Developing a Model and Questionnaire for Predicting Intention to Use Job Boards: A Jobseeker-Oriented Research on the E-Recruitment Adoption in Iran

Table 1 shows the variables influencing intention to use e-recruitment and job boards that have been examined in previous studies.

The variables of this study have been chosen by referring to the e-recruitment literature and models utilized in jobseekers’ behavioral intentions studies. Given the e-recruitment context of Iran and the authors’ own first-hand experience with jobseekers’ behaviors and concerns for more than three decades collectively, these variables are also the most relevant for this study. The operational definition and theoretical reference for each variable are detailed in Table 2, which would be the basis for designing the research instrument (the SEDA-IUQ) and then extracting and examining latent factors influencing intentions to use job boards among Iranian jobseekers.

Based on a comprehensive literature review on e-recruitment services and several well-established theories and models within technology and e-recruitment adoption, a conceptual framework is designed to outline the hypothesized relationships between observable variables affecting the intention to use the job boards. Furthermore, the framework will serve as a theoretical foundation for the research model, encompassing the latent factors extracted subsequently. Fig. 1 signifies our conceptualized research framework.

Figure 1

Given the demonstrated scalability and versatility of the technology acceptance model (TAM), it is a well-researched model to explain most human behaviors (Davis, Bagozzi and Warshaw, 1989). TAM is favorably rigorous in behavior prediction in information technology (IT) adoption for employment-related issues; it is a potentially helpful common foundation for research into the behavioral intentions of jobseekers in the face of online job search and e-recruitment services. Moreover, TAM has been utilized by researchers to investigate e-recruitment adoption intentions and also has found support in the accurate prediction and compelling explanation of user behaviors across a variety of settings (Kaur and Kaur, 2022; Nadlifatin, et al., 2022; Khalid, et al., 2021; Nguyen, 2021; Irawan, Adiputra and Arshanty, 2021; Candra, et al., 2020; Grimaldo and Uy, 2020; Woon and Singh, 2019; Selvanathan, et al., 2019; Banerjee and Gupta, 2019; Chang and Kim, 2018; Siew, et al., 2018; Zhang, et al., 2018; Arsanti and Yuliasari, 2018; Mahmood and Ling, 2017; Alsultanny and Alotaibi, 2015; Brahmana and Brahmana, 2013; Kashi and Zheng, 2013; Goldberg and Allen, 2008).

Hence, TAM could be an ideal theoretical foundation for our conceptual framework intended to be applied to an original and unprecedented research context. Serving as the framework's core, the TAM (Davis, Bagozzi and Warshaw, 1989) encompasses the relationships between perceived usefulness (PU), perceived ease of use, attitude, and intention to use.

A model of the antecedents of perceived ease of use (Venkatesh and Davis, 1996) is utilized to outline the effect of jobseekers’ self-efficacy on the perceived ease of use. Signaling theory (Spence, 1903) and models proposed by Cober, et al. (2003), Allen, Mahto and Otondo (2007), and Kroustalis (2009) imply the relationships between job board content and esthetics as predictors for PU. Brand equity theory (Keller, 1993; Cober, et al., 2004; Allen, Mahto and Otondo, 2007) inspires to propose the impact of reputation on the intention to use. Ultimately, given the critical role technical capabilities and web functionalities play in shaping the perceptions of today's tech-savvy applicants, the attitude toward job board functionality construct is incorporated into the framework.

The main instrument for collecting primary data in this study is a survey questionnaire. The questionnaire consists of three sections as follows:

The first section consists of a brief note about job boards and research, while asking the respondent if they have used a job board before.

Authors decided to design, validate, and employ a researcher-developed questionnaire for several reasons. Firstly, all the questionnaires and items suggested in e-recruitment literature mainly focused on the perspective of management, behavioral, and psychological sciences and lacked a focus on computer science and IT; secondly, due to the fascinating speed of technology growth, the user's expectations are constantly evolving, so the conventional technology adoption measures are lacking in terms of technical terms for job boards and employment websites.

Hence, using the opinion of practitioners and academics in the e-recruitment and human resource (HR) management fields and based on the conceptual framework, from the perspectives of behavioral sciences, human–computer interaction, and especially the quality of the website, the relevant items have been constructed to create the proposed questionnaire. Appendix A lists all of the survey items in the SEDAIUQ questionnaire.

The face and content validity of the questionnaire was confirmed through the opinions of several academic and HR experts and the calculation of the Content Validity Index (CVI) and Content Validity Ratio (CVR) coefficients. Construct validity was assessed using exploratory factor analysis (EFA) of the data obtained from the distribution of the questionnaire among 200 jobseekers. As a result, Kaiser-Meyer-Olkin (KMO) values for all items were higher than 0.7, confirming the adequacy of this index for each item of the questionnaire. Bartlett index was also significant (14,939.39) at P ≤ 0.01, indicating the adequacy of the matrix.

The reliability of the questionnaire was tested using Cronbach's α in a preliminary study among 30 participants, leading to values above 0.6 for all variables. Hence, the reliability of the questionnaire was confirmed. Moreover, Cronbach's α was estimated separately for each of the extracted latent factors after the final implementation of the questionnaire.

As shown in Table 3, Cronbach's α values were higher than the minimum cutoff score of 0.7, indicating that the questionnaire has the necessary reliability.

A Google Docs-based online questionnaire was distributed via several professional social networking web-sites, online communities, and emails to potential Iranian participants to collect data at different times and platforms. Participation in the survey was completely anonymous, voluntary, and random, based on ethical considerations. Since this study sought to generalize the results from a total statistical population, and Morgan's table suggested 384 as the number of samples, data collection was ceased after collecting 500 completed questionnaires.

The only requirement for completing the questionnaire was the experience of using a job board. If the respondent had no experience, he/she would not be allowed to complete the questionnaire, in order to avoid misleading data and obtain the most reliable data possible. Incomplete questionnaires and unwillingness to answer the questionnaires were considered exit criteria, and after separating incomplete questionnaires, 447 questionnaires entered the statistical analysis, leading to a response rate of 89% for the study. Respondents were 82.6% male, 89.8% employed, and most were in the age range of 31–40 years.

EFA through Statistical Package for the Social Sciences (SPSS) was used to extract and identify the factors affecting the intention to use job boards among Iranian jobseekers. The following steps were taken for that purpose: 1)

The possibility of factor analysis on the data: Since factor analysis requires a large sample size, the adequacy of sampling was tested using the Kaiser–Meyer–Oaklin index (0.86) and the significance of the Bartlett's sphericity test (chi-statistic doubled to 14,939.30 at the level of P ≤ 0.01); thereby, results indicated that the conditions for factor analysis are met.

Since hypotheses H₁–H₈ were proposed for the relationships between observable variables, they should be modified for the extracted latent factors. First, direct impacts of each of the four factors on the dependent factor (reputation-seeking intention to use) were hypothesized as H’₁–H’₄:

H’₁: Job board technological innovation positively affects its reputation-seeking intention to use.

H’₂: Job board content accessibility positively affects its reputation-seeking intention to use.

H’₃: Job board practical utility positively affects its reputation-seeking intention to use.

H’₄: Job board functional adequacy positively affects its reputation-seeking intention to use.

Furthermore, extending on the literature review elaborated in Section 2.4 (hypothesis development for observable variables) and the items clustering in each factor (see Section 4.1), six hypotheses are proposed as follows:

Esthetics, content, and PU: As hypothesized in Section 2.3.1, both esthetics and content are significant antecedents of PU; hence, the fifth and sixth hypotheses are formulated as follows:

H’₅: Job board technological innovation positively affects its practical utility.

H’₆: Job board content accessibility positively affects its practical utility.

In order to test the modified hypotheses for the latent factors, the path analysis method has been used. The perquisites for using path analysis are elaborated below. As shown in Table 5, a linear relationship between predictor and dependent variables was confirmed by forming the correlation matrix of latent factors.

Furthermore, Table 6 shows a Durbin–Watson value of 2.02, which satisfies the condition (the ideal range 1.5–2.5); hence, the error term is independent and fulfills the assumption. Finally, the normality of the data was evaluated and satisfied by the Kolmogorov–Smirnov test (see Table 7).

Several fit indices (e.g., Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA), Incremental Fit Index (IFI), Normed Fit Index (NFI), Non-Normed Fit Index (NNFI)) have been used to ensure the model is appropriate to test hypotheses regarding causal relationships between latent factors. As shown in Table 8, all of these fit indices are acceptable.

After approving the measurement model of research constructs (latent factors), the research model was further evaluated by estimating the structural relationships among the research constructs using structural equations modeling in LISREL software. Fig. 2 also depicts the results of the structural model analysis.

Figure 2

As shown in Table 9, all the hypothesized paths, except for that linking content accessibility and functional adequacy to reputation-seeking intention to use (H’₂ and H’₄, respectively) were significant (P < 0.01). Hypotheses H’₁ and H’₃ were supported, and the results confirmed that both technical innovation and practical utility had a significant positive effect on reputation-seeking intention to use job boards. Furthermore, the results confirmed that practical utility perceptions toward a job board depend on technical innovation, content accessibility, and functional adequacy of job boards, supporting the hypotheses H’₅, H₆’, and H₇’. Fig. 3 shows the standardized estimate coefficients diagram for the model.

Figure 3

The goodness-of-fit statistic of the research model is presented in Table 10. Comparison of fit indices with their corresponding recommended values provided evidence of a good model fit.

While our findings support IT adoption literature and previous studies anchored on jobseekers’ behavioral intentions toward e-recruitment systems and services, this study represents some novel insights into e-recruitment adoption through its exploratory design leading to several multidimensional factors that significantly depict antecedents of behavioral intentions of jobseekers.

A strong correlation between the two variables of job-seeker's (computer) self-efficacy and perceived ease of use in the functional adequacy factor is in line with the model of the antecedents of perceived ease of use (Venkatesh and Davis, 1996). Furthermore, as Chang and Kim (2018) found, accuracy, security, self-efficacy, and perceived ease of use had a positive effect on PU, which is in line with our finding that content accessibility and functional adequacy positively affect practical utility. Moreover, our findings support previous studies indicating that jobseekers’ perceptions of technological capabilities and functionalities and self-efficacy in using job boards are important competence-related variables, in turn affecting their adoption intentions (Khalid, et al., 2021; Nguyen, 2021; Woon and Singh, 2019; Zhang, et al., 2018; Lin, 2010; Cober, et al., 2004).

Content accessibility, functional adequacy, and technological innovation were found to have a significant impact on practical utility, which in turn influences reputation-seeking intention to use. This finding supports previous studies by indicating that when jobseekers perceive easier and catchier access to accurate, relevant, and timely information and more self-efficacy and technology empowerment and innovation associated with adoption, they feel more determined to exploit job boards (Meah and Sarwar, 2021; Chen, Warden and Liuu, 2021; Woon and Singh, 2019; Priyadarshini, Sreejesh and Jha, 2019; Roy-Chowdhury and Srimannarayana, 2013; Williamson, et al., 2010; Kroustalis, 2009; Cober, et al., 2004). The results are similar to those of Poudel (2018), revealing that usefulness, technology self-efficacy, perceived enjoyment, ease of use, and attitude toward websites drive jobseekers to apply for jobs through online recruitment.

Consistent with the theory of brand equity and previous studies suggesting reputation, trust, image, and organizational brand as predictors of behavioral intentions toward e-recruitment websites (e.g., Cober, et al., 2004; Williamson, et al., 2010; Allen, Mahto and Otondo, 2007; Lin, 2010, Mahmood and Ling, 2017; Irawan, Adiputra and Arshanty, 2021; Schaarschmidt, Walsh and Ivens, 2021; Rahman and Patra, 2020; Candra, et al., 2020), our findings support this contention and extend the e-recruitment literature by revealing multidimensionality of reputation-seeking intention to use factor. The strong correlation between reputation and intention to use circuitously highlights the impact of network effect, social influence, and trust in decision-making.

Consistent with TAM (Davis, 1985; Davis, Bagozzi and Warshaw, 1989) and previous studies (e.g., Siew, et al., 2018, Candra, et al., 2020; Khalid, et al., 2021; Nadlifatin, et al., 2022), this study revealed a robust relationship between PU and ease of use, which are being clustered under one factor named practical utility that, in turn, affects e-recruitment adoption intentions.

Furthermore, this study revealed the strong influence of attitude on job pursuit intention, since attitude items were reflected in nearly all the latent factors influencing behavioral intentions identified through EFA. This finding is in line with the TAM2 (Venkatesh and Davis, 2000), which omits the attitude construct in their model because PU and perceived ease of use are two fundamental attitudinal beliefs of TAM that lead to a positive or negative attitude in an individual.

Technological innovation and practical utility revealed a substantial effect on reputational-seeking intention to use. While the latter finding is widely supported in the previous studies (Kaur and Kaur, 2022; Nadlifatin, et al., 2022; Nguyen, 2021; Candra, et al., 2020; Selvanathan, et al., 2019; Chang and Kim, 2018; Zhang, et al., 2018; Kashi and Zheng, 2013; Brahmana and Brahmana, 2013), none of the primary models of technology acceptance and previous studies have theoretically incorporated and empirically tested technology innovation as an influential factor on adoption intention variables such as the intention to use, attraction, or attitude toward e-recruitment services. A possible explanation could be the unprecedented pace of technology advancements and digital transformation that makes innovation an integral and inevitable part of each E-business (including e-recruitment) service provision.

As Lin (2010) argued, e-recruitment adoption models without considering jobseekers’ perceptions would be incomplete and potentially misleading. In support of his argument, this study develops a research model that successfully applies several well-established models and theoretical frameworks in the e-recruitment adoption realm to investigate online job search behavior in a novel context.

While previous studies have only examined the effect of one-dimensional variables such as PU and ease of use or reputation, content, or esthetics on the attitude or behavioral intentions toward e-recruitment and job search websites (e.g., Cober, et al., 2004; Allan, Mahto and Otondo, 2007; Goldberg and Allen, 2008; Kroustalis, 2009; Lin, 2010; Williamson, et al., 2010; de Goede, van Vianen and Klehe De Geode, 2011; Gregory, Meade and Thompson, 2013; Kashi and Zheng, 2013; Brahmana, and Brahmana, 2013; Alsultanny and Alotaibi, 2015; Huang and Chuang, 2016; Mahmood and Ling, 2017; Zhang, 2018; Siew, et al., 2018; Arsanti and Yuliasari, 2018; Grimaldo, and Uy, 2020; Schaarschmidt, Walsh and Ivens, 2021; Nadlifatin, et al., 2022), using an EFA approach, this study has revealed the multidimensionality of the latent factors discovered through its exploratory approach. This phenomenon implies that, for instance, a combination of perceptions of ease of use and usefulness and attitudes toward job boards (as practical utility factor), rather than just each of these variables independently, directly influences the intention to use job search websites.

Based on a comprehensive study of technology acceptance models and previous studies in the e-recruitment adoption, this study reveals that due to the rapid technological change, the dimensions of jobseekers’ perceptions and attitudes toward e-recruitment systems are also evolving rapidly. Nowadays, users make objective and subjective multidimensional evaluations when deciding or providing feedback on attitudes and intentions to use technologies (particularly e-recruitment systems and job boards).

Overall, this study contributes to both theory and application. At the conceptual level, the research findings support technology acceptance models in online recruitment and show their applicability to jobseekers’ reactions to websites and the need to update and consider factors based on innovation and technology growth that reflect users’ expectations and jobseekers. This study highlights the role of the technological innovation factor, which consists of innovation in design, services, and content, in the perceptions of reputation-seeking intention to use, reflecting the importance of innovation and creativity and timeliness in technology deployment, which could boost user intentions toward e-recruitment ventures.

From an e-recruitment perspective, this study makes a significant theoretical contribution by developing, validating, and implementing a researcher-developed questionnaire, the SEDA-IUQ, that incorporates behavioral science and HR management perspectives with ideas based on the website quality and human–computer interaction. The SEDA-IUQ can serve as a comprehensive and multi-perspective evaluation of behavioral intentions toward job boards, overcoming the limitations of previous scales and questionnaires.

This study offers valuable insights into the entire e-recruitment industry, including corporate recruitment websites, third-party job boards, and job search platforms. It directly contributes to the e-recruitment and third-party job boards to make them more effective and valuable for their users, highly customer driven, and constantly addressing the concerns and demands of jobseekers in a global context.

For practitioners, particularly e-recruitment service providers, this study could be helpful because it addresses the factors influencing the intention of jobseekers for online job applications, enabling them to design and administer job boards and corporate recruitment websites effectively. This study suggests that jobseekers prefer attractive job boards with an innovative approach to design, service delivery, and content to other job search resources. Our results also indicated that practical utility plays a vital role in determining behavioral intentions toward job boards; hence, e-recruitment service providers should seek to provide plenty of useful features, services, and capabilities.

Another crucial practical finding was that jobseekers’ behavioral intentions toward job boards may be driven by how they perceive the technological innovation and practical utility associated with job boards. Indirect effects on behavioral intentions were found through content accessibility and functional adequacy. Therefore, job board developers and administrators are projected to focus on technological innovation (in design, content, features, and services), practical utility, functional adequacy, and content accessibility to differentiate themselves from competitors in the market.

Employing an innovative and creative approach to design and development coupled with presenting practical, accessible services and facilities may raise awareness and interest among jobseekers. These efforts may significantly affect jobseekers’ behavioral intentions and encourage them to exploit job boards, leading to a steady flow of qualified applicants.

Results from this study have shed light on the online job search behavior of today's jobseekers who consider several variables (such as esthetics, usefulness, ease of use, self-efficacy, reputation, and attitude) simultaneously when forming perceptions and intentions toward the usage of online recruitment systems. Specifically, website esthetics exceed the traditional concepts of esthetics (limited to design) and address innovation from several dimensions of design, services, content, and facilities. Both E-recruitment system developers and HR experts can utilize our findings to improve the effectiveness of their activities.

Despite the comprehensiveness of the proposed model and its empirical support and fascinating insights that this study offers to the e-recruitment and job boards’ adoption literaturę. The findings should be carefully considered, given the limitations in concepts, methods, scope, and sample size. First, the study utilized a cross-sectional measurement method, limiting the ability to observe and make definitive conclusions about causal relationships between research variables. Therefore, longitudinal data could provide a more transparent basis for the transient causal relationships suggested by the proposed research model. The second limitation is the reliance on survey data through self-report measures. While self-reporting in this type of study is widespread, future studies that collect data through various sources and methods could produce critical insights and implications for the e-recruitment adoption context.

The present study focused only on the context of one country, so to make the research results more generalizable, it is recommended to apply the research model to other countries and cultures, leading to cross-cultural studies. Furthermore, despite the comprehensiveness of the model, the variables studied in the present study cannot explain all the influencing variables in the attitude and intention to use job boards, which can provide an opportunity for further and complementary studies. Despite these limitations, this study contributes to e-recruitment research by taking a crucial step toward filling the gap in this popular, but scarcely investigated research field in the Iranian e-recruitment context.