Australasian Journal of Educational Technology
2012, 28(4), 580-598

Empirical validation of the importance of employees’ learning motivation for workplace e-learning in Taiwanese organisations
Hsiu-Ju Chen and Chia-Hung Kao
I-Shou University, Taiwan E-learning systems, adopted by organisations for employee training to enhance employees’ performance, are characterised by self-directed, autonomous learning. Learning motivation is then of importance in the design of e-learning practices in workplace. However, empirical study of the alignment of e-learning with individual learning needs and organisational goals is an area with limited research. This study intends to clarify the importance of learning motivation in employees’ e-learning systems use behaviour, based on the information systems success model. Data from one hundred and eighty-five employees who used e-learning systems in their work environment were gathered in Taiwan and analysed with PLS. Results showed that employees’ learning motivation, reflecting their learning needs and strengths, influenced perceived usefulness and satisfaction with e-learning, and their use of the systems, which enhanced their task performance. The results validated the importance of employees’ learning motivation and the need for alignment of employees’ learning needs and organisational goals in e-learning training. The clarification can help facilitate an organisation’s human capital management, and contributes to further advancement of the information systems success model.

Introduction
In a competitive environment, finding ways to use employee training and learning to help enterprises to adapt to changes the external environment is an important issue. With the prospect of cost-effective investment in e-learning training, many enterprises have adopted e-learning systems for employee training to assist in their human capital management in recent decades (e.g. Wang, Wang & Shee, 2007). Through use of elearning systems, employees can transfer what they have acquired from the training to their jobs and thereby increase their productivity (e.g. Chen, 2010). This helps employees’ renewals of knowledge and skills while also reducing knowledge gaps between what the organisations have and what they need in keeping competitiveness. However, e-learning in workplaces still remains a fragmented, complex, and challenging area (Wang, 2011). The impact of organisational training contexts on new entrants’ e-learning training has been verified, but e-learning in workplaces is still confronted with a highly complex set of factors, such as learners, activities, outcomes, etc. (Chen, 2012; Collin, 2006; Wang, Ran, Liao & Yang, 2010). The alignment of the learning with employees’ individual learning needs and organisational goals in a systemic way also lacks clarification in existing studies (Collin, 2006; Wang et al., 2010). E-learning provides learning more learner-centred than instructor-oriented, and workplace learning involves adult learning that emphasises employees’ rational

Chen and Kao

581

motivation in learning to meet their needs. Learning motivation which indicates human stimulated needs driving individuals to act to meet those needs, is then of importance in the design of e-learning practices for employees’ self-directed learning in work environments (Cross, 1982; Houle, 1979; Rubenson, 1991). Consequently, it raises an important issue clarifying the impact of learning motivation on employees’ elearning training. The information systems (IS) success model, which captures both the technological dimension and the human dimension of human use of information systems/services, provides a systematic theoretical foundation for investigating employees’ learning outcomes from e-learning systems (DeLone & McLean, 2003; Seddon, 1997; Wang et al., 2007; Wu & Wang, 2006). This study is therefore theory based on a model for empirical clarification of the impact of learning motivation in employees’ e-learning training. It provides a theoretical basis reflecting the impact of employees’ other perceived net benefits involved in e-learning system usage (Seddon, 1997; Seddon & Kiew, 1996). This study therefore adopts this model to clarify the importance of learning motivation in employees’ self-directed, autonomous e-learning training for productivity. The clarification seeks to improve understanding of employees’ elearning use in the changed training environment in workplaces, and also the alignment of the learning with employees’ learning needs and organisational goals in e-learning training.

Literature background
E-learning systems for employee training E-learning refers to learning experiences gained through use of information technology, and focuses on the broadest view of learning that goes beyond the traditional learning paradigms (Rosenberg, 2006; Tsai, Shih & Feng, 2008). It is also characterised by self-directed, autonomous learning, which refers to the process whereby learners systematically achieve learning goals by themselves (Markus & Wurf, 1987; Schunk & Zimmerman, 1994). Through e-learning systems, learners can study course contents in an independent manner. They can also decide when to study, the sequence of the content to study, and the amount of time to spend for selfeducation, without time and space barriers (e.g. Blake & Butcher-Green, 2009; Newton & Doonga, 2007). Recently, e-learning systems have been increasingly adopted by organisations for employee training for cost reduction reasons. In workplaces, employees are adults who have good self-concept, and usually they have a clear understanding of their learning needs and can learn independently (Eggen & Kauchak, 1994; Slavin, 1994). Employees are then expected to increase their performance through undertaking selfdirected, autonomous e-learning training. However, current development of elearning tends to focus on technical issues of design, and most e-learning applications may not perform well in motivating users to learn (Wang et al., 2010). Conflicts between an organisation’s aim to invest in e-learning systems for human capital management and employees’ needs for renewal of knowledge and skills may then arise if the alignment of the learning with individual learning needs and organisational goals is not achieved (Wang et al., 2010). Learning motivation has been shown to be important in learning behaviour and learning outcomes in a largely independent environment (e.g. Chen & Chih, 2011;

582

Australasian Journal of Educational Technology, 2012, 28(4)

Houle, 1979; Tempelaar, Gijselaers, Van de Loeff & Nijhuis, 2007). It refers to learners’ activating force to choose learning goals and perform in a way that will achieve these goals; it also indicates the extent to which their needs will be met by performing the activity (Cross, 1982; Rubenson, 1991). Task performance indicates the outcomes that organisations expect from employees when performing their tasks (Borman & Motowidlo, 1993; Kirkpatrick, 1994; Wang et al., 2010). Thus, to clarify the alignment of e-learning with employees’ learning needs and organisational goals in self-directed, autonomous e-learning training, this study is motivated to validate the impact of learning motivation upon employee use of e-learning systems in workplaces. The IS success model To indicate the success of information systems (IS), DeLone and McLean (1992) conducted a comprehensive review of IS success literature and proposed the IS success model. Ten years later, they proposed an updated IS success model recognising ecommerce environments, and the basic structure is similar to that of its original model (DeLone & McLean, 2003). However, to clarify the combined process and causal explanations of the IS success model of DeLone and McLean (1992), Seddon (1997) proposed another adaptation of the model. He took part of the model into the process and variance model: the partial behavioural model of IS use and the IS success model, and then linked the two models with a construct: the individual, organisation, and societal consequences of IS use. The model is shown in Figure 1.
Partial behavioural model of IS use
Expectations about the net benefits of future IS use IS use (a behaviour, not a success measure) Individual, organisation, and societal consequences of IS use (not evaluated as either good or bad) Observation, personal experience, and reports from others

1. Measures of information and system quality Feedback (partial basis for revised expectations)

2. General perceptual measures of net benefits of IS use

3.Other measures of net benefits IS use Net benefits

to:
System quality Perceived usefulness Individuals

Organisations Information quality User satisfaction

Society

IS (design) success model

Figure 1: The re-specified IS success model of Seddon (1997)

Chen and Kao

583

The adapted IS success model by Seddon (1997) explains the relationship between the process and variance models in the original IS success model, while still reflecting the essential values of the original model by DeLone and McLean (1992). Besides, it adds the dimension of perceived usefulness, reflecting users’ perceived instrumental value of information systems, and indicates the possibility of other constructs to improve the use of the system through perceived usefulness and user satisfaction (Seddon, 1997; Seddon & Kiew, 1996). This study is thus theoretically constructed based on the model for empirical clarification of the impact of learning motivation in employees’ elearning training. Characterising the importance of learning motivation in employees’ e-learning Cognitive psychologists elaborate that the occurrence of learning lies in changes of the inner psychological structure of learners who choose to give meaning to the learning experience; an environment only provides potential stimulus to encourage learning (Bandura, 1977). Thus, learners play a key role in learning. A learner’s inner psychological structure refers to a learner’s cognition of natural and social world and exists in the form of symbols. When new experiences change a learner’s inner psychological structure, learning occurs. In order to make learning occur, it is helpful to provide general principles in an appropriate context to match learners’ inner psychological structure and assist them in applying what they have learned to new problems (Bandura, 1977; Gredler, 1992; Griffin & Griffin, 1996; Roblyer, 2004). E-learning provides modern learning that is more learner-centred than lecturercentred (Lee, Yoon & Lee, 2009; Tsai et al., 2008). Employees thus have greater control over their learning/training process through e-learning. By using various forms of symbols, e-learning systems may provide simulation courses for employee training; they may also provide courses introducing general rules or specific knowledge for employees to learn. And through the use of these systems, trainees may give meaning to their learning experience; by so doing, they can change their inner psychological structures. Learning then occurs and assists trainees to apply what they have learned to their tasks (Gredler, 1992; Bandura, 1977; Griffin & Griffin, 1996; Roblyer, 2004). In workplaces, e-learning systems are expected to provide general principles to match employees’ inner psychological structure for training. Employees can then use the systems to learn independently and autonomously, and decide on when to use them to accomplish their learning goals. When they give meaning to the experiences, their inner psychological structure changes and learning occurs. Thus, if employees are willing to accept the systems and take responsibility for their training, the change in their inner psychological structure aids their training/learning transfer to work, as expected in employee training (Chen, 2010; Chen, 2012; Wexley & Latham, 1981). Motivation indicates human stimulated needs that drive individuals to act, in order to meet those needs (Cross, 1982; Houle, 1979; Rubenson, 1991). It also reflects an individual’s expected value for an activity. The higher the expected value is, the stronger the degree that needs can be met by performing that activity. In andragogy, learning motivation also drives an adult’s voluntary learning and active participation in learning/training activities (Houle, 1979; Tempelaar et al., 2007). It refers to an activating force leading individuals to choose learning/training goals and perform in a way that will achieve these goals. With expectations by reason of the learning value, the force, i.e. the strength of behaviour, is determined (Cross, 1982; Rubenson, 1991).

584

Australasian Journal of Educational Technology, 2012, 28(4)

In Seddon’s (1997) IS success model, it indicates the possible impact of other net benefits of individuals, organisations, and society on users’ perceived usefulness and user satisfaction, which motivate them to use information systems/services. The study by Seddon and Kiew (1996) also demonstrates empirically that user involvement, reflecting users’ opinions about the relevance of the system to their own goals, influences their perception of the system as useful and satisfying. Later, Fraser and Salter’s study (1995), being similar to that of Seddon and Kiew’s (1994), provides consistent results. User involvement refers to users’ subjective psychological state that reflects the importance or relevance of specific or general information systems to them. Learning motivation is also a subjective psychological state which indicates users’ expected value in performing learning activities, and drives adults’ voluntary learning and active participation in training activities (Boshier, 1978; Houle, 1979; Pintrich, 1987; Pintrich & DeGroot, 1990). As a result, if employees are willing to take responsibility for their own learning, their perceptions of the system’s usefulness and user satisfaction increase and motivate their further use of e-learning systems that aids their training/learning transfer to work (Chen, 2010; Chen, 2012; Wexley & Latham, 1981). Therefore, based on the IS success model, this study intends to empirically clarify the impact of learning motivation on the users’ perceived usefulness and user satisfaction in their e-learning system use.

Research hypotheses
The connection of the technological dimensions and human dimensions of e-learning systems: Information quality, system quality, perceived usefulness, and user satisfaction In general, e-learning systems provide content for user learning through various media. However, current developments tend to focus on technical issues of design regarding information technology (Wang et al., 2010). From the socio-technical viewpoint, both the technological dimensions (i.e., system quality and information quality) and the human dimensions (such as perceptions of usefulness and user satisfaction) should be captured in the IS success model (Wu & Wang, 2006). Hence, based on the re-specified IS success model (Seddon, 1997; Wu & Wang, 2006), the following hypotheses are proposed. H1: H1a: H1b: H1c: H1d: Employees’ perception of technological quality is significantly associated with their perceived system use motivation of e-learning systems. Employees’ perception of information quality is significantly associated with their perceived usefulness of e-learning systems. Employees’ perception of information quality is significantly associated with their user satisfaction with e-learning systems. Employees’ perception of system quality is significantly associated with their perceived usefulness of e-learning systems. Employees’ perception of system quality is significantly associated with their satisfaction with e-learning systems.

Perceived usefulness and user satisfaction in motivating e-learning system use Through e-learning systems that implement and support training, employees gain knowledge, skills and a new attitude regarding their jobs. Without understanding the benefits (i.e. perceived usefulness and user satisfaction), employees may not use the system. They could be kept from obtaining important information or knowledge, and

Chen and Kao

585

be left with no clear direction on how to move on in their jobs. In the IS success model, both usefulness and user satisfaction are shown to be important perceived benefits that motivate users’ system use (Chen, 2012; Wang et al., 2007; Wu & Wang, 2006). Furthermore, perceived usefulness, indicating users’ extrinsic motivator for instrumental value in using computer technology, impacts upon user satisfaction (Davis, Bagozzi & Warshaw, 1992; Venkatesh, 2000). Therefore, hypotheses about perceived usefulness, user satisfaction, and system use are proposed as follows. H2: Employees’ perceived system use motivation and use of e-learning systems are significantly interrelated. H2a: Employees’ perceived usefulness is significantly associated with their use of elearning systems. H2b: Employees’ user satisfaction is significantly associated with their use of elearning systems. H2c: Employees’ perceived usefulness is significantly associated with their user satisfaction of e-learning systems. Perceived net benefits aroused by e-learning system use Literature has shown that employees who are well trained also have good job performance (Whitfield, 2000). Although the evaluation of employees’ training outcomes, or the means by which the effect (or value) of trainees’ training on the business or environment is measured, such as through increased sales and improved productivity, is important, the objective measurement of net benefits of information systems, such as cost reduction or increased sales, may be lacking because of environmental intervention and intangible system impact (Kirkpatrick, 1996; Wang et al., 2010; Wu & Wang, 2006). Therefore, users’ perceptions are commonly adopted as a measurement of the beneficial consequences from system use, in empirical studies based on the IS success model (Chen, 2010, 2012; Wang et al., 2007; Wang et al., 2010; Wu & Wang, 2006). Task performance indicates the outcomes that organisations expect from employees when performing their tasks (Borman & Motowidlo, 1993; Kirkpatrick, 1994; Wang et al., 2010). Employees’ task performance is then not only adequate to be measured as an outcome of employees’ use of e-learning systems but also an adequate goal that organisations quest for in employees’ e-learning training. Therefore, the following hypothesis is proposed. H3: Employees’ use of e-learning systems is significantly associated with their task performance.

The impact of learning motivation Finally, Seddon’s (1997) IS success model indicates the possible impact of other net benefits on users’ perceived usefulness and user satisfaction, which motivate them to use the system. If employees are willing to take responsibility for their learning, their perceptions of e-learning system’s usefulness and user satisfaction increase and motivate their further use of the system. Therefore, based on the IS success model, this study intends to attest to the impact of employees’ learning motivation on their perceived usefulness and user satisfaction with e-learning systems.

586

Australasian Journal of Educational Technology, 2012, 28(4)

In addition, learning motivation drives adults’ voluntary participation in training activities (Boshier, 1978; Houle, 1979). It refers to an activating force leading individuals to perform learning activities and thus can affect users’ use of e-learning systems directly (Pintrich, 1987; Pintrich & DeGroot, 1990). Therefore, this study also clarifies the impact of learning motivation on systems usage. The following hypotheses are proposed. H4: Employees’ learning motivation improves their perceived system use motivation of e-learning systems. H4a: Employees’ learning motivation improves their perceived usefulness of elearning systems. H4b: Employees’ learning motivation improves their user satisfaction of e-learning systems. H5: Employees’ learning motivation is positively associated with their use of elearning systems. The conceptual research model for this study is shown in Fig. 2.
Learning motivation H4a Information quality H1b H1c System quality H1d User satisfaction H2b H1a Perceived usefulness H4b H2a H5 Individual beneficial consequence of IS use H3 Task performance

H2c

Use

Figure 2: The conceptual research model

Research method
Measurements In order to validate the impact of employees’ learning motivation in their use of elearning in workplaces, this study adopted the survey method. The construct measurements included: learning motivation, information quality, system quality, perceived usefulness, user satisfaction, system use, and task performance. All construct measures and operational definitions were based on existing instruments and literatures. All measurements were self-reported by employees, using a Likert seven-point scale ranging from (1) strongly disagree to (7) strongly agree. For measuring task performance, the measurement of users’ perception was also adopted for the lack of objective measurement of net benefits from the information systems (Wu & Wang, 2006). Among different assessments of performance measures, self evaluation by the employees was still acceptable and a valuable reference (Campbell, Dunnette, Lawler & Weick, 1970; Shore & Thornton, 1986), and was therefore adopted

Chen and Kao

587

in this study. The operational definitions for each construct are described below, and pertinent questions are listed in Appendix A. Information quality and system quality The general definitions and measures of both information and system qualities in traditional IS studies were adopted in this study. Information quality indicates the users’ perception of completeness, adequateness, and clarity of the information and format in the output of e-learning systems (Wu & Wang, 2006; Rai, Lang & Welker, 2002). System quality was defined as users’ perception of easy operation, responsiveness and stability in using the systems (Chen, 2010; Wang et al., 2007). Perceived usefulness, user satisfaction, and system use The measures of the constructs of perceived usefulness, user satisfaction, and system use were mainly based on the studies of Seddon and Kiew (1996), Rai, Lang and Welker (2002), and Wu and Wang (2006). The operational definition of perceived usefulness refers to users’ perception of using e-learning systems to improve their job performance. User satisfaction was defined as users’ overall satisfaction with elearning systems, and system use indicates users’ use and dependence on e-learning systems. Self-evaluated task performance With the diversity of courses provided by e-learning systems, this study adopted task performance to measure the benefits of e-learning systems use. Task performance is defined as the outcomes that organisations expect from employees when performing their tasks (Borman & Motowidlo, 1993; Kirkpatrick, 1994). The measurement was based on Borman & Motowidlo (1993), and was self-evaluated by employees (Shore & Thornton, 1986; Campbell et al., 1970). Learning motivation Many studies have clarified the origins of learning motivation and also empirically verified learning motivation of adults (e.g., Boshier, 1978; Cross, 1982; Kao, 2009; Pintrich & Schunk, 2002; Rubenson, 1979). While most measurements measure adult motivation for continuous education, e-learning systems adopted by organisations are for employee training. Therefore, in exploring the impact of learning motivation in employees’ use of e-learning systems, this study adopted the three constructs: professional advancement, learning achievement and influence of others, based on definitions from the literature. The measurements adopted in this study were mainly based on studies by Boshier (1978), Morstain and Smart (1974), and Cross (1982). The questionnaire was established prior to a formal survey. The pre-study results from twenty respondents indicated that the measures could be easily understood. The questionnaire was then used in the formal survey. The questions are shown in Appendix A.

Data collection
To gather data, questionnaires were issued to organisations that had publicly announced their investment in e-learning systems, or had applied governmental subsidies to adopt e-learning systems in Taiwan, such as banks, governments, hightech manufacturers, etc, in a period over two months. Because some respondents indicated that they did not use e-learning systems at work, only those organisations whose employees indicated that they actually used e-learning systems in their jobs

588

Australasian Journal of Educational Technology, 2012, 28(4)

were targeted. In the questionnaire, respondents were asked to answer questions based on their most frequently used e-learning system. In order to attract valid respondents, this study provided valuable coupons for drawing lots. Of about three hundred questionnaires issued, a total of one hundred and eighty-five were validated. Descriptive statistics are given in Table 1. Table 1: Descriptive statistics (N=185)
Description Male Female Years of computer Less than 5 years use 6-10 years Above 10 years Education High school College/ university Graduate school (above) Organisation Below 100 employee numbers 100-499 500-1999 2000 or more Age 20-30 21-40 41-50 Above 50 Job category Administration/ sales Technical/ engineering Organisation Traditional manufacturing industry High-tech manufacturing Financial services Government General services Others Total number Gender Frequency 106 79 21 62 102 20 117 48 30 30 16 109 50 48 42 45 94 91 39 18 72 23 28 5 185 Percent 57.3% 42.7% 11.4% 33.5% 55.1% 10.8% 63.2% 25.9% 16.2% 16.2% 8.7% 58.9% 27.0% 25.9% 22.7% 24.3% 50.8% 49.2% 21.1% 9.7% 38.9% 12.4% 15.2% 2.7% 100%

The average level of use of e-learning systems by respondents was about five (5.02) on a 7-point Likert scale (Table 2). The results indicate that these respondents had experience in using e-learning systems in workplaces and were valid respondents for this study. Most of the respondents indicated that they used e-learning systems providing courses on job information (63.2%); others used e-learning systems providing courses about job technology (23.8%), a general introduction to the job (9.2%), or language (3.8%). Data on the purposes for using e-learning systems is presented in Table 3. Table 2: Respondents’ use of e-learning systems (7-point Likert scale; N=185)
Response 1 Totally disagree 2 Disagree 3 Slightly disagree 4 5 Slightly agree 6 Agree 7 Totally agree Frequency 0 6 9 45 58 50 17 Percent 0.0% 3.2% 4.9% 24.3% 31.4% 27.0% 9.2%

Chen and Kao

589

Table 3: Respondents’ purposes for using e-learning systems (N=185)
Purpose Job knowledge Job general introduction Job technology Language Frequency 117 17 44 7 Percent 63.2% 9.2% 23.8% 3.8%

Data analysis
Having a limited sample size, PLS (partial least square) was adopted for data analysis as it was less demanding on sample size (Beaton, Lings & Gudergan, 2008; Hsieh, Rai & Keil, 2008). PLS is a commonly accepted data analysis method. It adopts the bootstrapping technique for re-sampling, and the partial least square method for coefficient estimation (Chin, 1998; Gefen, Straub & Boudreau, 2000). The sample size compared to the largest path number in research model should be at least five to ten times larger (Chin, 1997; Majchrzak, Malhotra & John, 2005). The sample size of this study was one hundred and eighty-five and the largest path number of the model was eleven. Therefore, it was adequate to adopt PLS for data analysis. This study adopted SmartPLS 2.0 for data analysis (Ringle, 2005). The measurement and structural models were both evaluated, and each construct was modeled to be reflective in data analysis. Convergent validity and discriminate validity Convergent validity and discriminate validity were first provided to verify construct measures (Komiak & Benbasat, 2006). Cronbach’s alpha and the underlying factor structure were also justified. Convergent validity Convergent validity refers to the consistency with which multiple items measure the same construct. Unidimensionality, the average variance extracted (AVE), and the composite reliability (CR) are adequate indicators in understanding convergent validity of measurements (Bagozzi & Yi, 1988; Steenkamp & Van Trijp, 1991). They, as well as Cronbach’s alpha, were provided in this study. About unidimensionality, factor loading (>0.5) and t-value (>1.96) of items were both required. The results of factor loading, as well as AVE, CR and Cronbach’s alpha, are given in Table 4. The results showed that all the constructs had AVE values higher than 0.5, and all CR values were higher than 0.7 (Chin, 1998; Fornell & Larcker, 1981; Hair, Babin, Money & Samouel, 2003). All question items had acceptable loadings (>0.5) and t-value (>1.96). The results showed the commonly acceptable convergent validity of the measurements. Besides, all Cronbach’s alpha values were also higher than 0.7, and showed the reliability of all measurements. Discriminant validity In measuring discriminant validity, average variance extracted (AVE) and crossloading could be adopted to understand discriminant validity (Kerlinger & Lee, 2000). The correlation between different constructs should be lower than the square root of the variance extracted from the individual construct (Chin, 1998; Fornell & Larcker, 1981). In addition, the factor loadings belonging to the same construct should be

590

Australasian Journal of Educational Technology, 2012, 28(4)

higher than those of different constructs (Chin, 1998). The results of the AVEs are shown in Table 5. They showed that the square roots of the AVEs of all constructs were higher than their correlation coefficients with other constructs. The results of factor loadings are shown in Appendix B. They showed that each item loaded higher on its principal construct than on other constructs. Overall, the results suggested good measurement properties. Table 4: Average variance extracted (AVE), composite reliability (CR), Cronbach’s alpha, and factor loading/weight of construct measurement
Constructs Information IQ1 quality (IQ) IQ2 IQ3 IQ4 IQ5 System SQ1 quality (SQ) SQ2 SQ3 SQ4 Perceived PU1 usefulness PU2 (PU) PU3 User satisf- US1 action (US) US2 Profession PA1 advancePA2 ment (PA) PA3 Learning LA1 achieveLA2 ment (LA) LA3 Influence of IO1 others (IO) IO2 IO3 Use (USE) USE1 USE2 USE3 Task perfTP1 ormance TP2 (TP) TP3 TP4 AVE CR alpha IQ SQ PU US PA LA IO USE TP t-value 0.66 0.91 0.87 0.82 28.48 0.84 27.20 0.84 34.95 0.84 38.56 0.72 18.19 0.72 0.91 0.87 0.87 36.63 0.85 34.76 0.85 41.12 0.83 27.92 0.80 0.92 0.88 0.88 28.18 0.87 40.97 0.93 75.71 0.86 0.92 0.84 0.93 92.91 0.92 56.34 0.79 0.92 0.87 0.89 36.87 0.89 41.71 0.88 40.67 0.85 0.94 0.91 0.93 76.00 0.92 60.44 0.91 54.63 0.71 0.88 0.79 0.75 17.86 0.86 29.05 0.91 53.83 0.86 0.95 0.92 0.93 72.64 0.94 93.21 0.92 75.99 0.71 0.91 0.87 0.75 14.90 0.88 33.16 0.83 26.43 0.92 71.93

Table 5: Construct correlations and square root of average variance extracted (AVE)
Const. IQ SQ PU US PA LA IO USE IQ 0.81 a SQ 0.66 0.85 a PU 0.65 0.59 0.89 a US 0.65 0.64 0.69 0.93 a PA 0.50 0.45 0.53 0.45 0.89 a LA 0.52 0.48 0.60 0.59 0.57 0.92 a IO 0.41 0.45 0.56 0.50 0.59 0.61 0.84 a USE 0.53 0.56 0.67 0.60 0.47 0.56 0.57 0.93 a TP 0.58 0.55 0.69 0.58 0.65 0.65 0.60 0.64 a: Indicates the square root of average variance extracted (AVE) of the construct. TP

0.85 a

Chen and Kao

591

Path analysis With adequate convergent and discriminant validity, the hypotheses were then empirically tested. However, learning motivation was composed of the sub-constructs of professional advancement, learning achievement and the influence of others; it was thus measured by the repeating indicators from the first-order constructs (Diamantopoulos, Riefler & Roth, 2008; Wiley, 2005). The results of the SmartPLS analysis are shown in Figure 3. The results indicated that excluding H1c, all other hypotheses were supported. The results showed that employees’ use of e-learning systems had a significantly positive association with their self-evaluation task performance; thus, H3 was accepted. The results indicated that employees gained beneficial consequences from their use of the systems in performing tasks. Besides, perceived usefulness and user satisfaction were shown to have significant influence on system usage, and perceived usefulness had significant impact on user satisfaction. Therefore, H2a, H2b, and H2c were accepted. Furthermore, learning motivation was shown to have a significant influence on not only perceived usefulness and user satisfaction, but also system usage. Therefore, H4a, H4b, and H5 were accepted. Finally, perceived usefulness was shown to be influenced by information quality, but not by system quality, and user satisfaction was significantly influenced by both information and system qualities. Thus, H1a, H1b, and H1d were accepted, while H1c was rejected. The results empirically validated the impact of other supposed net benefits to individuals, organisations, or society in the IS success model. They showed employees’ learning motivation, reflecting their learning needs strengths in acting to meet the needs, increased their motivation for e-learning systems use, i.e. perceived usefulness and user satisfaction. Use of the systems enhanced their task performance, something sought by organisations. The validation indicated the alignment of e-learning systems with organisational goals, indicated by employees’ task performance, and employees’ individual learning needs, indicated by the strengths of their learning motivation.
Learning motivation Information quality Perceived usefulness 56.66% 0.33*** User satisfaction 59.89% 0.39*** 0.16* 0.35**

0.34***

*

0.27***

Individual beneficial consequence of IS use
Task performance 40.90%

0.19* 0.15 System quality 0.24**

Use 52.27% 0.20*

0.64**

* |t|>1.96, p2.58, p3.29, p