that people cite your work. Downstream-use evidence shows what they did with it. Both matter, but adjudicators often find the second more concrete.

Software and code

If you published a tool, check its repository on GitHub, GitLab, or similar for forks, stars, and issues opened by outside users. Search for projects that list your package as a dependency, and for forum posts or papers describing its use. A fork count alone is thin evidence; a citing paper or a forum thread describing an application is stronger.

Datasets

If you released a dataset, check the hosting repository (e.g., Zenodo, Dryad, GEO, an institutional archive) for a download or usage counter, and search for papers that cite the dataset's DOI.

Clinical guidelines and patents

Search Google Patents for patents citing your publications, and search PubMed or guideline repositories (e.g., a national or specialty society's guideline library) for your work cited in clinical or technical guidance.

Media and trade coverage

Search for news articles, trade publications, or professional-society newsletters describing your work's adoption, not just its publication.

Capturing the evidence

For each item, save a PDF or screenshot showing the full URL and today's date, or an official usage report if the platform provides one. Note the access date on the file itself, not just in your tracker, since documents can be separated from their metadata during assembly.

Building the impact index: a worksheet you can hand to a letter writer or include as an exhibit summary

Once citation data, independent-citation samples, peer review records, and downstream-use evidence are collected, consolidate them into a single reference sheet — the impact index. This is a working document, not an exhibit itself, but it is what you or a letter writer will use to locate the right exhibit for every factual claim in a recommendation letter or petition statement.

Table layout

Use one row per discrete claim:

Claim Supporting evidence type Exhibit number Source/database Date verified
Method X adopted by three independent labs Annotated citing-paper list Exhibit 9 Scopus 2026-01-10
40 independent citations of paper Y Citation report Exhibit 11 Web of Science 2026-01-12
Reviewer for Journal Z, 2022–2025 Editor confirmation letter Exhibit 15 Journal Z editorial office 2026-01-14
Dataset downloaded 1,200 times Repository usage report Exhibit 18 Zenodo 2026-01-16

Assigning exhibit numbers

Number exhibits sequentially in the order they appear in the petition's exhibit list, not in the order you collected them. Build the exhibit list first, then populate this worksheet with matching numbers so the two documents never drift apart. If you renumber exhibits later, update this sheet immediately — do not leave stale numbers.

Cross-referencing

In the cover letter and table of contents, cite exhibit numbers directly next to each factual claim (for example, "independent citation by three research groups, Exhibits 9–11"). Give the worksheet itself to any letter writer so their draft references the same exhibit numbers you use elsewhere in the filing.

Naming, formatting, and filing the supporting documents

File naming

Use a consistent scheme across every exhibit: Exhibit-[number]-[source]-[documenttype]-[YYYY-MM-DD].pdf. Example: Exhibit-14-Scopus-CitationReport-2026-01-15.pdf, or Exhibit-22-GitHub-RepoForkCount-2026-02-03.pdf. The date is the date you captured the evidence, not the publication date of the underlying work. Keep names short but unambiguous — an officer should be able to match a file name to the impact index worksheet without opening it.

Converting and dating screenshots

For any web-based evidence (Scholar profile, repository stats, dataset usage dashboard), capture a full-window screenshot that shows the URL bar and, where possible, an on-screen date or timestamp. Print or export directly to PDF rather than pasting into a word processor. If the source has no visible date, add a single caption line beneath the image stating the access date and URL, and keep that caption inside the same PDF page.

Assembling the combined exhibit set

Once individual PDFs are finalized, merge them in exhibit-number order into one combined file or binder, with each exhibit starting on a new page. Add PDF bookmarks matching each exhibit number and title, and paginate the combined file sequentially (either whole-document page numbers or exhibit-internal numbering referenced in your table of contents).

Undated, uncredited, or unsourced printouts are a recurring weakness in this kind of evidence — every document should show, on its face, what it is, where it came from, and when it was captured.

Common failure modes in impact documentation

Before assembling the final exhibit set, check the record against the errors that recur most often in this kind of petition.

Raw counts with no context

A citation total or h-index on its own does not show what the citations did. If the impact index worksheet has a row with a number and no linked narrative or annotated citing-work sample, treat it as incomplete.

Self-citation mixed into the total

Any count that has not been filtered to exclude the petitioner's own citations and co-author citations invites a request for the underlying methodology. Keep the filtered, independent-citation figure separate from the raw database total, and label both.

Undated, unsourced screenshots

A screenshot with no visible URL, no access date, and no indication of which database or repository it came from is hard to verify and easy to question. Every image-based exhibit needs both pieces of information on the page itself, not just in a caption elsewhere.

Claims with no exhibit behind them

A sentence in a cover letter or personal statement asserting adoption, influence, or recognition must point to a specific exhibit number. An unattached claim reads as assertion rather than documentation.

Letters that repeat the CV

A recommendation letter that lists positions and publications without referencing specific documented evidence duplicates material already in the record instead of adding to it.

These are the gaps that most often generate a Request for Evidence asking for substantiation of claims already made.

We use Google Analytics on our public pages to see how visitors find and use the site. It stays off unless you accept, and it's never linked to your case, your answers, or your documents. Privacy.