988368

Introduction 
Traffic accidents in Iran have a high prevalence and severity [1]. Traffic accidents due to deviating from the road due to human error due to sleepiness, resulting in high mortality [2]. Railway transport accidents can have serious consequences, including injuries and fatalities. According to the US Department of Transportation, human error is a significant factor in many train accidents [3]. In the US, railway fatalities totaled 893 in 2021, 20% higher than the revised total of 744 in 2020 and the highest since 2007. Nonfatal injuries totaled 5781, showing an increase of 4%. Of the 617 trespass-related fatalities, 94% were attributed to persons other than railroad employees. Meanwhile, 11 employees were killed while on duty, matching the 2020 count. There were 6 train passenger fatalities, up from two in 2020 [4]. Train accidents can be caused by a variety of factors, including equipment failure, mechanical failure and human error. Although railway transport is increasing around the world, despite the adoption of various safety measures, railroad accidents, and fatalities have been rising in many countries. Railway disasters have increased in developing countries, while the opposite is true in Europe. According to statistics, 74% of global railway disasters have occurred in Africa, Asia and South/Central America combined [5]. The information obtained from industrial accidents indicates the high contribution of human errors (70%-90%) in many catastrophic accidents, such as the Amtrak train accident (2017 and 2018, USA) [6, 7, 8]. The leading causes of such errors can also be seen as the application of incorrect mental processes, such as neglect, inattention, forgetfulness, and carelessness by people [9-11].
With the existence of increasing technological advancements, manpower is still considered the most critical element in working systems. [9] Therefore, human errors are possible in many occupations, including control room operators [12, 13]. These operators monitor and control various processes using advanced and modern hardware, such as closed-circuit televisions or visual displays [14, 15]. They must evaluate large amounts of data while monitoring and controlling such complex and dangerous processes to make effective and critical decisions to achieve system goals [16]. 
In the railway industry, the work processes of traffic control room operators are such that they exchange extensive information during their work shifts and have a high workload [13, 17]. Furthermore, the occurrence of catastrophic events like the London train collision (England, 1999) and the Neyshabour train accident (Iran, 2004) [18] and the investigations carried out in the field of rail accidents in different countries, including India and England, at different periods indicate the high role of human error in the occurrence of many of these accidents [17-20]. 
Accordingly, many researchers have investigated the probability of human errors among employees working in the railway industry by using various human error techniques [21-24]. Having a lot of flexibility and applicability, learning and using quickly and easily, as well as having a comprehensive and structured approach are among the advantages of some of these techniques. However, these techniques mainly focus on job tasks rather than individual people’s errors [25].
Therefore, due to the lack of a specific tool to estimate the probability of human errors of individuals in railway traffic control rooms (RTCRs) and also the necessity of having a tool in this field according to those mentioned above, the present study prepares a standard, valid, and reliable tool to measure the probability of monthly human errors among the RTCR employees to take a step toward preventing human error and, subsequently, the occurrence of accidents.

Materials and Methods
In this mixed-method research, in the first step after forming an experts panel, an initial questionnaire containing 67 questions in two parts (the first part for the employees of the central control room and the second part for the employees of the RTCR in different regions of the country) using some data collection methods and experts’ opinions were designed. To check and determine the face and content validity of this questionnaire, the members of the expert panel, and in line with its reliability estimation, 35 employers working in the RTCR participated. The participation of panel members in this study took place in three rounds between November 2021 and February 2022. After the investigations, the final questionnaire was designed with 67 questions without removing any of the questions. The implementation process of this research is presented in Figure 1.
Forming an expert panel 
To ensure the validity and reliability of the questionnaire, the questionnaire should be tested by several experts. The members of the panel of experts should also be selected from among the experts who are active in the field of the questionnaire content so that correct and accurate judgments can be made possible [26]. Based on this, 15 experts were selected according to the predetermined objectives of the study and asked to comment on each of the questions. These people have been selected considering their experience, expertise, and knowledge in this field. The Mean±SD age and work experience of these experts were 39.00±4.175 and 10.53±2.800, respectively (Table 1).



Collecting information
The questions of this questionnaire were designed using methods and some data collection tools, such as literature reviews and existing data, observing the work duties of the employees, as well as interviewing each of the employees and consulting with experts and the primary researchers of this research during several sessions. Also, in this regard, this study used all the opinions of the members of the expert panel and available scientific resources.

Instruments validity
Determining face validity
In the present study, the face validity of the questionnaire questions was measured in qualitative and quantitative ways.

Qualitative face validity
To perform qualitative face validity, the initially designed questionnaire was given to seven expert panel members consisting of three university assistant professors in the field of occupational health and safety from Shahid Beheshti University Medical of Sciences, two experienced employees working in the central control room, and two employees from the RTCR of the Tehran District (Iran) with more than 15 years of experience. Then, they were asked to comment on the questionnaire questions and their appearance. After receiving the opinions of each of these people, the necessary changes were made to the questions, and their qualitative face validity was done.

Quantitative face validity
In this step, the quantitative method of impact score was used to reduce and eliminate inappropriate terms and determine the importance of each of these terms. In this method, to determine the quantitative face validity of the questionnaire, the questions were provided to the members of the expert panel and they were asked to evaluate the appearance of each of the questionnaire questions in terms of simplicity, relevancy, and clarity, and comment on each of them according to the purpose of the research [27]. For this purpose, the questions presented in the first, second, and both parts of the questionnaire were respectively provided to the experienced employees of the central control room (n=5), the experienced employees of the Tehran District the RTCR (n=5), and university assistant professors (n=5) to comment on the features of the questions. 

Determining content validity
To check the content validity, two indexes of content validity ratio (CVR) and content validity index (CVI) were used according to the following steps (with the participation of the same participating expert panel members in determining face validity).

Step 1: Determining the CVR index
Ten experts (five university faculty members and five employees of the central control room) were asked about the first part of the questionnaire, and ten other experts (five university faculty members and five employees of the control room of the Tehran District, Iran) were asked about the second part of the questionnaire to comment on the importance and necessity of all the questions in each part. Three ranges, namely necessary, useful but unnecessary, and unnecessary were used to measure experts’ opinions. After collecting comments, CVR values for each question were calculated according to Equation 1 and compared with Lawshe’s table [28]. Since the opinions of ten experts were used in each section of the questionnaire, the questions with numbers >0.62 were accepted. In Equation 1, “ne” is equal to the number of experts who chose the “necessary” option, and “N” is equal to the total number of experts.



Step 2: Determining the content validity index
The CVI value of the questions was determined using the method of Waltz and Basel (1981) and according to Equation 2 [29]. In this regard, the questionnaire was sent to each member of the panel of experts, and they were asked to comment on each of the three criteria of simplicity, relevancy and clarity based on a 4-point Likert scale (1=unrelated, 2=somewhat related, 3=related, and 4=completely related). If the CVI score of the questions is >0.79, the content validity of the questions will be accepted [26 ]. In Equation 2, “n” is equal to the total number of experts who selected the options “completely relevant” and “relevant,” and “N” is equal to the total number of experts.

2. CVI=n⁄N

Instrument reliability
Determining the reliability of the questionnaire
To measure the reliability of the questionnaire, this tool was distributed in two stages with a time interval of 12 days in February 2022 among 35 employees of the central control room and Tehran District (Iran) control room. The selection was made considering the level of experience and expertise of these individuals. The Mean±SD age and work experience of these employees were 31.03±4.462 and 8.86±2.522, respectively (Table 1). Subsequently, using the weighted kappa coefficient calculated in the Stata software, version 13, the reliability of this questionnaire was calculated.

Results
In this research, after face validity, none of the questions were removed, and only five questions were edited. In addition, considering that all questions had CVR >0.62 and CVI >0.79, the content validity of the questions was also confirmed. The CVI values of the questionnaire were 0.9, 0.9 and 0.92 in terms of simplicity, relevancy, and clarity, respectively. Also, its CVR value was 0.87 (Table 2).


The reliability of the questionnaire was also proved by obtaining the values of 73.71%, 87.14% and 80.31% for the minimum, maximum and average percentages of agreement between 67 questions (Table 2).
Ultimately, a final questionnaire consisting of two parts was designed (Appendix 1), in which in the first part, 43 questions specific to 12 job positions working in the central RTCRs and in the second part, 24 questions specific to six job positions working in other RTCRs were presented (Table 3).

Appendix 1.
Part 1: Monthly probability of human error questionnaire for RTCR employees
Please select the answer that best describes the calculated amount of human error among RTCR employees for the previous month. Please select the most appropriate answer (questions 9-1 and other questions related to your job position). If you are uncertain about the answer, please select the closest one. Your answers will remain confidential and do not affect your activities or process.
Job position: Education level:   Job experience:   Marital status:   Age:   
Questions 1-9: General 
Questions 10-12: Head of department
Questions 13-14: Shift supervisor
Questions 15-17: Communication expert
Questions 18-23: Automatic train control expert
Questions 24-25: Line expert
Questions 26-29: Track and movement expert
Questions 30-31: Sign expert 
Questions 32-33: Local branch expert
Questions 34-36: Passenger expert
Questions 37-39: Traction force expert
Question 40: Cargo expert
Questions 41-43: Deputy department
Question 44: Other




                

Part 2. Human error probability questionnaire for RTCR employees
The following questions concern the probability of human error among control center staff during the past month. Please indicate the correct answer that best indicates the estimated probability of human error determined by you. If you are not sure about the answer, please select the closest one. Your responses will remain confidential and will not affect your processes or activities. Try to answer all questions related to your job position.
 Job position:   Education level:   Job experience:   Marital status:   Age:   
Questions 1-3: Director of rolling stock control
Questions 4-9: Assistant chief for control centre
Questions 10-12: Director of control for movement and operation
Questions 13-15: Rail stock control in the wagon domain
Questions 16-19: Control on the railway way
Questions 20-24: Director of the dispatcher







Among the questions distributed in the first part, nine are public and completed by all central RTCR employees.

Discussion
Considering that railway lines carry out a large volume of public transportation [30], the occurrence of human error by the RTCR operators can lead to substantial financial and life losses [13]. On the other hand, considering various factors, such as the absence of systems to avoid human errors [31], the ineffectiveness of the laws and regulations [32] and the payment of irreparable financial and life losses following the incidence of human errors [13], it is obligatory to design and apply different techniques to increase safety level at workplaces [33]. Therefore, according to the mentioned points, in this study, a reliable tool was presented to measure the probability of monthly human errors among the RTCR employees. This questionnaire is designed in two parts. The first (with 43 questions) and second (with 24 questions) parts of this questionnaire can measure, the probability of human error among employees in the railway traffic central control rooms. In this study, CVI and CVR as the most reliable measurement methods [26] were used to check the validity of the questionnaire. The CVI values of this questionnaire in terms of simplicity, relevance, and clarity were equal to 0.9, 0.9 and 0.92, respectively, and its CVR value was estimated to be equal to 0.87. The reliability of this tool was also investigated with the participation of 35 employees working in the central control rooms and railway traffic in the Tehran District. They answered the questions in two stages, with a time interval of fewer than two weeks. According to the results obtained from this answering and applying and calculating the weighted kappa coefficient, the minimum, maximum, and average percentage of agreement between the questions were 73.71%, 80.31% and 87.14%, respectively; reliability. The study by Mahdinia et al. investigated the effects of work pressure, mental workload, human-system interaction, and environmental distraction on three types of human errors (slip, mistake, and error) in steel industry workers. In this study, a questionnaire was used as a valid and reliable tool to measure the human errors of the participants. Notably, the internal consistency coefficients (Cronbach α) of the three parts of the human error questionnaire used in this study were 0.78, 0.88 and 0.84, respectively and the answers to the questions were measured based on a 0-5 Likert scale [34]. The answers to the designed questionnaire, by Lee et al., are measured in a range of 0-5. In another study, the relationship between sleep, sleep environments at work, and the human errors of train drivers, the human errors of these drivers were measured based on their judgment and by mentioning several questions and the relationship with other desired components (sleep and sleeping places at work) [35]. In another research conducted by Rowland et al., the questionnaire tool was used to investigate emergency care workers’ views about the types of human errors and the factors influencing human errors that affect the safety of patients in the pre-hospital emergency care environment [36]. The present study tried designing an approved and appropriate questionnaire by forming an expert panel consisting of university faculty members and experts in RTCRs as well as collecting the available information. According to the stated content, along with the use of standard methods of identifying and evaluating human errors, the use of other measurement tools, such as questionnaires or software can help to reduce or prevent the occurrence of human errors in many work environments, especially places with high job importance such as control rooms.

Conclusion
According to the results of this research, the present questionnaire has good validity and reliability. It can be used to measure the probability of monthly human error among employees of RTCSs. By using this tool, in addition to examining the human errors of the employees themselves, it is possible to examine the impact of other factors on the human errors of these employees, thereby reducing the probability of human error. Therefore, other researchers are advised to design tools and software in addition to using this tool to reduce the probability of human errors in susceptible and risky jobs such as control rooms.

Study limitations
Although the validity and reliability of the results are statistically acceptable, if it were possible to increase the number of participants in the study, it is predicted that the results would be obtained more accurately.

Ethical Considerations
Compliance with ethical guidelines
This study was approved by the Ethics Committee of Shahid Beheshti University of Medical Sciences (Code: IR.SBMU.PHNS.REC. 1400.039). Each participant entered the study after completing the written consent form to conduct this research. An informed consent form was obtained from all participants. All participants signed the informed consent form and were explained the procedure of research before the study onset.

Funding
This research is extracted from master’s thesis of Soleiman Ramezanifar, approved by the Department of Occupational Health and Safety at Work, School of Public Health and Safety, Shahid Beheshti University of Medical Sciences and was financially supported by Shahid Beheshti University of Medical Sciences, Tehran, Iran (Project No.: 1400/29255).

Authors' contributions
All authors contributed equally to preparing this article.

Conflict of interest
All authors declared no conflict of interest.

Acknowledgments
The authors thank the cooperation of the honorable Director General of the Railway Movement of the Islamic Republic of Iran, the members of the expert panel, and the heads and employees of the RTCRs for their sincere appreciation.

References

1. Zare H, Abdollahi M, Poursadeghiyan M, Kasiri N. Epidemiological study of fatal road accidents in eastern Iran in a five-year period. Health in Emergencies and Disasters Quarterly. 2022; 8 (1): 47-54. [DOI:10.32598/hdq.8.1.452.1]
2. Askari A, Hosseinpour R, Bakhtiari M, Sepehr P, Ghodrati Torbati A, Salehi Sahlabadi A, et al. Investigating the relationship between subjective and interpretive drowsiness with lane departure in simulator driving. Health in Emergencies and Disasters Quarterly. 2024; 9(4):255-64. [DOI: 10.32598/hdq.9.4.374.2]
3. Szabo JC. Enhancing our rail safety: Current challenges for passenger and freight rail [Internet]. 2014 [Updated 6 March 2023]. Available from: [Link]
4. United States National Safety Council. Railroad deaths and injuries [internet]. 2023 [Updated 28 July 2024]. Available from: [Link]
5. Bang HN, Miles LS, Gordon RD. Challenges in managing technological disasters in Cameroon: Case study of Cameroon’s worst train Crash-the Eséka train disaster. International Journal of Disaster Risk Reduction. 2020; 44:101410. [Link]
6. Ramezanifar S, Ramezanifar E, Sahlabadi AS. Investigating the probability of human error in Iranian Industrial Control Rooms: A systematic review. International Journal of Occupational Hygiene. 2022; 14(1):77-91. [Link]
7. Jahangiri M, Hoboubi N, Rostamabadi A, Keshavarzi S, Hosseini AA. Human error analysis in a permit to work system: A case study in a chemical plant. Safety and Health at Work. 2016; 7(1):6-11. [DOI:10.1016/j.shaw.2015.06.002] [PMID]
8. O’Toole R. The midwest rail plan: A disaster waiting to happen [Internet]. 2021. [Updated 28 July 2024]. Available from: [Link]
9. Zare A, Rad SY, Dehghani F, Omidi F, Mohammadfam I. Assessment and analysis of studies related human error in Iran: A systematic review. Journal of Health & Safety at Work. 2017; 7(3):8. [Link]
10. Mohammadfam I, Amid M, Hajiakbari M. [The study of human errors in de-mining operations using the CREAM. Technique. Journal of Military Medicin. 2016; 17(4):241-7. [Link]
11. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence of adverse events and negligence in hospitalized patients. New England Journal of Medicine. 1991; 324(6):370-6. [DOI:10.1056/NEJM199102073240604] [PMID]
12. Sembiring N, Tambunan M, Febriani. Human error analysis on production process of door products with SHERPA and HEART method. Paper presented at: 1st International Conference on Industrial and Manufacturing Engineering. 16 October 2018; Medan City North Sumatera, Indonesia. [Link]
13. Ramezanifar S, Ramezanifar E, Khadiv E, Sahlabadi AS, Eskandari D, Namdari M. Applying the HEART Technique to identify and assess the human error in the Central Railway Traffic Control Room. Journal of Health and Safety at Work. 2022; 12(3):595-616. [Link]
14. Li X, McKee DJ, Horberry T, Powell MS. The control room operator: The forgotten element in mineral process control. Minerals Engineering. 2011; 24(8):894-902. [DOI:10.1016/j.mineng.2011.04.001]
15. Han SH, Yang H, Im DG. Designing a human-computer interface for a process control room: A case study of a steel manufacturing company. International Journal of Industrial Ergonomics. 2007; 37(5):383-93. [DOI:10.1016/j.ergon.2006.12.006]
16. Yang CW, Hsieh TL, Lin SF, Lin CJ, Teng HM, Chiu YF. Operators’ signal-detection performance in video display unit monitoring tasks of the main control room. Safety Science. 2011; 49(10):1309-13. [DOI:10.1016/j.ssci.2011.04.010]
17. Grozdanovic M, Janackovic GL, Stojiljkovic E. The selection of the key ergonomic indicators influencing work efficiency in railway control rooms. Transactions of the Institute of Measurement and Control. 2016; 38(10):1174-85. [DOI:10.1177/0142331215579948]
18. Jahangiri K, Ghodsi H, Khodadadizadeh A, Yousef Nezhad S. Pattern and nature of Neyshabur train explosion blast injuries. World Journal of Emergency Surgery. 2018; 13:3. [PMID]
19. Kyriakidis M, Pak KT, Majumdar A. Railway accidents caused by human error: Historic analysis of UK railways, 1945 to 2012. Transportation Research Record. 2015; 2476(1):126-36. [DOI:10.3141/2476]
20. Aher SB, Tiwari DR. Railway disasters in India: Causes, effects and management. International Journal of Reviews and Research in Social Sciences. 2018; 6(2):125-32. [Link]
21. Sun Y, Zhang Q, Yuan Z, Gao Y, Ding Sh. Quantitative analysis of human error probability in high-speed railway dispatching tasks. IEEE Access. 2020; 8:56253-66. [DOI:10.1109/ACCESS.2020.2981763]
22. Karami F, Ghiyasi S, Soltanzadeh A. [Identification and evaluation of human errors in locomotive maneuvers using the SHERPA Technique: A case study in a rail operation repair and development project (Persian)]. Iranian Journal of Ergonomics. 2021; 8(4):1-14. [DOI:10.30699/jergon.8.4.1]

23. Jin J, Li K, Yuan L, Wei G. A human error quantitative analysis of the railway ATO system based upon improved CREAM. Paper presented at: 2019 IEEE Intelligent Transportation Systems Conference (ITSC). 27-30 October 2019; Auckland, New Zealand. [DOI:10.1109/ITSC.2019.8917181]
24. Catelani M, Ciani L, Guidi, G, Patrizi G. An enhanced SHERPA (E-SHERPA) method for human reliability analysis in railway engineering. Reliability Engineering & System Safety. 2021; 215:107866. [DOI:10.1016/j.ress.2021.107866]
25. Stanton NA. Human factors methods: A practical guide for engineering and design. London: CRC Press; 2017.[DOI:10.4324/9781351156325]
26. FallahMadvari R, Farhang Dehghan S, Laal F, Abassi Balochkhaneh F, Fallah Madvari A. [Designing and Validating a Questionnaire for the assessment of staff satisfaction with occupational health, safety, and environmental management system status (Persian)]. Journal of Occupational Hygiene Engineering. 2020; 7(1):31-40. [DOI:10.52547/johe.7.1.31]
27. Monazam M, Laal F, Sarsangi V, Fallahmadvari R, Najafi K, Fallahmadvari A. [Designing and determination of validity and reliability of the questionnaire increasing the duration of using the hearing protection device by workers based on BASNEF model (Persian)]. Journal of Ilam University of Medical Sciences. 2018; 25(6):21-8. [DOI:10.29252/sjimu.25.6.21]
28. Lawshe CH. A quantitative approach to content validity. Personnel Psychology. 1975; 28(4):563-75. [DOI:10.1111/j.1744-6570.1975.tb01393.x]
29. Waltz, CF, Bausell BR, Nursing research: Design statistics and computer analysis. Philadelphia: F.A. Davis Company: 1981. [Link]
30. Atrkar Roshan S, Daneshvar S. [Estimation of economic losses caused by fire in the Tehran Subway Line 1 stations (RS and LPS) using scenario planning technique (Persian)]. Journal of Environmental Science and Technology. 2017; 19(4):335-46. [Link]
31. Khandan M, Vosoughi S, Azrah K, Poursadeghiyan M, Khammar A. Decision making models and human factors: TOPSIS and ergonomic behaviors (TOPSIS-EB). Management Science Letters. 2017; 7(2):111-8. [Link]
32. Salehi Sahlabadi A, Azrah K, Mehrifar Y, Poursadeghiyan M. Evaluation of health risk experienced by urban taxi drivers to whole-body vibration and multiple shocks. Noise & Vibration Worldwide. 2024:09574565241252992. [DOI: 10.1177/09574565241252]
33. Ebrahimi MH, Abbasi M, Khandan M, Poursadeghiyan M, Hami M, Biglari H. Effects of administrative interventions on improvement of safety and health in workplace: A case study in an oil company in Iran (2011-2015). Journal of Engineering and Applied Sciences. 2016; 11(3):346-51. [Link]
34. Mahdinia M, Mohammadfam I, Mirzaei Aliabadi M, Aghaei H, Soltanian AR, Soltanzadeh A. The mediating effect of workers’ situation awareness on the relationship between work-related factors and human error: a path analysis approach. International Journal of Occupational Safety and Ergonomics. 2022; 28(3):1958-66. [PMID]
35. Lee DW, Kim SJ, Shin NY, Lee WJ, Lee D, Jang JH, et al. Sleeping, sleeping environments, and human errors in South Korean male train drivers. Journal of Occupational Health. 2019; 61(5):358-67. [DOI:10.1002/1348-9585.12059] [PMID]
36. Rowland M, Adefuye A. Human errors and factors that influence patient safety in the pre-hospital emergency care setting: Perspectives of South African emergency care practitioners. Health SA Gesondheid. 2022; 27:1798. [DOI:10.4102/hsag.v27i0.1798] [PMID]