Journal of Forensic Science and Medicine

ORIGINAL ARTICLE
Year
: 2018  |  Volume : 4  |  Issue : 4  |  Page : 203--212

Research on subjective bias cognition effect in handwriting identification


Bing Li, Tiantian Ma 
 Key Laboratory of Evidence Law and Forensic Science, Ministry of Education, Collaborative Innovation Center of Judicial Civilization, Institute of Evidence Law and Forensic Science, China University of Political Science and Law, Beijing, China

Correspondence Address:
Prof. Bing Li
Fada Institute of Forensic Medicine and Science, No. 26, Houtun South Road, Qinghe Xiaoying Area, Haidian District, Beijing
China

Abstract

This article starts from the point of view that handwriting examination is dependent on empiricism and by analyzing the core steps of handwriting identification explains that it might be influenced by subjective bias. In practice, examiners cannot avoid biases in decision-making; instead, we must accept the existence of subjective bias in handwriting identification and then discuss its impact; for instance, feature selection in the process of comprehensive evaluation, which involves a comparison of the number and quality of similarities and differences between a questioned sample and the references. While we conclude that comprehensive evaluation is the most important step in the identification process, industries in China do not stipulate explicit and transparent criteria for it, making it hard to numerically quantify the characteristics of handwriting identification. In this article, forensic examiners' opinions on handwriting identification were obtained through a survey. One finding was that most handwriting examiners believe that handwriting identification is subject to subjective bias. In addition, they believe that the subjective cognition of handwriting identification can somehow help actively produce the correct opinion; before the examination, most handwriting examiners think that they should understand the context and so on. Finally, through the questionnaire, which contained variations such as the same case with different background information, different cases with the same background information, and the same case with or without context, it was concluded that handwriting identification does have certain subjectivity. However, which kind of factors influence this subjectivity is not presently clear. Furthermore, it is difficult to control uncertainties when forming an opinion on identification. An alternative way is to perform scrutiny after the formulation of handwriting opinions; for example, internal and external reviews such as appearing in court.



How to cite this article:
Li B, Ma T. Research on subjective bias cognition effect in handwriting identification.J Forensic Sci Med 2018;4:203-212


How to cite this URL:
Li B, Ma T. Research on subjective bias cognition effect in handwriting identification. J Forensic Sci Med [serial online] 2018 [cited 2019 May 20 ];4:203-212
Available from: http://www.jfsmonline.com/text.asp?2018/4/4/203/248700


Full Text



 Current Practices in Handwriting Identification



The “experience” of handwriting identification

According to cognitive psychology, intuition differs from logical and conscious reasoning, whereas the ability to recognize or understand things is a way of thinking that is opposed to logical analysis and conscious thinking. In this sense, intuition is a sufficient experience, and the acquisition of this experience is the result of the accumulation of intuition. To a large extent, experience and intuition are synonymous.[1]

Experience is the knowledge individuals accumulate in the process of study and work. It has certain universality and plays an important role in an individual's understanding of an object. Therefore, experience is an important factor that affects individuals' understanding of initiatives. When an individual screens, extracts, and processes information regarding an object, the existing knowledge and work experience also participate in this process and influence the formation of the individual's new knowledge.[2] For example, the price of real estate in China has been increasing in recent years. As a result, many sellers have denied having signed contracts. Therefore, some examiners have conducted casework related to the signatures in the contracts. After having worked on such cases, when an examiner encounters a similar case, he/she might have a subjective prejudgment and believe that the seller has a purpose and then conduct the examination under this prejudgment. Accordingly, when handwriting examiners go about their work, their subconscious may play a certain role.

The learning of professional skills is divided into three stages, namely cognitive, related, and autonomous. These three stages are a part of the constant process of progress, from the unfamiliar to the familiar and from a skill to a habit. The autonomy phase refers to the process in which the central cognition begins to launch tasks when we become proficient in an area.[3] In the long-term accumulation of their own work, handwriting examiners gradually enter the autonomous stage of skill learning. When they come into contact with cases that seem familiar or which involve handwriting similar to what they have previously encountered, they will have some spontaneous responses. For example, when dealing with an imitation handwriting case for the first time, a handwriting examiner must pay attention based on what he/she has learned. After a certain number of comparisons, he/she may dare to say if it is a case of imitation. However, after many years of empirical experience, it is possible to make a quick and accurate judgment of imitation.

This automated response to skills based on self-knowledge learning and work experience is human nature; it is not specific to handwriting examiners. In many kinds of empirical-based work, the accumulation of work experience will lead to some subjective automated responses. For example, some experienced doctors only need to check the pulse for a minute or two to diagnose the illness. After many years of teaching, teachers will know very well what sort of knowledge is obscure and hard for students to grasp. Similarly, experienced investigators can quickly target the suspect and solve the crime.

Expert opinion is a kind of scientific evidence that should satisfy the requirements of neutrality, scientific basis, and absolute impartiality. As explained before, expert opinions are based on examiners' knowledge accumulation, work experience, and intuition. Whether this kind of guidance of the subconscious is positive or negative constitutes the focus of controversy. This controversy may also lead to doubts or questions about handwriting identification.

Procedures of handwriting identification and current issues in China

General steps involved in handwriting identification

Handwriting identification can be roughly divided into three core steps as follows:

Acceptance and hearing a caseIdentification and interpretationWriting a report of expert opinion.

The identification process of step can be roughly divided into three steps as follows:

Respective tests

These analyze the questioned sample and references separately. First of all, the examiners have to analyze all the characteristics, from quantity to quality, in order to decide whether the sample meets the qualification conditions. Second, the control samples must be adequate; there should be enough samples from the same period and in the same condition as the questioned sample in order to determine whether the handwriting is normal or an imitation or disguised handwriting. The main point of this step is to carry out the analysis of the questioned material and that of references separately.

Feature comparison

This is a summary of the first step, involving careful comparison of the characteristics of the questioned writing and references by the examiners.

Comprehensive evaluation

This is the key step, involving analysis of differences and similarities and a determination of whether the questioned and control samples reflect the same person's writing habits. The comprehensive evaluation is to determine the nature of the differences and similarities in handwriting characteristics that have been found and recorded. The main task is to analyze whether the similarities or differences in writing characteristics are intrinsic.

However, there are many factors that cause nonintrinsic features, also called “nonintrinsic” similarities. The examiner needs to interpret it.

The current situation and problems in the comprehensive evaluation of handwriting identification in China

Owing to the lack of an adequate theoretical base, there are still some questionable issues. The first relates to the characteristics of handwriting. There are various ways to classify handwriting features. Overall, there are two main categories, which can be divided into eight subcategories with 11 types of handwriting characteristics. There is no clarity about the value of each handwriting feature, and between similarities and differences, which are more valuable. The second point is the core step in identification – comprehensive evaluation. In a handwriting examination, there will certainly be many points of similarity, and there will be many differences as well. However, we have to determine which are intrinsic and which are nonintrinsic by comparing the similarities with the differences. In addition, how many intrinsic matching points can be found to reach the identity? There are no uniform rules.

In summary, the core process of examination involves analysis and comparison. Similarities and differences can either be intrinsic or nonintrinsic. There are no objective uniform standards; each examiner has his/her own considerations. This can be because of erroneous results due to examiners' carelessness. The lack of objective standards in the industry and laymen's misunderstanding of experts' opinions may also be caused by this.

 Domestic and Foreign Research Status on the Issue of Subjectivity in Handwriting Opinions



Domestic research status

Regarding this topic, although there is relevant literature available, most perspectives still remain at the theoretical level. Handwriting identification is discussed in terms of Marxist epistemology, the scientific principles of handwriting identification, and the basis of cognitive psychology. Some literatures even directly mention that “in the handwriting identification, human factors play a certain dominant role,” without explaining what the “human factors” are and analyzing their causes.[4] There are also some studies that only contain piles of paragraph theory. It is necessary to determine the source and role of examiners' subjective bias, which requires practical data or real cases for demonstration.

When it comes to subjective bias in handwriting identification, most of the literatures directly demonstrate the scientific nature of handwriting identification while avoiding the main question. From the writing habits formed by dynamic stereotypes, to the philosophical thinking of individualization, and elaboration of the principles of handwriting identification, it seems that handwriting identification backed by scientific principles should not be subject to subjective biases.

In this case, we need to explore whether the subjective bias of handwriting identification has positive or negative effects, so as to provide suggestions for handwriting examiners as well as to make expert opinions more neutral and credible.

Foreign research status

Foreign researchers are doing a better job than our own. However, although they realized that the forensic sciences were confronted with serious criticism with respect to cognitive bias (e.g., Miller, Risinger, and Dror) and put Miller's suggestions into practice, no good general procedures have been implemented for decreasing the challenge of cognitive bias in most laboratories. There is still a need for procedural changes in the forensic sciences.

An experienced examiner who has conducted experimental research and analysis is trying to determine whether handwriting examiners can be affected by the information of the case and make mistaken handwriting testimony. The experimental process is briefly described as follows: 12 students were trained in handwriting identification. After passing the test, they were divided into two groups of six. The samples for each group were the same. (The samples did not belong to the suspects; the students were just provided with various contexts). For Group 1, the police had eyewitnesses to prove that the suspect had committed the crime and there was a sample belonging to the suspect and for Group 2, there were three suspects and their samples but no other evidence. The results of the experiment were surprising. Only one of the six participants in Group 1 offered a negative opinion, four offered a positive opinion, and one presented no opinion. The six participants of Group 2 presented negative opinions. The final conclusion was to keep away from irrelevant context; there is still a need for procedural changes in forensic sciences.

The authors accept that background information should be obtained when accepting the case. We believe that the handwriting examiner is free from any influence and can make correct conclusions with respect to his/her opinion. After reading the conclusions of this article, we also look forward to our own experimental results.

Thinking about the subjective bias in handwriting identification

To sum up, dynamic stereotyping, individualization theory, and other referred theories make handwriting identification feasible and practical. However, there are still many doubts. From the illustration above, the questions mainly stem from two aspects. First is the fact that handwriting examination mainly relies on the accumulation of knowledge and experience, which means that examiners may be affected by the materials and samples. Second is the lack of objective quantification standards and quantitative comparison of results; instead, human factors account for a very important part.

In more and more fields, quantification seems to represent a robust scientific basis. However, in the practice of handwriting identification, the implementation of the same recognition activity depends to a large extent on professional experience. The brevity of the process makes it impossible for subjectivity in handwriting identification to be effectively constrained by a quality monitoring system. Many people believe that the lack of specific quantification criteria in the individualization process makes the final decision unpredictable.

The issue of subjective bias in handwriting identification has received a lot of attention. Although various disciplines are subjective in nature; for example, the lack of a uniform objective standard for medical treatment, with doctors' individual opinions forming the basis of decisions, none of them has been questioned as much as handwriting identification. The subjective bias of handwriting examiners may be derived from the unconscious accumulation of experience during their studies and work. This is a situation that anyone can encounter during learning and working. Can you reject the scientific nature of handwriting identification only because of subjectivity? Can the credibility of the opinion be considered low just because it is subjective? Is it because of subjectivity that examiners fail to be neutral? We need empirical research to answer these questions.

 the Questionnaire Survey



Design

The design of the questionnaire was based on Baidu.com's “questionnaire research” service. The design process was simple and easy to operate. The topic could also be easily set. After completing the questionnaire design, a link was easily generated, which could be sent through the likes of QQ and WeChat to any mobile or PC terminal. The questionnaire could be filled and submitted simply by accessing the Internet. The submitted questionnaire results were displayed on the designer's home page.

From the interpretation above, the authors came up with several factors that may affect the subjective bias of handwriting examiners. The main issues are work experience, knowing in advance whether the context is forming a prejudgment, attitude toward expertise work, court appearances, and so on. It is mainly from these factors that we designed some basic questions.

In order to avoid psychological preconceptions, one examiner had to complete only one questionnaire and one comparison. Handwriting identification is time-consuming and laborious. It is not expected that an expert will become bored as obtaining a single test paper. It is also necessary to control the number of questions and the way to attempt them. Only in this way can conclusions be more scientific and objective.

Hence, the final task was to complete two comparable questionnaires and contrast two handwriting cases. One questionnaire was divided into two parts, all of which had multiple-choice questions. Each questionnaire contained only two handwriting cases, which were clear and easy to understand. The first questionnaire was issued to examiners, students, and interns in this field as a control group, whereas the second questionnaire was only issued to professional examiners and compared with the first questionnaire.

The first part only set 11 simple multiple-choice questions (the first questionnaire had 13 questions and the second questionnaire had 11). The first ten questions had only one possible answer, whereas the last questions had more than one possible answer among the multiple choices and the item was left blank.

The second part was a case analysis. Cases 1 and 2 were designed separately. One was a blank case without background information and the other was a reverse case with strong information. Both questionnaires contained the same cases; the only difference was in the context.

The design of the first part of the questionnaire

In general, the structure of a questionnaire should not be too complex. It should be easy to read and understand and there should be clarity regarding options. Overly long questionnaires not only induce psychological boredom in interviewees, but can also affect the authenticity of the data and the reliability of the results. In general, questionnaires that take 20–30 min to complete are ideal.

There are many factors that affect the subjective bias of an examiner, including work experience, life experience, knowledge and skills, professional ability, or specific information. We selected some representative factors and compiled the following multiple-choice questions.

The first part contained 11 questions covering the examiner's age, professional background, work experience, professional titles, courtroom appearances, selection preferences regarding handwriting characteristics, cognitive attitudes to the context, and so on. These questions covered most aspects of handwriting identification.

The questionnaire was issued via mobile terminals such as WeChat or QQ. To filter out invalid questionnaires, we set some age, education, and post options. The starting point was basic; the impact of age in our daily life and work is enormous, and the authors wanted to explore whether these two factors have some influence on handwriting identification and can screen valid questionnaires.

Case design of the questionnaire

In order to enable the questionnaire to be completed quickly and prevent emotional dissatisfaction among interviewees, the materials provided included not only magnified images of the questioned and control samples, but also the comparison tables of handwriting features. Considering that the originals of all the cases could not be sent to every expert, the signatures of the two selected cases were already worked out under the magnifier and microscopic observation. They all had moderate writing speed and normal writing lines and had used ink gel pens. There was no possibility of copying. In such a case, there was no need to test the original. Only a careful analysis of the comparison tables of handwriting features was required to arrive at a conclusion.

The signatures used in the two questionnaires were exactly the same, and both cases had actually been completed. Cases 1 and 2 had the following characteristics: (i) the casework was conducted more than twice by different laboratories, indicating some controversies; however, considering the difficulty of the two cases and the fact that they were very similar, they provided credible data for use in questionnaire surveys and analysis; (ii) the control samples were in very good condition and had high credibility; and (iii) the case setting of the first questionnaire was inductive, whereas that of the second questionnaire was relatively neutral. In this way, by comparing different cases from the same questionnaire and the same cases from different questionnaires, a more convincing conclusion could be drawn.

Case 1 in the first questionnaire was a blank case without any background information. There was only a questioned sample, control samples, and the comparison table of handwriting features. Case 2 was inductive. While Case 1 in the second questionnaire contained basic information and was not inducible, Case 2 was designed to be a blank case without any context.

Reliability and validity test of the questionnaires

Each questionnaire must be tested for reliability and validity.

Reliability analysis of these questionnaires

Most reliability tests use the retest-reliability analysis method, which involves testing the same group repeatedly at different times (usually 14–48 days apart). The reliability of the questionnaire can thus be analyzed according to the degree of difference between the two results. However, if one examiner were to take two tests, he/she would form a prejudgment before the second test, which could affect the effectiveness of the questionnaire. Therefore, the authors did not use the retest-reliability analysis method and instead used SPSS software version 22.0 for Windows (IBM company, New York, US).

From the statistical point of view, a reliability coefficient between 0.6 and 0.7 indicates that the internal consistency of the scale is within the acceptable range, between 0.7 and 0.8 indicates better internal consistency, and between 0.8 and 0.9 indicates very good internal consistency.

Before performing the reliability test, each item needs to be scored. Accordingly, to the question “How important do you consider the role of forensic expertise in the conviction of the judge?” interviewees had to respond with one of the following options: “Very important,” “Important in some cases but not always,” “Almost as important as other evidence,” and “Not very important,” which were scored as 4, 3, 2, and 1, respectively.

For the question “Is there any influence of subjective bias on the expert's opinion?” the possible answers were “Definitely, because expert opinion is the result of subjective opinion,” “Has a partial effect,” “Uncertain if it will be affected,” and “Will not be affected by subjective bias,” which were scored as 4, 3, 2, and 1, respectively.

For the question, “Do you think expert opinion is credible keeping in mind subjective biases?” the possible answers were “Completely credible; examination has its own scientific nature,” “Minimally credible,” “Not credible,” and “Not at all,” which were scored as 4, 3, 2, and 1, respectively.

For the question “If the expert opinion is influenced by the subjective bias of the examiner, what role do you think this influence plays?” the possible answers were “Promotes the examiner to form the correct opinion,” “Influences the examiner very slightly,” “Cannot say,” and “Impedes the examiner from forming the correct opinion,” which were scored as 4, 3, 2, and 1, respectively.

For the question “Do you think understanding the context is necessary for handwriting identification?” the possible answers were “Very necessary,” “Every case is different; it is necessary sometimes,” “Not easy to say,” and “Unnecessary; knowing the context will cause prejudice,” which were scored as 4, 3, 2, and 1, respectively.

For the question, “In your professional career, how often do you present your expertise in the courtroom? The possible answers were “Often,” “A few times,” “Never,” and “I am not a professional examiner,” which were scored as 4, 3, 2, and 1, respectively.

After the credibility scores of each item were checked, the internal consistency reliability test was conducted. The results showed that the Cronbach's alpha coefficient of the internal consistency of the survey data was 0.776 [Table 1]. These data indicated that the questionnaire had high internal consistency and reliability.{Table 1}

The validity of this questionnaire

Validity refers to the degree to which a tool really measures what it is meant to assess. In this survey, the objective was to determine whether the questionnaire could address questions about subjective bias in handwriting identification.

Validity refers to the extent to which a scale can accurately measure the desired variables. It is used to consider whether the design of the questionnaire is scientific and whether it can effectively measure true variables. In general, it can be divided into external validity, internal validity, structural validity, and other evaluation indicators, of which factor analysis is a test of internal validity. When using factor analysis, test validity is reflected by two test indicators: (i) Kaiser–Meyer–Olkin (KMO) value, which must generally be >0.6 and (ii) Bartlett's test of sphericity which mainly demonstrates the correlation coefficient between test items; when the results have a significance <0.05, data validity can be effectively measured. When both of the above criteria are satisfied, it indicates that the data validity is good.

As demonstrated by the results of the validity test depicted in [Table 2], the KMO value of the survey data was 0.797. As this met the requirement of more than 0.6, the KMO value test was passed. At the same time, the results of Bartlett's test showed that the approximate Chi-square value was 264.005, and the significance level was 0.000. The results of Bartlett's test also met the requirement with a significance level below 0.05. Therefore, it can be seen that the validity of the questionnaire data structure was good, and the questionnaire design met the statistical criteria.{Table 2}

 Results and Analysis of Questionnaire Survey



Analysis of valid questionnaires

After retrieving the questionnaires, a basic analysis of the first part, mainly based on age, work position, and working years, revealed that both the first and second questionnaires were valid. Work experience of <6 years was considered “internship;” this group largely included the 20–30-year-old examiners or interns. Interviewees over 40 years old were defined as very experienced experts and so on. The first questionnaire received 149 valid responses, whereas the second received fifty valid responses.

The first questionnaire was sent to handwriting examiners, interns, and students with handwriting knowledge and assistant examiners. The second questionnaire was only sent to a group of handwriting examiners. After the questionnaires were issued and recycled, we received good feedback.

Basic details regarding interviewees

Regarding the first questionnaire, the 149 valid responses were from 49 handwriting examiners, seven assistant examiners, 51 interns or students, and 39 others with no experience; these data are presented in [Chart 1].[INLINE:1]

The second questionnaire received fifty valid responses from seven assistant examiners, 41 professional examiners, and two interns from forensic laboratories.

Analysis of the first part

Subjective questions about handwriting identification

[Chart 1] shows that about 60% of the examiners believed that expert opinions are definitely subjective and 35% believed that some are subjective. None of the examiners believed that opinions are completely unaffected by subjectivity. It can be concluded that almost all the examiners in this survey believed that handwriting examination involves subjective influence.

Moreover, among the positive attitude group, comprising 60% of the examiners, about 75% believed that this sort of subjectivity can enable the formulation of correct opinions, and that background information must be carefully considered before conducting handwriting identification. Only a few examiners felt that only questioned samples and references are required.

In short, it can be concluded that handwriting examiners prefer to think that subjective bias, which is a part of the process, enables the formulation of accurate opinions. Essentially, they believe that it is necessary to understand the context when conducting an examination [Chart 2].[INLINE:2]

Status of forensic experts' courtroom appearances

[Chart 3] depicts that the appearance rate of the examiners is not high; about 60% never go to court and 40% rarely appear in court.[INLINE:3]

Comprehensive analysis of cases

Analysis of questionnaire 1: Comparison of the examination of the same cases by examiners with different levels of experience

Consider the results of the identification of Case 1, with an overall accuracy rate of 75.17%.

There were 112 people who believed that the samples were from the same person, 25 who tended to identify, 11 made an exclusion, and one tended to be negative. Since the overall data contained a blank group of “students without experience,” the accuracy rates of each group were compared and analyzed. As the number of assistant examiners in this group was small–only seven people–there was no statistical significance, but the authors put the data together and found a very obvious regularity.

The accuracy rates for the conclusions of the case for the professional examiners, assistant examiners, and interns or professional students were 93.88%, 71.43%, and 60.78%, respectively. The results are depicted in [Chart 4].[INLINE:4]

From the above illustrations and data, it is possible to analyze the trend of the identification accuracy of professional forensic experts and interns and experienced and inexperienced examiners. The accuracy rate gradually declined in the following order: forensic experts, assistant examiners, students, and interns. Of course, these data were expected. The difference in accuracy rates from nearly 100% to roughly 60% also demonstrates that forensic science is scientific and that nobody can do it without special knowledge. This is also an example of how handwriting identification relies on the expert's work experience.

Analysis of questionnaire 1: Analysis of different handwritings by the same population

A comparison within the same professional group demonstrated that the accuracy rate of Case 1 was very high and that of Case 2 was relatively low. There might be two possibilities as follows:

The complexity of handwriting: The signature strokes of Case 1 were few and easy to analyze, especially for laypeople or assistant experts with less experience. Fewer ink lines lead to finding similarities easily (such as the Chinese name “丁 一.”), and differences are harder to find

In Case 2's signature, the strokes were more numerous and relatively more complicated. There are some similarities and some differences. The layman does not understand the sequential steps of the test. At this time, it is relatively difficult for the layman to take a few disordered samples. It is difficult to arrive at a conclusion without the arrangement of the background.

Case 1 was the first to appear in the questionnaire. When we first start doing something, we always have more enthusiasm. For example, mornings are usually very efficient. The same is true for a questionnaire. When an examiner receives the first case, he/she will pay more attention to it. Therefore, the accuracy rate of Case 1 was expected to be slightly higher than that of Case 2.

The facts of the context also have different influences on different groups of people. Regardless of the skill qualification of an examination, a positive result requires a higher degree of psychological conviction of the individual. If there is greater doubt and caution, it is not easy to draw a positive result. Professional examiners have the greatest changes in their degree of inclination, which means that, when they perform examinations, they must not only take into account the interference of the case, but must also combine professional experience with knowledge and be more cautious.

From the above data, it can be fully explained that the facts of the case have a certain influence on the examiner's examination and opinion, and the degree of influence is inversely proportional to the examiner's professionalism. Case 2 had a set of details, and the information was clear. The induction was very strong, which caused many interviewees, including professional experts, to be affected. From the statistical point of view, in handwriting identification, the facts of the case can affect handwriting examiners' decisions, and this influence does not necessarily lead to the correct answer.

Well, the influence of the casework that was mentioned above on the handwriting experts has nothing to do with the examiners' work experience. The authors tried to make a histogram for only the professional examiner. I believe that this figure is more persuasive.

The distribution of professional examiners according to work experience was as follows: 3–five years: seven people; 6–8 years: 35 people; and 9 years or more: seven people. According to this distribution, the number of people who made correct identifications was one person, 11 people, and five people, respectively. The difference in the total number of people at each stage makes it difficult to compare the ratios, so the ratio base is the total number of people in each stage.

The accuracy rate of professional examiners' conclusions regarding Case 1 was almost 100%, which would indicate high handwriting identification skill, but the accuracy rate regarding Case 2 was very low. From [Chart 5], it can be concluded that the facts of the case have a certain influence on forensic examiners, but this effect also shows a regularity. In other words, it is inversely proportional to the experience of the examiners; the longer the work experience of the examiners, the lower the impact. The less the expert's work experience, the greater the impact. This conclusion should also confirm the conclusion of the previous paragraph.[INLINE:5]

To sum up, the facts of the case affect the examiner's opinion to some extent, and the degree of influence is inversely proportional to the examiner's years of experience.

Analysis of questionnaire 2: The selection of handwriting characteristics in the identification process

When [Chart 6], [Chart 7], [Chart 8], [Chart 9] are considered based on the results, it can be clearly observed that forensic experts are irregular when faced with the characters of the same signature. With regard to the complexity and variety of handwriting characteristics, it is difficult to establish numerical standards. The characteristics of the ink lines of each Chinese character are not the same, and the differences in the writing habits of each individual bring great difficulty. This also somehow proves that handwriting identification is not suitable for establishing numerical standards.[INLINE:6][INLINE:7][INLINE:8][INLINE:9]

In short, handwriting examiners have no regularity in selecting features of handwriting. Therefore, the author believes that the feature comparison of handwriting identification is temporarily not suitable for establishing numerical standards.

Comparative analysis of questionnaire 1 and questionnaire 2 – Analysis of the handwriting in Case 1 by the same group

The signature in questionnaires 1 and 2 as well as in the control samples was exactly the same. The only difference was that questionnaire 1 had no context, whereas questionnaire 2 only had a simple noninductive context: the suspect himself had not recognized that he had signed this agreement. However, witnesses testified that the suspect had signed the agreement. The description of the context only demonstrated that there was a controversy, not definite proof. Therefore, the examiners' opinions were almost the same, and the accuracy rate was more than 90%. Therefore, a simple noninductive context does not affect the examiner's opinion.

Comparative analysis of questionnaire 1 and questionnaire 2 – Analysis of the handwriting in Case 2 by the same group

For Case 2, which was the same in questionnaires 1 and 2, the results of the identification were beyond expectation as the only difference was that of context. Questionnaire 1 had an inductive description as follows: “And there is other evidence and witnesses can prove that the pension is claimed by others.” Questionnaire 2 was a blank case without any context. Therefore, the fact is that the strong inducement has a relatively high influence on the handwriting identification from the view of examiners.

 Discussion



Issues of feature selection in comprehensive evaluation of handwriting identification

Examiners must carefully use comparative test methods, if necessary, and keep the basic knowledge of various disciplines, which will affect the final results or opinion, in mind. Therefore, it is very important to strictly control the qualification of handwriting examiners.

In my opinion, the analysis and evaluation of the value and nature of a handwriting feature is not a question of the repetition rate of a given Eigenvalue but rather consists of a series of incalculable factors that may depend mainly on the degree of fluency in writing and the complexity of the letter. When the degree of fluency declines, the quality of differences and similarities may decrease, and, when the complexity decreases, the degree of reproducibility becomes higher, so this is not a simple numerical process. Many disciplines and fields use numerical quantification standards to express their scientific nature, but I do not think that all disciplines need them. For handwriting identification, establishing a numerical standard is still difficult to achieve, and we do not need to suffer for it.

From the above analysis, it can be seen that, in the process of handwriting identification, it is unrealistic to make the comprehensive assessment of handwriting identification numerical. However, examiners who have less work experience are indeed prone to experiencing difficulties, and they are vulnerable to subjectivity (such as in terms of facts). In order to avoid such a problem, the author believes that a more effective examiner training system should be established, and training systems from other professions, such as those of lawyers and doctors, should be referenced.

The postgraduate physician standardized training is an important course for medical students. Being an effective way to cultivate the necessary habits in clinicians, it is a process clinicians must undergo. The main training mode is “5 + 3,” that is, after 5 years of medical undergraduate education, there is a 3-year standardized training for resident physicians (the practice involves a department rotation system, that is, the need for practical work in each department, experiencing the tasks of each department and, in the end, the completion of an anonymous evaluation by senior doctors in each department). After the assessment of clinical practice ability, residents must pass the national unified examination, after which a unified national certificate can be obtained.

There are many similarities between handwriting examiners and doctors, and a large amount of basic knowledge and rich work experience are indispensable to both professions. According to the “Administrative Measures for the Registration of Forensic Expertise” in China, it can be concluded that the following two criteria need to be met: (i) senior professional or related titles; more than 5 years of relevant work experience with associated industry-practicing qualifications, undergraduate or higher degree, or relevant college diplomas and (ii) To apply for an experience (or skill)-based forensic examination business, more than 10 years' experience in related professional work, and strong professional skills.

Although there are rules concerning work experience and related professional titles, the main problems stem from the differences: the lack of a national unified qualification test and of unified and authoritative professional bodies to perform evaluations. Is the requirement of “more than 10 years' experience” too much? What are the “stronger professional skills”? Can our handwriting examination industry also establish standardized training like physician training? For example, after 1 or 2 years of unified training after graduation, after some time, the assessment is performed. Finally, the same certificate is issued. Only after obtaining the certificate, can examiners be deemed adequately competent.

At present, owing to various factors, the control of the actual capabilities of various types of professional examiners, including handwriting identification, is still insufficient. What applicants have is “relevant” knowledge, work experience, and skills. Most of the “entry” examinations are based on the laws, rules, and regulations related to forensic examination.

To assess whether examiners have sufficient handwriting examination capabilities and whether they are fully qualified for handwriting expertise, the only feasible way is to introduce a strict and uniform training and assessment program.

Whether to understand the context when accepting casework

According to the data analysis of Case 1 and Case 2, it is concluded that the expert should avoid understanding the relevant case when conducting the examination as the background information of the case has a greater influence on handwriting examiners than do the samples.

Through comparison and analysis of the cases of questionnaire 1 and Questionnaire 2, the conclusion is that, while context is not necessarily inductive, it is important to focus on avoiding context with strong inducement.

However, through questionnaire 1, it was found that more than 73% of forensic experts believe that subjectivity can promote professional examiners to formulate correct expert opinions, and that more than 75% of professional examiners believe that it is necessary to understand the relevant information when performing an examination. The key point at this time is that the factual information that the handwriting examiners think must be understood is precisely what might be inductive.

Based on all the above points, this article makes a case for forensic examination to include relevant mandatory systems for forensic science sectors and forensic evaluators. In other words, handwriting examiners should avoid focusing on highly inductive information before going about their job. It is necessary to prevent examiners from directly contacting persons who have a direct interest in the case. Such a rule can avoid incalculable problems in the examination, especially the subjective bias of the examiner. This issue has not been strongly demonstrated theoretically and practically, and it is difficult to avoid. The qualification assessment system for forensic personnel is not perfect; all that can be done is to introduce this series of mandatory provisions to enhance the accuracy of identification and to improve the scientific basis and impartiality of handwriting identification from the aspects that can be seen.

 Conclusion



At this point, the data analysis of the questionnaire survey has been completed, and some conclusions regarding the subjective bias of handwriting examiners have been drawn. Every kind of science has its own characteristics. The complexity of knowledge structures and the difficulty of establishing numerical quantification standards in the identification process have formed its own characteristic empirical science: handwriting identification. Research on the neutrality and objectivity of handwriting identification is still far from sufficient. This is just a beginning. I hope this question will receive more attention and will be the basis of more in-depth studies by handwriting examiners and related researchers. According to the statistics and discussion of the two questionnaire surveys in this article, it is concluded that handwriting examiners are affected by information other than the questioned and control samples. The information that needs to be avoided is strong inductive context.

The simplest and most direct solution is to prevent handwriting examiners from being exposed to information that is too inductive. Which factors beyond the facts of a case can affect handwriting examiners' opinions should be the focus of continued efforts. However, in reality, the boundaries of inductive cases are still indistinguishable, and we need to continue to sum up experience through practice.

Financial support and sponsorship

This research was funded by Youth Scientist Program of CUPL Science Research Project (2016), Grant No. 16ZFQ82009.

Conflicts of interest

There are no conflicts of interest.

References

1Zhijin Z, Xiaochuan Z, Chang L. Review of intuition research. Prog Psychol Sci 2005;(6):43-9.
2Zhibin Z. Intuition and deviation control in handwriting identification. J Yunnan Police Coll 2014;(3):94-8.
3Anderson J, Yulin Q, Translated. Cognitive Psychology. 2012 Edition. Beijing: Beijing University of Posts and Telecommunications Press; 2012. p. 26.
4Xiaotian F. Questionnaire Survey in Social Survey. 2014 Edition. Beijing: China Renmin University Press; 2014. p. 11.