Graphology: An Introduction
(This is my first post anywhere on anything related to tech or tools, so, I apologize if there should be a certain format to it.)
Introduction
Graphology is the analysis of the physical characteristics and patterns of handwriting with attempt to identify the writer, indicate the psychological state at the time of writing, or evaluate personality characteristics. It is generally considered a pseudoscience. — Wikipedia
Even though the art itself is a few centuries old, it wasn’t until 1967 when James U. McNeal Published his work “Graphology: A New Marketing Research Technique” in the Journal of Market Research in 1967 that popular opinion and scientific curiosity made their way into the field. It was followed by several attempts to find a mathematical basis for treating graphology as a standard for assessing human emotion or behavior. My team and I, as a part of our University project report, went through roughly 43 research papers from 1967 i.e. McNeal’s work to the ones published in 2019. Most of them have been published in reputed journals but none of them could conclusively establish graphology as a science of studying behavior.
Most of them, including McNeal’s original work, established a fairly high degree of accuracy though, the coefficient of correlation between 0.5 and 0.7.
Psychology is difficult to quantify
It is common practice in psychology to presume that phenomena of interest can and should be represented numerically, and that inferences should be based on mathematical analysis of those numerical representations…
Based on our review, we argue that the limitations of quantification are in direct conflict with the motives that drive its use. We further argue that this is primarily because quantification in psychology has historically been, and continues to be, a generally unreflective practice.
— Excerpts from Kathleen Slaney’s and Scott Neufeld’s Work “Quantification in Psychology: Critical Analysis of an Unreflective Practice” published in 2016 in the Journal of Theoretical and Philosophical Psychology
We made a project
Using the currently available data and methodology employed by graphologists, we have tried to create an AI agent that works as a graphologist with a rudimentary knowledge of the same. Our project achieved a high degree of correlation under the conditions we had set (approx. 0.67).
The coefficient of correlation is high though.
The mathematical correlation of 0.5–0.7 is a pretty high degree of influence between two given factors or sets, especially on a subject as fluid as psychology.
So, based purely on statistical analysis (we made, my team and I came up with the conclusion that graphology can in fact be considered as a tool for studying human behavior but with limited application i.e. it cannot be and must not be treated as a pure science as of now. Probably evidence derived both empirically and those derived from graph theory and its application in forensic analysis of handwriting may generate a database large enough to generate generalized results.
Yet again, I fear that this might not be able to achieve a 100% generalization.
