The Charles Smith Blog reports on the handwriting examination section of the Oregon State Police Crime Laboratory closing

I have been a long time fan of The Charles Smith Blog. It is firmly rooted in my Feedly RSS reader. It should be one of your regulars too. It always provides for great summaries of the forensic science stories that are in the news.


For example here is his great treatment on the closing of the Oregon State Police Handwriting section:

STORY: “Oregon State Police close handwriting lab after investigation of bias, sloppiness,” by reporter Brian Denson, published by The Oregonian on February 14, 2014.

PHOTO CAPTION:  “Handwriting examiners in the Portland Metro Forensic Laboratory, in Clackamas County, used video spectral comparators for the painstaking process of matching suspects’ known writing samples to those believed to have been used in the commission of crimes. The unit is now shut down.”

GIST: “Oregon State Police officials have quietly shut down their handwriting analysis unit, investigated for lapses in quality control, and laid off their last two examiners at a time when fewer crime cases require the work. The department now farms out handwriting analyses to the Washington State Patrol’s forensics laboratory and, in some cases, FBI examiners, according to Lt. Gregg Hastings, an OSP spokesman. Hastings confirmed that the Questioned Documents Unit, the formal name of the handwriting examination group, formally closed on Dec. 7, 2012, nine months after OSP officials suspended the unit’s examinations. Oregon State Police had faced the possibility that the findings of its two full-time examiners had caused problems with criminal prosecutions. This forced them to confront a whodunit under the roof of their own forensics lab. The Oregonian first reported on the controversy in late 2012, after state police discovered that one of its two handwriting examiners committed a serious error in a suspected murder-for-hire case. Internal reviews of work in the handwriting examination unit, obtained by the newspaper through open-records requests, detailed allegations of bias, sloppy work, and dishonesty………State police officials looking into miscues in the handwriting unit in 2012 identified 45 cases that required outside reviews by qualified handwriting examiners. They notified prosecutors handling the cases. Hastings said this week that it appeared no criminal cases were significantly altered or harmed by the findings. Killing the Questioned Documents Unit prompted the layoffs of its last two full-time analysts: Ron Emmons and Christina Kelley. The two had been on paid administrative leave in the $66,000-a-year jobs until state police formally closed their unit.”

According to several sources, the very first use of handwriting examination in the courtroom happened in 1792 trial Goodtitle Drevett v Braham reported at 100 Eng Rep 1139 (1792). As a science endeavor, the validity of handwriting examination for conclusions as to source has been majorly under attack ever since the 2009 National Research Counsel National Academy of Science report “Strengthening Forensic Science in the United States: A Path Forward.” We find in the NAS report the following apt summary of the discipline as typically practiced in the United States today:

The examination of handwritten items typically involves the comparison of a questioned item submitted for examination along with a known item of established origin associated with the matter under investigation. Requirements for comparison are that the writing be of the same type (handwritten/cursive versus hand printed) and that it be comparable text (similar letter/word combinations). Special situations involving unnatural writing are forgery (an attempt to imitate/duplicate the writing of another person) and disguise (an attempt to avoid identification as the writer). The basis for comparison is that handwriting/handprinting/numerals can be examined to obtain writing characteristics (also referred to as features or attributes). The characteristics are further classified into class characteristics (the style that the writer was taught), individual characteristics (the writer’s personal style), and gross/subtle characteristics. Specific attributes used for comparison of handwriting are also referred to as discriminating elements, of which Huber and Headrick have identified 21. Comparisons are based on the high likelihood that no two persons write the same way, while considering the fact that every person’s writing has its own variabilities. Thus, an analysis of handwriting must compare interpersonal variability—some characterization of how handwriting features vary across a population of possible writers—with intrapersonal variability—how much an individual’s handwriting can vary from sample to sample. Determining that two samples were written by the same person depends on showing that their degree of variability, by some measure, is more consistent with intrapersonal variability than with interpersonal variability. Some cases of forgery are characterized by signatures with too little variability, and are thus inconsistent with the fact that we all have intrapersonal variability in our writing.

Scientific Interpretation and Reporting of Results
Terminology has been developed for expressing the subjective conclusions of handwriting comparison and identification, taking into account that there are an infinite number of gradations or opinions toward an identification or elimination. Several scales, such as a five-point scale and a nine-point scale, are used by questioned document examiners worldwide.
The nine-point scale is as follows:
1. Identification (a definite conclusion that the questioned writing
matches another sample)
2. Strong probability (evidence is persuasive, yet some critical quality
is missing)
3. Probable (points strongly towards identification)
4. Indications [that the same person] did [create both samples] (there
are a few significant features)
5. No conclusion (used when there are limiting factors such as disguise,
or lack of comparable writing)
6. Indications [that the same person] did not [create both samples]
(same weight as indications with a weak opinion)
7. Probably did not (evidence is quite strong)
8. Strong probably did not (virtual certainty)
9. Elimination (highest degree of confidence)

Boy doesn’t that all sound wonderfully subjective?

All you have to do is to consider the following example and ask did this come from one person or more than one?

The above comes from the FBI review article “Handwriting Examination: Meeting the Challenges of Science and the Law.” There is clear variation in the four samples. Is the variation due to the person or due to it being different people? This is a basic question. This is not even close to the more complex question that the typical handwriting examiner is asked to make and offers in court every day which is the attribution to a particular source. When dealing with a true unknown sample whose source is truly not known such as the typical seized and submitted sample (e.g., a bank robbery note), the source of this variation as illustrated above drives the validity of the opinion as to the capability of attributing it to a sole source (or not). In this particular case because it is not an unknown, meaning that the person writing these four “Samantha Scott Smith” entries was seen by the examiner, it was easy for them to conclude that it was simple intra-personal variation (same person, just “natural” variation”), and not inter-personal variation (coming from two or more people). This is Figure 4 of the review article and is entitled “Four signatures written by the same individual, demonstrating variation.”

The major tenets that the theory of handwriting analysis can be used to attribute source stands upon is as follows:

  1. that each person’s handwriting is unique (They even claim that unlike DNA, identical twins will write differently), and
  2. that a given person’s handwriting is also relatively stable and changes little over time.

The ultimate thought is a sample of a person’s handwriting can be compared to that of a similar style handwritten document (meaning cursive to cursive and print to print) to determine and authenticate the written document’s writer; if the writing styles “match,” it is likely that one person wrote both documents. Unlike other parts of Question Document Examination where chemistry is involved, ultimately the typical handwriting expert is making a value judgment of “sufficient similarity” based upon that examiner’s training, knowledge and experience. Although there are published standards for this interpretation such as ASTM E2290 – 07a (Standard Guide for Examination of Handwritten Items), even that standard calls for a very high degree of subjectivity and variability.

In the typical crime laboratory, the analysts are not using more empirical methods such as the CEDAR-FOX system. Even when the FBI wanted to defend the science of handwriting analysis it relied not exclusively upon the human interpretation as much as the human interpretation guided by the CEDAR-FOX system as one can see in the review article cited above.

In the typical crime laboratory, it is the lack of relying on more objective means of evaluation such as the CEDAR-FOX system, the lack of the proof of the tenets that supposedly ground handwriting analysis, and the lack of objective criteria that is not subject to inter-rater or intra-rater variability that makes the opinion expressed very vulnerable to bias. Hence why the state of Oregon ultimately did the correct thing and abolished the handwriting unit in its state crime laboratory.

Despite all of this, in the last major examination of its validity under the Daubert standard, in United States v Prime, 431 F . 3d 1147 (C.A. 9 , 2005), it survived scrutiny. Certainly this forensic science discipline is subject to legitimate attack in the courtroom.

Although he was talking about graphology, what best sums up the truth of handwriting analysis as practiced in a good amount of crime laboratories can be summed up by a quote from Barry L Beyerstein, Ph.D when he said “[T]hey simply interpret the way we form these various features on the page in much the same way ancient oracles interpreted the entrails of oxen or smoke in the air. i.e., it’s a kind of magical divination or fortune telling where ‘like begets like.”

Leave a Reply

Your email address will not be published. Required fields are marked *