Factors Determining Intention to Continue Using E-HRM

Abstract

The development of information technology has promoted organizational transformation through the utilization of an electronic information system. This research aimed to identify factors that influence continuous intention to use E-HRM. This empirical research applies the Technology Acceptance Model and Cognitive Model for identifying significantly impacted areas of continuous intention to use E-HRM in a highly dynamic environment. The data were collected using questionnaires delivered directly to respondents. The sample was 100 employees of ESQ Group selected through random sampling. The variables used were subjective norms (X1), perceived behavioral control (X2), perceived innovativeness (Y1), cognitive absorption (Y2), satisfaction (Y3), and continuous intention to use E-HRM (Y4). Statistical analysis using Structural Equation Modelling (SEM) with Smart PLS was applied. The results revealed that behavioral control (X2) did not influence the continuous intention to use E-HRM (Y4) and that cognitive absorption (Y2) and satisfaction (Y3) did not significantly influence continuous intention to use E-HRM (Y4). Subjective norms (X1) significantly influenced both perceived innovativeness (Y1) and continuous intention to use E-HRM (Y4), perceived behavioral control (X2) significantly influenced both perceived innovativeness (Y1) and cognitive absorption (Y2), and perceived innovativeness (Y1) significantly influenced both satisfaction (Y3) and continuous intention to use E-HRM (Y4).

1. Introduction

The rapid development of the internet in the last decade, which has transformed every aspect and changed the way organizations do business and web empowerment for Human Resource Management is one of the newest developments. The Human Resources (HR) function has undergone dramatic changes due to the rapidly growing use of information technology (Lee & Xuan, 2019; Chulanova et al., 2019; Haerani et al., 2020). HR activities can now be conveyed, not only by HR professionals but also by increasing use of information technology. The rapid development of the internet has driven the implementation and application of Electronic Human Resource Management (E-HRM). According to Olivas-Lujan et al. (2007), HRM department using information and communication technology (ICT) is becoming an increasingly important phenomenon commonly referred to as E-HRM. This is basically a decentralized HR function for management and employees. At the same time, the use of E-HRM, which has various functions, can be carried out with the presence of technology to develop its full potential (Lin, 2011; Umar et al., 2020). In general, E-HRM is believed to provide several key benefits for organizations, such as increasing efficiency and reducing costs associated with HRM, facilitating the shift in the role of HR to a more strategic level, and improving client services, from handling administrative entities using paper and pencil and HR tasks. On a labor-intensive basis, E-HRM applications can be used in carrying out various functions such as recruitment, selection, performance management, and payroll administration (Omran & Anan, 2018; Reina & Scarozza, 2020)

However, organizations have not yet been to feel that an E-HRM system can help them make better HR decisions. The use of ICT in human resource services has become an important strategy to gain a competitive advantage. HR can claim to create a competitive advantage and align functions more closely with company strategy by creating added value for managers and employees through a more effective flow of information in the ‘market space.’ In addition, Grant and Newell (2013) stated that effective human resource management can make a significant contribution to organizational performance and that human-resource-related problems are central to the creation of a sustainable competitive advantage.

On the other hand, literature studies have shown that one of the main determinants of continued intention to use an information system is user satisfaction (Ha & Park, 2013). Increasing customer satisfaction by meeting their needs and requirements has long been an important concern in the field of marketing and consumer behavior (Bhattacherjee, 2001). In the area of Information Systems (IS), user satisfaction with technology is also an important factor affecting the adoption and continued use of technology (Bhattacherjee, 2001). Apart from user satisfaction, trust is a major determinant of adoption and sustainable use of e-commerce (Venkatesh et al., 2011) and internet banking (McNeish, 2015). Other factors that influence the use of electronic-based services are system usability, security, and self-efficacy (Bhattacherjee, 2001; Bhattacherjee & Premkumar, 2004; Hsu & Chiu, 2004; Venkatesh et al., 2011).

In summary, most of the literature on IT-based information systems adoption (including E-HRM) and usage focuses on early adoption. Only a few have examined post-implementation intentions and behavior (Bhattacherjee, 2001; Bhattacherjee & Premkumar, 2004). Previous studies of Bhattacherjee (2001), Bhattacherjee and Premkumar (2004), have argued that the factors on an individual’s decision to use technology can vary from the initial adoption phase to the next phase of use and finally to the post-implementation phase. This variation results from different experiences of individuals in the perception of and product users (Karahanna et al., 1999; Venkatesh et al., 2011). Therefore, there is a need to study the use of technology in the post-implementation phase. Consequently, users could be more cautious when adopting E-HRM. Therefore, to fill research gaps and to assist organizations in understanding users’ continuation use intention, this study examines the factors that influence continuous intention to use E-HRM. By extending the exception confirmation mode (ECM) to E-HRM, we can validate model values in the context of new emerging technologies. This practice is in line with the philosophy of building knowledge. ECM also provides a solid basis for the development of this research model. Research in the E-HRM field is relatively new. Similarly, Ruel et al. (2007) underlined that the topic of E-HRM will certainly not be out of date and that it’s full potential is still being anticipated; therefore, academic involvement in the topic needs to develop. Ruel et al. (2007) further emphasized that research on E-HRM is still in the “youth phase” and is still under theory. Besides, Strohmeier and Kabst (2009), Gregeby and Hugosson (2017) highlighted that despite the widespread use of E-HRM systems, few researches have focused on this issue and the extent to which users accept E-HRM.

In a review of the literature, Maier et al. (2013) stated that almost a third of the studies reviewed were exploratory and did not have a specific theoretical basis to guide E-HRM research. Also, Maier et al. (2013) emphasized that it is essential to expand the literature on E-HRM and its value by using more empirical techniques. This argument is in line with Bondarouk et al. (2009), Strohmeier and Kabst (2009), stating that previous research in the E-HRM field has been criticized for a general lack of theory. In terms of scientific activity, however, surprisingly few researches have been done on the impact of IT on HR. Schalk et al. (2013) stated that empirical evidence is limited to strategic considerations taken into account in the decision-making process around the application of E-HRM because it is a prerequisite for E-HRM to be successfully used for strategic purposes. Previous research has identified various factors and results of E-HRM research regarding job relevance, ease of use, usability, attitudes, trust, usability, quality, user support, preferred HR roles, HRM effectiveness, appreciation of E-HRM applications, and language standardization. However, this research has focused on the early adoption and use of users. Little research-oriented attention has been devoted to investigating the post-adoption use of E-HRM by leveraging Technology Continuity Theory (TCT). The vital role that user continuity plays is critical to the success of E-service providers. Besides, there is a dearth of research on the factors influencing the intention to continue using E-HRM. Most of the studies on E-HRM were carried out by various researchers in developed countries (the United States and Europe) (Galanaki et al., 2019; Holm, 2020; Poba-Nzaou et al., 2020). The study is limited to Malaysia, an environmentally, economically, and technologically very different country from developed countries. Large gaps in the market environment and management between developed and developing countries (such as Indonesia) may lead to different findings on E-HRM.

In an attempt to respond to gaps in the literature, this study the factors that influence users continued use of E-HRM by trying other alternative models. We proposed that the intention of sustainable use of E-HRM is influenced not only by satisfaction but also by other factors such as the perception of innovation and other factors. This research was then useful in explaining the possibility of continuing the current user experience with E-HRM in the future. The research contributes both theoretically and empirically to our understanding of the determinants of intentional sustainable use of E-HRM in companies. Starting from these ideas, the purpose of the research was to how the factors that influence the continuation of E-HRM adoption by system users. This study aims to investigate the key factors affecting sustainability intentions in the E-HRM system with a model based on Innovation Diffusion Theory (IDT), Expectation Confirmation Theory (ECT), and Planned Behavior Theory (PBT). It aimed to investigate the success factors that influence the continued adoption of E-HRM and to investigate the relationship between the independent and dependent variables.

2. Literature Review

2.1. E-HRM as a Human Resource Management Information System

According to Alzhrani (2020), Human Resource Information System is a digital system that provides up-to-date and accurate data for control and decision-making purposes. Meanwhile, according to Knauer et al. (2020), Human Resource Information Systems is software that has a database and allows for inputting, storing, and manipulating data from employees in the company. This allows for global visualization and access to important information from employees. This means that human resource management is changing to a more sophisticated digital form and is useful for storing, collecting, and informing data for control and decision-making purposes (Alzhrani, 2020; Knauer et al., 2020; Nurlina et al., 2020; Sumaryati et al., 2020). The human resource information system must be good in order to present information that is: timely, accurate, concise, relevant, and complete.

The term E-HRM first appeared in the late 1990s when e-commerce came to the business world. E-HRM also means administrative support from the HR function to use internet technology and improve working conditions and employee performance (Gardner et al., 2003). There are three levels of E-HRM: (1) E-HRM operations related to administrative functions such as payroll and employee personal data; (2) E-HRM relational related to business process support, recruitment, training, performance management, and so on; (3) Transformational E-HRM is related to strategic HR activities such as knowledge management and strategic reorientation. The adoption of E-HRM in organizations has been debated on three main aspects. First, E-HRM can increase efficiency by reducing HR transaction costs and the number of employees. Second, E-HRM can replace physical capabilities by using digital resources. Finally, the effective use of an integrated E-HRM system can change the HR business model by freeing up HR executives to provide strategic value to businesses that they could not do previously. This study tries to develop things related to the behavior of companies implementing E-HRM.

First is subjective norms refer to the perception of social pressure to execute or not execute that behavior, including the pressure of family, friends, and other important people (Ajzen, 2020). Subjective norms are always considered to be the most conflicting element in the theory of planned behavior. Some researchers have found that subjective indicators play an important role in forecasting the intentions (Kolvereid et al., 2006; Luc, 2018), and some researchers completely ignore the indicators (Peterman & Kennedy, 2003; Veciana et al., 2005). Related to the Diffusion of Innovation Theory, subjective norm variables can be operationalized as media/internet, social structure, social norms, heads of organizations/institutions, experts in IT, and colleagues who influence the adoption of E-HRM use to individuals (the context of this research is teaching).

The second is Perceived Behavioral Control (PBC). Ajzen (2002) defined perceived behavioral control as the perceived acceptance or difficulty of performing the behavior. Perceived behavioral control can be antecedent for the actual levels of control (Armitage & Conner, 2001). So, Behavioral Perceived Control is determined by a number of beliefs about the existence of factors that can facilitate or complicate the implementation of the behavior displayed (Hongdiyanto et al., 2020). The definition of Perceived Behavioral Control is operationalized to a person whose ability to use E-HRM and perceived in relation to his or her abilities.

The third is Perceived Innovativeness, defined by Tominc (2019) as ideas, practices, or objects that are perceived / recently felt by individuals. In the context of E-HRM research, the definition is operationalized as the perception that an individual is associated with more value for E-HRM compared to conventional HR management, as measured by the innovation attribute; relative advantage, compatibility, complexity, testability, and observability (Galanaki et al., 2019). For the context of innovation in information technology (including E-HRM), Karahanna et al. (1999) developed perceptions of usability, images, compatibility, ease of use, testability, visibility, and reliability.

The fourth is Cognitive Absorption, which means the condition in which the individual is more deeply involved when interacting with the software (Saadé, & Bahli, 2005; Barnes et al., 2019; Ghasemaghaei, 2020). Cognitive Absorption, indicated by the five indicators: temporal dissociation, focused immersion, increased enjoyment, control, curiosity. When operationalized in the context of E-HRM, cognitive absorption is operationalized into a state where educators are more deeply involved when interacting with E-HRM, and the visible signs are fully concentrated.

The fifth is Satisfaction; in the context of the Expectation Confirmation Theory defined by Oliver (1980; 1993), satisfaction is a pleasant or positive emotional state resulting from an assessment of the performance of a product or service. If it adjusted to the context of the definition of E-HRM, in that case, it could be operationalized with a pleasant or positive emotional state of the E-HRM user (teacher) after seeing the benefits and abilities of E-HRM. Since satisfaction is a person’s sense of response to what is the target of consumption, the indicators of satisfaction des-cribed in this study are consistent with teacher perceptions.

Sixth is Continues Intention, which means a behavioral tendency to continue using information systems (Swar et al., 2017).

2.2. Hypotheses

This study examined every domain of the Continues Intention to use E-HRM and then investigate whether each of the determinants would be affecting the user to continue using E-HRM. The investigation is based on the following hypotheses:

H1: Subjective Norm (X1) is significantly and positively related to Perceived Innovativeness (Y1).

H2: Subjective Norm (X1) is significantly and positively related to Continues Intention to use E-HRM (Y4)

H3: Perceived Behavior Control (X2) is significantly and positively related to Perceived Innovativeness (Y1).

H4: Perceived Behavior Control (X2) is significantly and positively related to Cognitive Absorption (Y2).

H5: Perceived Behavior Control (X2) is significantly and positively related to Continues Intention to use E-HRM (Y4)

H6: Perceived Innovativeness (Y1) is significantly and positively related to Satisfaction (Y3)

H7: Perceived Innovativeness (Y1) is significantly and positively related to Continues Intention to use E-HRM (Y4)

H8: Cognitive Absorption (Y2) is significantly and positively related to Satisfaction (Y3)

H9: Cognitive Absorption (Y2) is significantly and positively related to Continues Intention to use E-HRM (Y4)

H10: Satisfaction (Y3) is significantly and positively related to Continues Intention to use E-HRM (Y4)

3. Research Methods and Materials

This study was designed to answer the formulated research problems and research objectives and to test hypotheses using a quantitative approach. This research was explanatory research, which seeks to explain the relationship between variables through hypothesis testing. Based on the data collected, this research was categorized as survey research. According to Ketokivi (2019), the survey research approach was a data collection and measurement process to produce quantitative descriptions of several aspects of population studies. Some research groups also argued that cross-sectional surveys are the right method for research problems and support the use of information technology, especially E-HRM systems. Because it is currently a global phenomenon that does not only shows an upward trend but also gives a stagnant impression. For more details, this research was required. The research approach used in this research was quantitative, which process starts from theory, which in turn, uses deductive logic hypotheses that are empirically expressed accompanied by measurement and operationalization of the variables. Furthermore, generalizations were based on statistical analysis to conclude the research findings. Based on the purpose of obtaining this explanation, the units of analysis in this study were the following perceptualities: (1) Subjective Norms (X1); Perceived Behavioral Control (X2); Perceived Innovativeness (Y1); Cognitive Absorption (Y2); Satisfaction (Y3), and; Continuous Intention to Use E-HRM (Continuance Intention) (Y4).

Population refers to the entire group of people, events, or things of interest that the researcher wants to investigate. The sample is part of the population (Sekaran, 2003). The sampling method in this study was purposive sampling. The reason is that the data obtained were in accordance with the research objectives and comparable to previous studies. The sample criteria must use an information system related to the E-HRM task, and they must work in the operational department and the company. The number of samples in this study was following the analysis plan. Based on Hair et al. (2010), the appropriate minimum sample size in this study is 100 respondents. Based on that opinion, the number of respondents used in this study was at least 100 people.

Likert scale was used to measure social phenomena. It was explicitly determined by researchers and referred to as research variables. With a Likert scale, the variables to be measured were translated into dimensions. The dimension was then translated into measurable indicators, which were in the form of questions or statements answered by the respondents. The scale form in the questionnaire used in this study applied a Likert scale with a standard scale of 1 to 7. Each respondent was asked to provide an assessment question with a choice of answers to the Likert scale as follows: (1) Strongly disagree: score 1; (2) Disagree: score 2; (3) Disagree: score 3; (4) Neutral: score 4; (5) Somewhat agree: score 5; (6) Agree: score 6; (7) Strongly agree: score 7.

This study used two types of data: primary data and secondary data. Primary data are collected directly for further analysis to find solutions to the problems studied (Sekaran, 2003). The primary data of this study were collected from questionnaires sent to respondents. Questioners were used to measure IS success. Secondary data collected by researchers are published in statistical media and other journals, and information available from whatever published or unpublished sources, available both inside and outside the organization, all of which may be of use to the researcher. This study used secondary data as additional information to assist in research analysis.

Data collection in this study was carried out by a survey method by distributing a list of questions directly in the form of a questionnaire. Respondents from ESQ Group Jakarta employees filled out the questionnaire. Before the questionnaire was administered, organizations were firstly asked about their willingness to become respondents and allowed their staff to fill out the questionnaire. The next step was to send a questionnaire package consisting of 30 questionnaires.

The explanation of research variables and building indicators is as follows: (1) Subjective Norms (X1); X1.1 Opinions in the mass media or on the internet recommend the use of E-HRM; X1.2 As an employee, the work environment presses me to use E-HRM; X1.3 It is mandatory for employees to conform to the latest teaching methods; X1.4 Leaders in my company encourage the use of E-HRM; X1.5 The opinion of IT experts supports the positive use of E-HRM;

(2) Perceived Behavioral Control (X2); X2.1. I can complete my study program operating E-HRM; X2.2. For me, I can use E-HRM when I need it; X.2.3. The software in E-HRM is compatible with the software that I use; (3) Perceived Innovativeness (Y1); Y1.1 Experience has been tried several times, showing: E-HRM facilitates the implementation of work processes; Y1.2 Experiences were tried several times, showing: E-HRM to improve the quality of HR management processes; Y1.3 Experiences were tried several times, showing: E-HRM makes HR management processes more effective; Y1.4 Experience has been tried several times, showing: E-HRM increases my teaching productivity; Y1.5 Experience of trying several times: E-HRM is more fun for planning HR management; (4) Cognitive Absorption (Y2); Y2.1. Peak time, when I move on to E-HRM, I end up using as much time as I want; Y2.2 When I use E-HRM, I pay no attention to anything other than E-HRM; Y2.3 When using E-HRM, I am fascinated with what I do; Y2.4 I am happy to work with E-HRM; Y2.5 My curiosity is getting higher when interacting with E-HRM; (5) Satisfaction (Y3) Y3.1 I am very comfortable with the overall E-HRM that I use; Y3.2 I am very happy with the overall E-HRM that I use; Y3.3 I intend to continue using E-HRM rather than discontinuing its use; (6) Continuous Intention to Use E-HRM (Continuance Intention) (Y4); Y4.1 I intend to continue using E-HRM rather than discontinue use; Y4.2 My goal is to continue using E-HRM rather than using any alternative means; Y4.5 If I can, I would like to stop using my E-HRM. Based on the research concept model and the operationalization of each variable, the built research model is shown in the Figure 1.

Figure 1: Research Model and Research Variable Indicators

Furthermore, the validity test was carried out by item analysis. Each value obtained for a series of items was correlated with the total value of all items of a variable. The correlation test used Product Moment Correlation, with the minimum requirement for an item that is considered valid is a value of r ≥ 0:30 (Raymaekers & Rousseeuw, 2019). The Reliability test is intended to determine the extent to which the measurement results remain consistent if measurements are made twice or more on the same statement using the same measuring instrument. The reliability test in this study used the Cronbach α technique, in which an instrument can be said to be reliable if there is a reliability coefficient of Cronbach ≥ 0.6 (Sekaran, 2003).

Table 1 in the third column shows the value of the Pearson correlation coefficient, which is the calculated value of the validity test results for each question on each variable. It appears that the Subjective Norms (X1) are as many as 5 question items; each value of the Pearson correlation coefficient shows a value greater than r table = 0.3246. Thus the five questions on the variable (X1) are accepted as valid, and so for the other variables. Meanwhile, the reliability of the instrument proven in each variable was analyzed using the Cronbach’s Alpha technique. The reliability of the instrument calculating all variables is shown in Table 2 below.

Table 1: Validity Test Results for Each Indicator

Table 2: Results of Instrument Reliability Investigation

The tool used was a SmartPLS as part of a multiple regression model. In path analysis, the association coefficients in the multiple regression model that are connected to each other are estimated at once. In SEM, path analysis is performed with various latent variables (LV), which are usually perceptual variables that cannot be measured directly. In SEM, each LV score is computed as a weighted average of a series of variables, usually referred to as manifest or indicator variables measured directly. Models can consist of a single multi-indicator combination and the LV indicator in SEM.

While many conditions for convergence exist for the calculation of LV scores in SEM, there are many approaches to SEM; the quantitative methods liter-ature often classifies SEM approaches into two main types: covariance and variant-based (Cheung & Chan, 2005; Ntiamoah et al., 2019; Alaloul et al., 2020). The latter is also known as the component-based approach or PLS for SEM (Wong, Lim, Lim, & Hong 2019), where PLS usually stands for “smallest box.” SEM-based PLS has several main advantages over Covarian-based SEM, as follows: (a) it always seems to produce solutions, even in complex models; (b) it does not require variables to meet the criteria for parametric analysis, such as multivariate normality and large sample size; and allows for parameter estimation in models with formative and moderating effects of LV. One of the disadvantages of PLS-based SEM is that it usually does not produce a suitable index, which is useful in assessing the overall fit between a model with multiple LVs and the dataset used in SEM analysis. In addition, the PLS-based SEM structural model involves reflexive and formative variables or indicators. As shown in the figure, structural equation modeling measures in the SmartPLS software were adopted from the previous study (Sarstedt & Cheah, 2013).

4. Results and Discussion

In this study, random sampling was selected from 1 predetermined agency, ESQ Jakarta research place. The number of samples taken was 100 permanent employees-HRM. In analyzing the data from the results of this study, the Structural Equation Modeling (SEM) test was used to prove the research. In the SEM test, there were 3 steps of analysis: (1) testing the relationship between indicators and latent / construct variables (outer model or measurement model); (2) testing the relationship between latent / construct variables (structural model), and (3) testing model fit. (1). Subjective Norm (X1) consists of 5 questions.

The data shows the distribution of sample data based on gender. Of the 100 selected respondents, 59 people were male (59%), and 42 people were (42%). Out of 100 respondents, the educations were Bachelor Degree, followed by Master Degree, Senior High School, Diploma 3, Diploma 1, educational level of Diploma 2 and empty. Based on the data on respondent age distribution, most of the respondents were at the youngest age interval between 29–34 years old (37%). The second-highest proportion was between 23–28 years old (35%), and the third-highest was 41–46 years old (20%). At least eight people were between 40–46 years old (8%).

Following the research concept framework image, and the research hypothesis image, the number of observed variables or the number of latent/construct variables was six. The total indicators were 26 questions (26 manifest variables). The 6 constructs were: (1) Subjective Norms (X1) consists of 5 questions; (2) Perceived Behavioral Control (X2) consists of 3 questions; (3) Perceived Innovativeness (Y1) consists of 7 questions; (4) Cognitive Absorption (Y2) consists of 5 questions; (5) Satisfaction (Y3) consists of 3 questions; (6) Continuous Intention to Use E-HRM (Continuance Intention) (Y4) consists of 3 questions. Analysis of the relationship between latent/construct variables in the SEM model is nothing but testing the structural model in path analysis. The inner model proved the proposed research hypotheses. In this study, the analysis used Bootstrapping using SmartPLS software. The results of running calculations with Bootstrapping produced the output in Figure 2 below, and the results of several stages of analysis are described below.

Figure 2: The Output Results the Initial Inner Model Show Path Coefficient Value

Figure 2 above explains the results of the Bootstrapping calculation to test the inner model that illustrates the research hypotheses in the SEM model simultaneously. The value of direct effects is the path coefficients on the structural model and appears in the SEM image. The direct effects value ranges from −1 to +1. If the value of positive direct effects is close to +1, it is said to have a strong and parallel direct effect, but if the value of negative direct effects is close to −1, it is said to have a strong and opposite direct effect (Hair et al., 2010), The results of the path analysis which explain the direct effects of the constructs to the other constructs are as follows:

[1] Subjective Norms (X1) have a significant effect on Perceived Innovativeness(Y1) with a path coefficient of py1x1 = 0.156 and p-value = 0.031. Thus the hypothesis is proven.

[2] Subjective norms (X1) have a significant effect on the Continuous Intention to Use E-HRM (Y4) with a path coefficient of py5x1 = 0.248 and p-value = 0.004. Consequently, the hypothesis is proven.

[3] Perceived Behavioral Control (X2) has a significant effect on Perceived Innovativeness (Y1) with a path coefficient of py1x3 = 0.156 and p-value = 0.168. Thus the hypothesis is not proven.

[4] Perceived Behavioral Control (X2) has a significant effect on Cognitive Absorption (Y2) with a path coefficient of py2x3 = 0.524 and p-value = 0.000. As a result, the hypothesis is proven.

[5] Perceived Behavioral Control (X2) has no significant effect on Continuous Intention to Use E-HRM Continue Using E-HRM) (Y4) with a path coefficient of py5x3 = 0.510 and p-value = 0.000. Thus the hypothesis is proven.

[6] Perceived Innovativeness (Y1) has a significant effect on Satisfaction (Y3) with the path coefficient of py4y1 = 0.334 and p-value = 0.002. Consequently, the hypothesis is proven.

[7] Perceived Innovativeness (Y1) has a significant effect on Continuous Intention to Use E-HR-Mintention (Y4) with a path coefficient of py5y1 = 0.485 and p-value = 0.000. Thus the hypothesis is proven.

[8] Cognitive Absorption (Y2) has a significant effect on Satisfaction (Y3) with the path coefficient, namely py3y2 = 0.494 and p-value = 0.000. As a result, the hypothesis is proven.

[9] Cognitive Absorption (Y2) has no significant effect on Continuous Intention to Use E-HRM (Y4) with a path coefficient of py4y2 = −0.040 and p-value = 0.708. Therefore, the hypothesis is not proven.

[10] Satisfaction (Y3) has a significant effect on the Continuous Intention to Use E-HRM (Y4) with a path coefficient of py4y3 = 0.197 and p-value = 0.136. Consequently, the hypothesis is not proven, even though there is a small positive effect.

In this study, 10 research hypotheses were proposed; seven of them were proven. One hypothesis had an insignificant positive effect, and two hypotheses were unproven. At the stage of testing the fit of the model or the suitability of the model, there are 5 types, among others by looking at the coefficient of determination (R square), f square, q square, the standardized root mean square residual (SRMR), and the NFI value. The initial stage of testing the fit of the model is to determine the coefficient of determination (R square). The value of R2 explains the percentage value of the effect of exogenous and/ or endogenous variables on other endogenous variables, and the value of R2 only exists for endogenous variables. Several exogenous or endogenous variables are said to have a significant effect on other endogenous variables if they show a value of R2 > 0.20 (Hair et al., 2010).

The results of calculating R2 are shown in Figure 3, which shows that Continuous Intention to Use E-HRM Y4) has the largest R² of 0.441. It can be said that the connected exogenous/endogenous variables affect. Furthermore, for the Satisfaction (Y3) R² (0.399) and the Perceived Innovativeness (Y1) R² of 0.224, it can be said that the connected endogenous variable has an influence. On the contrary, the Cognitive Absorption (Y2) only has R² of 0.189. This shows that the connected exogenous / endogenous variable has a weak effect. The variables that have not been studied still occupy a dominant position. This can be seen in Figure 4.

Figure 3: R2 Output Results (Model Fit)

Figure 4 shows that Continuous Intention to Use E-HRM (Y4) has the largest unexplored R² of 0.510. It means that the endogenous variables that have not been studied are still dominant. Furthermore, for the Satisfaction (Y3), the R² which has not been studied was (0.424), and for Perceived Innovativeness (Y1), the R² which has not been studied was 0.419. This means that the endogenous variable that has not been studied was still dominant. On the contrary, Cognitive Absorption (Y2) only has R² of 0.479, meaning that the endogenous variable which has not been studied yet is still greater than the one studied.

Figure 4: Model Match Output Results (2)

5. Conclusions

The results of the discussion show that variables whose indicator constructs need to be changed if used for further research are Perceived Innovativeness and Cognitive Absorption. Variables that can be relied on for further research without changing the indicator constructs is Satisfaction. Some relationships between variables that have a significant effect include: Subjective Norms (X1) have a positive effect on Perceived Innovativeness (Y1), Subjective Norms (X1) have a positive effect on NMM E-HRM (Y4), Perceived Behavioral Control (X2) has a positive effect on Perceived Innovativeness (Y1), Perceived Behavioral Control (X2) has a positive effect on Cognitive Absorption (Y2), Perceived Innovativeness (Y1) has a positive effect on Satisfaction (Y3), Perceived Innovativeness (Y1) has a positive effect on Continues Intention to Use E-HRM (Y4), Cognitive Absorption (Y2) has a positive effect on Satisfaction (Y3) and Satisfaction (Y3) has a (weak) positive effect on Continues Intention to Use E-HRM (Y4), Cognitive Absorption (Y2) has no effect on Continues Intention to Use E-HRM (Y4), Perceived Behavioral Control (X2) has no effect on Continues Intention to Use E-HRM (Y4).

The E-HRM system is seen by users as something that is needed and useful in supporting the HR management process but has not yet reached a satisfactory level for the user. It is necessary to review some of the models that have been developed by previous researchers because the form and quality of information technology are currently changing rapidly. If the E-HRM system that is being implemented has maximally used its capabilities, then, in that case, another factor that can be worked on to increase satisfaction is to increase the user’s HR capability by continuing to provide training so that the quality of HR management comparatively gets better.

Regarding the conclusions, the suggestions put forward are as follows: because the coefficient of determination of the Satisfaction variable only reaches R2 = 39.9%, to explore other factors that have the potential to provide the power of influence to continue using E-HRM is suggested. Considering that the E-HRM system that has been implemented is seen by users as something that is indeed useful in supporting the HR management process but has not yet reached a satisfactory level, the institution is recommended to continues to review the quality of the system for its sustainable maximized use as well as to increase the quality human resource users with various training.