Management and Educational Perspective

Abstract The current research was conducted with the aim of providing a model of return on investment in the training of employees in the petrochemical industry. This research was based on an empiricist philosophy, fundamental from the goal point of view, and with a deductive-inductive approach with a mixed research method. The research data in the qualitative part was collected through interviews with 17 management professors (theoretical experts) and managers of the petrochemical industry (experimental experts) who were selected by purposive sampling. Also, the indicators identified by fuzzy Delphi method with the participation of qualitative department experts were validated and prioritized by fuzzy Swara method. Finally, the final research model was presented. To achieve the objectives of the research, by analyzing the interviews with experts, a set of useful indicators of return on investment in employee training was identified. The main categories of this research are indicators of return on investment in employee training, grouped in the form of eight main categories: organizational infrastructure, strategic alignment, training and development of human resources, incompatibility of human resources, employee empowerment, process-oriented strategy, return on investment to Company, and gain competitive advantage. The findings have shown that training and development of human resources is an underlying variable that affects the empowerment of employees and process-oriented strategy. In the meantime, it is important to strengthen the organizational infrastructure and create strategic alignment to achieve return on investment in employee training. By implementing the mentioned items, the return of capital to the company and gaining a competitive advantage will be achieved. The ranking of indicators of return on investment in employee training with the Swara Fazi method showed that the support factors of senior management, informing employees of the results of training goals, continuous training with wide applications are of the highest importance.
Extended abstract
Introduction
In order to find out the adequacy of the training and improvement programs of the organization's employees, their effectiveness should be evaluated. One of the important factors in this field is to pay attention to the return of training investment, which is considered as an important model in evaluating the effectiveness of employee training. In recent years, paying attention to the rate of return on investment in organizations has become a fundamental issue for managers (Imani, Rezaei, & Sharifi, 2020).
Despite the importance of return on investment in education, this issue is still unknown in the organizational field. This is due to issues such as the lack of evaluation tools, the inability of evaluators to involve investors, and the inability to implement inclusive approaches to education. On the other hand, few academic studies have been done in this field, and a deep research gap is observed in this field. In this regard, the National Petrochemical Company of Iran, which is the most efficient developmental organization in the field of petrochemicals in West Asia, faces the same problem of measuring and evaluating the practical results of training courses, which are often very expensive.
Therefore, the question that this research is trying to find an answer for is, what elements and components does the investment return process in the training of petrochemical industry employees include, what conditions are it affected by, and what are the quantitative and qualitative criteria for evaluating the results of the training?
 
Theoretical framework
The return on investment of training and measuring the effectiveness of training was first introduced as a model in the United States Journal of Development and Education in 1959 by Donald Kirkpatrick. This important model includes four levels or aspects of training; the first level is comprehensive and measures the feelings of the participants in relation to the attractiveness, usefulness and relevance of training to their needs; the second level is learning and the effect of training in terms of progress in Skills, knowledge and attitudes, and in other words the realization of learning goals are measured; At the third level, the actual effect of training, i.e. behavioral changes after learning, is measured; and finally, at the fourth level, the final results of training are analyzed. Due to the increasing importance of evaluating the effectiveness of training, Jack Phillips developed the model and added a fifth level to it and called it ROI (Urbancova, Vrabcova, Hudakova & Petr, 2021), which stands for return on investment. At this level the monetary value of the employee training program is evaluated and the information (quantitative and qualitative) of the fourth level is covered in monetary form.
Phillips recommended that ROI studies be conducted only for programs based on a comprehensive needs assessment. Effectively, measuring return on investment training is a fundamental need for organizations; otherwise, organizations will never understand how valuable their efforts are. Do they use the budget and time efficiently? Is training profitable? or how (and whether) organizations can improve their performance? (Nemec, P. B, 2018.)
Return on investment can be in the form of a number of criteria, including:
_ Added value resulting from activities due to skill acquisition and improvement; greater flexibility among employees who can perform a wide range of tasks; reducing general costs for the company, such as more efficient use of existing facilities, reducing consumption costs and reducing human resource costs; greater ability to innovate in terms of adopting new technology and introducing better forms of work organization;
This means that companies should be aware of a wide range of factors for measuring the return of employee training investments and create a tool to measure them (Nemec, P. B, 2018.)
Research Methodology
The current research is of a mixed type (qualitative and quantitative) and includes two main parts of model design and model validation. In the first part, in order to design the model of return on investment in the training of petrochemical industry employees, the qualitative method of grounded theory was used according to Strauss and Corbin's point of view, using MaxQDA 20 software; and in the second part; the validation of the model, the indicators' expertise validation using the fuzzy Delphi method was discussed by coding in MATLAB environment. Finally, in order to determine the importance of return on investment indicators in employee training, Swara Fazi method and MATLAB software have been used.
Theoretical experts include university professors in the field of human resource management who have scientific publications in the form of books and articles in this field. Experienced experts also include petrochemical industry managers who have at least ten years of executive experience in this industry. Theoretical and purposeful snowball sampling method was used to select the interviewees and the number of interviewees reached 17 until the saturation level was reached.
The main tool of research data collection in the qualitative part was the semi-structured interview with six open questions. In the second part, based on the identified categories, a seven-level Delphi questionnaire was used. Finally, indicators were prioritized using Swara's questionnaire. To check the validity and reliability of the data, Holstein's coefficient and P-Scott's index were used.
Research Findings
To analyze the data collected in semi-structured interviews, the three-step coding method of Strauss and Corbin; under the title of open coding, axial coding and selective coding, was used. In the open coding stage, 356 codes were identified, with 6 paradigms, 8 main categories, and 47 indicators; which was calculated based on the researcher's perception and inference from the conducted interviews using the grounded theory method. The fuzzy Delphi approach has been used to validate the identified indicators and screen the final indicators. The seven degree range has been used for the expert's viewpoint Fuzzied. Finally, 42 indicators were selected during two rounds of the experts' point of view, and the criteria of return on investment in employee training were grouped into eight main categories, which are: organizational infrastructure, strategic alignment, training and development of human resources, incompatibility of human resources, employee empowerment, process-oriented strategy, return of capital to the company, and gaining competitive advantage.
Conclusion and Discussion
The present research was conducted with the aim of providing a model of return on investment in the training of employees in the petrochemical industry. Based on the obtained results, it was determined that the organizational infrastructure component was identified as the causal condition, and the strategic alignment component was identified as the background condition affecting the return on investment in employee training. This is mentioned in the results of the study of Phillips (2012), and from this point of view, it is consistent with the results of the present study. Also, the component of training and development of human resources as a central phenomenon and the component of incompatibility of human resources as intervening and inhibiting conditions were identified. This is in line with the results of the study of Poudel (2022). The results showed that the components of employee empowerment and process-oriented strategy are effective as necessary strategies and measures for the return of investment in employee training. This is also mentioned in the results of the study of Annabil (2017), and this point of view is consistent with the results of the current research. Finally, it was determined that the components of return on investment to the company and gaining competitive advantage are positive consequences of return on investment in employee training, which are consistent with the results of the study of Shabahang, Hendricks, Hayek, Ryer, & Lauler (2021).
Also, the ranking of indicators of return on investment in employee training with the Swara Fuzzi method showed that the factors of support of senior management, informing employees of the results of training goals, continuous training with wide applications are of the highest importance.
Based on this, some practical suggestions are presented to the managers of the petrochemical industry for the return of investment in employee training: if the company is professional and the processes are mature and the management capabilities are improved; which leads to the elimination of unnecessary rules and regulations, the managers can provide the necessary training to the employees. Also, the managers of the petrochemical company, having faith in the topics of employee training and providing the necessary equipment for employee training along with improving the software and hardware infrastructure and finally creating an environment with training facilities in the company, can achieve the design of a distinctive human resources management system. In general, continuous training with wide applications leads to reducing conflict among employees and increasing self-confidence in them. Also, with the improvement of communication between managers and employees and the support of senior management, we can hope for an increase in the learning of human resources. What the relevant managers of the company need in order to return the investment is to review the mechanism of recruitment, training and improvement of employees, increasing the skills of employees, and employing expert and elite employees.
 

Abstract Abstract This research was done with the aim of providing a systematic model of employee training using artificial intelligence. The research method is a combination of qualitative and quantitative methods. The statistical community in the qualitative section includes an unlimited number of experts familiar with the subject, and the statistical sample in this section is 20 people selected by the snowball method. The statistical population in the quantitative section includes all the specialists and experts related to the research subject in an unlimited number, of which 384 people were selected as a sample by a simple random method. Researcher-made questionnaires with confirmed validity and reliability were used to collect the data of this research. Data analysis in this research was done at two levels of descriptive statistics and inferential statistics. To complete the Delphi process in this research, Kendall's coefficient was used with the help of SPSS software. In order to test the research model, structural equation technique was used through Smart-PLS statistical software, and the results show the appropriate fit of the model. The findings of the research show that the inputs of the model include 1- educational data, 2- personal information, 3- educational needs, 4- user feedback, and 5- data of the work environment. The model process also includes 1- determining the needs and goals, 2- collecting data, 3- pre-processing the data, 4- training the artificial intelligence model, 5- evaluating and improving the model, 6- implementing and deploying, and 7- monitoring and updating. Finally, the outputs of the model include 1- individual feedback, 2- educational suggestions, 3- monitoring and follow-up, and 4- support and guidance. Extended abstract Introduction In 1950, Alan Turing, an English mathematician, wrote an article entitled "Computing Machines and Intelligence" and posed the question: "Can machines think?" This led to further exploration of the use of machines to support human decision-making, with the formation of a workshop in 1956. The workshop was organized by John McCarthy, an American mathematician, who focused on the "study of artificial intelligence" (Frehywot, 2023). As science and technology advance, since 2013, when Frye and Osborne estimated that nearly half of US jobs are at risk of high automation, AI has been at the top of policymakers' agendas, and now the consensus is that AI creates fundamental changes in the labor market. With the use of artificial intelligence, many skills that were important in the past become automated; many jobs are also obsolete or transformed; and artificial intelligence is increasingly used (Tuomi, 2018). Artificial intelligence includes various related technologies, often supported by machine learning algorithms, whereby it achieves set goals through supervision (human-guided) or unsupervised (autonomous machine) (Walsh et al., 2019). Today, most experts believe that the implementation of smart technology will dynamically transform work environments. There are applications of artificial intelligence in all industries and professions, and human resource management is no exception to this rule. Therefore, for the organization to remain relevant and maintain its competitive advantage, it is necessary to adopt new technological developments (Kaushal et al., 2023). Today, the surprising speed of innovation in business processes and technology requires that the employees of organizations continuously acquire new skills and be able to adapt to changing practices. Thus, educational needs become more personalized (Ford et al., 2017). Employee skill development was once done entirely by on-the-job supervisors; but now with the increasing demand for new skills, technological advances have enabled training and development on mobile platforms, such as smartphones and laptops (Maity, 2019). Artificial intelligence should be applied to organizational learning to help companies solve their training challenges. When recruiting and hiring new employees, the main challenge is to quickly and effectively make them fully aware of and understand the organization's internal policies and procedures. Machine learning features have been incorporated into various HR software systems to facilitate greater efficiency (Iqbal, 2018). Even more comprehensively, the use of various artificial intelligence technologies can help companies develop a learning organizational culture and avoid the common training design model based on traditional competency model analysis (Chen, 2023). According to the above, considering the development of digital technology, especially artificial intelligence, and the increasing demand for personalized training, the past training methods are no longer able to meet personal needs, and the adoption of artificial intelligence-based training can effectively fill the shortage of personalized training. Therefore, the main question that the researcher seeks to answer in this research is: "What is the system model of employee training using artificial intelligence?" Theoretical foundations Artificial intelligence is often defined as a computer system with the ability to perform tasks normally associated with intelligent beings. The first explicit definition of artificial intelligence came in a funding proposal to the Rockefeller Foundation in 1955. This definition states that "any aspect of learning or any other characteristic of intelligence can originally be described so sufficiently precise that a machine can be built to simulate it." This initial definition quickly led to deep discussions. In practice, the early developers of artificial intelligence interpreted intelligence and thinking as the mechanical processing of logical statements (Tuomi, 2018). In another definition, artificial intelligence is defined as "making a machine behave in a way that would be called intelligent if a human behaved." Although artificial intelligence was defined in 1955, it has recently gained worldwide recognition due to the technological revolution. Artificial intelligence is discussed as non-human intelligence designed to perform specific activities and tasks (Kaushal et al., 2023). McCarthy describes artificial intelligence as the science and engineering of building intelligent machines through algorithms or sets of rules, that the machine follows to mimic human cognitive functions, such as learning and problem-solving (Frehywot, 2023). This definition provides a combination of AI capabilities and what it is. Human resource management has undergone an early revolution with the help of artificial intelligence, which has gradually affected human resource operations. These functions, which were already performed entirely by humans, are being recreated with the help of a computer assistant. HR functions such as performance appraisal, learning and development, and talent acquisition are some of the areas where artificial intelligence has been introduced (Kaushal et al., 2023). By means of a variety of artificial intelligence technologies, it can be more comprehensive to help companies to form a learning organization culture that uses the traditional instructional design model based on traditional gap analysis. A customized curriculum can comprehensively test and locate staff levels through technical tools, and intelligently promote customized courses (Jia et al., 2018). In the process of education, AI technology can help learners automatically record learning data. Employees can simply enter learning objectives, archives and learn key points, and the course is automatically completed by the AI teacher (Jia et al., 2018). Methodology The current research is applicable in terms of its purpose. In terms of the method of data collection, it is descriptive-correlation in nature. The statistical population of this research in the qualitative part is made up of experts related to management issues (human resource management and system design) with a master's degree or higher in Sistan and Baluchistan province. In the quantitative part, all the employees of the departments of Sistan and Baluchistan province who have a bachelor's degree or higher, formed the statistical population of this research. The statistical population is considered unlimited in both qualitative and quantitative sections. Therefore, the sample size in the qualitative section is considered to be 20 people, selected by snowball method. In the quantitative part, Cochran's formula was used for unlimited communities with an error level of five percent, and the sample size of 384 people was determined and selected by simple random method. The tool of data collection in this research was researcher-made questionnaires. SPSS and Smart-PLS statistical software were used to analyze the data of this research. Findings The value of Cronbach's alpha for all variables of the research questionnaire is more than 0.7. Therefore, the research questionnaire has acceptable reliability. Also, the composite reliability value of all variables is higher than 0.7, which indicates the internal consistency of the reflective measurement model. All values related to convergent validity are also above 0.5, which indicates the similarity or internal validity of the reflective measurement model. Also, the reliability of the data collection tool and the quality of the reflective measurement model have also been confirmed. Conclusion Considering the importance and necessity of organizations to use artificial intelligence in the process of human resources management, a model for training employees using artificial intelligence was presented in this research, which has three parts according to the system mode: input, process and output. Model inputs include 1- educational data, 2- personal information, 3- educational needs, 4- user feedback, and 5- workplace data. The process of the model includes 1- determining the needs and goals, 2- collecting data, 3- pre-processing the data, 4- training the artificial intelligence model, 5- evaluating and improving the model, 6- implementing and deploying, and 7- monitoring and updating. The outputs of the model include 1- individual feedback, 2- educational suggestions, 3- monitoring and follow-up, and 4- support and guidance.