Examinations
Handwriting expert opinion scale: what does "probably wrote" mean?
The SWGDOC terminology for forensic document examiners has nine levels, with identification at the top and elimination at the bottom. "Probably wrote" sits third from the top: the evidence points rather strongly toward the known writer but falls short of a "strong probability", where the evidence is very persuasive. A proposed Department of Justice text would require every qualified handwriting opinion to come with an explanation of its limiting factors.

What scale do handwriting examiners use for their conclusions?
The SWGDOC terminology sets nine levels: identification, strong probability, probable, indications, no conclusion, and four mirror levels down to elimination. Each level has a definition and model wording for the report.
SWGDOC published the Standard Terminology for Expressing Conclusions of Forensic Document Examiners, ver. 2013-2. Its purpose clause says the terminology "is intended to assist forensic document examiners in expressing conclusions or opinions". The document adds that the appropriate definitions "could be quoted in or appended to reports".
The nine levels run in order. On the side of common authorship: identification, strong probability, probable and indications. In the middle: no conclusion. On the side of different authorship: indications did not, probably did not, strong probability did not and elimination. Each level comes with an example sentence about a fictional John Doe.
Other scales exist. A 2021 presentation to the American Academy of Forensic Sciences by Mara L. Merlino and colleagues studied the "nine-level American Society for Testing and Materials (ASTM) opinion measure" alongside scales with 3, 5 and 7 levels. A 2018 study in the Journal of the American Society of Questioned Document Examiners used a 4-point scale of support.
What does "probably wrote" mean?
That the evidence points rather strongly toward the known writer but falls short of a strong probability. It is the third level of nine, below identification and strong probability and above indications.
SWGDOC defines the probable level as evidence that "points rather strongly toward" common authorship while stopping short of the level above it. The model sentence reads: "It has been concluded that the John Doe of the known material probably wrote the questioned material".
The standard does not translate the word into a percentage. No level of the scale carries a number, and the proposed Department of Justice language bars examiners from stating a level of certainty that is "absolute or numerically calculated". A report that says "probably wrote" gives a verbal position on a ranked scale, and its meaning comes from the definition.
A probable opinion is a qualified opinion. The SWGDOC text says that when "the opinion is less than definite, there is usually a necessity for additional comments", and that opinions may be qualified "when the evidence falls short of an identification or elimination."
How is "strong probability" different from an identification?
An identification means the examiner has no reservations whatever. Strong probability means the evidence is very persuasive, but some critical feature or quality is missing, so the examiner stops short of an identification.
For identification, SWGDOC writes that "The examiner has no reservations whatever" and is certain that the writer of the known material wrote the questioned material. The model wording is "It has been concluded that John Doe wrote the questioned material". The standard adds a rule for this level: "An identification is not qualified."
Strong probability is reserved for evidence that is very persuasive while "some critical feature or quality is missing". The model sentence reads: "There is strong probability that the John Doe of the known material wrote the questioned material".
The proposed Department of Justice language lists conditions for an identification. The examiner finds "significant characteristics in agreement, both in quality and quantity", "There are no fundamental differences to suggest another writer", "There are no significant limitations with the items examined", and "Unexplained characteristics are far outweighed by the combined effect of agreement in all other details." When a limitation prevents an identification, the opinion is reported at a qualified level.
What does "indications" or "may have written" mean?
The weakest positive opinion. There are few significant features, but those present agree with the known writing. SWGDOC calls it a very weak opinion and says it should always carry additional limiting words.
The model sentence for this level is "There is evidence which indicates (or suggests) that the John Doe of the known material may have written" the questioned material. SWGDOC adds: "There should always be additional limiting words or phrases" when the opinion is reported.
The proposed Department of Justice text uses the same step under the heading of a qualified "did write" opinion, with three strength terms: "Strong Probability Did Write," "Probably Did Write," and "Indications May Have Written." It describes the qualified opinion as one based on characteristics in common where "a limitation(s) exists which prevents an identification."
When does an examiner report no conclusion?
When limiting factors leave the examiner unable to lean either way. SWGDOC calls no conclusion the zero point of the confidence scale and names disguise and a lack of comparable writing among the causes.
"This is the zero point of the confidence scale," the SWGDOC definition says. The model sentence reads: "No conclusion could be reached as to whether or not the John Doe of the known material wrote" the questioned material. Other names for the level are totally inconclusive and indeterminable.
The proposed Department of Justice language describes the same outcome as an opinion that "no determination can be reached whether the items being compared were or were not prepared by the same writer(s)." It adds that the opinion "is usually reached because of significant or numerous limiting factors, such as a lack of comparability or lack of clarity and detail in the submitted items," and that it "requires an explanation of limiting factors."
A no conclusion result says nothing in favor of either side. It records that the material did not support an opinion, and under both texts the report names the limiting factors, such as a lack of comparable known writing or of clarity and detail in the items.
What do "probably did not write" and elimination mean?
They mirror the positive side. Probably did not means the evidence points rather strongly against common authorship. Elimination is the highest degree of confidence that the known writer did not write the questioned material.
The four levels below no conclusion are indications did not, probably did not, strong probability did not and elimination. SWGDOC writes that indications did not "carries the same weight as the indications term", and that strong probability did not "carries the same weight as strong probability on the identification side of the scale".
For probably did not, the model sentence reads: "It has been concluded that the John Doe of the known material probably did not write the questioned material". SWGDOC notes that some examiners prefer "It is unlikely that the John Doe of the known material wrote the questioned material", and calls "unlikely" merely the Anglo-Saxon equivalent of "improbable".
Elimination is defined as "the highest degree of confidence expressed by the document examiner in handwriting comparisons" that the writings are by different people. The proposed Department of Justice language mirrors this with "Strong Probability Did Not Write," "Probably Did Not Write" and "Indications May Not Have Written", and an elimination that requires significant disagreement and no significant limitations.
Why must a qualified opinion explain its limits?
Because the level alone does not tell the reader what held the examiner back. Both SWGDOC and the proposed DOJ text require an explanation of limiting factors whenever the opinion falls short of identification or elimination.
The proposed Department of Justice text says of the qualified did write opinion: "This is a less than definitive opinion and requires an explanation of limiting factors." It says the same of no conclusion. As examples of limiting factors it gives "a lack of comparability or lack of clarity and detail in the submitted items".
SWGDOC gives disguise and the lack of comparable writing as examples for the no conclusion level, and it asks for additional comments whenever the opinion is less than definite. A report that states a probable or indications opinion without naming the limitation gives the reader the level and leaves out the reason for it.
A lawyer reviewing the report can compare the stated limitations with the material the examiner received: copies or originals, how many known signatures, from which years, written in what circumstances.
What may a Department of Justice examiner not say about a handwriting match?
Under the proposed DOJ text, an examiner may not exclude all other writers, may not state an absolute or numerically calculated certainty, and may not claim that the method has a zero error rate or is infallible.
The three statements not approved read as follows. An examiner may not claim that two samples came from the same writer "to the exclusion of all other writers." "An examiner may not state a level of certainty in his/her opinion that is absolute or numerically calculated." And "An examiner may not state that the method used has a zero error rate or is infallible."
The document limits its own reach. "If adopted, this document will apply to Department of Justice personnel who perform examinations" and provide expert testimony on handwriting evidence. It states that it does not imply that different language used by Department personnel was "incorrect, indefensible, or erroneous", that its examples "are not intended to serve as precedent for other forensic laboratories", and that testimony practice "may be dependent upon the precedent set by the judge or locality".
The approved list itself has a five-part structure: identification, qualified did write, no conclusion, qualified did not write and elimination. The qualified categories hold the graded terms, so the DOJ structure covers the same nine positions as SWGDOC.
Do examiners agree on what each level means?
Not fully. A 2021 AAFS study found that the boundaries between opinion levels overlap and that examiners were sometimes unclear how to apply the scale language. A 2018 study reported no consensus on which scale to use.
Merlino, Freeman, Dahir and Edwards had examiners rate opinion strength on scales with different numbers of levels and on sliding scales of 0-20 and 0-100. They reported that "the boundaries between levels of opinion strength overlap", that "the range of opinion strengths varied according to the number of levels on the scale (3, 5, 7, 9)", and that examiners "were unclear in some instances as to how to apply opinion strength measure language". Their conclusion: "more research is needed to determine how the scale levels are understood".
Caligiuri, Mohammed, Lanners and Hunter wrote in 2018 that there is "no consensus within the discipline" on whether to use a 4-, 5- or 9-point scale and that "Reliability studies have not been conducted." They chose a 4-point scale to keep down the number of inconclusive opinions, which they describe as common with more complex schemes such as SWGDOC's. With strong or very strong support as the threshold, their four examiners were right 80 percent of the time on same-writer pairs and 97 percent on different-writer pairs.
Is the nine-level scale being replaced?
A proposed standard would report how strongly the findings support one proposition over another. A 2024 presentation by a Los Angeles Police Department examiner showed nine support levels and a table mapping the old terms.
Miriam Angel of the Los Angeles Police Department presented the proposed scale on June 25-26, 2024. The proposed scale has nine levels: extremely strong, strong, moderate and limited support for a common source, the same four for different sources, and equal support. The examiner assesses the findings under both propositions, the same writer and a different writer.
A transition table maps each SWGDOC conclusion to a support level, and the slides note that the mapping runs in one direction only. Limitations in the material, according to the slides, "move the probability/probabilities toward Neutral."
The idea is older than the presentation. In 2004 Norbert Koller, Kai Nissen, Michael Rieb and Erwin Sadorf published a 172-page German work on probability conclusions in handwriting opinions, listed by the National Criminal Justice Reference Service as NCJ 205916. Its abstract states that expert opinions "should not be made as categorical statements implying absolute certainty" and that the work discusses standardization of the probability scale.
What does IAE do when a handwriting report gives a qualified opinion?
An IAE examiner prepares a report on your questions about the questioned signature or writing, or a review of the opposing examiner's report, in up to 5 business days after the format and cost are confirmed.
The IAE handwriting page lists the questions: whether the person named on a will, contract or check signed it, and whether the signature shows signs of simulation or tracing. Under ANSI/ASB Standard 070 (2022) the examiner checks whether questioned and known writing are originals and comparable, weighs similarities and dissimilarities together, considers alternative interpretations and bases no conclusion on one feature alone.
A review of the opposing report checks whether it states each opinion with its basis and the facts or data considered, as Rule 26(a)(2)(B) requires, and whether the examiner followed Standard 070 on originals, comparability, distortion and alternative explanations. Absent a stipulation or court order, a rebuttal report is due within 30 days after the other side's disclosure under Rule 26(a)(2)(D)(ii). Since December 1, 2023, the party offering a handwriting opinion must show the court it is more likely than not that the opinion rests on sufficient facts or data and a reliable method reliably applied.
Send the report under review or the questioned document, the known writing with a note on the source and date of each item, and your questions through the Request a quote button on the service page. The secretariat confirms receipt the same day and, within one business day, the format, the cost, the deadline and the examiner from the registry. You approve the candidate and the CV before work starts, and cases with tight deadlines have a 48-hour rush.
Registry members passed the collegial attestation of the council under the published Criteria Matrix, and the Member ID in the report can be checked by QR code. The examiner confirms in the document that he or she has no interest in the outcome, and materials are not passed to third parties without your consent. The court weighs the report with the other evidence. Neither IAE nor the examiner guarantees the outcome, and IAE is not a law firm and does not advise on case strategy.
Sources
- SWGDOC Standard Terminology for Expressing Conclusions of Forensic Document Examiners, ver. 2013-2 (NIST)
- U.S. Department of Justice, Proposed Uniform Language for Testimony and Reports for the Forensic Handwriting Analysis Discipline
- Angel M. (Los Angeles Police Department), Communicating Forensic Findings: Forensic Document Examination, presentation, June 25-26, 2024 (NIST)
- Merlino M. L., Freeman T. M., Dahir V. B., Edwards C., J3 Measuring and Communicating Opinion Strength in Forensic Handwriting Examination: Concepts, Findings, and Future Directions, AAFS 2021
- Caligiuri M., Mohammed L., Lanners B., Hunter G., Kinematic Validation of FDE Determinations about Writership in Handwriting Examination: A Preliminary Study, Journal of the American Society of Questioned Document Examiners, vol. 21, no. 1, June 2018
- Koller N., Nissen K., Rieb M., Sadorf E., Probabilistische Schlussfolgerungen in Schriftgutachten (2004), NCJ 205916, abstract (OJP)
Read also
All articles
Can a notarized signature be challenged as forged, and what does the handwriting examiner have to show?
A notary's stamp does not make a signature unchallengeable, but it changes the proof. In New York a certificate of acknowledgment raises
Block capitals: can a handwriting expert identify who hand printed a note or a form?
Anonymous notes, address labels and filled-in forms are often written in block capitals, and the writer may expect capitals to hide his or her
Signed on a signature pad: can a handwriting expert examine an electronic handwritten signature?
A signature written with a stylus on a pad or tablet is still handwriting, and a handwriting examiner can compare it with known signatures. The imageExpert examinations
Ready to discuss your case?
Send the materials: the secretariat will confirm the format, cost, deadline and the specialist within one business day.

