Abstract
Background
In the realm of contemporary management, decision-making emerged as a pivotal aspect, particularly within the human resource (HR) domain, profoundly influencing an organization’s culture and efficacy. The integration of high-quality data and artificial intelligence (AI) based automation in the recruitment process became very important in driving data-driven decision-making processes while hiring for any position.
Objective
This study sought to deeply explore the landscape of AI-based decision-making in the recruitment process while investigating the adoption of AI-enabled data-driven approaches within recruitment practice. Additionally, it aimed to propose a strong and healthy model aimed at significantly enhancing organizational effectiveness.
Methods
Smart PLS tool stood as the cornerstone of the analytical approach, offering a healthy framework for testing and validating our proposed model. To achieve the objectives, a two-fold approach was employed. Firstly, a comprehensive review of existing literature was conducted, encompassing seminal works and contemporary research on AI in HR and recruitment decision-making. Secondly, a small-scale survey was administered across diverse organizations representing IT and manufacturing industries. The survey focused on evaluating the current trends in AI adoption within the recruitment process and its impacts on decision-making, specifically emphasizing the utilization of assisted intelligence tools for routine activities and reporting upkeep. This dual-method approach provided insights into the practical challenges faced by organizations in adopting end-to-end decision-making automation.
Results
The combination of information gathered from a small survey and extensive research showed a common pattern: widespread use of AI tools mainly for routine activities categorized as assisted intelligence, offering partial solutions for the recruitment day-to-day tasks. However, it underscored the growing necessity for automation intelligence that encompasses holistic end-to-end decision-making processes in recruitment practice. The utilization of the tool Smart-PLS for analyzing the data from the small-scale survey of 50 respondents further enhanced the depth and reliability of the findings. For declaring the model’s fitness, the Smart PLS tool was used by providing various indicators and techniques, such as goodness-of-fit measures and bootstrapping procedures, to ensure the accuracy and reliability of the model.
Conclusions
In the evolving landscape of increased automation, striking a balance between AI-driven decision-making in recruitment and human intervention holds paramount importance, especially in decisions affecting employees. The utilization of the Smart PLS tool not only enhanced the credibility of our findings but also underscored our commitment to employing state-of-the-art methodologies in advancing knowledge within our field. This study advocates for the adoption of an AI-enabled data-driven approach, emphasizing its potential to optimize recruitment practice and strengthen organizational effectiveness.
Keywords
1. Introduction
Artificial intelligence (AI) now plays a significant role in human decision-making processes in the workplace, including employment, treatment assignment, and investigation processes, to mention a few. This process is referred to as “AI-based decision-making,” in which humans make judgments in these circumstances using their own knowledge and the recommendations of an AI-based algorithm (such as data-driven models, knowledge-based models, etc.). This Man-Machine partnership is providing great help to HR professionals to make decisions that are aligned with data reports. AI is enabling organizations to analyze past situations, check current trends, and forecast future uncertainty to survive in this competitive world.
There is no commonly adopted definition of Artificial Intelligence. It is the science and engineering of creating intelligent machines, particularly clever computer programs. “The science and engineering of constructing intelligent machines” is how John McCarthy, who first used the word in 1956, defines artificial intelligence (AI) (Artificial Solutions Industry Insight Homage to John McCarthy, the Father of AI). It is a computer-enabled system, let’s say a robotic system that is intended to process information to produce results, much like how human labor in the organization does, by learning, sorting problems, and making judgments. It has been suggested that AI is the study of how to use computers to replicate human intelligence. It was in the late 90 s that the use of AI in management began. Initially, it was limited to the use of computers and electronic data maintenance but now the easy use of the internet has changed the working pattern.
The process of choosing a course of action, obtaining data, and weighing potential options is known as decision-making. By gathering relevant data and identifying options, a good decision-making process helps to make more deliberated, considered decisions. Human resource decision-making is more concerned with the most efficient methods of investing in people. This includes identifying the use of money and resources concerning increased organizational effectiveness and employee performance. The role of HR professionals is expanding to strategic HR and towards HR business partners. Considering that the scope of making a decision has enhanced. Now, it is not limited to the decisions related to time tracking and attendance, employee development, performance management, compensation and benefits, expense management, and employee retention but it involves decisions for business growth and expansion, risk mitigation, skill development, funds management, HR data, and analytics.
1.1. AI-based organizational decision making in recruitment
Over a span of seven years across various organizations, the author recognized the vital role of AI in HR practices. During recruitment processes, decisions often relied on intuition rather than structured approaches. In teams, challenges like duplicated job postings, disorganized applicant tracking, and unmanaged interview scheduling were common. These issues led to less effective hiring and procedural errors, hurting organizational performance. AI emerged as a valuable solution, simplifying job postings, organizing applicant data, and supporting data-driven hiring decisions.
AI has reduced the workload of the HR admins through AI interventions which are keeping records of years that can be available with a single click. It is well-accepted that processing millions or billions of records is not doable, manually. Human minds would not be able to link the connections among data elements, which are seen to be essential in every decision-making. AI in decision-making is purely concerned with a data-driven approach. For HR decision-making, analytics has been considered a tool that helps to get clear ideas about past events which is important to address while making any decisions. HR analytics is data-driven insights that are developed by HR leaders, which include filtration of data and analyzing it to achieve organizational objectives. In the domain of recruitment, AI-based decision-making leverages HR analytics to streamline processes, identify patterns, and make informed hiring decisions, thereby optimizing recruitment outcomes and organizational success.
2. Literature review
The role of artificial intelligence in human resource management (HRM) practices. The referred literature provided insights from various studies and research papers on the benefits, challenges, and ethical considerations of integrating AI technologies in HR functions and decision-making.
2.1. Historical context and future directions
Artificial intelligence from a historical perspective [1] contextualized the evolution of AI, providing historical insights into its progression and outlining pivotal milestones. This historical perspective served as a foundation for envisioning future trajectories and potential advancements in AI-driven talent acquisition strategies, emphasizing the need for continued research and ethical considerations.
The literature provided a comprehensive overview of the integration of artificial intelligence into human resource management practices, highlighting its benefits, challenges, and ethical considerations. It underscored AI’s effectiveness in optimizing HR processes such as candidate screening and workflow automation while cautioning against over-reliance on AI technologies. The review emphasized the importance of fairness, transparency, and accountability in algorithmic decision-making within HR contexts, alongside continuous monitoring and refinement of AI systems. Furthermore, it acknowledged the successful application of AI in fundamental HR processes like selection, hiring, and management, with recognition of potential disparities in AI adoption among HR managers. Challenges and opportunities in effectively leveraging AI for decision-making processes, including legal and ethical accountability, were explored, alongside its impact on recruitment practices. By delving into recruiters’ perspectives, strategic implementations, technological innovations, global perspectives, and emerging frameworks, the review provided valuable insights into AI’s evolving role in HRM. Ultimately, it underscored the significance of responsible AI deployment, ongoing research, and ethical considerations in shaping the future of talent acquisition and HR practices, aiming for improved efficiency, fairness, and effectiveness in organizational decision-making and workforce management.
One of the key studies mentioned was the experimental study by John J. Lawler and Robin Elliot on the application of AI-driven expert systems in HRM [2]. The study explored how AI-based expert systems aided in decision-making, problem-solving, or knowledge management within HR functions. The paper provided empirical findings or insights from the experiment, elucidating the effectiveness, limitations, and implications of employing AI-based expert systems in HRM practices. Another study by Ali B. Mahmoud recognized the efficacy of AI-based systems in optimizing HR functions such as candidate screening, talent management, and workflow automation [3]. Despite acknowledging AI’s benefits, the study cautioned against over-reliance on AI technologies. The conclusion stressed the importance of human oversight and judgment in ensuring effective, ethical, and empathetic HR practices, advocating for a balanced approach integrating AI as a supportive tool with human capabilities in decision-making. Alina Kö Chling and Marius Claus Wehner’s study emphasized fairness and equity in algorithmic decision-making within HR contexts [4]. It highlighted the critical importance of developing and deploying AI systems prioritizing fairness, transparency, and accountability to mitigate discriminatory outcomes. The conclusion advocated continuous monitoring, auditing, and refinement of algorithms to align with ethical and legal standards, promoting fairness and non-discriminatory practices in HR decision-making. Anna Karmanska’s research underscored the advantages of leveraging HR analytics for evidence-based decision-making. It highlighted how HR analytics empowered organizations to optimize talent acquisition, retention, and development strategies using data-driven insights [5]. The conclusion encouraged organizations to embrace HR analytics as a crucial tool for achieving organizational goals and gaining a competitive edge in talent management and workforce optimization. Bhagyalakshmi R et al delved into the realm of HRM, exploring the utilization of AI in HRM procedures [6]. Their research, conducted through a sample of 140 HR employees, illuminated the successful application of AI in fundamental HR processes such as selection, hiring, and management. Factor analysis revealed essential components derived from AI’s integration into HRM, emphasizing unification, automation, channelization, and position. The study also established substantial support for a confirmatory factor analysis (CFA) model, demonstrating the significant influence of AI on HRM functions. Notably, the research underscored the potential disparities in AI adoption among HR managers based on their income categories, indicating the evolving importance of AI applications in HRM. Khan AI et al examined the burgeoning data generation landscape and the challenges associated with effectively leveraging AI for decision-making processes [7]. Identifying various data sources such as social media networks, business/transaction systems, administrative/government systems, and ubiquitous systems, the study shed light on the immense potential of these sources in prediction and decision-making using AI. Moreover, the research explored challenges in implementing effective decision-making strategies while harnessing the potential of these new data sources. Giuffrida I provided comprehensive insights into the legal and ethical considerations surrounding the accountability and liability associated with AI decision-making [8]. The paper extensively discussed challenges in establishing liability within the AI ecosystem, emphasizing the need for explicability in AI decision-making processes. By scrutinizing the complex landscape of AI’s impact on sectors such as government, finance, and healthcare, the research illuminated the critical considerations and potential ramifications of assigning AI legal personhood. Jaiswal et al. delved into the challenges and opportunities inherent in AI implementation within multinational corporations [9]. Their study highlighted key skill sets essential for sustainable AI adoption and practical tools for human resource development. By emphasizing the unusual pace of technological change and the need for reevaluating human resource development strategies, the research laid the groundwork for advancing employee learning and competencies in the era of AI. The paper on trust in AI-assisted decision-making presented a systematic review of empirical methodologies for evaluating trust in AI-assisted decision-making [10]. By meticulously analyzing 83 relevant papers, the research offered guidelines for enhancing trust evaluation in AI-assisted decision-making processes.
Bilal HMOUD and Varallya iLASZLO’s exploration critically analyzed AI’s role in automating aspects of recruitment and selection while examining its impact on traditional HR roles. It weighed the advantages and limitations of AI-driven HR practices, discussing the potential for efficiency gains and preserving human-centric elements in decision-making processes [11]. The study aimed to provoke thoughtful discussions among HR professionals and stakeholders about the transformative potential of AI in HR.
2.2. AI applications and challenges in talent acquisition
Digitalization in talent acquisition: A case study of AI in recruitment [12] meticulously explored how AI technologies were integrated into talent acquisition processes [12]. This case study outlined practical applications, such as AI-powered resume screening, candidate matching, and interview scheduling, emphasizing their role in enhancing efficiency and reducing biases. It highlighted challenges related to algorithmic biases and ethical considerations, urging for responsible AI deployment in recruitment practices. Literature Research – Will AI Technology Replace Humans In Talent Acquisition [13] critically examined the potential implications of AI replacing human roles in talent acquisition [13]. It discussed the shift toward automated processes and the consequent impact on recruitment professionals. The study deliberated on ethical dilemmas, including biases embedded in AI algorithms and the need for human oversight to ensure fair and effective recruitment outcomes.
2.3. Recruiters’ perspectives and perceptions of AI
Recruiters’ perception of artificial intelligence in Recruitment [14, 35] provided nuanced insights into recruiters’ attitudes toward AI-driven tools in talent acquisition [14]. These studies delved into the evolving role of AI at various stages of recruitment and uncovered discrepancies in perceptions among recruiters. While some embraced AI’s efficiency in screening and matching candidates, others expressed concerns about AI’s potential to replace the human touch in decision-making.
The role of AI in talent acquisition evolving and at which points/aspects of the TA lifecycle is AI being used [15] offered a comprehensive analysis of AI’s deployment across the talent acquisition lifecycle [15]. It detailed AI’s integration in sourcing, candidate assessment, and onboarding, elucidating its evolving role and impact on recruitment efficiency and candidate experience.
2.4. Strategic implementations and business case for AI in HR
The business case for AI in HR with insights and tips on getting started (Guenole & Feinzig, IBM Watson Talent) and the benefits of eHRM and AI for talent acquisition [17, 18] underscored the strategic advantages of AI adoption in HR practices [16–18]. They highlighted AI’s capacity to mitigate biases, expedite decision-making, and optimize resource allocation within talent acquisition, thereby substantiating the business case for AI integration in HR strategies.
2.5. Technological innovations and AI impact
The impact of AI writing assistance on job posts and the supply of jobs on an online labor market [19] investigated the repercussions of AI-driven writing assistance in job postings on online labour markets [19]. This study assessed how AI influenced the language and content of job descriptions, affecting the quality of job posts and subsequently altering the supply and demand dynamics in online recruitment platforms.
2.6. Global perspectives and industry-specific impacts
AI and the manufacturing and services industry in India [20] provided a contextualized examination of AI’s implications within the Indian manufacturing and services sectors [20]. It delineated industry-specific impacts, such as AI-driven automation in skill-specific recruitment and its influence on talent acquisition strategies within these sectors.
2.7. Emerging frameworks and conceptual models
Structuring AI resources to build an AI capability: A conceptual framework [21] introduced a conceptual framework delineating a strategic roadmap for organizations to build AI capabilities in talent acquisition [21]. This framework emphasized the need for structured resource allocation, skill development, and governance mechanisms to effectively leverage AI in HR practices.
2.8. Augmenting decision-making-integrating AI into organizational processes
Al-Surmi et al. conducted research aiming to enhance operational efficiency through strategic alignment of marketing and IT initiatives, presenting a decision-making framework validated with a SEM model (structural equation modelling) and data from 242 managers across sectors, highlighting IT strategy’s mediating role and organizational structure’s moderating impact on the relationship between marketing strategy and performance [22]. Ali K. Dogru1, Burcu B. Keskin discussed AI in operations management, emphasizing its extensive applicability, benefits, challenges like data security and workforce upskilling, and the importance of a holistic approach to AI integration [23]. Tania Babina, Anastassia Fedyk, Alex He, and James Hodson explored AI’s impact on firm growth and innovation, revealing significant economic effects and the association between AI investments and higher growth, innovation, patenting, and research & development spending across sectors [24]. Bäck, Hajikhani, and Suominen examined job advertisements’ utility in tracking AI adoption, revealing increasing demand for AI skills across industries [25]. Barnea A advocated for AI’s integration into decision-making processes, envisioning AI’s augmentation of analysts’ reports to bolster management judgment [26]. Lee BC et al. focused on decision-making models for AI-generated recruiting interview systems, emphasizing internal recruitment processes’ significance [27]. Harver provided a balanced perspective on the benefits and limitations of recruitment tech & AI in talent acquisition processes [28]. Icrunchdata launched a job posting platform with AI technology, promising real-time job matching [29]. Jarrahi MH explored the synergy between AI and human decision-making within organizations, advocating for AI’s augmentation of human contributions [30]. Various authors contributed significant insights into diverse aspects of AI’s applications and implications. For instance, Jennifer Johansson and Senja Herranen explored AI’s impact on traditional recruitment practices in HRM, shedding light on how AI-driven algorithms and machine learning are transforming the recruitment landscape [31]. Jingtao Fan, Lu Fang, Jiamin Wu, Yuchen Guo, and Qionghai Dai delved into the intersection between brain science and AI, elucidating how insights from neuroscience inform the development of AI technologies, particularly in cognitive computing and machine learning [32]. Furthermore, Mishra, Rodriguez investigated the challenges and solutions in recruitment processes within industries, proposing an AI-driven methodology harnessing natural language processing and machine learning techniques to enhance candidate selection [33]. Scott Mondore, Shane Douthitt, and Marisa Carson addressed the optimization of HR analytics to drive positive business outcomes, emphasizing the strategic alignment of HR initiatives with organizational objectives [34]. Stankiewicz, Bureau, and Hancock delved into recruiters’ perceptions of AI-based tools in recruitment, highlighting the potential of AI to improve the recruitment process by reducing bias and increasing efficiency [35]. Kuziemski M et al. conducted a comprehensive analysis of automated decision-making systems in the public sector, exploring the legal and policy implications of AI integration in governance frameworks [36]. Through case studies on immigration control in Canada, employment services optimization in Poland, and digital service personalization in Finland, the study emphasized the need for standardized frameworks to evaluate AI’s impact on public services and the increasing demands on governments to manage technological advancements accountably. Maria José Sousa’s work on HR Analytics Models for Effective Decision-Making focused on predictive analytics models in HRM, aiming to enhance decision-making related to workforce planning, performance management, and talent acquisition [37]. The paper highlighted successful implementation examples and underscored HR analytics models’ importance as strategic tools aligning with organizational objectives. Stone M et al. identified a gap in AI’s utilization in guiding strategic marketing decisions, emphasizing the necessity of exploring AI’s potential in strategic marketing choices [38]. Duan Y et al. conducted a comprehensive study on challenges inherent in AI-based decision-making systems, providing insights and research recommendations for IS scholars and practitioners, addressing historical perspectives and emerging challenges in integrating AI into decision-making processes, particularly in the era of Big Data [39]. Collectively, these studies underscored the need for continued research and attention to AI’s evolving role in HRM, and decision-making systems.
3. Research gap
After a detailed review of papers, a pattern of AI was observed and divided into three aspects; assisted intelligence, augmented intelligence, and automation intelligence. The intelligence which focused on automating basic tasks was called assisted intelligence. For instance, machines for attendance record-keeping. With augmented intelligence came symmetry. It was a two-way process. The aim was for machines to learn from human input. Humans, in turn, based their decision accuracy on intelligent information. Some of the payroll software was a great example of it. Automation intelligence was, simply put, including machines automating the complete process, with humans out of the loop. For instance, autonomous robots and self-driving cars. There was no such AI tool that had been developed to make automated decisions for HR functions. There were certain tools that automatically worked and took the decision of onboarding and document verification. This was on a limited aspect; the current need was to focus on the development of a widely acceptable end-to-end automated tool that could make a decision of any HR functions that were aligned with data and led to organizational effectiveness.
The application of AI was rapidly evolving, offering promising solutions to streamline processes and improve decision-making in recruitment. AI tools in recruitment encompassed various stages, from resume screening and candidate matching to interview scheduling and onboarding. However, while some existing tools automated certain aspects, such as candidate assessment and document verification, there remained a notable gap in the development of comprehensive AI solutions capable of making end-to-end decisions across the recruitment lifecycle. The current need was for AI-driven systems that leveraged data insights to enhance candidate sourcing, selection, and retention, ultimately driving organizational effectiveness and competitive advantage in talent acquisition. As AI continued to advance, there was a growing opportunity to develop and implement innovative AI technologies tailored specifically for recruitment, revolutionizing the way organizations identified, attracted, and onboarded top talent.
The existing literature lacked a comprehensive exploration of the intricate relationships among key components such as job posting and advertisement, application management, decision-making processes in hiring, quality of hire, and process agility, along with their consequential impacts on organizational advantage or disadvantage. This study aimed to address this gap by meticulously examining the interplay among these components and developing a comprehensive model to explore their relationships and impacts on organizational outcomes.
3.1. Research problem
The traditional way of making decisions in recruitment relies on human judgment. Professionals depend on their intuition, which they’ve developed over years of experience, using only a small amount of data. They trust their experience and gut feeling to distinguish good candidates from bad ones, high performers from low performers, and safe choices from risky ones.
Relying on human intuition for decisions might not be the best approach. Our brains have many cognitive biases that affect our judgment. Over thousands of years, we’ve developed quick decision-making shortcuts, called heuristics, to avoid processing too much information. However, these aren’t always enough. Data-driven decisions can suffer from problems like poor data quality, isolated data, lack of organizational support, misunderstanding of data use, and difficulty in finding relevant data. AI-based decision-making in recruitment refers to the use of AI by businesses to use datasets with AI to help make decisions that are quicker, more accurate, and more consistent.
3.2. Research objectives
Explore the landscape of AI-based decision-making in recruitment by analyzing the adoption and usage of AI-enabled data-driven approaches through small-scale survey. Develop an AI-driven decision-making model for recruitment by synthesizing existing literature, best practices, and frameworks to enhance organizational effectiveness.
3.3. Research questions
Does AI use for job posting improve the quality of decisions? Does AI for the management of applications improve the quality of decisions in the recruitment process? Does the AI-enabled decision to hire improve the quality of decisions? Does AI used for job posting improve process agility? Does AI for the management of applications improve process agility? Does AI-enabled decision to hire improve process agility? Does AI-based quality of decision and process agility lead to relative outcomes?
4. Model
Model: Artificial intelligence-based recruitment practice. Source: Authors.
The model proposed has various constructs and variables which have been derived from extensive literature encompassing job posting & advertisement, management of applications, decision of hire, process agility, quality of decision and the relative outcomes. The recruitment process has evolved significantly with the advent of online job platforms and AI-driven technology, facilitating efficient job posting and advertisement across multiple channels. Application management, conducted either manually through spreadsheets or via automated application tracking tools, plays a crucial role in candidate selection. The decision to hire is informed by various factors, including task assignments, activity monitoring, and interview automation. Process agility, measured by indicators like speed, alignment, innovation, and collaboration, enables organizations to adapt swiftly to changing recruitment dynamics. Quality of decision-making, assessed through criteria such as information clarity, systematic results, and automation, ensures effective candidate selection.
The model described above served as a representation of artificial intelligence-based recruitment practices. Job posting and advertisement were critical aspects of talent acquisition, enabling organizations to attract candidates. Previously, this was accomplished by placing job ads in newspapers and magazines. However, over the past decade, numerous online job platforms emerged, replacing traditional methods. In the current landscape, there were many job portals, and organizations strived to ensure maximum visibility across all platforms. AI-driven technology now facilitates posting job requirements across multiple platforms with a single click, streamlining data management and tracking. Key metrics such as reach to prospective candidates, application response rate, and promotion activity costs served as indicators for measuring the effectiveness of AI-enabled job posting and advertisement.
Management of applications was once again a crucial factor directly impacting recruitment. Organizations often employed a mixed approach, wherein applications were managed through spreadsheets, involving manual collection, segregation, and screening. Alternatively, application tracking tools or other software solutions automate these processes. In the construct of application management, indicators such as collection, segregation, and screening were utilized for measurement.
The decision to hire was an essential task in the recruitment process where an organization determined which candidate to offer a job position. This decision was typically made after a thorough evaluation of candidates’ qualifications, skills, and experience, and fit with the company culture and job requirements. Modern recruitment practices incorporate several key elements to inform the decision to hire, including assigning relevant tasks to candidates, monitoring their task activity, automating personal interviews, and scheduling these interviews efficiently and these factors had been used to measure the decision to hire.
Process agility in recruitment practice referred to the ability of an organization to quickly adapt and respond to changes, challenges, or opportunities in the recruitment process. It involved being flexible, dynamic, and responsive in managing various aspects of recruitment, such as sourcing candidates, screening resumes, conducting interviews, and making hiring decisions. In the model, process agility was measured by 4 indicators; quicker process, aligned processes, process innovation, and collaboration. Where, process innovation meant regularly assessing and refining the hiring process based on insights from recruiters, candidates, and hiring managers, and collaboration meant the act of individuals or groups working together towards a common goal or objective. It involved sharing ideas, resources, and responsibilities to achieve mutual benefits or outcomes. It could lead to improved collaboration between HR, hiring managers, and other stakeholders, fostering innovation through the exchange of diverse perspectives.
Quality of decision in recruitment practice referred to the effectiveness and accuracy of the decisions made during the hiring process. It encompassed the ability of hiring managers or recruitment teams to select the most suitable candidates for a given position, based on relevant criteria and considerations. Indicators of this construct were clarity of information, systematic results, and mundane process automation where, clarity of information focused on improved clarity of information for decision-making, systematic results in terms of quality focused on improved ability to compare and contrast results and mundane process automation referred to the faster decisions due to automation.
Based on the proposed model, the following hypotheses have been formulated: Artificial intelligence used for job postings improves the quality of decisions. AI for the management of applications in recruitment improved the quality of decisions. AI-enabled decision to hire improves the quality of decision. AI used for job posting has improved process agility. AI used for the management of applications improved process agility. AI-enabled decision to hire improved process agility. AI-based process agility leads to relative advantage. AI-based quality of decision leads to relative advantage.
5. Methodology
In conjunction with the foundational knowledge gathered from an extensive literature review, this study aimed to augment insights by conducting a small-scale survey targeting HR practitioners. This survey served as a primary research tool to gather firsthand insights, opinions, and experiences regarding AI-based decision-making within Recruitment practice. The survey was designed to elicit the practical implementations observed by HR professionals in utilizing AI-enabled data-driven approaches.
The method of data collection employed was snowball sampling. Participants were approached through community groups on LinkedIn and WhatsApp. The selection process ensured that participants came from organizations with more than 100 employees in the IT segment and a minimum of 50 employees in the manufacturing segment. The data collection process took almost 2 months. The questionnaire was sent as a link, and participants responded through the same link.
The survey comprised demographic questions and Likert scale statements. Smart PLS software was used as a tool for variance-based structural equation modelling (SEM) using the partial least squares (PLS) path. By engaging with practitioners directly, this survey provided real-time insights, complementing the theoretical understanding derived from the literature review. The combination of these methodologies sought to provide a comprehensive view of AI’s integration into HR decision-making processes, enhancing the depth and applicability of the study’s findings.
6. Analysis
The collected data from 50 respondents within varying designations (ranging from executive to deputy manager) showcased a broad spectrum of experience in the HR domain, with an average experience of 11.5 years and a range spanning from 2 to 23 years. The respondents represented organizations primarily from the Manufacturing and IT industries, spread across cities including Ahmedabad, Pune, Vadodara, and Gandhinagar in the states of Gujarat, India and Maharashtra, India.
An impressive 100% of the organizations utilized automation tools for recruitment purposes, with 60% employing Application Tracking Systems (ATS) to manage job applications from diverse sources. Notably, 80% of organizations used a single tool for filtering, processing, shortlisting, scheduling, and tracking applications, whereas the remaining 20% utilized multiple tools and manual methods.
Measurement model.
The transition from manual management to automated processes has led to significant improvements in efficiency, with 80% of respondents citing enhanced ease in managing application data, interview scheduling, and centralized data storage. Additionally, the application of AI tools garnered an average rating of 4.2 in improving job posting responses compared to manual recruitment practices.
Regarding application segregation and screening, 60% of respondents relied on keyword-based segregation, while 40% employed manual reviews, telephone screenings, and automated matching. Surprisingly, communication of interview availability with candidates was predominantly done manually by recruiters (75%), despite available automation tools.
While certain AI tools provided facial expression analysis during interviews, only 10% of respondents found this feature useful. Moreover, the reliability and support received from automation tools exhibited a varied landscape, with an average reliability rating of 62.1%, ranging from 20% to 90%.
Several key variables were examined to gauge the impact of technology integration within the workplace. Cronbach’s alpha test was employed to assess the reliability and internal consistency of these variables.
In addressing the evolving landscape of talent acquisition, our study introduced a comprehensive AI-driven recruitment model designed to optimize the hiring process and achieve relative outcomes. This model encompassed essential variables such as job posting and advertisement (JPA), management of applications (MOA), decision to hire (DOH), quality of decision (QOD), process agility (PA), and relative outcome (RO).
Our model incorporated constructs such as reaching prospective candidates, application response rate, and cost of promotion activity to ensure effective candidate engagement and outreach. Additionally, variables including collection, segregation, and screening streamlined application management, while constructs like relevant assignment of tasks and monitoring of task activity enhanced task allocation and oversight throughout the recruitment process. Moreover, our model emphasized the importance of automation and systematization, with constructs focusing on the automation of personal interviews (PI), scheduling of PI, and clarity of information, leading to systematic results and quicker processes. Furthermore, variables like mundane process automation and collaboration promoted process innovation and alignment, ultimately facilitating relative advantages in recruitment outcomes. To examine the relationships among the variables, several statements for each construct were drafted and analyzed.
By introducing this AI-driven recruitment model and its associated variables and constructs, we aimed to provide a comprehensive framework for optimizing recruitment processes, enhancing decision-making quality, and ultimately achieving relative advantages in the talent acquisition process.
7. Results
The average variance extracted (AVE) values achieved were 0.658,0.659, 0.686, 0. 584, 0.649, and 0.780, surpassing the acceptable AVE threshold of 0.50. This indicates that all the research constructs exhibit reliability and validity, as substantiated by the statistical values.
Figure 2 above illustrates the resulting framework using the PLS algorithm to assess Cronbach’s alpha and correlation, while the figure presented factor loadings of items for each construct, specific measurements were undertaken for each construct. Job posting and advertisement were gauged through three items, designated as JPA1, JPA3, and JPA5. The management of application was assessed through three items, labeled MOA 1, MOA 2, and MOA 3. Decision to hire was measured by 4 components labeled as DOH 2, DOH 4, DOH 5, and DOH 6. Similarly, Quality of decision was appraised with three items, denoted as QOD 1, QOD 2, and QOD 4, and Process agility was evaluated with four items, encompassing PA 1, PA 2, PA 3, and PA4. Whereas Relative outcome was measured by three items, designated as RO 1, RO 2, and RO 3.
Conceptual model.
8. Discriminant validity-Fornell-Larcker criterion
In our analysis, we utilized the Fornell-Larcker criterion to assess discriminant validity. This criterion compared the square root of the average variance extracted (AVE) for each construct with the correlations between constructs. Meeting this criterion confirmed that each construct captured unique variance, ensuring the reliability of our measurement model.
Construct evaluation.
JPA 2, JPA 4, DOH 1, and DOH 3 were excluded from the analysis because their factor loading was below 0.5.
Table 3 displayed a matrix that illustrated the relationships and distinctiveness among four categories. The diagonal values (Decision Of Hire: 0.811, Job Posting & Advertisement: 0.812, Management Of Application: 0.828, Process Agility: 0.764, Quality Of Decision: 0.805, Relative Outcome: 0.883) represented the self-relation of each category, indicating their individual reliability and distinctiveness. The off-diagonal elements showed the discriminant validity between pairs of categories. These values measured how distinct each category was from the others, with higher diagonal values and lower off-diagonal values indicating better discriminant validity.
Discriminant validity.
Table 4 presented R-square values for three categories: process agility (0.274), quality of decision (0.426), and relative outcome (0.539). These values indicated the proportion of variance in a dependent variable explained by each category in a regression model. Process agility explained at 27.4%, quality of decision at 42.6%, and relative outcome the highest at 53.9%. These percentages reflected the strength of the relationship between each independent variable and the dependent variable, with higher values indicating a stronger explanatory power in the model.
R-Square.
Table 5 summarizes results from statistical hypothesis testing, examining relationships between job posting & advertisement, management of application, and decision of hire with process agility and quality of decision, and process agility and quality of decision with relative outcome.
The table showed that all tested relationships were statistically significant, as indicated by their original sample values (ranging from 0.1 to 0.549), high T statistics (ranging from 0.525 to 4.856), and low P values (all 0.049 or lower). Each hypothesis in the table was marked as “Supported,” indicating that the data provided strong evidence for these relationships. This suggested a significant impact as well as a notable relationship between the variables.
Summary of the results derived from statistical hypothesis testing.
9. Discussion
The test results for hypotheses 1 to 8 indicated a positive and significant influence, statistically ranging between a p-value of 0.001 and 0.049, all below 0.05 at a significance level of 5%. All hypotheses were substantiated.
H1, H2 & H3: Examining the correlation of AI-based job posting & advertisement, management of application, and decision to hire with quality of decision; the research results showed a regression coefficient (β) of 0.1, 0.169, 0.538, an observed t-value of 0.525, 0.813, 4.448, with a p-value of 0.049, 0.041, 0 respectively, meeting the statistical acceptance criteria.
H4, H5 & H6: Examining the correlation of AI-based job posting & advertisement, management of application, and decision to hire with process agility; the research results showed a regression coefficient (β) of 0.505, 0.328, 0.289, an observed t-value of 3.237, 2.183, 1.915, with a p-value of 0.001, 0.029, 0.049 respectively, meeting the statistical acceptance criteria.
H7: AI-based process agility led to relative outcomes. The results of the analysis showed that the regression coefficient (β) was 0.549 and the t-value was 4.856, with a p-value of 0.000. With these statistical values, the results of H7 were accepted as significant.
H8: Examining the impact of AI-based quality of decision and relative outcome; The results of the analysis showed a regression coefficient (β) of 0.265, a t-value of 2.401, and a p-value of 0.016. Thus, H8 was accepted as significant based on statistical values.
The model posited that enhanced job posting & advertisement, management of application, and decision to hire resulted in agile processes and heightened decision quality. This, in turn, fostered relative advantages for organizations, enabling them to outshine competitors in talent attraction and make well-organized, data-driven recruitmentdecisions.
10. Conclusion
The contemporary landscape of talent acquisition evolved significantly, driven by advancements in technology and innovative recruitment practices. Job posting and advertisement transitioned from traditional methods to embrace the vast array of online platforms, facilitated by AI-driven technologies that streamlined data management and tracking. Similarly, the management of applications transformed, with organizations adopting a mix of manual and automated processes to enhance efficiency and effectiveness. Decision-making in hiring became more sophisticated, integrating elements such as task assignments, monitoring, and automation of interviews to ensure informed selections.
Moreover, process agility emerged as a crucial component, enabling organizations to adapt quickly to changing recruitment dynamics and collaborate effectively to drive innovation. Quality of decision-making was paramount, emphasizing clarity of information, systematic results analysis, and automation to expedite decision-making processes.
In recruitment processes, decisions often rely on intuition rather than structured methods. Common challenges within teams included duplicative job postings, inefficient applicant tracking, and poorly managed interview scheduling. Our model aims to assist organizations in improving decision-making, streamlining processes, and enhancing overall organizational effectiveness. The study focuses exclusively on manufacturing and IT organizations, which may restrict the generalizability of findings across other sectors. Additionally, the sample size of 50 participants could be expanded to get the comprehensiveness and applicability of the study’s findings, offering valuable insights for a wider range of industries and organizational sizes. Future research could explore diverse industries and larger participant pools to further enrich our understanding and applicability of AI in HR decision-making.
Overall, the model underscored the interconnectedness of enhanced job posting, application management, decision-making, process agility, and decision quality, positing that improvements in these areas yielded relative advantages for organizations. By leveraging these advancements, organizations could distinguish themselves in talent acquisition, make informed and data-driven recruitment decisions, and ultimately gain a competitive edge in attracting top talent.
Footnotes
Acknowledgments
The authors have no acknowledgment.
Author contributions
CONCEPTION: The conception of the research idea and overall article design were led by Ritu Purohit.
METHODOLOGY: Ritu Purohit developed the methodology.
DATA COLLECTION: The article includes primary data collection.
INTERPRETATION OR ANALYSIS OF DATA: Ritu Purohit conducted the analysis of the collected data, interpreted the findings, and drew conclusions based on the results obtained.
PREPARATION OF THE MANUSCRIPT: Ritu Purohit took the lead in drafting the manuscript, including writing the introduction, methods, results, and discussion sections.
REVISION FOR IMPORTANT INTELLECTUAL CONTENT: Ritu Purohit critically reviewed and revised the manuscript for important intellectual content, ensuring the accuracy and coherence of the scientific arguments and conclusions.
SUPERVISION: The supervision of the research project and manuscript preparation was provided by Dr. Tanushri Banerjee, who served as the corresponding author. Dr. Banerjee provided guidance, oversight, and expertise throughout the research process, ensuring the integrity of the study and the accuracy of the manuscript’s content.
Author Biographies