Next Article in Journal
Measuring Biases of Visual Attention: A Comparison of Four Tasks
Next Article in Special Issue
Migration and Adaptation as Indicators of Social Mobility Migrants
Previous Article in Journal
Postnatal Acetaminophen and Potential Risk of Autism Spectrum Disorder among Males
Previous Article in Special Issue
Motivational Conflicts and the Psychological Structure of Perfectionism in Patients with Anxiety Disorders and Patients with Essential Hypertension
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:

Formalized Computer-Aided Handwriting Psychology: Validation and Integration into Psychological Assessment

IHS Institute for Handwriting Sciences, 8046 Zurich, Switzerland
Hochschule für Philosophie, 80539 Munich, Germany
Author to whom correspondence should be addressed.
Behav. Sci. 2020, 10(1), 27;
Received: 23 October 2019 / Revised: 22 December 2019 / Accepted: 30 December 2019 / Published: 3 January 2020
(This article belongs to the Special Issue XVI European Congress of Psychology)


In contrast to traditional researches that involve a manual, non-quantitative, and subjective way of performing handwriting analysis, in the current research, a special computer-aided method of revised handwriting analysis is used. It includes the detection of personality traits via manual quantitative registration of handwriting signs and their automated quantitative evaluation. This method is based on a mathematical–statistical model that integrates multiple international publications on the evaluation of handwriting signs. The first aim is the validation of the revised method against the 16 Personality Factor Questionnaire Revised (16PF-R), which is performed as a self-report personality test by test persons and was developed and researched empirically by Raymond B. Cattell et al. A second aim is the development of an integrated model for assessment including handwriting analysis: when both methods come to the same result on a certain scale, then the construct can be accepted with higher reliability; in contrast, when results are contradictory, they should be regarded as a limitation of each method and raise awareness in the researchers, as these contradictions are a precious source of additional information regarding the complexity, ambiguity, and context specificity of personality traits.

1. Introduction

Personality questionnaires based on self-assessment began evolving about a century ago and are still the most common instruments in personality assessment [1,2] for the the time and effort they require and their research economics [3]. In some cases, self-assessment is often the only access to a test person [4]. The 16 PF-R Personality Questionnaire is one of them and is still frequently used today, especially in career counselling and in business for employee testing and selection. Even though questionnaires are nowadays being tested with scientific methods and are improved in terms of performance data and robustness, certain problems of self-assessment questionnaires remain: systematic biases regarding retrospective questions about experiences and frequencies [5], subjectivity of test persons, dependency on context (e.g., purpose of test), tendencies to social desirability (e.g., in the hiring process), and tendencies to the middle or extreme by answers to questionnaire items and ambiguity of understanding questionnaire items.
This is why nonverbal tests are getting more popular. Handwriting psychology, historically known as graphology in its old from, represents a nonverbal testing system for personality traits. It is traditionally used for recruiting personnel, in medicine, forensic, and some other areas. There are certain advantages of this method for psychological assessment:
  • Large number of personality traits can be assessed in one procedure.
  • Social desirability is totally excluded.
  • Writing by hand is a normal and natural activity for most people that excludes the influence of testing situations.
  • Handwriting samples can be obtained independently of time, place, and physical presence of the test persons.
  • Psychological development of test persons can be assessed when several handwriting samples produced in different ages are available.
  • The language of the handwriting sample (valid for European languages) has no influence on the evaluation process.
As with every method, there are also disadvantages, which are mainly valid for the traditional way of handwriting analysis [6,7,8,9,10,11,12]: it can only be performed by specially educated experts and it is a non-standardized, complex, and time-consuming interpretation process with little transparency, recognition, and acceptance of the method in the field of psychology, due to contradictory validation results.
The aim of this study is to use the advantages of handwriting analysis as a nonverbal, additional test system in the framework of personality assessments and to avoid the disadvantages of the traditional method by using the revised, computer-based, and standardized evaluation of handwriting signs with the HSDetect program. By using the HSDetect program, an automatic, computer-based evaluation of 378 personality traits and behavior patterns is done after a manual registration of 544 handwriting signs in a standardized protocol. The content of a handwritten text which is used for the registration of the handwriting signs can be freely chosen by the test persons. In contrast to the traditional method of handwriting analysis, any information in the text or indications that would reveal something about the writer’s personality plays no role in the registration process, since the evaluation of the personal traits with HSDetect is automated and therefore takes place without any interference of an expert. With the revised method, it is not possible for an expert to know any connections between handwriting signs and personality traits included in the program, and therefore the expert cannot, as is common with the traditional method, select suitable personality traits manually from evaluation tables according to her/his experience. Details of the operating principles of the HSDetect program will be presented in Section 2.3 of this article.
In general, the authors are of the opinion that the number of research works with positive validation results is large enough not to disqualify the method and not to reduce their results to “anecdotal evidence” [13] (p. 82). It might be true that there is not enough validation information regarding handwriting analysis, but there is as well not enough scientific evidence and information either to reject it as an assessment method, especially, taking into consideration that the quality of validation studies done so far suffer from “significant methodological negligence” [8] (p. 191) and “many of these studies could be criticized methodologically in terms of measurement of both personality and graphology” [14] (p. 80).
The proposed, revised method with the HSDetect program opens new perspectives due to a formalized approach to handwriting analysis. It implements a clear and literature-based matching of handwriting signs to psychological traits by means of mathematical–statistical modelling and computer-based evaluation. The formalized procedure enables an easy integration of handwriting analysis as an additional useful method into the common personality assessment. It helps to avoid the biased influence of questionnaire tests based on self-evaluation. As a natural step before integration, a validation of the computer-aided method of handwriting analysis has to be done. In the current study, validation against the popular and well-validated test 16 Personality Factor Questionnaire Revised (16PF-R) was selected [15], based on the revised edition of the frequently used 16PF Questionnaire by Raymond B. Cattell, Maurice Tatsuoka, and Herbert Eber.
In summary, the objectives of the present study are as follows:
  • Development of a formal validation procedure of computer-aided handwriting analysis.
  • Application of this procedure by means of the 16PF-R.
  • Development of an integration model for personality assessments including handwriting analysis.

2. Materials and Methods

2.1. Project Overview

The project covered both the validation study and the model for integrating handwriting analysis into the 16PF-R testing procedure. The last ensures more reliable results for 16PF-R scales. The project consisted of the following steps:
  • Step 1: each test person performed the German version of the 16PF-R test and supplied a handwriting sample (free text, one A4-sized page) produced independently in their mother language.
  • Step 2: 16PF-R scales were modelled with matching personality traits from the database of HSDetect based on the scale description provided by the developers of the 16PF-R [15]. This model was generated by the authors, and these traits of HSDetect were used for evaluation and validation (cf. steps 5 and 6).
  • Step 3: 16PF-R was evaluated by the first author according to the test manual.
  • Step 4: handwriting signs of the handwriting samples (originals, no photocopies) were registered manually by the second author (qualified and experienced handwriting expert) with a standardized handwriting protocol consisting of 544 handwriting signs per test person.
  • Step 5: handwriting signs obtained with the protocols in step 4 were evaluated algorithmically with the HSDetect program. The result was a list of 378 personality traits and behavior patterns with percentages for each test person.
  • Step 6: validation of the 16PF-R test results against modelled scales of step 2 was done.
  • Step 7: computer-aided handwriting analysis could easily be integrated into psychological assessment to make the evaluation of psychological constructs more objective.

2.2. Participants and Material

The study included 58 participants characterized by three different mother languages and different ages and education levels. Of the participants, 22.4% (13) were male, and 77.6% (45) female; 12 test persons were younger than 30 years of age, 21 were between 30 and 45 years old, 16 between 46 and 60, and 9 over 60 years old.
The test persons were recruited in different areas (work, through colleagues, at events, and via newspaper ads). A required criteria for taking part in the experiment was the ability to write fluently by hand in one’s mother language and the willingness to complete the 16PF-R-questionnaire in German or English. Among the 58 test persons, 44 were German-speaking, 13 Russian-speaking, and 1 English-speaking.

2.3. Modelling of Handwriting Evaluation with HSDetect

Traditional validation studies of handwriting analysis have two major problems: the first one is that the procedure of personality trait evaluation used was mostly manual and subjective, although often the experts demonstrated statistically good agreement. The second problem is that the researchers took for practical reasons just small subsets of the handwriting variables and selected a very restricted set of personality traits [14,16,17]. The improved validation procedure proposed in this study and discussed in the current article is based on the computer-aided handwriting analysis called HSDetect and solves both problems. The system is based on the following principles:
  • Consolidation and harmonization of different handwriting analysis methods and schools to avoid biased results.
  • Formal presentation of all handwriting signs, personality traits, and the relations between them.
  • Quantitative registration of handwriting signs and evaluation of personality traits.
  • Assurance of evaluation objectivity and reliability.
In HSDetect, the handwriting signs are evaluated manually by experts, and the corresponding psychological traits are calculated algorithmically. Besides the algorithms of the handwriting analysis, HSDetect includes two databases: the database of the handwriting signs and psychological traits with connections between them (the handwriting analysis model) and the database of the evaluation of subjects’ handwriting samples—statistical results that serve as norming data and as a basis for different statistical studies.
Both the handwriting signs and the values of the psychological traits are presented on a continuous scale from 0 to 1. Each psychological trait is mathematically modelled as a function of several handwriting signs. The relations are complex, many-to-many, which means that a handwriting sign relates to several traits, and a typical trait is a function of several signs. A detailed description of the mathematical model of HSDetect can be found in earlier publications [18,19]. The main mathematical principles and an example are explained below:
The trait value of i-th trait ti is modelled by the following function:
ti = yiα ∙ ri1−α
where yi shows the strength of the i-th trait; ri stands for the reliability of the of the evaluation of the trait level; α is a parameter that in the case of this experiment was set to 0.6 on the basis of empirical trials. The strength of a trait yi is the function of the handwriting signs related to this trait:
yi = ∑ aij ∙ xj,
where xj is the value of the handwriting sign registered manually by the expert. The value of xj can range from 0, which means the sign is not present, to 1, which indicates that the sign is always strongly present in the handwriting. The parameter aij is a weight, which is defined on the basis of the statistical evaluation of multiple sources (publications): the more the sources reporting about an existing relation between the i-th trait and the handwriting j-th sign, the higher the value of aij. The sum is calculated over the set of Mi handwriting signs related to the i-th trait.
The reliability ri depends on the number of handwriting signs (ni) that are really present (that means xj > 0) in the analyzed handwriting sample. These signs are a subset of Mi. If we assume that the probability of an error, when we evaluate a trait on the basis of just one handwriting sign, is µ, then the expression for ri is as follows:
ri = 1 − µi,
Assuming µ = 0.25, the value of ri reaches a very high value of 0.9 with eight handwriting signs.
The last step is the normalizing of the trait value ti. That is necessary to make different traits comparable to the 16PF-R scales. In a handwriting sample, not every sign is present with the same frequency, which of course depends on language and writing conventions. That could lead to some biases in the trait values, and as a result, some traits would always have higher values than other traits. Normalization avoids this effect. A database of several hundreds of handwriting samples, which were previously evaluated with HSDetect, is used for normalization. Since the distribution of values for most traits is not normal, the following simple formula has to be used instead of the standard z-normalization:
zi = (yi − Yimin)/(Yimax − Yimin),
where zi is the normalized value of the i-th trait; Yimin and Yimax are the correspondingly minimal and maximal absolute values of the i-th trait (derived from the databank).
Below, the theory of the above-described mathematical–statistical basis of the HSDetect program is illustrated by an example. In Table 1, the mapping of the trait “responsiveness” (empathy, consideration) with the connected handwriting signs as reflected in expression (2) is presented.
Two examples of handwriting samples (cuttings) with high responsiveness are shown in Figure 1.
It can be clearly seen how different handwriting signs affect the trait level in the model. In practice, it is very rare that all of the possible signs per trait are present in a handwriting sample. Vice versa, often “contradictory” signs, those that point to an opposite, negative pole of the trait, are present as well. In Table 2, the model of the negative pole (“unresponsiveness”) is shown.
Two examples of handwriting samples (cuttings) with low responsiveness from our experiment are shown in Figure 2.
Since, in one sample, it is very likely to find signs related to both sides of a pole of a trait, it is always necessary to look at the difference between the plus and minus poles, which is calculated in expression (5). Figure 3 presents the result of the evaluation of both poles of the trait “responsiveness” and their difference. This an example of 6 out of 58 test persons. All test persons were evaluated with expression (4).

2.4. Scale Modelling of 16PF-R with Handwriting Analysis

In many validation studies with handwriting analysis and psychometric tests, modelling of test scales or factors is insufficiently implemented in the sense of oversimplification [9,20]. In the current project, each scale of the test was modelled by a thoroughly selected set of psychological traits based on the description of the 16PF-R test’s authors [21], which is presented in Table 3. The values of these traits were evaluated by handwriting analysis with HSDetect. Therefore, the path looked as follows: 16PF-R test scales -> corresponding psychological traits from HSDetect. In many former corresponding validation studies, the authors would directly map handwriting signs onto questionnaire test scales [16,17,22]. The number of manually involved signs is as well strongly restricted. The indirect integration of handwriting signs by modelling of the questionnaire test scale though traits of the HSDetect program is the uniqueness of this experiment and seems to be a smarter approach to effective validation results. This approach is practically possible only with a computer-aided procedure.
As already mentioned, in handwriting analysis, the complex relations between signs and traits could lead to unspecific results. Therefore, it is important to include into the modelling process not only the positive pole of a trait of a questionnaire test scale but also the negative one. For instance, for scale A (“warmth”), both already mentioned poles, “responsiveness” and “unresponsiveness”, are necessary, or for scale C (“emotional stability”), in addition to “emotional stability”, also the negative pole, “emotional instability”, must be taken into consideration.
Therefore, the formal presentation of a modelled scale value is as follows:
S = ∑zi+/n + −∑zj/n,
where S is the scale value; zi+ is the value of the i-th trait of the positive pole of the scale; zi is the value of the corresponding i-th trait of the negative pole; n+ and n are the numbers of traits of the positive and negative poles. The positive pole represents the high scores of a test scale [15,21], and correspondingly, the negative pole represents the low scores of a test scale.
Each scale of the 16PF-R is modelled with expression (5). The traits presented in Table 3 were calculated with expressions (1)–(4) for the evaluation of each test scale. The quality of the modelling of the scales was verified with Cronbach’s Alpha [23]. To estimate the Cronbach’s Alpha value. we used the norming data of HSDetect with several hundreds of handwriting samples and their evaluations.

2.5. Method of Validation

For validation of the computer-aided handwriting analysis, the calculated 16PF-R scales were compared with the results of the questionnaire test to verify the agreement between them. It is very important how a comparison procedure is designed. Often, researchers only use standard correlation as a simple and easy applicable method [24]. According to the authors’ opinion, this method is inappropriate for multi-factor psychometric instruments like handwriting analysis.
First, both methods are not really comparable, as handwriting analysis is a foreign-assessment method, and 16PF-R is a self-assessment method. It is well known that a gap between self- and foreign-assessment exists, which should not be underestimated [25,26,27].
Secondly, although the data of both 16PF-R and handwriting analysis are quantitatively generated, a reasonable interpretation cannot be so mathematically exact for individual test persons (as occurred in our experiment). Only statistically calculated variables (like mean or deviation) of large sample sizes can be purely mathematically compared.
Thirdly, the nature and number of items in 16PF-R and handwriting analysis are so different that the two types of results can only be formally compared with a system that reduces the complexity of these differences. Besides, stochastic influences of different factors on both systems are high.
To allow a reliable agreement check, the results of the 16PF-R-factor presentation of both the test and the handwriting analysis were mapped on a scale with three intervals: positive (strong domination of the positive scale pole), negative (strong domination of the negative scale pole), and neutral (absence of strong domination of either pole). The zone boundaries were individually defined for each scale according to statistical data.
In the 16PF-R questionnaire, each question is mapped on one scale, and each answer (with a choice of three variants) has its weight (for 16PF-R, the weights are 1, 2, or 3). That is typical for a questionnaire test. Thus, each test scale has its minimal and maximal scores depending on the number of related questions. For instance, scale A ranges from 9 to 27, since there are 9 questions related to it. For the zone-based comparison, the negative zone ranges from 9 to 15, the neutral zone from 15 to 21, and the positive zone starts at 21 and finishes at 27. Also, for scale G, the negative zone is from 11 to 18, the neutral zone from 18 to 26, and the positive zone from 26 to 33. This assumes that the distribution is close to uniform, which is often the case.
In the same way, zones for the handwriting model are defined. However, the boundaries are not predefined but depend on the results of the evaluation of all subjects. Predefined zone boundaries could result in strong biases: their theoretical values can change from −1.0 to +1.0. Practically, the distribution for each trait and, correspondingly, every scale might vary very strongly and it is far from being uniform.

2.6. Method of Integration

Traditionally, psychological assessments prefer the use of direct self-reports in form of standardized questionnaires, whereas some researchers prefer the use of indirect, nonverbal testing methods, as they are interested in test results without the effects of language interpretation problems, cultural differences, and faking while filling out questionnaires [25,28].
From the perspective of the test persons, it is often not easy to fill out questionnaires, as they start reflecting about “consistency–inconsistency” or the “organizations’ expectations” during the answering process [25] (p. 448). Other mentioned obstacles with answering questionnaire tests are the use of extreme versus middle responses, the “lack of ‘ambiguous’ as a response option”, and the lack of “information about the context” for some items [25] (p. 446).
The mentioned disadvantages of self-assessment questionnaires can be compensated by the integration of handwriting analysis in the assessment. The integration enables the improvement of evaluation objectivity of the psychological constructs, which are presented in the current research by 16PF-R scales: when both 16PF-R and handwriting analysis come to the same result on a certain scale despite different response styles of the test persons, then this construct can be accepted with higher reliability. When results are contradictory, then they should be regarded as a limitation of each method and raise awareness in the researchers, as these contradictions are a precious source of additional information regarding the complexity, ambiguity, and context specificity of personality traits.

3. Results

The summary of the evaluation of 58 test persons for each scale is shown in Table 4. It shows the numbers of agreements, neutral results, and contradictions (Contr.). Agreement per test person means that the results are in the same zone, whereas contradiction means that they are in opposite zones. Neutral represents the combinations with adjoining zones.
To check statistical significance, binomial distribution with three outcomes can be used. For 58 test persons and the standard significance level of 0.05, the critical number of outcomes is 25. From Table 4 we can conclude that 14 of 16 scales showed a good statistical agreement. No scales showed statistically significant disagreement.

4. Discussion and Conclusions

In general, it can be concluded that the zone-based comparison instead of a simple correlation showed reasonably positive results between 16PF-R and handwriting analysis. The assumption that people who can be labelled as “perfectionists” (Q3) on one method should also appear “perfectionists” on all other methods is grounded in a one-dimensional view of the human nature that holds it is impossible for people to be a “perfectionist” and a “non-perfectionist” at the same time. A variety of different contexts can be identified that illustrate that personality traits are not uniform and one-dimensional across personality. In addition, method variance [29] due to distinctions between the methods discussed in Section 2.6 has to be considered as a factor not to be neglected. Therefore, the strengths and limitations of each assessment method (questionnaire self-assessment test and foreign-assessment handwriting analysis) should be recognized and appreciated. The different response styles of the 16PF-R test and the handwriting analysis in combination can provide useful information about personality traits, as both methods together can consistently illuminate the full scope of any personality construct.
On the basis of our findings, future research could confront the test persons with the self-concepts determined by the results of the 16PF-R and the handwriting analysis and find out which concepts in which specific contexts fit best to the self-perception of the test persons. With these results, the 16PF-R scales modelled with traits from handwriting analysis can be differentiated context-specifically. This will contribute to the differentiation of the 16PF-R scales and to the improvement of scale modelling (cf. Table 4), while taking into account the dynamic and stable aspects of self-perception.
The presented unique approach of revised, computer-assisted handwriting analysis can also be implemented in big-data research, as new connections between handwriting signs and personality traits could be determined in future studies, considering the more than 500 recordable handwriting signs in the HSDetect database and results from various verbal and nonverbal personality tests.

Author Contributions

Both authors have equal contributions to the article. That was a mutual research. All authors have read and agreed to the published version of the manuscript.


This research received no external funding.

Conflicts of Interest

The authors declare no conflict of interest.


  1. Weiner, I.B.; Greene, R.L. Handbook of Personality Assessment, 2nd ed.; John Wiley & Sons: New York, NY, USA, 2017; ISBN 978-1-119-25890-2. [Google Scholar]
  2. Stemmler, G.; Hagemann, D.; Amelang, M.; Spinath, F.M. Differentielle Psychologie und Persönlichkeitsforschung, 8th ed.; Kohlhammer: Stuttgart, Germany, 2016; ISBN 978-3-17-025721-4. [Google Scholar]
  3. Mummendey, H.D.; Grau, I. Die Fragebogen-Methode; Chap. 1.3; Hogrefe: Göttingen, Germany, 2008; ISBN 978-3-8409-1948-7. [Google Scholar]
  4. Schmidt-Atzert, L.; Amelang, M.; Fydrich, T.; Moosbrugger, H. Psychologische Diagnostik, 5th ed.; Springer: Berlin, Germany, 2018; ISBN 978-3-662-58054-7. [Google Scholar]
  5. Kahneman, D. Thinking, Fast and Slow; Part II; Penguin: London, UK, 2012; ISBN 978-0-14-103357-0. [Google Scholar]
  6. Ben-Shakhar, G.; Bar-Hillel, M.; Bilu, Y.; Ben-Abba, E.; Flug, A. Can graphology predict occupational success? Two empirical studies and some theoretical ruminations. J. Appl. Psychol. 1986, 71, 645–653. [Google Scholar] [CrossRef]
  7. Tett, R.P.; Palmer, C.A. The validity of handwriting elements in relation to self-report personality trait measures. Personal. Individ. Differ. 1997, 22, 11–18. [Google Scholar] [CrossRef]
  8. Klimoski, R.J.; Rafaeli, A. Inferring personal qualities through handwriting analysis. J. Occup. Psychol. 1983, 56, 191–202. [Google Scholar] [CrossRef]
  9. Fox, J. A correlational Analysis between Handwriting Characteristics and Personality Type by the Myers-Briggs Type Indicator. Ph.D. Thesis, University of the Rockies, Denver, CO, USA, 2010. [Google Scholar] [CrossRef]
  10. Barnes, E.M.; Six, J.; Hawkins, R.C., II. PsychoGrapho: A double-blind cohort study of Jungian typology and handwriting analysis. J. Psychol. Res. 2016, 6, 50–56. [Google Scholar] [CrossRef][Green Version]
  11. Furnham, A.; Gunter, B. Graphology and personality: Another failure to validate graphological analysis. Personal. Individ. Differ. 1987, 8, 433–435. [Google Scholar] [CrossRef]
  12. Eysenck, H.J.; Gudjonsson, G. An empirical study of the validity of handwriting analysis. Personal. Individ. Differ. 1986, 7, 263–264. [Google Scholar] [CrossRef]
  13. Driver, R.W.; Buckley, M.R.; Frink, D.D. Should we write off graphology? Int. J. Sel. Assess. 1996, 4, 78–86. [Google Scholar] [CrossRef]
  14. Furnham, A.; Chamorro-Premuzic, T.; Callahan, I. Does graphology predict personality and intelligence? Individ. Differ. Res. 2003, 1, 78–94. [Google Scholar]
  15. Cattell, R.B.; Cattell, A.K.; Cattell, H.E.P. 16PF Fifth Edition Questionnaire. Champaign; Institute for Personality and Ability Testing: Champagne, IL, USA, 1993. [Google Scholar]
  16. Gawda, B. Lack of evidence for the assessment of personality traits using handwriting analysis. Pol. Psychol. Bull. 2014, 45, 73–79. [Google Scholar] [CrossRef][Green Version]
  17. Kacker, R.; Maringanti, H.B. Personality Analysis through Handwriting. GSTF J. Comput. 2012, 2, 94–97. [Google Scholar] [CrossRef]
  18. Chernov, Y. Psychological Handwriting Analysis. System Analysis and Computer-Aided Applications in Psychology and Forensics; Genesis: Moscow, Russia, 2011; ISBN 978-5-98563-259-0. [Google Scholar]
  19. Chernov, Y. Formal validation of handwriting analysis. In Handwriting Research: Validation & Quality; Chernov, Y., Nauer, M.A., Eds.; Epubli: Berlin, Germany, 2018; pp. 37–68. ISBN 9783746720388. [Google Scholar]
  20. Gavrilescu, M.; Vizireanu, N. Predicting the big five personality traits from handwriting. EURASIP J. Image Video Process. 2018, 1. [Google Scholar] [CrossRef][Green Version]
  21. Cattell, H.E.P.; Schuerger, J.M. Essentials of 16PF Assessment, 1st ed.; John Wiley & Sons: New York, NY, USA, 2003; ISBN 978-0-471-47413-5. [Google Scholar]
  22. Bushnell, I.W.R. A Comparison of the validity of handwriting analysis with that of the cattell 16PF. Int. J. Sel. Assess. 1996, 4, 12–17. [Google Scholar] [CrossRef]
  23. Cronbach, L.J. Coefficient alpha and the internal structure of test. Psychometrika 1951, 16, 297–333. [Google Scholar] [CrossRef][Green Version]
  24. Sedlmeier, P.; Renkewitz, F. Forschungsmethoden und Statistik für Psychologen und Sozialwissenschaftler, 2nd ed.; Pearson Higher Education: München Harlow, Germany, 2013; ISBN 978-3-86894-131-9. [Google Scholar]
  25. König, C.J.; Merz, A.-S.; Trauffer, N. What is in applicants’ minds when they fill out a personality test? Insights from a qualitative study. Int. J. Sel. Assess. 2012, 20, 442–452. [Google Scholar] [CrossRef]
  26. Ziegler, M.; MacCann, C.; Roberts, R.D. (Eds.) New Perspectives on Faking in Personality Assessment; Oxford University Press: New York, NY, USA, 2012; ISBN 978-0-19-538747-6. [Google Scholar]
  27. Holden, R.R. Underestimating the effects of faking on the validity of self-report personality scales. Personal. Individ. Differ. 2008, 44, 311–321. [Google Scholar] [CrossRef]
  28. Amelang, M.; Schäfer, A.; Yousfi, S. Comparing verbal and non-verbal personality scales: Investigating the reliability and validity, the influence of social desirability, and the effects of fake good instructions. Psychol. Beiträge 2002, 44, 24–41. [Google Scholar]
  29. Campbell, D.T.; Fiske, D.W. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol. Bull. 1959, 56, 81–105. [Google Scholar] [CrossRef] [PubMed][Green Version]
Figure 1. Handwriting samples with high responsiveness.
Figure 1. Handwriting samples with high responsiveness.
Behavsci 10 00027 g001
Figure 2. Handwriting samples with low responsiveness.
Figure 2. Handwriting samples with low responsiveness.
Behavsci 10 00027 g002
Figure 3. Example of trait Responsiveness for the first six subjects.
Figure 3. Example of trait Responsiveness for the first six subjects.
Behavsci 10 00027 g003
Table 1. Handwriting signs for trait “responsiveness”.
Table 1. Handwriting signs for trait “responsiveness”.
NoHandwriting SignParameter aij
1Connection form—garlands or threads0.193
2Normal or strong right slant0.130
3High fullness of letters (especially middle zone)0.105
4Pasty stroke formation0.084
5Last letters have lighter pressure0.072
6Letter size in words is tapering0.060
7Diacritic marks are bound to the letter and are irregular0.043
8Lead-out stroke of last letters is long ascending or horizontal0.038
9Larger middle zone0.038
10Connected handwriting0.038
11Legible signature0.035
12Irregular handwriting0.029
13Ovals open at top0.027
14Wide letters0.018
15Legible handwriting0.018
16Small letter size0.018
17Diacritic marks—stronger pressure, emphasized, heavy0.013
18Average word spacing0.009
19Return stroke of lower zone is rising to the right of stem0.009
20Left margin is wide0.009
21Narrow line spacing0.009
22Rising line direction0.009
Table 2. Handwriting signs for trait “unresponsiveness”.
Table 2. Handwriting signs for trait “unresponsiveness”.
NoHandwriting SignParameter aij
1Thin, sharp stroke formation0.145
2Angular connection form0.145
3Emphasized last letters0.082
4Elongated or threading letter form0.095
5Uneven, not controlled margins0.065
6Narrow letters (small width)0.063
7Smaller middle zone0.063
8Horizontal line direction0.063
9Slant in words becomes smaller0.060
10Narrow margins0.050
11Thin handwriting form0.038
12Vertical or left slant0.057
13Small letter size0.038
14Narrow word spacing0.019
15Slow speed0.019
Table 3. Scale modelling of the 16 Personality Factor Questionnaire Revised (16PF-R) with handwriting analysis.
Table 3. Scale modelling of the 16 Personality Factor Questionnaire Revised (16PF-R) with handwriting analysis.
ScalesScale LabelPositive PoleNegative Pole
AWarmthresponsiveness, compulsiveness, need of contacts, lack of reserveunresponsiveness, reserve, keep distance from the outer world
BReasoningshrewdness, logical thinking, abstract thinkinginertia of thinking, absence of logic, concrete thinking
CEmotional Stabilityemotional stability, stability under stress, poiseemotional instability, cannot keep stability under stress, irritability
EDominancedominance, persuasiveness, conflicted character, self-assuredness, obstinacysoftness, diplomacy, inability to persuade, self-uncertainty
FLivelinesslack of reserve, infantilism, naturalness, vanity, lack of restrains, loquaciousnessmoderation, maturity, reticence, artificiality, reserve, inconspicuousness
GRule-Consciousnessconventionality, resignation, politeness, responsibility, resistanceeccentricity, irresponsibility, independence, impoliteness, lack of resistance
HSocial Boldnessextroversion, vanity, interpersonal skills, impudenceunsociability, inconspicuousness, diffidence, withdrawnness
ISensitivityemotionality, sensibility, sentimentality, receptivenessgood judgement, cynicism, insensitivity, insusceptibility
LVigilanceskepticism, carefulness, mistrustnaivete, imprudence, enthusiasm
MAbstractednesscreativity, idealism, imaginativeness, poor judgement, inattentivenessgood judgement, lack of imagination, reasoning, lack of creative ability, attentiveness
NPrivatenessinsincerity, tactfulness, nondisclosuresincerity, tactlessness, indiscretion
OApprehensionconscience, melancholia, self-uncertainty, anxiety, inner conflictintegrity, unconcern, unscrupulousness, self-assuredness, zest for life
Q1Openness to Changecuriosity, dynamics, versatility, fussiness, solidity, lack of resistanceresistance, conservatism, lack of curiosity, narrow field of interests
Q2Self-Relianceindependence, individualism, self-sufficiency, self-dependencedependency, cooperativeness, need of contacts
Q3Perfectionismneatness, meticulousness, methodicalness, zealousnessnegligence, disorderliness, inaccurateness, scrappiness
Q4Tensionirritability, impatience, nervousness, self-control giving way to excitementpatience, quietness, poise, control of initial impulses
Table 4. Result summary of zone agreement of 16PF-R and HSDetect (n = 58).
Table 4. Result summary of zone agreement of 16PF-R and HSDetect (n = 58).

Share and Cite

MDPI and ACS Style

Chernov, Y.; Caspers, C. Formalized Computer-Aided Handwriting Psychology: Validation and Integration into Psychological Assessment. Behav. Sci. 2020, 10, 27.

AMA Style

Chernov Y, Caspers C. Formalized Computer-Aided Handwriting Psychology: Validation and Integration into Psychological Assessment. Behavioral Sciences. 2020; 10(1):27.

Chicago/Turabian Style

Chernov, Yury, and Claudia Caspers. 2020. "Formalized Computer-Aided Handwriting Psychology: Validation and Integration into Psychological Assessment" Behavioral Sciences 10, no. 1: 27.

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop