Journal

How many citations do you need for EB-1A, and what does an h-index prove?

No regulation and no passage of the USCIS Policy Manual sets a minimum number of citations for EB-1A. Citations appear in two places: under the original contributions criterion, where research "highly cited relative to others' work in that field" may show significance, and in the final merits determination, where the Manual names a high h-index for the field. Seven AAO decisions issued between 2015 and 2024 record how directors and the AAO compared such figures.

A researcher from behind sits among stacked journals in a library carrel

Is there a minimum number of citations for EB-1A?

No. 8 CFR 204.5(h)(3) lists ten criteria and asks for at least three, and none of them names a citation count. The Policy Manual says that none of its listed citation factors is required.

The regulation, 8 CFR 204.5(h)(2), defines extraordinary ability as "a level of expertise indicating that the individual is one of that small percentage who have risen to the very top of the field of endeavor." Under paragraph (h)(3) the petition shows sustained national or international acclaim through a one-time achievement or through evidence that meets "at least three of the following" criteria. Original contributions of major significance is criterion (v), and authorship of scholarly articles is criterion (vi). Neither text mentions citations.

The AAO said so in a decision of April 5, 2021 (In Re: 14274123) on a vehicle manufacturer's petition for an engineer: "Neither the statute nor the regulations require the Petitioner to submit citatory evidence for this criterion." The criterion there was original contributions. The Director had run a Google Scholar search and set the engineer's figures against scientists whose citations were "numbered in the thousands." The AAO called that comparison of cumulative citations an improper application and did not consider the Google Scholar data further. The appeal was dismissed.

For the second step, footnote 45 of Volume 6, Part F, Chapter 2 of the Policy Manual states that "none of the listed factors are required."

Where do citations count in an EB-1A petition?

Under criterion (v), as proof that a particular contribution has major significance, and in the final merits determination, as proof of standing at the top of the field. Criterion (vi) on scholarly articles does not ask for citations.

The Policy Manual reads criterion (vi) in two parts. "First, USCIS determines whether the person has authored scholarly articles in the field." Second, it decides whether the publication qualifies as a professional publication, a major trade publication or a major media publication. The criterion text in the chapter does not mention citations, so an article with none can still meet it.

Criterion (v) is where citations of a particular paper come in. The chapter's example is published research that has provoked widespread commentary on its importance from others working in the field, "and documentation that it has been highly cited relative to others' work in that field, may be probative of the significance of the person's contributions to the field of endeavor."

The third place is the final merits determination, the second step after the criteria are counted. Its list of examples includes evidence "that the total rate of citations to the person's body of published work is high relative to others in the field, such as the person has a high h-index for the field." Depending on the field and the comparative data, the Manual says, such evidence may indicate that the person is "among the small percentage at the top of the field."

What is the h-index, and how does the Policy Manual describe it?

A researcher has an h-index of 10 when 10 of their publications have at least 10 citations each, but not 11 with at least 11. J. E. Hirsch proposed the index in a 2005 paper.

Footnote 44 of the chapter calls the h-index "an example of a tool for measuring a researcher's output and impact." It is based on the highest number of the researcher's publications that have been cited at least that same number of times. The footnote's example is an h-index of 10: 10 publications "that have 10 or more citations each" "(but not 11 publications with at least 11 citations each)."

The footnote cites J. E. Hirsch, "An Index to Quantify an Individual's Scientific Research Output," Proceedings of the National Academy of Sciences, vol. 102, iss. 46, p. 16569 (2005). The preprint was submitted to arXiv on August 3, 2005. Its abstract proposes the index h, "defined as the number of papers with citation number higher or equal to h, as a useful index to characterize the scientific output of a researcher."

The same footnote sets a condition: "When relying on any of these tools, a petitioner should explain its methodology and significance with supporting documentation because the sources for the data and the duration of time involved in the calculation can impact the actual h-index value." A bare number with no database and no date does not meet it.

Which database should the citation figures come from?

The Policy Manual names none. It asks the petitioner to explain the source and the time period, because both change the h-index. One database and one retrieval date for every comparison let an officer check the figures.

Google Scholar computes the h-index automatically. When Google launched Scholar Citations on July 20, 2011, its announcement listed three metrics: "the widely used h-index, the i-10 index, which is the number of articles with at least ten citations," and total citations. "We compute each metric over all citations as well as over citations in articles published in the last five years," the post said, and "These metrics are automatically updated as we find new citations to your articles on the web." Profile owners can add missing articles and merge duplicates.

A figure that keeps updating needs a date. In a decision of April 2, 2024 (In Re: 28963434), a computer scientist in medical imaging and computer vision filed percentile statistics from OpenAlex. The AAO noted that the evidence "is undated" and that his own graph covered work through some point in 2023, which "postdates the petition filing date in 2022." Eligibility is judged as of filing.

The AAO also tried to check the numbers: "We attempted to reproduce the Petitioner's results on OpenAlex," but "we were unable to do so." He had searched the broad field of computer science, while OpenAlex offered a narrower category, Computer Vision and Pattern Recognition, that matched his work. He claimed a place among the top 0.14% of researchers and later the top 0.11%, and the AAO noted that the two figures conflicted. The material did not count toward criterion (v).

Are citations of one article and a career total judged differently?

Yes. In 2019 and 2024 decisions the AAO tied criterion (v) to citations of a particular article compared with similar articles and placed cumulative comparisons of whole careers in the final merits determination.

In the 2024 OpenAlex decision the AAO placed comparisons of whole-career citation rates of individuals in the second step: such a comparison "is more appropriate to the final merits determination." Under criterion (v), "A more apt metric under this criterion would be comparing citations to one of the Petitioner's" published articles with single papers of other researchers. The AAO read the Manual's phrase about research "highly cited relative to others' work" as directed at individual articles.

The decision of February 19, 2019 (ID# 2091154) concerned an electrical engineer, a senior staff engineer in silicon photonics. The Director compared his 1,310 total citations, with 121, 64 and 63 for his three most cited articles, against three researchers with 12,665, 8,419 and 21,459. The petitioner argued that the Director "imposed a threshold number of total citations." The AAO answered that a comparison of cumulative citations with others in the field "is often more appropriate" in the final merits determination, while citations of a particular article may be relevant under criterion (v).

On the articles themselves, the AAO found that the petitioner "did not articulate the significance or relevance" of the 121 citations of his top article. One citing article did not distinguish his work "from the other 424 cited papers." A book chapter "devoted three sentences and used two figures in citing to his 2005 article" and was not about his research. The AAO accepted judging and scholarly articles, rejected criterion (v) and dismissed the appeal with two criteria met.

How do you compare a citation record with others in the field?

With figures for the same narrow field, from the same source and date, read with the source's own caveats. The AAO has discounted percentile tables that did not answer the criterion's question and a ranking among 16 researchers.

In the 2019 photonics case the petitioner filed Clarivate Analytics baseline citation rates and percentiles by publication year, and claimed a place "among the top 0.08% of researchers in the field in terms of citation impact." The AAO did not find the figures inflated, but held that such a ranking "does not automatically establish majorly significant contributions to the field." The comparison it named instead "would be to compare the Petitioner's citations to other similarly, highly cited articles." A footnote quoted Clarivate's documentation: "citation rates should not be interpreted as representing the central tendency of the distribution."

In a decision of November 16, 2022 (In Re: 22143529), a senior operational analyst who modelled air transportation demand relied on a chart from a 2015 web article: "papers in the field of engineering average approximately 5 citations," so citations of her 2007 paper were nearly 10 times the average. She also said that a keyword search in Google Scholar placed the paper among the papers on aviation demand forecasting with the highest citation counts, and the AAO noted that she had not submitted the results of that search. Her reports from the Russian eLibrary website covered only authors affiliated with her former employer and her doctoral institution, and the list of total citations ranked her "out of only 16 other researchers." The appeal was dismissed.

In a decision of September 8, 2015 (AAO, September 8, 2015), a researcher working on cancer biomarkers showed 50 citations of his 2008 articles as of filing, and the record included another 2008 article with 253 citations as of the same date. The AAO wrote that "the record lacks evidence that this number is indicative of a significant influence within the field." He did not meet three of the ten criteria, and the appeal was dismissed.

What happens when an h-index is compared with the leaders of a field?

In a July 3, 2024 decision the petitioner met three criteria, yet the AAO dismissed the appeal at final merits. The Director had set his h-index of 8 against leaders with h-indices in the 70s, 60s and 30s.

The petitioner in In Re: 31672763 was a postdoctoral research fellow in biomedical engineering working on optical imaging, measurement and calibration. The AAO accepted three criteria: judging, original contributions and scholarly articles. The case turned on the second step.

The Director found that "citation count leaders had over 90,000 and 70,000 citations of their work" and that leading researchers had "h-index ratings in the 70's, 60's and 30's." According to the decision, "With his last publication in 2022, the Petitioner had an h-index of 8 in that year." The Director also noted a zero citation count across all his publications during 2022 and that his most cited works were published when he was a doctoral student, before his Ph.D. in 2019.

The AAO wrote that "The h-index in the field shows the activeness of the researcher and the importance of the research" and cited Hirsch. It found that the petitioner had not shown that the Director failed to consider the record as a whole or erred in law or policy, and the appeal was dismissed.

Does an early-career researcher need a high citation count?

Less so. Footnote 45 of the Policy Manual calls the citation factor less relevant for people early in their career, who may show acclaim through other evidence in the record.

The footnote reads: "This factor is less relevant for beneficiaries early in their career, as such persons have had less time to accumulate citations but may nevertheless have garnered acclaim and risen to the small percentage at the top of the field." The proof of that acclaim then comes from other evidence in the record.

The footnote does not remove the comparison. In the 2024 optical imaging case the Director compared a researcher who received his Ph.D. in 2019 with citation leaders of the whole field. A table of researchers who received doctorates in the same years and work in the same subfield, drawn from the same database, answers the career-stage point with numbers.

Can Altmetric scores replace citation figures?

Not by themselves. In an August 22, 2022 decision the AAO found that Altmetric profiles of a transplant surgeon's five articles did not rank the response to his work against other articles in the same field.

The petitioner in In Re: 21218536 "did not provide citation figures for his published work." He filed Altmetric profiles whose "Attention Scores" ranked his articles against "all research outputs," "outputs of the same age," and "outputs of the same age and source." The AAO wrote that the profiles "do not appear to rank the response to the Petitioner's work in relation to other articles in the same field of research."

The decision also states who carries the burden when citations are set aside: "If the Petitioner contends that citations are not a reliable indicator of that significance, then the burden remains on him to establish significance and the major significance of his contributions by other means." Criterion (v) was not met, and the appeal was dismissed.

Is the citation standard the same for O-1A and EB-1B?

For O-1A the Policy Manual uses the same examples: a total citation rate high relative to others in the field, or a high h-index. The EB-1B article criterion asks for scholarly journals with international circulation.

Volume 2, Part M, Chapter 4 of the Policy Manual covers O-1A beneficiaries in sciences, education, business or athletics. Section C.4, "Evaluating the Totality of the Evidence," lists evidence that "the total rate of citations to the beneficiary's body of published work is high relative to others in the field" or that "the beneficiary has a high h-index." Its footnotes repeat the definition of the h-index and the note that early-career beneficiaries "have had less time to accumulate citations." Journal impact factor appears in the same list.

For outstanding professors and researchers, 8 CFR 204.5(i)(3)(i)(F) accepts "Evidence of the alien's authorship of scholarly books or articles (in scholarly journals with international circulation) in the academic field." The text of that paragraph does not mention citations either.

What citation evidence should go into the petition?

A dated printout as of filing, copies of citing works showing how your research is used, comparison figures from the same database and subfield, and a written explanation of the method, as footnote 44 asks.

Each item below answers a gap that the AAO named in the decisions above: an undated and unreproducible percentile table in 2024, comparators outside the petitioner's subfield in 2022 and 2024, citing works that did not single out the petitioner's paper in 2019, and a count with no explanation of its significance in 2019.

  • Print the citation profile before filing, with the database and the retrieval date shown on the page.
  • Attach copies of the citing articles and mark the passages that discuss your work.
  • Give comparison figures for researchers in the same subfield and at the same career stage, from the same database and date.
  • For any percentile ranking, file the provider's methodology and the exact search parameters, so that an officer can reproduce the result.
  • Explain in a letter or brief what the numbers mean for the field, as footnote 44 of the Policy Manual requires for the h-index.

How does an article in IAE Review become citable?

An IAE Review article is available for citation from its first day of publication through a permanent URL, in open access under CC BY 4.0, with no article-processing charge. Whether others cite it depends on them.

IAE Review publishes online on a rolling basis, without fixed issues, as soon as peer review, revision and technical preparation are complete. Each article has a permanent page and PDFs in English and Russian, and all articles are free to read, download and cite. Authors keep copyright.

Research manuscripts go through double-anonymous review by at least two independent reviewers, in-house or invited, whose competence and conflicts of interest are checked before assignment. The anonymized main text is submitted with names, affiliations and self-citations removed. The initial decision comes within 7 business days, and the decision after review within 60 calendar days of submission; revisions requested by reviewers extend the term. Under its editorial policy, the journal will announce external indexing only after such indexing has been obtained.

IAE membership is not required and does not affect editorial decisions. Manuscripts in Russian or English go through the submission form as DOCX, PDF or LaTeX. Whether an article or its citations meet a criterion in a petition is decided by USCIS.

Sources

Read also

All articles

Journal

Want to publish in IAE Review?

Submit your manuscript in Russian or English as DOCX, PDF or LaTeX through the submission form. Initial decision within 7 business days.

Manuscript

Received

Initial decision within 7 business days

Article typeOriginal research
Reviewers2, double-anonymous
DOI10.68034 via Crossref