Effect of management support, training, and user involvement on system usage and satisfaction In Kuwait

Rouibah K., and Hamdy H.I., and Al-Enezi M. Z., (2009). Effect of management support, training, and user involvement on system usage and satisfaction in Kuwait. Industrial Management & Data System, Vol. 103, No 9; pp. 338-356 Effect of management support, training, and user involvement on system usage and satisfaction In Kuwait Kamel Rouibah * College of Business Administration Kuwait University, Hosny. I. Hamdy College of Business Administration, Kuwait University Majed Z. Al-Enezi Kuwait Fund For Arabic Development * Correspondence author: krouibah@cba.edu.kw Accepted to be published in INDUSTRIAL MANAGEMENT AND DATA SYSTEMS Purpose Design/methodology/approa ch Findings Research limitations/implications Practical implications What is original/value of paper The purpose of this paper is to investigate organizational factors and human motivations affecting IS/IT usage and user s satisfaction in an Arabic country The study develops a research model that links three organizational factors (top management support, availability of training, and user's involvement) to IS/IT usage and enduser's satisfaction via the mediation of TAM beliefs (usefulness and ease of use) by 382 IS users. The model was examined through the LISREL. The results indicate that IS/IT usage and user satisfaction is largely influenced by perceived usefulness. Among the organizational factors, top management support was found to have the strongest effect on IS/IT usage and user s satisfaction, followed by availability of training and user involvement. This study focuses on perceptions of respondents from the public organizations who have high experience with IS/IT. Theses perceptions are based on a one-time survey. For better validity a longitudinal study with additional variables is needed. This study is useful for practitioners who may learn how to increase and boost system usage and user's satisfaction in the Arab region. Contributes to highlights motives of IT/IS usage and user' satisfaction in the Arab world, and to educate researchers

Type of paper: Research paper about how IS/IT is perceived by end-users in Kuwait. Keywords Top management support, user training, user involvement, system usage, user's satisfaction, empirical study 1 INTRODUCTION The role of Information Systems and Information Technology (hereafter IS/IT) in providing businesses with a competitive edge has been the subject of much debate. However, it has been argued that IS/IT s utilization, not the IS/IT solution, is what provides the competitive advantage to businesses. While most previous research has been carried out in the technologically developed world, see literature review done by Legris et al., 2003, little is known about IS/IT acceptance in the Arab region. In fact, of the 124 research studies in IT adoption covered in three literature review papers that appeared in leading MIS journals between 1991 and 2004, no study took place in the Arab world. We define IS/IT acceptance in this paper as the potential user s predisposition toward personally using a specific system and being satisfied about it. In order to adequately evaluate acceptance, multiple constructs were considered. Usage and user satisfaction are commonly used to measure information success (DeLone and McLean, 1992; Briggs et al., 2008; Rouibah, 2009; Lee et al., 2009). User satisfaction was found to be the most widely used surrogate of computing success (Srite et al., 2007; Lee et al., 2009). Past studies on IS/IT acceptance advocate the existence of cultural and management differences between Western and Arab cultures (Rouibah, 2009) and other cultures (McCoy et al., 2007). Most of the findings on the relationship between IS/IT usage and its satisfaction are based on research in technologically developed countries (e.g. DeLone and McLean, 1992; Igbaria et al., 1997; Gelderman, 1998; Gahtani and King, 1999; Adamson and Shine, 2003; Lee et al., 2009), with the exception of a few studies in less developed countries (Khalil and Elkordy, 1999; Aladwani, 2002; Al-Gahtani, 2004; Rouibah, 2009). Previous studies have also found a wide variation in the predicted effects in various studies with different types of users, systems and cultures (Legris et al., 2003; McCoy et al., 2007). This study aims to study factors that influence IS/IT acceptance (determinant and consequences of use) in Kuwaiti public organizations. While there are differences between private and public organizations, so far much of the empirical research in the area of IS/IT and its impacts on organizations have occurred in the private sector of well developed countries. This study targets several benefits. First, among other Arabic countries, Kuwait has achieved the highest IS/IT utilization. Therefore, this study would orient decision makers on how to achieve sustainable benefits from IS/IT investment. A recent study has shown a substantial growth (9.4%) in IT investment in the Gulf region in the last three years and is expected to reach $7.9 billion by 2009 (Research Institute for Digital Economy, 2006). Second, although reasons for this investment vary such as e-government initiatives and promotion of e- commerce, but the overriding belief suggested that knowledgeable workers will be more productive by using IS/IT. Third, the study aims to outlines the conditions under which IS/IT usage and satisfaction could be increased to generate significant performance gains (Mahmood and Mann, 2000), and avoid failure (Anonymous 2008). Fourth, the complexity of IS/IT acceptance has became an important issue for academic researchers and practitioners over the world (Legris et al., 2003). Fifth, it is expected that this study would advance the knowledge about IS/IT acceptance in the Arab region. It is also expected 2

that it will contribute to educate IS/IT researchers about how IS/IT is perceived by end-users in a region where less is known about IS/IT acceptance. Section 2 discusses the relevant literature review on IS/IT acceptance, the foundation of the research model, and the current state of IS/IT in the Arab world. Section 3 introduces the proposed research model. Section 4 presents the research methodology. Section 4 describes the results and their implications while the last section, Section 5, summarizes the findings and points out future research directions. 2 BACKGROUND AND MOTIVATION 2.1 Related studies on IS/IT usage and satisfaction research outside the Arab world Over the last three decades, IS/IT usage and user's satisfaction has received a lot of attention by academics. Their studies can be grouped in the four categories. One research stream developed measurement for IS/IT usage and validation (e.g. Franz and Robey, 1986; Davis 1989; Moore and Benbasat, 1991), user satisfaction (Baroudi and Orlikowski, 1988; Doll and Torkzadeh, 1991) and new theories for satisfaction (Briggs et al., 2008). A second stream focused on the relationship between usage and user satisfaction but was not grounded in well-established theories (Gelderman, 1998; Khalil and ElKordi, 1999). A third stream focused on literature review (Au et al., 2002; Briggs et al., 2008). The fourth stream investigated the antecedents of user satisfaction using well known theories such as TAM and its variations (Davis, 1989), IS success model (DeLone and McLean, 1992), and applied it to different technologies, including MS office software (Igbaria et al., 1997; Aladwani, 2002; Al-Gahtani, 2004; Liu et al., 2006), banking systems (Adamson and Shine, 2003), instant messaging (Rouibah, 2009), application service provider (Kim and Kim, 2008), and open sources (Lee et al., 2009). With regard to the last category, the authors studied different factors that affect satisfaction; however, very few studies linked external factors to current usage and satisfaction. Among these studies, only Rouibah (2009) and Lee et al., (2009) studied the effect of external variables on user satisfaction via the mediation of beliefs and system usage. The link between usage and satisfaction, including the possibility of causality in either direction, has been examined by several studies (Al-Gahtani and King, 1999; Anandarajan et al., 2002; Al-Gahtani, 2004; Rouibah, 2009, Lee et al. 2009). Although satisfaction was strongly correlated with system usage, the mutual effect between constructs needs to be investigated since correlation does not imply causality (Hair et al., 1998). Rouibah (2009) suggested three models for the relationship between satisfaction and usage: 1) satisfaction and usage are not related; 2) satisfaction influences usage; and 3) usage influences satisfaction. The third model assumes that the more end-users use the system, the more they will be inclined to be satisfied. Rouibah (2009) found support for the third model, while Lee et al., (2009) found support for the second one. Thus, further studies are required to investigate the causality direction between the two constructs. While some past studies focused solely on the relationship between IS/IT usage and user satisfaction, others extended the relationship to other external factors that may have an impact on the two constructs. By investigating these extensions, several models were proposed. The technology acceptance model is among the most commonly used (Davis, 1989) because of its understandability and simplicity (Legris et al., 2003). Central to this model are two mediating factors: perceived ease of use and perceived usefulness, which have been reported to mediate the effect of several external factors on IS/IT usage and user satisfaction (Rouibah, 2009; Lee et al., 2009). 3

Previous studies have examined several independent variables that affect IS/IT acceptance including attitudinal variables (cognitive style, anxiety, and favorable attitude toward IS/IT) (Venkatesh and Davis, 2000; Venkatesh et al., 2003) and organizational factors (top management support, training, and user involvement) as determinants of IS/IT acceptance. However, attitudinal variables received more attention than organizational factors. A recent study (Kim and Kim, 2008) about determinants of application service provider satisfaction revealed that organizational factors (top management support, training, and user involvement) emerged to be of high importance for user satisfaction. The literature review on user satisfaction carried by Mahmood et al., (2000) revealed that user involvement, perceived usefulness, user experience (similar to availability of user training), and organizational support are the most determinants of user satisfaction. Regard the objective of this study, the literature review of past studies shows the following pitfalls: (i) Although training and organizational support received much attention (e.g. see Igbaria et al., 1995, 1997), user involvement was under studied (Kujala, 2003; Petter, 2008). (ii) Even though it is assumed that user involvement in the design of IS/IT would lead to successful IS outcomes, much of the empirical research so far has been unable to demonstrate its benefits, except few studies (Winston and Benjamin, 2000; Petter, 2008). (iii) Very few studies examined the simultaneous effect of organization factors (training, top management support and user involvement) on IS/IT acceptance. And the relative impact of user involvement in presence of other variables (top management support and training availability) remains unknown. Hence, how and how much the user involvement in IS/IT contributes to the usability and success of the final product needs further studies (Kujala, 2003). This will be a major concern in the current study. 2.2 Arab culture and related IS/IT studies in the Arab world Review of literature on IS/IT in the Arab world indicates that there are extensive empirical studies on IS/IT acceptance; see a literature review done by Rouibah (2009). Meanwhile, few studies have investigated the cause and consequences of IS usage. Khalil and Elkordy (1999) only explored the relationship between end-user satisfaction and system usage in Egypt; however, their study did not include external factors. Aladwani (2002) studied the effect of organizational actions (management advocacy and internal computing support) on end-user satisfaction via attitude in Kuwait. Al-Gahtani (2004) examined the impact of individual, technological and organizational factors on PC usage and end-user satisfaction in Saudi Arabia. While Al-Gahtani s (2004) study included organizational support and availability of training, user involvement was excluded. Although these studies have been useful in shedding light on IT/IS usage in the Arab world simultaneous effect of the different paths among studied variable was not a matter of their concerns. Arab culture has a strong patriarchal, tribal, and collectivist strand. Therefore, high-level management support is even more critical than it would be in Western culture where there is not such a strand. Such support is required to ensure user involvement and user training. 3 THE RESEARCH MODEL AND RESEARCH HYPOTHESIS 3.1 The Research Model The research model (depicted in Figure 1) provides the theoretical ground of the study. This model is motivated by the extension of TAM (Adamson and Shine, 2003), motivational model of PC usage (Igbaria et al., 1995; 1997), and information system success (Lee et al., 2009). The model investigates 4

the relationship between seven variables. These variables were included based on a review of past IS/IT literature in the West as well as the characteristics of Arab culture. The important factors are: top management support (e.g. Kim and Kim, 2008); availability of training (e.g. Mahmood et al., 2000; Gallivan et al., 2005); and user involvement (Mahmood et al., 2000; Petter, 2008). H1.d H2.d H1.c Top management support H1.a Perceived Usefulness H4.b H1.b H2.a H4.a Availability training/ resources H3.a H2.c H4.d Usage H5 User Satisfaction User involvement H2.b H3.b Perceived ease of use H4.c H4.e H3.c Figure 1 H3.d The Research Model 3.2 The Research Hypotheses A set of hypotheses is generated/proposed for each of the construct in the model as stated below and shown in the research model in Figure 1. The relationship behind each hypothesis is shown in Table 1 and is supported by previous literature. Relationships between variables H1a. Availability of organizational support will directly and indirectly affect PU H1b. Availability of organizational support will directly affect PEOU. H1c. Availability of organizational support will directly and indirectly affect IS usage H1d. Organizational support will directly and indirectl affect user s satisfaction Support from past literature Igbaria et al., (1997); Al-Gahtani and King (1999); Karahanna et al., (1999); Lewis et al., (2003) Karahanna et al., (1999); Igbaria et al., (1997); Anandarajan et al., (2002); Lewis et al., (2003) Karahanna et al., (1999); Straub et al., (2001); Anandarajan et al., (2002) [indirect effect]; Al- Gahtani (2004); Al-Gahtani and King (1999) [indirect effect via PU] Al-Gahtani and King (1999) [indirect effect]; Mahmood et al., (2000); Anandarajan et al., (2002) [indirect effect]; Aladwani (2002); Al-Gahtani (2004), Kim and Kim (2008) H2a. Availability of training will directly and indirectl David and Bostron (1993); Igbaria et al., (1995); affect PU Taylor and Todd (1995); Igbaria et al., (1997) H2b. Availability of training will directly affect PEOU Taylor and Todd (1995); Venkatesh (2000) H2c. Availability of training will directly and directly Igbaria et al., (1995); Igbaria et al., (1997) [indirect affect IS usage effect]; Al-Gahtani and King (1999); Al-Gahtani (2004); Gallivan et al., (2005) H2d. Availability of training will directly and Al-Gahtani (2004) indirectly affect user s satisfaction H3a. User involvement will directly and indirectly affect PU. H3b. User involvement will directly affect PEOU. Winston and Benjamin (2000) Frantz and Robey (1986); Jackson et al., (1997), Kujala (2003) 5

H3c. User involvement will directly and indirectly Igbaria et al., (1997); Jackson et al., (1997) [indirect affect IS usage effect] H3d. User involvement will directly and indirectly Khalil and ElKordy (1997); Mahmood et al., 2000; affect user s satisfaction Palanisamy and Sushil 2001; Winston and Benjamin 2008; Petter (2008), Kujala (2007) H4a. PU will directly affect IS usage. Igbaria et al., (1997); Al-Gahtani and King (1999) H4b. PU will directly and indirectly affect user s satisfaction Adamson and Shine (2003); Mahmood et al., (2000); Liu et al., (2006); Rouibah (2008); Petter (2008) H4c. PEOU will directly and indirectly affect IS usage Igbaria et al., (1995); Igbaria et al., (1997); Rouibah (2009) H4d. PEOU will directly affect PU. Davis (1989); Taylor and Todd (1995) H4e. PEOU will directly and indirectly affect user s Adamson and Shine (2003); Liu et al., (2006). satisfaction Rouibah (2009) [indirect via usage] H5. IS usage will directly affect user s satisfaction. Gelderman (1998); Khalil and ElKordy (1999); Al- Gahtani (2004); Rouibah (2009) 4 RESEARCH METHOD Table 1. Hypotheses to be tested and selected references with findings The survey instrument a structured questionnaire was developed to measure individuals perceptions to IS/IT acceptance and the core factors that might influence the end-user s reaction to IS/IT acceptance. In building this research model, the authors followed a process of fourth steps. First, existing literature on IS/IT acceptance was thoroughly reviewed. Second, a brainstorming session was held with five IS/IT managers to highlight the main factors that affect IS/IT usage and their satisfaction. Third, the main characteristics of Arab culture were also identified. Thus, a subset of variables was identified, and a comparison between the results of the brainstorming and this list of variables was made. The study selected three external variables (see Figure 1) that were deemed most appropriate and measurable. Fourth, this study developed a questionnaire instrument which was checked by two faculty members for validation, completeness, and readability. 4.1 Measures A questionnaire was designed reflecting the seven constructs of the conceptual model. Items for measuring each construct were selected from prior studies. Top Management Support (TMS) was assessed with nine items from Lee and Kim (1992), User Training (UT) with eight items from Igbaria et al., (1995), User Involvement(UI) with seven items and Actual Usage (AU) with four from Franz and Robey (1986), Perceived Usefulness (PU) with eight and Perceived Ease of Use (PEOU) with three from Davis (1989), and User Satisfaction (US) with thirteen items from Baroudi and Orlikowski (1988). Items in the questionnaire were measured on a five-point Likert scale ranging from (1) strongly disagree to (5) strongly agree. Questions related to characteristics of the respondents such as gender, age, experience, level of education and job title were also included in the questionnaire. 4.2 Sample The sample size was determined according to the following formula: N = [(Z 2 ) * P(1-P)]/ W 2, where W 2 refers to the precision of the study estimated at 0.09; P=73% refers to proportion of public organizations (39 organizations) with IS/IT departments compared to the total of public organizations in Kuwait (53); Z refers to the critical value from the normal table which corresponds to 95% 6

confidence; i.e. answers from the sample are 95% close to the reality; N is the total of needed respondents. Based on the formula, the required sample size is 374; taking into account that some respondents would not respond the sample size was increased by 10%, which lead to 420 questionnaires. These questionnaires were distributed to 39 public organizations in Kuwait which are clustered into three categories, according to the type of their budget. The first group was the one with a public budget that depends totally on government funds to conduct business. The sample included 18 organizations with a public budget out of 20. The second group was the one with a dependent budget (i.e. organizations that depend partially on government funding). The sample included 8 out of 10 existing organizations. The third group consisted of organizations with an independent budget and the sample included 8 out of 9 of the exiting organizations. To guarantee homogeneity in the sample, three criteria were used to select an organization from the identified organizations chosen: (a) information systems have been used more than two years in order to ensure stability in IS usage; (b) high IS usage occurs through the existence of the organization s network; and (c) each organization uses IS to support the business process of at least three business functions. 4.3 Data Collection A random pilot sample was drawn, before questionnaire distribution, to ensure reliability and validity of the questionnaire. 100 subjects in public organizations were first selected to test the questionnaire and the necessary analysis, editing and corrections were made accordingly. The organizations were contacted to ease the communication between the researchers and the potential respondents, and several visits were made to organizations to explain the purpose of the study, ensure the confidentiality of the data collected and get approval to distribute the questionnaires. A stratified sampling procedure was employed. Allocation to each strata was made according to their representation in the studied population (i.e. percentage of employees in Kuwaiti public sector). Therefore, among the 420 questionnaires, 216 were in organizations with a public budget, 102 within organizations with a dependent budget, and 102 within organizations with an independent budget. Among the 420 questionnaires, only 383 were completed and returned, yielding a response rate of 90.9%, where 51.3% represent respondents from organizations with a public budget, 25.9% from a dependent budget and 22.8% from an independent budget. Table 2 provides demographic data about the respondents. The overall reliability coefficient Cronbach s α was 93.6%, which reflects an acceptable consistency measure of the entire scale of the questionnaire (see for details Hair, et al., 1998). Table 2. Demographic data Public budget Independent Dependent Total Number % Number % Number % Number % Sample size 196 51.3 99 25.9 87 22.8 382 100 Gender Male 100 26.2 60 15.7 53 13.9 213 55.8 Female 96 25.1 39 10.2 34 8.95 169 44.2 Education leve Less than Bachelor 79 20.7 27 7.1 24 6.3 130 34 Bachelor 102 26.7 58 15.2 59 15.4 219 57.3 More than Bachelor 5 3.9 14 3.8 4 1 33 8.6 Position Strategic level 12 3.1 3 0.8 5 1.3 20 5.2 Management level 88 23 35 9.2 35 9.2 108 41.4 Operational level 96 25.1 61 16 47 12.3 204 53.4 7

4.4 Data analysis Structural Equation Model (SEM) technique was applied using LISEREL through the following steps: (i) assess and fix the model before SEM is applied; (ii) modify and create the best measurement model; and (iii) analyze and test the model paths. Before applying LISREL, the authors assessed the convergent and discriminant validity of the model, which was achieved through SPSS factor analysis. As shown in Table 3, factor analysis resulted in seven factors that accounted for 70% of the variance, and, guided by Hair et al., (1998), each of the factors met the base criteria for retention as follows: Items defining the various factors all had communalities greater than 0.50; Extracted factors accounted for greater than 50% of the variance in their sets of items; All factors had Eigen-values greater than 1.0; All item loadings were greater than 0.50; and Of the 52 items in the original list of items, twenty were deleted because they don't satisfy the above criteria. Table 3 shows the items retained and their factor loading. 4.5 Convergent and Discriminant Validity Testing the model included estimation of the convergent and discriminant validity. Convergent Validity is measured by Composite Reliability and Average Variance Extracted (AVE). Composite Reliability is the extent to which a variable or a set of variables is consistent in what it is intended to measure. Hair et al., (1998) recommended 0.50 and above as evidence of composite reliability. Table 4 shows that the composite reliability of all constructs varying from 0.768 to 0.942 and are, therefore, within the range of the recommendations. Average Variance Extracted (AVE) is a measure of reliability of constructs (Hair et al., 1998). It reflects the overall amount of variance accounted for by the original variables (indicators). A high AVE reflects a good representation of the original variable of the corresponding latent constructs. As a guideline, an AVE above or equal to.50 is considered acceptable (Hair et al., 1998) and all AVEs in Table 4 range from 0.521 to 0.871 and meet the recommendation. Discriminant validity is applied to assess the inter-correlation between the measures of the model. If the items associated with a measure correlate more highly with each other than with items associated with other measures in the model, the measure is determined to have adequate discriminant validity. The shared variance (r 2 ) among variables, obtained by squaring the correlations (r) in Table 4, are less than values of AVE which ensures a discriminant validity of the model. 4.6 Assessing the Goodness of Fit of the structural model Once the convergent and discriminant validity of the model were achieved, the model was tested for its goodness of fit. The literature suggests that for a good model fit, chi-square/degree of freedom (χ 2 /DF) should be less than 3.0, goodness-of-fit (GFI), normed fit index (NFI), non-normed fit index (NNFI) and comparative fit index (CFI) should be greater than 0.90, adjusted goodness of fit index (AGFI) should be greater than 0.80, and Root Mean Square Residual (RMSR) should be less than 0.10 8

(Hair et al., 1998). All of these indices were used as measures for goodness of fit. The data was fitted to several competing models, and measures of goodness of fit for the final model were obtained as follows: χ 2 /DF = 3.5, NFI=0.95, NNFI=0.95, CFI=0.96, RMSR=0.05, GFI=0.81 and AGFI=0.78. As can be seen, GFI and AGFI are less than the recommendations. However, they are still valid and reflect a reasonable accepted model (Hair et al., 1998). Table 3. Composite Reliability, Factor Loading and Extracted Variance Constructs / Indicator (appendix 1) Number Composite Factor of items deleted Reliability 1 Loading Variance Extracted 2 Training (TR) 4 0.887 0.600 TR1: Availability of internal training 0.685 TR2: Clarity of end-users role and objectives before t 0.696 TR3: Availability of adequate course material during t 0.738 TR4: Availability of IT support after training 0.693 TR5: Training programs precede effective IS usage 0.816 Top Management Support (TMS) 1 0.942 0.671 TMS1: TMS is much interested in IS usage rate 0.721 TMS2: TMS frequently mentions the various problems and matters related to IS development and operation 0.713 TMS3: TMS Makes and effort to provide stable funding for the system development and operation activities 0.812 TMS4: TMS tries to take part in deciding in what order the ISs should be implemented 0.771 TMS5: TMS tries to encourage the user departments to use IS 0.730 TMS6: TMS is favorable to computerization of tasks 0.752 TMS7: TMS is much concerned with the performance of IS operation 0.740 TMS8: TMS persists to have latest technology 0.733 User Involvement (UI) 5 0.931 0.871 UI1: End-users guide and direct the process of specifying and clarifying the input requirements and details for this system 0.907 UI2 End-users guide and direct the process of specifying and clarifying the output requirements and details for this system 0.884 Perceived Ease of Use (PEOU) 1 0.830 0.709 PEOU1: I find system flexible to interact with 0.650 PEOU2: I find system easy to use 0.701 Perceived Usefulness (PU) 2 0.910 0.614 PU1: Using the system improves my performance PU2: Using the system is useful in my job PU3: Using IS increases my productivity PU4: Using IS would enhance my effectiveness 0.619 0.687 0.750 0.774 9

PU5: IS enables me to accomplish tasks more quickly 0.704 PU6: IS supports critical aspects of my job 0.671 Actual Usage (AU) 1 0.768 0.521 AU1: Extend to which end-user actually use the IS compare to his original expectations 0.771 AU2: Extent to which the IS assist end-user to perform the job better 0.741 AU3 : Extend the end-user uses the reports or output provided by the IS 0.672 User Satisfaction (US) 7 0.880 0.711 US1: Completeness of the output information 0.748 US2: Reliability of output information 0.795 US3: Accuracy of output information 0.818 US6: Degree of training provided by development staff to users 0.721 US8: Users' feelings of participation 0.778 US10: Relevancy of output information to end-users 0.832 1 Composite reliability = sum of standardized loading) 2 / [(Sum of standardized loading) 2 + Sum of indicator measu ( λ i ) 2 / [( λ i ) 2 + δ i ] 2 Average Variance Extracted (AVE) = (sum of squared standardized loadings) / (sum of squared standardized load Sum of indictor measurement error) = ( λ 2 i ) / ( λ 2 i + δ i ) Table 4. Correlation structure between constructs [ a Composite reliability, b AVE, c Correlation, d Square correlation] 1 2 3 4 5 6 7 0.942 1-Training 0.887a 0.600 b 2- Top management support 0.65 c 0.42 d 0.671 3-User involvement 0.35 0.12 0.36 0.13 4-PEOU 0.72 0.72 0.51 0.52 5-PU 0.53 0.57 0.28 0.35 6-Usage 0.32 0.34 0.10 0.12 7-Users Satisfaction 0.39 0.15 0.41 0.17 0.931 0.871 0.44 0.19 0.48 0.23 0.28 0.07 0.32 0.10 0.830 0.709 0.52 0.27 0.31 0.10 0.43 0.18 0.910 0.614 0.62 0.38 0.61 0.37 0.768 0.521 0.55 0.30 0.880 0.711 4.7 RESULTS Although the analysis revealed significant positive associations between different latent constructs (see Table 4), SEM was employed to explore the causal interrelationship between variables of the model. The results of the multivariate test of the structural model are presented in Table 5 and Figure 2. The path coefficients (γ from external variable to latent and β from latent to latent) were examined. Prediction of perceived usefulness and perceived ease of use: Top management support, availability of training, and user involvement have a direct effect on beliefs (PEOU and PU); therefore, H1b, H2b, 10

and H3b are supported. PEOU has no effect on PU suggesting rejecting of Hd4. However, the three external stimuli don t exert any indirect effects on PU via PEOU, since PEOU has no effect on PU. Thus, H1a, H2a, and H3a are only partially supported. Results also indicate that among the three external stimuli, top management support is the strongest determinant of PU followed by user involvement. On the other hand, availability of training programs is the strongest determinant of PEOU followed by top management support. Prediction of current usage: Top management support, availability of training, and user involvement have no direct effect on IS usage; instead, they exert an indirect effect on current usage through PU. This result provided partial support for H1c, H2c, and H3c. The results also reveal that PU (H4a) has a direct effect on IS usage, but PEOU (H4c) has no effect on IS usage. It implied that only PU had a mediation between the three external stimuli and current usage. Among antecedents, PU has the strongest effect on current usage, followed by top management support. Prediction of user satisfaction: PU had direct and indirect significant influence on user' satisfaction. IS usage had also direct effect on user' satisfaction. This result provided support for H4b and H5. PEOU had a direct influence on satisfaction, but the indirect influence was not significant. Thus, H4 is only partially supported. Top management support, training and user involvement have only an indirect effect on user s satisfaction. This result provided partial support for H1d, H2d, and H3d. It implied that the PEOU, PU and Usage had a mediation effect on the relationship between the three external stimuli and user' satisfaction. It can be also seen that PU exerts the strongest total effect on user satisfaction followed by IS usage and top management support. As shown in Figure 2, the model in this study explained 64% of the variance in PEOU, 42% in PU, 39% in usage, and 43% in user' satisfaction. Table 5. Direct, indirect and total effects of the structural equation model Variables PEOU PU Current Usage User satisfaction Top Management support 0.40* (t=6.64) 0.32** (t=4.91)* Training 0.41** 0.22* (t=6.58) (t=3.39) Involvement 0.15* 0.28* (t=3.31)* (t=5.47) PEOU 0.12 ns 0.04 ns (t=1.35) (t=0.54) PU 0.60** (t=9.17) Direct Indirect Total Direct Indirect Total.27 (t=0.28) ns 0.19** (t=4.54).17 (t=1.28) ns 0.13* (t=3.27).06 (t=0.98) ns 0.17* (t=4.98) 0.19* (t=4.54) 0.13* (t=3.27) 0.17* (t=4.98) - 0.18* (t=3.06) - 0.60** (t=9.17) 0.16 (t=0.70) ns 0.23** (t=5.79) 0.11 0.18* (t=1.20) ns (t=4.60) 0.22 0.17* (t=1.05) ns (t=5.58) 0.30* (t=3.98) 0.23* (t=5.58) 0.18* (4.60) 0.17* (t=5.58) 0.17 0.18* (t=0.85) ns (t=3.06) 0.20* (t=4.09) 0.50** (t=7.88) Usage 0.34** (t=4.31) - 0.34* (t=4.31) R 2 0.64 0.42 0.39 0.43 * significant relationship; ** the strongest effect, ns: not significant relationship 11

4.8 Discussion Information usage This study revealed that three external variables have only an indirect influence on current IS, through the mediation of PU. The indirect influence is mediated by the role of usefulness. This means that availability of training, management support and user involvement contribute to increase IS usage only if the system is perceived as useful. This result confirmed the previously discussed theoretical expectation (see Table 1). In particular, studies have found that organizational support influences IS usage (Karahanna et al., 1999; Straub et al., 2001; Anandarajan et al., 2002; Al-Gahtani, 2004); availability of training influences IS usage (Igbaria et al., 1995; Igbaria et al., 1997; Al-Gahtani and King, 1999; Al-Gahtani, 2004; Gallivan et al., 2005); user involvement influences IS usage (Igbaria et al., 1997). Usefulness exerts the strongest effect on current usage followed by top management support and user training. Moreover, our study was in agreement with Al-Gahtani (2004) regarding the effect of management support on current usage which was found to be larger than that of availability of training. Unlike our results, Al-Gahtani and King (1999) in UK (a study performed in a Western culture) found, however, that user training has more effect on system usage than does management support. Our findings also extend those of Petter (2008) who identified three strategies for managing user expectations. These strategies included leadership (part of top management support), user involvement, and trust. Unlike our findings, Rouibah (2009) found that ease of use of instant messaging in Kuwait exerts the strongest effect on current usage, while Igbaria et al., (1995) found user training to exert the strongest effect on current usage followed by usefulness for PC usage in the USA. Differences between our results and previous studies could be explained by the type of used systems (general IS, instant messaging, and PCs), type of settings (social usage, private vs. public organization) and culture (Arab vs. Western). In addition, our results confirm the indirect effect of "facilitating conditions" of the unified theory of acceptance and use of technology on system usage in the USA (Venkatesh et al., 2003). Venkatesh et al., (2003) measured facilitating conditions by availability of resources, knowledge, and assistance to use the system. Our results also extended this concept to also include user involvement, besides management support and availability of training. User' satisfaction This study found three variables (PEOU, PU and usage) have a direct influence on user satisfaction. Thus, it supports studies that focused on the influence of PEOU on satisfaction (Adamson and Shine 2003; Liu et al., 2006); PU and satisfaction (Mahmood et al., 2000; Adamson and Shine, 2003; Liu et al., 2006; Rouibah, 2008; Petter, 2008), usage and satisfaction (Gelderman, 1998; Khalil and ElKordy, 1999; Al-Gahtani, 2004; Rouibah, 2008). Results also revealed that management support, availability of training, and user involvement have only an indirect effect on user' satisfaction. Two paths mediate such effect. The first one via PU and usage, and the second via PEOU and Usage, leading to support and challenge findings of previous studies. The indirect influence of management support on user satisfaction is supported by Al-Gahtani and King (1999); Anandarajan et al., (2002). However, it contrasts with those who found a direct effect (Mahmood et al., 2000; Aladwani, 2002; Al-Gahtani, 2004, Kim and Kim, 2008). The indirect influence of availability of training on user satisfaction confirms those of Igbaria et al., (1997) and contrasted with those who found direct effect (Al-Gahtani and King, 1999; Al-Gahtani, 2004; Gallivan 12

et al., 2005). This study also found that user involvement indirectly influences user satisfaction. This is a new result and contrasts with those who find a direct influence (Mahmood et al., 2000; Palanisamy and Sushil, 2001; Winston and Benjamin, 2008; Petter, 2008). Differences between the results of this study and past studies could be explained by the following: (i) no past study integrated all the antecedents of user' satisfaction in one model, and different models with different mediating factors were used; (ii) the studied technology is different; (iii) different statistical analysis techniques were used (regression vs. simultaneous analysis via LISREL); and (iv) and different respondents were used (students vs. employees). The influence of antecedents of user' satisfaction, from the most to the least, are successively: usefulness, usage, management support, training and user involvement. Such a result provides an answer to Kujala (2003), who discussed benefits and challenges of user involvement management; and who called for future research to know how and how much user involvement contributes to user's satisfaction. Findings of our study also challenge those of Kim and Kim (2008) who found that the total effect of system support on user satisfaction via user involvement is the strongest effect. Differences could be attributed to the usage of two different models. Kim and Kim (2008) used user involvement as a mediator between external factors and satisfaction, while our study used ease of use, usefulness and current usage. Finally, while Adamson and Shine (2003) succeed to link TAM to user satisfaction, this study goes further and shows that three external factors are linked to TAM construct and user satisfaction. With regard to the type of relationship between IS usage and user satisfaction, this study found support for the third model, meaning that as system usage increases, user satisfaction increases too. This model is based on the belief that the more organization support is given to end-users in term of training and users' involvement, the more the system use will increase, which will, in turn, lead to increasing user's satisfaction. This finding is in line with Arab past studies (Khalil and ElKordi, 1999; Al-Gahtani, 2004; Rouibah, 2008). Impact of culture on system usage and satisfaction This study found a lack of influence between ease of use and usefulness, and between PEOU and current usage. This new result challenges what has been found in Western culture, especially about TAM findings (Legris et al., 2003). It may also reveal the distinctiveness of the public sector. However, these results support those of McCoy et al., (2007) who called for caution when applying Western technology adoption models across cultures. Our study also found top management support to exert the strongest effect on system usefulness. Arab culture is dominated by high power distance and employees perceive benefits of a system and use them with high frequency only when top management is engaged in this process and provided support (Jreisat, 1990; Atiyyah, 1993). The cultural difference between our results and those in Western culture could be attributed to the power distance of Hofstede (2008). The emphasis in high power distance societies (e.g. Kuwait) is on status, authority, and legitimacy. Unlike the USA (which ranked 40), Arab societies are high on power distance (Kuwait ranked 80) and low on egalitarianism. Such disparities explain why top management support plays an important role in perception of system usefulness, system usage and user' satisfaction. 13

H1.d: 0.16 ns H2.d: 0.17 ns H1.c:0.27 ns Top management support Availability training/ resources H1.a: 0.32* H1.b: 0.40* H2.a: 0.22* H3.a: 0.28* Perceived Usefulness R 2 =42% H4.d ns H2.c: 0.11 ns H4.a: 0.60* Usage R 2 =39% H4.b: 0.34* H5: 0.34* Satisfaction R 2 =43% H4.c ns User involvement H2.b: 0.41* H3.b: 0.15* Perceived ease of use R 2 =64% H4.e: 0.18* H3.c: 0.06 ns H3.d: 0.22 ns Figure 2. The Final Structural Model, note *p <0.05; ** p <0.01; ***p <0.001; ns: not significant 5 CONCLUSION AND FUTURE RESEARCH 5.1 Research contributions This study has achieved three contributions: (i) it develops a new model for system usage and satisfaction with three external stimuli, not all investigated before, and validated it in Kuwait; (ii) it shows the most factors that affect system usage and user satisfaction. This model offers public companies strategies on how to boost system usage and satisfaction among end-users. And (iii) it applies a rigorous statistical technique - the structural equation modeling based on LISREL, and the variance explained by this model is much wider than those reported by previous and similar studies (Al-Gahtani and King, 1999, Aladwani, 2002; Anandarajan et al., 2002) 5.2 Research Limitations Results of the study could be interpreted in light of two limitations. (i) The sample is composed of respondents only from the public sector, i.e. governmental ministries. Since the SMEs represent 88% in Kuwait, the study results may not apply to the private sector. (ii) The average experience of the respondents with the used systems is nine years; therefore, the findings of the paper have limited generalizability. In particular, previous studies have shown user's experience affects of end-users IS/IT adoption (Venkatesh and Davis, 2000). 5.3 Managerial and research implications Several managerial implications can be drawn for IS/IT practitioners. 14

First, the validated model could be used as a decision making tool in order to orient what companies need to emphasis in order to increase system usage and user's satisfaction. The message to organizations in Kuwait, or training centers or international companies willing to penetrate this region, is clear: (i) ensure top management support and commitment in order to foster an atmosphere that helps users perceive the importance of the system and enhances their system usability and satisfaction; (ii) ensure the contribution of user participation/involvement to increase system usefulness and system usability; and (iii) don't focus too much on the ease of use of new systems. Since usefulness is the most determinant of system usability, then if users believe systems are complex, the system benefits will not be affected. Second, results of the study could be used to consider investment in IS/IT and people as well as to consider factors to motivate end-users. Top management are responsible to ensure that end-users are adequately involved early enough in the system development or during outsourcing in order to achieve IS/IT success. The more end-users are involved in these two processes, the more their requirements have the chance to be included and thus increase usability of the systems. The importance of management support as a motivational stimuli has been emphasized by previous Arab studies. Atiyyah (1993) reported the reasons of a training dissatisfaction primarily centered on the absence of senior management presence during training. Jreisat (1990) indicated that without the involvement of the upper managers in the process of conceptualization and monitoring the implementation of changes, effort at organizational change would be futile. From a research perspective, this study has been found that user s satisfaction explains 43% of the total variations against 39% of system usage. Accordingly, 57% of user satisfaction and 61% of system usage remain unexplained. We suggest three research directions. First, there is a need to conduct a longitudinal study and to incorporate additional variables such as result demonstrability and triability (Moore and Benbasat, 1991), social norm in terms of different peers and co-workers influence (Karahanna et al., 1999; Lewis et al., 2003; Gallivan et al., 2005), and gender effect (Venkatesh and Morris, 2000). The aim is to assess the relative influence of these variables compared to the three external factors that were investigated in this study, and initiate further comparative studies. A second direction consists to investigate the relationship between system usage, user satisfaction and firm s performance across time as well as factors that lead to dissatisfaction Shaw et al., 2002). One potential reason for satisfaction/ dissatisfaction is to examine the expectation gap between end-users and system providers (Tesch et al., 2005). Does the user's satisfaction/ dissatisfaction affect company performance? REFERENCES Adamson, I. and Shine, J. (2003), "Extending the new technology acceptance model to measure the end-user information systems satisfaction in a mandatory environment: a Bank s treasury", Technology Analysis & Strategic Management, vol. 15 No4, pp. 441-455. Aladwani, A. M. (2002), "Organizational actions, computer attitude, and end-user satisfaction in public organizations: an empirical study", Journal of End-User Computing, Jan-Mar, pp. 42-49. Al-Gahtani, S.S. (2004), "Computer technology acceptance success factors in Saudi Arabia: an exploratory study", Journal of Global Information Technology Management, Vol.7 No1, pp. 5-29. 15

Al-Gahtani, Said, S. and King, M. (1999), "Attitudes, satisfaction and usage: factors contributing to each in the acceptance of information technology", Behaviour & Information Technology, Vol. 18 No 4, pp. 277 297. Anandarajan, M. Igbaria, M. and Anakwe, U.P. (2002), "IT acceptance in a less-developed country: a motivational factor perspective", International Journal of Information Management, Vol. 22, pp. 47-65. Anonymous (2008). Failure Rate, statistics over IT projects failure rate. Web: http://www.itcortex.com/stat_failure_rate.htm, accessed 22 nd of September 2008 Atiyyah, H.S. (1993), "Management development in Arab countries: the challenges of the 1990s Management", Journal of Management development Vol. 12 No 1, pp. 3-12 Au, N. Ngaib, E. W. T., and Cheng, T.C.E. (2002), "A critical review of end-user information system satisfaction research and a new research framework", Omega, Vol. 30, pp. 451 478. Baroudi, J. J. and Orlikowski, W. J. (1988), "A Short-form Measure of User Information Satisfaction". Journal of Management Information Systems, Vol 4 No 4, pp. 44-59. Briggs, R.O. Reinig, B.A. and Vreede, G.D. (2008), "The Yield Shift Theory of Satisfaction and Its Application to the IS/IT Domain", Journal of the AIS, Vol. 9 No 5, pp. 267-293 Davis, F. D. (1989), "Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology", MIS Quarterly, Vol. 13 No 3, pp. 319 340. Delone, W. H. and Mclean, E. R. (1992), "Information systems success: the quest for dependent variable", Information Systems Research, Vol. 3 No1, pp. 60-95. Doll, W. J. and Torkzadeh, G. (1991), "The Measurement of End-User Computing Satisfaction: Theoretical and Methodological Issues". MIS Quarterly, March, Vol. 15, No1, pp. 5-10. Franz, C. and Robey, D. (1986), "Organizational context, user involvement and usefulness of information systems", Decision Sciences, Vol. 17, pp. 329-355. Gallivan, M. J. Spitler, V. K., and Koufaris, M. (2005), "Does technology training really matter? A social information processing analysis of co-workers influence on IT usage in the workplace". Journal of Management Information System, Vol. 22, No 1, pp. 153-192. Hair, J. F. Anderson, R. E. Tatham, R. L., and Black, W. C. (1998), Multivariate data analysis. 5 th Edition, Prentice Hall., Hofstede G. (2008), Geert Hofstede cultural dimensions, Web: http://www.clearlycultural.com/geerthofstede-cultural-dimensions, accessed 2 nd October 2008 Igbaria, M. Guimaraes, T. and Davis, G. B. (1995), "Testing the determinants of microcomputer usage via a structural equation model", Journal of Management Information Systems, Vol. 11 No 4, pp. 87-114. Igbaria, M., Zinatelli, N. Gragg, P., and Cavaye, A. L. M. (1997). Personal computing acceptance factors in small firms: a structural equation models. MIS Quarterly, Vol. 21, No 3, pp. 279-305. Jreisat, J. E. (1990), "Administrative change and the Arab manager". Public Administration and Development Vol., 10, No 4, pp. 413-421. Karahanna, E. Straub, D. W. Chervany, N. L. (1999), "Information technology adoption across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs". MIS Quarterly, Vol. 23 No 2, pp. 183-213. 16

Khalil, O. E and Elkordy M. M. (1999), " The relationship between user satisfaction and system usage empirical evidence from Egypt". Journal of End-user Computing, Vol. 11, No2, pp. 21-28. Kim, G. and Kim, E.S. (2008), "An exploratory study of factors influencing ASP application service provider". The Journal of Computer Information Systems, Vol. 48, No3), pp. 118-124 Kujala, S. (2003), "User involvement: a review of the benefits and challenges" Behaviour & Information Technology Vol. 22, No 1, pp. 1 16 Lee, J. and Kim, S. (1992), "The relationship between procedural formalization in MIS development and MIS success: a contingent analysis". Information Management, Vol. 22, pp. 89-111. Lee, S.T. Kim, H. and Gupta, S., (2009), "Measuring open source software success". Omega, Vol. 37, pp. 426 438 Legris, P. Ingham, J. Collerette, P., (2003), "Why do people use information technology? A critical review of the technology acceptance model". Information & Management, Vol. 40, pp. 191 204. Lewis, W. Agarwal, R., and Sambamurthy, V. (2003), "Sources of influence on beliefs about information technology use: an empirical study of knowledge workers", MIS Quarterly, Vol. 27, No 4, pp. 657-678. Liu, Y. Chen, Y. and Zhou, C. (2006), "Determinants Affecting End-User Satisfaction of Information Technology Service", 2006 International Conference on Service. Mahmood, M. A. and Mann, G. J. (2000), "Special issue: impact of information technology investment on organizational performance",. Journal of Management Information Systems, Vol. 16, No 4, pp. 3-10. Mahmood, M. Burn, J.M. Gemoets, L.A. and Jacquez, C. (2000), "Variables affecting information technology end-user satisfaction: a meta-analysis of the empirical literature", International Journal of Human-Computer Studies Vol. 52, No 4, pp. 751-771 McCoy, S., Galletta, D. F., and King, W. R. (2007), "Applying TAM across cultures: the need for caution", European Journal of Information System, Vol. 16, pp. 81-90. Moore, G. C. and Benbasat, I. (1991), "Development of an instrument to measure the perception of Adopting an information technology innovation", Information Systems Research, Vol. 2, No 3, pp. 192-222. Palanisamy, R. and Sushil, R. (2001), "Empirically testing the relationships between user involvement, information waste, and MIS success". Journal of Services Research Vol.1, No 1, pp.1-35 Petter, S., (2008), "Managing user expectations on software projects: Lessons from the trenches. International Journal of Project Management (forthcoming) Research Institute for Digital Economy (2006), Web: www.alwatan.com.kw; accessed 15 th January 2006 Rouibah, K. and Hamdy, H., (2009), "Factors Affecting Information Communication Technologies Usage and Satisfaction: Perspective From Instant Messaging in Kuwait". Journal of Global Information Management 17(2); April/June, (forthcoming) Shaw, N. C. DeLone, W. H. and Niederman, N. (2002), "Sources of dissatisfaction in end-user support: an empirical study". ACM SIGMIS Database, Vol. 33, No 2, pp. 41-56. Straub, D. Karen, L. and Hill, C. E. (2001), "Transfer of information technology to the Arab world: a test of cultural influence modeling", Journal of Global Information Management, Vol. 9, No 4, pp. 6-28. 17