In Claim Raven's analysis, condition records tagged as quoting lay evidence had a 41.5% favorable-outcome share. Records tagged as not quoting it had a 22.0% share. That's a substantial difference, but it is not a measurement of what happens when a veteran rewrites a statement.
The distinction matters. The Board writes its explanation after reviewing the evidence. A quotation in that explanation may show why evidence mattered, or simply reflect how the judge chose to describe the case. Our comparison cannot separate those possibilities.
I rechecked the underlying counts for this article. They support a useful discussion about lay evidence. They don't support a promise that getting quoted doubles your chances.
What the numbers actually count
The comparison covers 77,151 condition records from 42,083 distinct Board decisions, checked on September 15, 2026. One decision can address several conditions and therefore contribute several records. These are not 77,151 different veterans or independent appeals.
| Extracted lay-evidence label | Condition records | Favorable outcomes | Favorable share |
|---|---|---|---|
| Quoted in the decision | 38,143 | 15,848 | 41.5% |
| Not quoted in the decision | 39,008 | 8,581 | 22.0% |
The difference is 19.5 percentage points. The table uses Claim Raven's shared outcome definition: grants, partial grants and mixed outcomes in the numerator; those outcomes plus denials and remands in the denominator. A remand sends an issue back for further work and is not counted as a grant. Dismissed, withdrawn, moot and unclassified outcomes are excluded from that denominator.
The larger database contains 101,518 condition records from the analyzed Board-decision subset. Another 24,367 records have no quoted/not-quoted label, so they are absent from this comparison. That missing information is one reason the table should not stand in for all Board appeals.
Our VA disability statistics reference explains the wider dataset and its outcome definitions.
A quotation is part of the Board's reasoning
A decision that discusses a statement in detail is different from a decision that merely lists it. But the difference is observed in the decision itself, after the evidence has been weighed.
That creates several possible explanations for the gap. Stronger claims may contain more useful statements. Judges may quote evidence that helps explain a favorable finding. Certain conditions may make observations from a veteran or family member especially relevant. The mix of conditions and other evidence may differ between the two groups.
The table does not tell us how much each explanation contributes. It also doesn't establish that a statement was missing, ignored or legally inadequate when the extraction says it was not quoted.
The original cross-tabulation by medical-opinion quality does not solve that problem. Those quality labels were also extracted from written decisions. Comparing records within a label is not the same as holding all other evidence constant, and it does not establish an independent effect of quoting a statement.
I would use this finding to ask how lay evidence appears in the Board's reasoning. I would not use it to estimate a reader's odds.
What lay evidence can add to a claim
Lay evidence is information from someone describing facts they know or things they observed, without needing specialized medical training. A veteran can describe symptoms and their effects. A spouse, coworker or fellow service member may be able to describe changes or events they witnessed. That is the basic distinction in 38 CFR § 3.159(a)(2).
The useful question is whether the statement helps establish a fact relevant to the claim. A detailed account of interrupted sleep, missed work or a witnessed injury may address something a brief treatment note doesn't explain. A statement cannot automatically supply a medical diagnosis or a complex medical link simply by describing symptoms confidently.
For example, a spouse may be able to describe when they first noticed headaches and what happened during the attacks. Whether PTSD caused those headaches is a different question that may require medical evidence.
Specific observations are more useful than rating language
Consider these two fictional examples:
My back is terrible and I deserve a higher rating.
During our grocery trips over the past two months, I have usually stopped walking after about ten minutes because of back pain. My wife has finished the shopping while I sit near the checkout. Before that period, I could usually finish the trip with her.
The second example gives the reader an activity, a time period, an approximate frequency and a change. Those details can be compared with other evidence. They do not make the statement automatically credible, and the example is not language to copy unless it accurately describes someone's own experience.
A useful statement usually answers a few ordinary questions:
- What did you experience or observe?
- When did it happen, and how often?
- What activity did it interrupt or change?
- How do you know? Were you there, or are you repeating what someone told you?
- Are the dates exact, or your best recollection?
Precision includes being honest about uncertainty. If you remember a season rather than a day, say that. If symptoms vary, describe the variation. A statement becomes less useful when it turns an estimate into a false exact date or describes every day as the worst day.
Start with the fact your evidence needs to explain
Before adding another statement, identify the question it is meant to answer. Is the gap an event during service, the history of symptoms, the effect on daily activities, or something else? Then identify who has firsthand knowledge of that fact.
That keeps the statement focused. It also helps separate an observation you can provide from a medical question you may need a clinician to address. VA's evidence requirements describe the broader evidence needed for different claim types.
If you need help organizing a witness account, our buddy-letter examples show several ways to describe firsthand observations. The goal is a clear, accurate account that answers a relevant question. Whether the Board later quotes it is outside your control.
Sources and limits
This is Claim Raven's descriptive analysis of structured, AI-extracted condition records from Board decisions. The aggregate counts were rechecked against the database on September 15, 2026 using the shared outcome taxonomy. That check verifies the calculation; it does not independently validate every extracted quotation label or outcome against the original decision.
The sample consists of appealed matters captured in our database, not a representative sample of initial VA claims. Multiple records can come from the same decision, and a decision can appear in both quoted and not-quoted groups for different conditions. Missing labels, extraction error, differences in condition mix and the Board's own selection of evidence for discussion limit interpretation. No causal effect or individual approval probability is estimated here.
I'm Landon, founder of Claim Raven and an Army veteran. Claim Raven sells claim-preparation tools; this article is an explanation of our analysis, not legal or medical advice. For help choosing a response in your own case, VA provides a directory of accredited representatives.
VA Disability Appeal Letter Samples: Three Review Templates
VA claim denied but favorable findings? Read the two together
Judge Variance at the BVA: 40 Points Apart on Similar Volume