INTRODUCTION

Graphology is the study of the relationship between handwriting traits and personality traits.  Graphologists claim that handwriting is an expressive movement and certain features relate to a person character allowing a trained[1] graphologist to evaluate and gauge characteristics such as loyalty, ambition, motivation, sincerity, reliability, long-term sexual compatibility, and in general the psychological health of an individual.  These personality characteristic can be interpreted from a person handwriting style by examining such handwriting characteristics as: spacing, slant, pressure, symmetry, legibility, connection of letters, and pace of writing.

            Graphology is a tool most commonly used in the employment field as a psycho-diagnostic tool.  There are advantages for employers to use graphologist over the more traditional psychologist evaluations.  The reports rendered by graphologist are cheaper and non-invasive.  This allows employers to use graphology without the knowledge of potential employees.  Graphologists require only a piece of handwritten text, referred to as scripts.  Ideally the scripts should be written on an unlined sheet of paper and contain a brief biographical statement of the subject.  Graphologist will normally only require the subject age, sex, handedness, and level of education.  However, they will render evaluation on just about written text.

            Graphologists claim that their interpretations are accurate and reliable.  There are several validation studies that say otherwise.  Whether or not graphology is of any value to employers or individuals seeking personal insight, is not the concern of this paper and it will be left to your own judgment whether any credence should be given to the field of graphology.  This paper focuses on whether or not an expert witness, trained in graphology, and offering expert testimony should be admitted under the current rules of evidence in the post-Daubert world.  It is clear from the many validation studies that graphological testimony should not be permitted. 

A BRIEF HISTORY OF GRAPHOLOGY

In 1871 Abbe Jean Hippolyte Michon published, the Practical System of Graphology, which gave the science of graphology its name.  Since then graphology has spread globally as a technique for testing personality.  Graphologist fall within one of three schools of philosophy.  The trait school, which Michon is associated with, claims that specific graphological traits such as the slant of writing reflect a specific personality trait.  The Gestalt school which evolved in Germany in the early 20^th century claimed that interpretations must be made on the writing as a whole relying on the interpretation and judgment of the graphologists.  In 1929 graphoanalysis was developed as a compromise between the two schools and is referred to as the holistic personality pattern, which combines both a specific one-on-one trait analysis with the more general intuitive approach.  Graphology, while explored as a possible legitimate predictor of character trait by the psychology profession, has now been discarded as valid.  Presumably, psychologist abandoned graphology due to the negative validation studies, lack of ethics in graphological associations, and low or non-existent educational standards required to practice.[2]

VALIDATION STUDIES

“In theory, astrology could be a science if reliability and validity were established, so could graphology, or phrenology.  The door to science’s house is open to anything reasonable, reliable and valid.  Much of modern science rests on premises that would have been laughed at within our century.  A good deal of medicine has become science only in the latest decades by using statistical techniques, and systematic studies and by improving methods of observing the operation of the living, uninvaded human body”[3]. The validation studies show that graphology is neither reliable, valid, or of any more assistance to the trier of fact that using their own judgments. 

            In the early 1970’s, a study group for graphology was initiated in the Netherlands by the Society of Industrial Psychology, consisting of validation research in the form of four related experiments.  Notable is the fact that of the ten scientists conducting the study, four were graphologist, two of which served on the board of the largest graphological society in the Netherlands.  The results failed to validate graphology as reliable.[4]

            The first study involved ten psychologist, ten graphologist, and 10 recent graduate in psychology with a course in graphology.  The study concluded that “graphological training, knowledge and experience; finds no support.  The null hypothesis cannot be rejected.”[5]  The results as to the graphologist found that their “deviations from chance expectation proved to be very slight, even when they were positive and exceeded the significance thresholds.  In particular, the discrepancies between the ‘practically certain’ judgments of graphologists were somewhat disturbing.  Other evaluation techniques likewise showed the reliability of the graphologist’s judgment to be very low.”[6]

            The second study repeated the first, however the time constraints put on the graphological analysis were removed and the judgment of 10 laymen were added (three medical students, three law students, three economic students and a geology student, with no training in graphology).  This study had similar results as the first with laymen having similar results as the other four groups and slightly better than chance.  The study found “most decidedly alarming are the great discrepancies between the graphologists in their judgments…, graphologists attain so little agreement either with themselves or with others…that graphology cannot qualify as a stable and reliable method.[7]

            The third study evaluated the judgment of three groups consisting of ten psychologist, ten graphologist, and ten laymen, theology students.  The task was to predict, from an analysis on the handwriting samples whether the person was energetic or weak. Energetic was defined as the ability to work independently, showing initiative and perseverance, and not easily discouraged with disappointment.  Weak was described as the opposite of this and personnel files containing reviews were used to create the specimens.  The study found that the graphologist’s judgments of the energy variable (energetic or weak) was not very reliable, that there were no differences between the three groups and the judgment of the graphologist have only a limited value as predictors of business personnel ratings.[8]

            The last study repeated the first, however removed the constraint of choosing either energetic or not and allowed the graphologists to make a complete review.  The study found that “as a diagnostic method graphology is not only low on validity, but also demonstrable poor in intersubject reliability” and that “graphologist did not distinguish themselves in their results from laymen”.[9]  All four studies found similar results: that graphologists are no better than the laymen from predicting personality characteristic, can not agree on their results, and are no better than chance.

            In 1986 Baruch Nevo edited a collection of works on graphology.[10]  A chapter of this collection contained a several studies. One study found that graphologist’s predictions compared against that of employee’s immediate supervisors can achieve positive, if low (i.e., between .2 and .3), validities.  However when the non-graphologist evaluate the same text, they achieve similar, if not higher results.[11]  In another study comparing graphological judgment against military personnel files found that the “element of graphoanalysis as the sole basis for military personnel selection is premature”.[12]  The study further concluded that lay persons and psychologists who were given scripts of individuals achieved greater agreement than morphological handwriting signs.”[13]  In a third study conducted by Baruch compared the ability to predict criminal behavior by using scripts of inmates against non-inmates.  This study found that “the accuracy of criminal predisposition ratings is much lower than the claims put forward by some practitioners who try hard to market their skills as know all graphologists.”[14] In another study the conclusion was that the “single signs do not reflect sufficiently close and consistent relation as to enable diagnosis of intelligence of a given individual.”[15]  All the studies had similar findings: Graphologist’s predictions are indiscernible from laymen, their results are inconsistent amongst themselves, and successful interpretations are little better than chance.  The fact that judgments are found slightly more correct than chance can be linked to the fact that most analyzed script contain some sort of biographical information.  Personnel files or short auto-biographical text allows anyone, graphologist or not, glean certain character traits not from any morphological assessment, but from mere intuition.  One is likely to guess that a person is ‘successful’ when the scripts examined contain evidence of that very success. “In particular, contrary to graphological theory, the script content may give information useful to inferences and prediction.  To the extent that evidence of validity can be produced it may be based on these ostensibly irrelevant areas.  In fact, it is not uncommon for handwriting analysts to ask for and to use script samples of an autobiographical nature.  And predicting future behavior from autobiographical material is clearly feasible.”[16]

                        Graphology fails even when compared against standard personality tests.  In a study attempting to validate graphologist judgments against the Edwards Personality Inventory it was found graphologists to be unreliable and inconsistent.  “No agreement was found between personality profiles as determined by the Edwards Personality Inventory and by assessments by professional graphologist.  Subjects were also unable to recognize their own evaluations prepared for them by the analyst from bogus evaluations selected at random.”[17]  Another study compared graphologist’s results against Cattell’s Sixteen Personality Factor Test, which purports to reflect what the subject knows of himself, yielded similar results. This study found a negative correlation between the profiles generated by the graphologist and the standardized test.[18]  This study also found that certification and years of experience also had no effect on the graphologists ability.

            When Graphologist judgments are measured against supervisor evaluations it was found that when their judgments are based on autobiographical scripts, graphologist can achieve positive yet insignificant validities.  However non-graphologist achieved similar validities.[19]  Another study measure the ability of graphologist to correctly guess the profession of the author yielded results showing the unreliability of graphology.  The subjects were nationwide renown, spent their entire career in one profession for more than ten years.  The group consisted of mathematicians, philosophers, artists, executives, engineers, architects, physicians, and jurists.  Five graphologist participated, all of which had authored books, appeared on television, and were considered prominent in their field.  However, none of the graphologists were able to correctly predict an author’s profession with any accuracy.  Their mean probability of making a correct prediction exceeded a chance prediction based on binomial distribution by only .06.[20] This same study also compared the rates of agreement and found that in 7 of the 10 cases the extent of agreement between graphologists failed to reach significance and that the remaining three whose agreement reached significance had worked by mutual consultation.[21]

            Yet despite the overwhelming evidence graphologist still persist that their field is valid tool to predict personality.  One defense is that one cannot expect a linear relation between graphometric indicators and personality variables because graphology does not operate in a linear pattern.[22]  Also that Graphologist who participate in studies are forced to contradict their professional standards, by making analysis on un-readable scripts or requesting the graphologist to predict psychological variables in which they are not fully conversant.[23] However the results are in.  It is clear that what ever validation study was attempted that none of the studies found accuracy significantly higher than chance predictions.  Further, they can’t agree.  None of the studies were able to show that graphologist reliably gave consistent results, whether or not they were valid.

ADMISSIBILITY

            It is easily conceivable that Graphological testimony might be relevant in any given case.  However, relevance does not automatically create an obligation to admit the testimony.  It is clear that graphology fails under the Frye test of general acceptance.[24]  There is not one major accredited scientific organization that contains a section for graphologists.  The American Academy of Forensic Sciences does not contain a section for graphologists.  The American Board of Forensic Document Examiners, The American Society of Questioned Document Examiners, and any of their sub-groups do not contain members trained in graphology, nor do they practice graphology.  The numerous validation studies above clearly show that graphology is not accepted as reliable or valid.  In Daubert v. Dow Pharmaceuticals, Inc.,[25] the Court listed several factors to determine whether the particular theory or technique would assist the trier of fact and be admissible under Rule 702 of the Federal Rules of Evidence.  Graphology has been tested and subjected to peer review and publications; however the results weigh against admissibility.  The potential rate of error is known and shown in the validation studies to be unreliable.  The art of graphology is not accepted by the scientific community and has attracted only minimal support and is subject to great criticism.