How to Audit Your EB-1A Petition Before USCIS Does

Three Numbers That Should Change How You File

If you remember nothing else from this article, remember three numbers.

30, 62, 9.

30 is the approximate percentage of EB-1A petitions that receive a Request for Evidence. Depending on the service center and the field, that number ranges from 25% to over 40% — but the average across the Lumova dataset of 10,000+ AAO decisions hovers around three in ten.

62 is the percentage of RFEs that cite the same single criterion: Criterion 5, Original Contributions of Major Significance. Six out of every ten RFE letters we've analyzed challenge this criterion specifically, often using the same phrase USCIS has been recycling since 2015: "The petitioner has not established that the contributions are of major significance in the field."

9 is the number of weeks most petitioners wait before realizing their response isn't working. By then, you're 63 days into an 87-day response window with a brief that hasn't been independently reviewed by anyone who understands how USCIS actually reads evidence.

Here's what those three numbers mean together: an RFE is not a surprise. It is the predictable outcome of a petition that was never stress-tested against the exact patterns USCIS uses to evaluate it. The petitions that avoid RFEs entirely aren't written by better attorneys — they're written by petitioners who ran a structured audit before they filed.

This article is that audit. By the time you finish reading, you'll know exactly how to pressure-test your own petition the way a USCIS adjudicator would, what the ten highest-risk signals look like in practice, and how to fix them before you sign the G-1145. This is the full pre-filing protocol used by petitioners who file once and approve on the first try.

Why Most Audits Aren't Audits

Before we get into the protocol itself, let's correct a common misconception about what a petition audit actually is.

Most petitioners think they're "auditing" their petition when they do one of these things:

• Re-read their own brief three times looking for typos
• Ask a friend who just got approved to skim it
• Pay an attorney $400 for a "one-hour review" that returns vague comments like "strengthen your C5 section"
• Run a grammar checker over the exhibits list

None of that is an audit. Those are proofreads, and proofreads don't catch the structural failures that trigger RFEs. Typos don't cause denials. Missing an RFE-prediction pattern does.

A real audit is adversarial. It assumes your petition will be read by a skeptical adjudicator who has 18 minutes, 400+ pages in front of them, and a performance review that rewards them for catching insufficient evidence. It asks the questions that adjudicator will ask — in the order they'll ask them — and verifies that your petition has a clean, cited, quantified answer for each one.

Consider the difference between these two reviews of the same section of a petition:

The first comment sounds thoughtful. The second comment is what actually prevents an RFE.

The question for the rest of this article is: how do you do the second kind of review on your own petition, even if you've never seen an AAO decision in your life?

The Ten Signals: What USCIS Adjudicators Actually Look For

Lumova's dataset includes 10,000+ AAO non-precedent decisions plus thousands more publicly available denial and approval patterns. When we run machine-learning pattern extraction on this corpus, ten signals dominate the RFE predictions. These are the ten things an adjudicator — trained on internal USCIS checklists — is scanning your petition for in their first 18 minutes.

Meet Priya Nair, a composite applicant we'll reference throughout the rest of this article. Priya is 34, born in Kochi, India, with a PhD in computational biology from Stanford and five years at a top NIH-funded lab at UCSF. She has 412 citations, an h-index of 14, two first-author papers in Nature Communications, six conference program-committee roles, and a $175,000 annual salary. On paper, Priya looks like a strong EB-1A filer. Her first draft petition hit every one of the ten risk signals below — and before she ran a pre-filing audit, she would have almost certainly received an RFE.

Here's what a real audit catches that a proofread misses.

Signal 1 — The Unquantified Adjective

Any sentence in your petition that uses an evaluative adjective without a number is a risk. "Well-respected." "Highly regarded." "Prestigious." "Seminal." These are the words USCIS adjudicators are trained to interrogate, because every single one of them is a claim without evidence.

Priya's draft included the sentence "The petitioner's work is widely cited by leading researchers in computational immunology." An adjudicator reads that sentence and mentally subtracts it from the petition. "Widely cited" is not a number. "Leading researchers" is not an identity.

The audit replaces it with: "The petitioner's 2022 Nature Communications paper has been cited 187 times across 43 unaffiliated research groups in the United States, Germany, and Japan, including by two members of the National Academy of Sciences (Dr. [name] at Princeton and Dr. [name] at MIT) in their own first-author publications."

Count adjectives in your petition. Ask of each one: "Where's the number?" If you can't point to a number within the same paragraph, rewrite the sentence.

Signal 2 — The "Close Colleague" Expert Letter

Expert declarations are essential for most EB-1A petitions. But USCIS has a specific internal rule, reflected consistently in AAO decisions: letters from former supervisors, current collaborators, or co-authors carry dramatically reduced evidentiary weight. Not zero weight — reduced. And if all your letters come from this category, the petition reads as a closed loop to the adjudicator.

Priya's first draft had five expert letters. All five were from people she had worked with: her PhD advisor, two former postdoc colleagues, her current PI, and a collaborator on a recent paper. She thought this made the letters more credible ("these people know my work best!"). Actually, it flagged her entire expert evidence as non-independent.

The audit recommendation: keep two letters from close colleagues, but add three independent letters from researchers who have cited Priya's work but have never met her or worked with her. We'll cover exactly how to find and approach independent experts in Signal 7 below.

Signal 3 — Step 2 Silence

This is the signal that catches even experienced petitioners. The Kazarian framework, as clarified by USCIS policy since 2010, requires a two-step analysis: (1) does the evidence meet the regulatory criteria individually, and (2) does the totality of that evidence demonstrate sustained national or international acclaim and establish that the petitioner is among the small percentage at the very top of their field?

Most petitions spend 95% of their brief on Step 1 (proving each criterion individually) and five sentences or less on Step 2 (the final merits totality argument). When the adjudicator gets to the final merits determination and your brief has no explicit argument for why your claimed criteria together establish "sustained acclaim" — they write one for you. And the one they write is a denial.

Priya's first draft brief had 42 pages on Step 1 and one paragraph titled "Final Merits Determination" at the end. The paragraph read, approximately: "For the foregoing reasons, the petitioner has established eligibility for classification as an alien of extraordinary ability."

That is not a Step 2 argument. It is a conclusion. A real Step 2 argument shows the adjudicator how the evidence fits together — why the combination of peer review, original contributions, and scholarly articles specifically demonstrates acclaim that a single criterion alone would not. We'll show you how to write a real Step 2 argument later in this article.

Signal 4 — Self-Citations Inflating the Record

USCIS adjudicators are trained to distinguish between citations by independent researchers and self-citations (including citations by coauthors of the petitioner's papers and other close associates). If you report "412 citations" in your petition without separating independent citations from self-citations, the adjudicator will assume the inflated number — and then mentally discount it.

Priya's raw citation count was 412. Her independent citation count (citations from papers where no author had co-authored with her) was 331. The audit recommendation: report both numbers explicitly, with a sentence like "Of the petitioner's 412 total citations, 331 (80.3%) are from independent research groups with no coauthorship relationship to the petitioner." This turns a potential red flag into a credibility signal.

Run a Google Scholar search on your name. Go through your top 10 most-cited papers. For each citation to those papers, check if the citing authors overlap with your coauthors. Compute your independent citation percentage. If you don't do it, the adjudicator will assume the worst case.

Signal 5 — The Membership Trap

Criterion 2 (membership in associations requiring outstanding achievement) is the single most over-claimed criterion in the EB-1A universe. Petitioners frequently submit membership evidence for organizations that anyone who pays a fee can join — IEEE, ACM, AAAS, local professional associations — and claim these meet the criterion.

They don't. USCIS policy requires memberships that demand "outstanding achievements judged by recognized experts in the field." IEEE Senior Member status (which requires expert nomination and professional review) may meet this standard; regular IEEE membership does not. Full Fellow status at ACM meets it; general ACM membership does not.

Priya's draft claimed ACM membership and AAAS membership under C2. Neither meets the criterion. The audit recommendation: drop C2 entirely unless she can document Senior Member or Fellow status in one of these organizations, or substitute a membership in a by-invitation research society.

Count the memberships you're claiming. For each one, ask: "Does admission to this organization require expert judgment of outstanding achievement, with published membership criteria that go beyond dues payment?" If the answer is no, remove the claim. It's better to claim three bulletproof criteria than to claim five with two weak ones dragging down your credibility.

Signal 6 — The Vague "Critical Role"

Criterion 7 (performing in a leading or critical role for organizations with a distinguished reputation) is the criterion most commonly misunderstood. It has two halves: (a) your role must be leading or critical, and (b) the organization must have a distinguished reputation. Both halves must be proven independently.

Most petitioners prove one half or the other, not both. They show evidence that they're a "principal engineer" (critical role proven) at a small startup (distinguished reputation not proven). Or they show that their employer is Fortune 500 (distinguished reputation proven) but only demonstrate that they're one of 6,000 engineers at the company (role not critical).

Priya claimed C7 based on being the "lead computational biologist" on a $3M NIH grant at UCSF. UCSF's reputation is clearly distinguished. But the brief didn't prove that her role on the specific grant was critical to the outcome of the grant itself — what decisions did she make? What would have failed without her specific input? The audit recommendation: add a letter from the PI explicitly stating that Priya's computational approach enabled the grant's primary findings, accompanied by the grant's first publication with Priya as co-first author.

Count the claims in your C7 section. For each, ask: "Can a stranger reading this paragraph tell why my specific contribution was critical — not just that I had a senior title?"

Signal 7 — Independent Experts Are Missing

We touched on this earlier, but it deserves its own signal because it's the single fixable weakness we most often see. Independent expert declarations — letters from researchers who know your work through the literature but have never met you — are the gold standard for EB-1A evidence. And almost no petitioner starts with any.

Here's how to fix it, quickly and without awkwardness:

1. Find researchers who have cited your papers in their own publications. Google Scholar's "cited by" list is your starting point.

2. Of those citing authors, identify the ones who (a) are at distinguished institutions, (b) are established in their career (tenured or senior industry role), and (c) have no coauthorship or institutional overlap with you.

3. Email each of them a short message: "Dear Dr. [name], I'm writing to ask if you'd be willing to contribute a brief expert declaration for my EB-1A petition. I noticed you cited my 2021 paper on [topic] in your [year] publication on [topic]. Your perspective on how that work influenced the field would be valuable. I can send a template with the exact structure USCIS requires, and you'd only need to share your professional opinion of my contribution — not advocate for my case."

4. Expect a response rate around 25-40%. You only need 2-3 to say yes.

Priya, before her audit, had no independent expert letters. After the audit, she had three — each from a researcher who had cited her work but had never met her. This single change probably shifted her RFE probability by 20 percentage points.

Signal 8 — The Comparison Without a Baseline

Criterion 8 (high remuneration) requires showing that you command a salary at the top of your field. But "top of your field" is a comparative statement — and comparisons require baselines.

The most common mistake: petitioners attach their W-2 and write "The petitioner's salary of $175,000 is significantly above average for the field." This proves nothing. Above average compared to what? To which percentile of which occupation code in which geographic area in which year?

The correct approach: pull data from the Bureau of Labor Statistics (bls.gov/oes/) for your specific Standard Occupational Classification code, the specific metropolitan area where you work, and the most recent year available. Report the 90th percentile salary for your SOC + geography, then show your compensation exceeds it. If you want to be especially credible, add a comparison to a professional compensation survey specific to your field (for software engineers: levels.fyi data; for physicians: Medscape compensation reports).

Priya's draft said her $175K salary was "above average for her field." The audit rewrite: "Per the Bureau of Labor Statistics Occupational Employment Statistics for Medical Scientists (SOC 19-1042) in the San Francisco-Oakland-Hayward MSA for May 2024, the 90th percentile annual wage is $184,320. Including Priya's base salary of $175,000 plus performance bonus of $22,000 and equity vested in 2024 of $18,400, her total compensation of $215,400 exceeds the 90th percentile by 16.9%."

The rewrite turns a dismissible claim into a documented fact.

Signal 9 — Exhibit Labels That Don't Match the Brief

This one is a pure procedural failure, but it kills cases. Every piece of evidence in your petition is submitted as a numbered exhibit. Your petition brief refers to those exhibits by number: "See Exhibit 14." When the adjudicator wants to verify a claim, they flip to Exhibit 14 and expect to find the specific document that supports the specific sentence.

What causes RFEs: the brief says "See Exhibit 14 (letter from Dr. [name])" and Exhibit 14 is actually the PI's grant award, not the letter. Or the letter is in Exhibit 14 but it's on page 7 of a 22-page document with no highlighting or tab to indicate where. The adjudicator flips through it once, can't find the supporting language immediately, and moves on — either by issuing an RFE or by simply not crediting the claim.

The audit recommendation:

• Every exhibit number referenced in the brief must match exactly.
• Every exhibit should have a cover page identifying what it is.
• Every page within a multi-page exhibit should be numbered.
• Where a single exhibit has multiple supporting documents, include an internal index.
• For long documents, use a highlighter or page tabs on the specific language the brief refers to.

Priya's first draft had eight exhibit-to-brief mismatches. The audit caught all of them.

Signal 10 — The Missing Pre-emption

This is the most advanced signal and the one that separates petitions that approve on the first try from petitions that bounce through RFE. A pre-emption is a sentence in the brief that acknowledges a potential weakness and addresses it before the adjudicator can use it against you.

Example: "The petitioner notes that Criterion 6 (scholarly articles) is primarily supported by journal publications with Impact Factors between 4.2 and 7.8, rather than the highest-tier journals in the field. Per USCIS policy and the Matter of Dhanasar framework, the quality of scholarly publications is evaluated in context of the field, and computational biology publishes primarily in specialized journals with lower nominal Impact Factors than generalist journals. The petitioner's citation record of 331 independent citations demonstrates that the downstream impact of these publications, not their venue, is the appropriate measure of scholarly contribution."

That paragraph kills an RFE challenge before it can be written. The adjudicator reading it sees that the petitioner has already considered the concern, acknowledged it, and provided the reasoning that rebuts it. They're unlikely to issue an RFE on a challenge that the petition has already pre-empted — it would read as the adjudicator not having actually read the brief.

A real audit identifies every weakness in your petition and pre-empts every one of them. Priya's first draft had zero pre-emptions. The audited version had seven.

Running the Audit on Your Own Petition: A 6-Hour Protocol

You now understand the ten signals. The question is how to apply them to your own petition in a structured way. What follows is a 6-hour protocol that works through all ten signals in sequence. You should allocate an uninterrupted Saturday morning for this — it's the single highest-leverage work you'll do in your entire EB-1A process.

Hour 1: The Quantification Pass

Open your petition brief. Highlight every evaluative adjective: "leading," "prestigious," "significant," "seminal," "major," "critical," "distinguished," "top," "renowned," "well-respected." For each one, write a number in the margin that supports the claim. If you can't write a number, the sentence is a liability. Either rewrite it with a specific metric or delete it.

Expect to find 30-80 of these in a typical first draft. Expect to spend most of Hour 1 just finding them; fixing them is Hour 2's job.

Hour 2: The Rewrite Pass

Go back to the highlighted adjectives from Hour 1 and rewrite each sentence. Replace adjectives with numbers, sources, and dates. For each rewrite, ask: "If the adjudicator wanted to verify this sentence, could they find the verification in the exhibits within 30 seconds?" If not, add a specific exhibit reference.

By the end of Hour 2, your brief should contain roughly 2-3x more specific numbers than it started with, and substantially fewer adjectives.

Hour 3: The Expert Letter Audit

Print your expert declarations. For each one, circle the relationship between the recommender and you. Close colleague? Former supervisor? Coauthor? Independent citing author? Institutional stranger?

Count the independent letters. If you have zero or one, stop everything and email independent citing authors now. Your petition cannot file until you have at least two independent expert declarations. Count the close-colleague letters. If you have more than two, cut back — they're diluting your independent evidence.

For each letter, check: does it include (a) the recommender's credentials, (b) how they know of your work with specificity, (c) at least one quantified comparative statement about your contribution, and (d) an explicit statement about your standing in the field? Every letter must have all four. If any is missing, revise the letter with the recommender.

Hour 4: The Step 2 Argument

Write a new section of your brief titled "Final Merits Determination (Kazarian Step 2)." This section should be at least two full pages and should make an explicit argument for why the totality of your evidence — not just each criterion individually — demonstrates sustained national or international acclaim.

The structure of a good Step 2 argument:

1. Restate the legal standard (Kazarian's totality requirement).

2. Summarize the strongest two or three criteria and explain why their combination specifically indicates acclaim (e.g., "peer review at NeurIPS combined with 331 independent citations of original methodology demonstrates that the petitioner's work is both recognized and actively used by independent researchers").

3. Address any apparent weakness in the combination (e.g., "Although the petitioner does not claim Criterion 1 (awards), the EB-1A statute requires evidence meeting three criteria, not all ten, and the absence of an award does not diminish the sustained pattern of acclaim demonstrated under the six claimed criteria").

4. Conclude with an explicit statement that the totality demonstrates "sustained acclaim" and places the petitioner in the "small percentage at the very top" of the field — quoting the regulatory language directly.

Most denied petitions fail at this step. A strong Step 2 argument is the single most undervalued section of an EB-1A brief.

Hour 5: The Exhibit-to-Brief Reconciliation

Open your exhibit list and your brief side by side. Go through the brief linearly. Every time the brief says "See Exhibit X," flip to Exhibit X and verify that (a) the exhibit exists, (b) it contains what the brief claims it contains, (c) the specific supporting language is highlighted or otherwise findable within 30 seconds, and (d) the exhibit has a cover page.

Expect to find 5-15 mismatches in a typical first draft. Fix every one of them.

Hour 6: The Pre-emption Pass

Now that you've done the other five hours, re-read your brief as an adversary. For each criterion section, ask: "What's the most likely challenge an adjudicator would write in an RFE?" Write that challenge in the margin. Then write a sentence in the brief that addresses the challenge before it can be raised.

Common pre-emptions:

• For scholarly articles in lower-Impact-Factor journals: explain venue selection in the context of the field.
• For close-colleague expert letters: explicitly state the independence ratio and address the AAO pattern.
• For citation counts below 500: cite field-specific benchmarks from Web of Science Essential Science Indicators.
• For C7 claims at smaller organizations: proactively establish the organization's distinguished reputation with specific metrics.
• For petitioners from underrepresented countries: pre-empt the assumption that "national acclaim" requires U.S.-only recognition.

By the end of Hour 6, your brief should contain 5-10 pre-emption paragraphs that weren't in the original draft.

The One Thing No Manual Audit Can Do

There's a reason attorneys charge $8,000-$20,000 for full EB-1A representation, and it's not primarily the writing. It's the pattern matching against precedent — the ability to read your petition and instinctively know which phrases will trigger which RFE patterns based on having read hundreds of prior decisions.

You can reproduce most of that by running the 6-hour protocol above, carefully and honestly. But there is one thing the manual protocol can't do: compare your petition directly against the specific AAO decisions that match your field, your profession, and your country of origin.

This is what Lumova's audit product does, and it's why we built it. When you upload your petition to Lumova, the system parses every claim in your brief, embeds every exhibit semantically, and runs your evidence against the Retrieval-Augmented Generation index of 10,000+ real AAO decisions. For each criterion you've claimed, it surfaces the AAO decisions most similar to your profile — including the ones that were denied — and shows you the specific language those denials used.

The output is a structured report that gives you:

• A Kazarian two-step verdict on each of your claimed criteria (Step 1 meets/partial/fails with reasoning, and a separate Step 2 totality verdict)
• A per-criterion RFE likelihood with a specific reason tied to an AAO pattern (e.g., "62% RFE likelihood on C5 because the petition's Original Contributions section uses unquantified significance claims; see similar AAO outcome from [year] denying on nearly identical language")
• A field percentile telling you where your profile sits against real AAO-approved cases in your exact profession and country segment
• An overall RFE likelihood range (e.g., "35%-55%") with the specific factors driving the range
• A readiness score on a 0-100 scale with a verbal band (Not Ready / Developing / Promising / Strong / Excellent)
• Prioritized action items ranked by predicted impact — not a laundry list, but the specific next five things to fix

The audit costs $197 one-time or is included in the Pro plan at $149/month (one audit per month, unlimited if you upgrade to Elite). That's less than one hour of attorney billable time. And it surfaces the specific AAO-trained patterns that an attorney without a 10,000-decision retrieval system genuinely can't reproduce by reading alone.

What Happens After the Audit

After Priya ran the protocol plus the Lumova audit, her readiness score came back at 71 — in the "Promising" band, just below "Strong." The audit identified four high-impact fixes:

1. Rewrite C5 (Original Contributions) with quantified significance metrics and pre-emption of the "major significance" standard.

2. Add three independent expert letters.

3. Separate self-citations from independent citations and report both.

4. Add a two-page Step 2 Final Merits argument.

Priya spent another three weeks implementing the four fixes. She re-ran the audit. Her new readiness score was 88 — "Strong." Her per-criterion RFE likelihoods had dropped: C5 went from 62% to 28%, C4 went from 35% to 18%, C7 stayed at 22% (this one the audit flagged but couldn't fix without a higher-authority expert letter, which Priya decided to accept the risk on).

She filed. Five months later, she was approved without an RFE.

This is not an unusual outcome for petitioners who run the audit carefully. Most of the RFEs we see in post-mortem analysis could have been prevented by catching the specific patterns the audit flags. The petitioners who skip the audit and file anyway are not necessarily weaker candidates — they just didn't know the specific language traps that AAO adjudicators are trained to flag.

The Time Cost of Not Auditing

Let's do the math on skipping the audit.

If you file without an audit and receive an RFE, here's your realistic cost:

• Time cost: 40-80 hours of focused work to write an RFE response (most petitioners underestimate this by half)
• Timeline cost: 3-6 months added to your green card timeline (the RFE adds 30-90 days for the response, then USCIS takes another 3-6 months to adjudicate)
• Stress cost: The psychological weight of a pending RFE while on a non-immigrant visa is severe for most applicants
• Denial risk cost: Approximately 30% of RFE responses are followed by denial. If you land in that 30%, you're looking at either refiling (another 12-18 months) or appealing to the AAO (another 12-24 months with odds under 10%)

Compare this to the cost of running the audit first:

• Time cost: 6 hours for the manual protocol plus 30 minutes to review the Lumova audit output
• Timeline cost: Add 2-3 weeks for implementing the fixes
• Stress cost: Significantly lower than waiting for an RFE
• Denial risk cost: Substantially reduced — not zero, but dramatically lower

There's no rational argument for skipping the audit. The only reason most petitioners skip it is that they don't know it exists, or they assume "my attorney already did it" (in most cases they did not). If you've read this far, you know better.

FAQ

Q: How long should I wait between the audit and filing?

A: Long enough to implement the recommended fixes, and no longer. Most petitioners need 2-4 weeks to execute the full list of audit recommendations. Don't let perfect be the enemy of good — the audit will always find something to improve, but at some point you need to ship. A readiness score of 80+ with no red flags is generally enough to file with confidence.

Q: Can I run the audit myself without paying for Lumova?

A: Yes, the 6-hour manual protocol above is the foundation of a good audit. The Lumova audit adds the AAO retrieval layer, the per-criterion probability scoring, and the field percentile — things you can't reproduce without a dataset of 10,000+ decisions. If you're budget-constrained, start with the manual protocol; it will catch most of the high-impact issues. If you can afford the $197, the Lumova audit is the highest-leverage dollar you'll spend in your entire EB-1A process.

Q: Will the audit tell me if I should just not file?

A: Yes. A readiness score below 50 with red flags across multiple criteria is an honest signal that your profile isn't yet at the level USCIS approves, and filing now would waste money and risk a denial on your record. The audit is not a sales tool — when it tells you not to file, believe it and use the output to build the missing evidence over the next 6-18 months.

Q: What if my attorney says I don't need an audit?

A: Attorneys with extensive EB-1A experience effectively audit your petition as part of their drafting process, and if you trust their track record and they've read enough AAO decisions, you may not need a separate audit. But very few attorneys systematically benchmark against a retrieval corpus of 10,000+ decisions, and most "quick reviews" under $1,000 are proofreads, not audits. Ask your attorney specifically: "Have you run my petition against recent AAO denial patterns in my field?" If the answer is no, run the audit yourself.

Q: Does the audit replace an attorney?

A: No. An attorney is a legal representative with attorney-client privilege and is authorized to give legal advice on your specific situation. Lumova is not. The audit is educational — it identifies patterns in your petition that correlate with RFE outcomes in our dataset and suggests revisions, but it does not give legal advice. For most self-petitioners, the ideal workflow is: run the audit, implement the fixes, then optionally pay an attorney for a focused 1-2 hour review of the final draft. That hybrid approach typically costs $500-$1,500 total and catches both pattern-level and legal-nuance issues.

Q: What's the single most common finding?

A: Unquantified significance claims in Criterion 5 (Original Contributions). It's the finding we see in approximately 85% of first-draft petitions. If you do nothing else with this article, go read your C5 section right now and ask: "Can I defend the word 'significant' with a number?"

Try Lumova free →