Scientific publications — trends 2026

US visas
1

Colleagues, I want to raise a topic that many, out of old habits, underestimate for EB-1A and partly for O-1 in academic and research profiles given the current realities of 2026.

I did some research on recent information and cases and my conclusion is simple: merely having scientific publications in 2026 is often no longer enough.

Scientific publications help satisfy the scholarly articles criterion (authorship of scientific articles), but at the final merits stage USCIS looks deeper — whether these articles have real impact on the field (citations, independent use, implementation, comparable standing with the top percentage of specialists in the profession).

Moreover, recent self-reported cases demonstrate this quite clearly.

Just recently there was a case of a faculty profile in engineering. Several criteria were met, but the final merits determination resulted in denial, and the applicant himself writes that he now sees the point of strengthening the case with new citations and stronger evidence of real field impact, i.e., influence on the professional field. (Reddit)

There is also a case where after an RFE USCIS even accepted three criteria, including original contributions, but at the final merits it still reverted to the logic “publications and citations are insufficient; the profile is typical for a researcher.” So the problem is not just to “meet three criteria,” but to convincingly show top-of-field level — i.e., the level of the upper echelon of the profession. (Reddit)

At the same time it’s important not to slide to the opposite extreme. There was a reverse example: an industry profile was approved after a NOID even with a not-astronomical number of citations (about 69 at filing and 85 later). But the case was won not by the numbers themselves, but by a strong combination of external measurable impact, judging, media, and critical role. So low or medium citation counts don’t always kill a case, but then the other evidence must be really strong and independent from each other. (Reddit)

My practical conclusion at the moment — for EB-1A, a paper without citations increasingly only serves as a formal closing of the criterion, but not as convincing evidence of recognition. For final merits, it’s not the existence or number of articles that decides, but the numerical metrics of that recognition — citations.

So what to do? The following options are emerging.

  1. increase citations (through collaborations, judicious self-citation, with external help)
  2. seek to bolster publications with reviews from prominent professionals in the field
  3. secure mentions of the scientific work in recommendation letters, where relevant
  4. don’t rush if you feel it’s insufficient — in 2026 haste can be costly

I’m interested in gathering fresh facts here. If you’ve had RFE, NOID, or approvals on EB-1A/O-1, I’d appreciate it if you share how the research side of your case fared.

3 Likes
2

About citations — subjectively, it seems to me their number should be measured in at least the hundreds, and there could be just one publication if it’s super-strong. It’s like an h‑index of 10 already being considered okay for EB1A. I agree that it’s not the number of papers that matters, but what’s behind them — USCIS looks at the real impact in the field when assessing merits.

6 Likes
3

@Vitaliy do you think there’s some number of those citations that would make an officer say “ok”? Or is simply having them enough? Say one or three per article? And do you need them for all articles or only selected ones? That citations help is clear, but where’s the “threshold of help” that needs to be crossed? What do you think? I get that there’s nothing objective or statistical to point to, but just humanly, how many do you think might be enough?

Also the logic of people preparing O-1 or EB-1: “I’m not from academia, I’m in business—what citations am I supposed to have? I’ll state that in the petition, or the officer should understand and not require citations if I’m not a scientist.” Do you think that won’t work or has stopped working? If so, hasn’t the system broken, and do we really have to play by broken rules here? Maybe it’s easier to argue that they shouldn’t demand citations if the manual doesn’t say so.

7 Likes
4

I have these thoughts:

  1. Having something is definitely better than having nothing.
  2. Performance metrics — raw counts and indices on Google Scholar (GS). Based on these you should try to maximize the indices so it looks as strong as possible. So a reasonable minimum is about 10 citations per paper, to make the profile i10 as high as possible.
  3. A healthy distribution of citations — it looks odd when five papers each have exactly 10 citations; mathematically that’s suspicious. Given USCIS’s increased scrutiny of anything unusual, I’d pay attention to that too.
  4. As discussed on Reddit — you’ll be compared to crowds of peers in your industry. The simplest route in applied fields for EB, IMO, is to show that scientific work is generally not practiced in your industry.

And again, use not just one but all available mechanisms. I wrote about recommendation letters, reviews, etc. You can probably also add indirect factors to cite — the high standing of the journals where the article was published; a plagiarism-check report for the article if the metrics are really strong; and so on. Also appealing to the fact that such requirements aren’t in the manual, even though you have citations, is probably reasonable too.

7 Likes
5

Yeah, citations really matter more than the number of papers — 1,000 citations is already a solid EB-1. But don’t panic if it’s less — I’ve seen cases where with a couple hundred citations and a decent h-index people sailed through. The main thing is that the distribution looks normal, otherwise it looks like gaming the numbers)

6 Likes
6

In my initial post I mentioned Reddit, where even lower citation counts (roughly 10+ per publication) were accepted, and on Telegram (tg) it was the same — people shared when a science post with a small number of citations was a hit.

6 Likes
7

I’ve seen a case where someone with an h-index of 8 got EB-1A approval without any problems — they just put the case together well. So an h-index of ten is a good benchmark but not a magic number. The main thing is to show that the impact is real, not just inflated numbers )

6 Likes
8

Listen, recommendation letters also matter a lot at the final stage — about 5–8 of them, with at least 3 from independent people who are top in the field. I’ve seen applicants with average metrics get through precisely because of well-put-together letters from the right recommenders, so don’t get hung up on the numbers only)

6 Likes