Structural Equation Analysis

This study integrated the theoretical approaches and empirical findings of research on DSS usage in local authorities in Egypt, and tested a structural equation model examining TAM and other organizational and human factors that supposedly increase DSS usage by CEO in local authorities in particular and other organizations in general. PU and PEU showed significant direct effect on DSS usage at 0.001 and 0.10 levels and in consequence confirmed all earlier cited studies about TAM (Davis, 1989; Igbaria, Zinatelli, Cragg, and Cavaye, 1997). As TAM proposes, both PU and PEU are important in technology acceptance and usage. However, their relative importance in the acceptance process has been shown to be different in previous studies. For instance, (Davis, 1993) found that usefulness dominated ease of use, whereas (Adams, Nelson, and Todd, 1992) found ease of use to be more influential than usefulness.

In understanding the main relationships in the creation of the overall model, consideration has been given to the dynamics of the relationship between the independent and latent variables, and their overall impact on usage as the chosen performance indicator. It is important to better identify what the latent variables are in the context of the underlying issues that were researched. PEU in particular considers what the user viewed to be the central issues in creating user friendliness of software, and whether the current DSS was causing a barrier to entry or facilitating further use.

PU is based on DSS users' key business performance measures for success and consequently, has direct implications on the independent variables that identify the overall usage.

Finally, DSS usage provides a performance indicator specific to software applications supporting the strategic decision-making process.

We will first consider SEM in terms of the effects of the latent variables on independent variables. The latent variables have emerged through many other variables in the context of perceived ease of use (PEU) and perceived usefulness (PU), as outlined above. Further analysis of the data (see Figure 2) in the study have shown that the top management characteristics, i.e., understanding of DSS and involvement in the process of design and development, influence the PEU while none of the other research constructs have any significant effect on PU. The results showed a strong direct and positive relationship between perceived usefulness from one side and DSS characteristics and PEU from the other. In addition, there was an inverse relationship between PU and both top management and external support characteristics. These results for PU are somewhat surprising because this author expected that top management characteristics and support would have a positive direct effect on PU. This expectation was supported by an earlier result of this research, showing a positive relationship between PEU and top management. One plausible explanation for this result might be that decision makers think that central government creates a barrier to any benefits of functionality that DSS could offer to them by centralizing most of the strategic decision-making, as was obvious from the interviews with CEO and IT managers.

Evidence for the assumption of mediation of PEU and PU between DSS usage and new external constructs raises some important implications for the construction of research models of information technology adoption phenomena. These new constructs need to be re-examined in different contexts to confirm their validity. It appears that the relative importance of PEU is higher than PU, although both were found to be important, suggesting that the usability of DSS is crucial for its actual use in SDM in local authorities in developing countries, particularly in Egypt. Consequently, it is assumed that DSS needs to be driven by business processes, and where possible, the ability to customize applications will support the need for ease of use. This is of relevance to other researchers in evaluating these two constructs and in considering other related issues such as the stage and environment of DSS implementation.

Implications for Practice: Perhaps the most significant implication of the findings of this research is the necessity of moving toward decentralization regarding strategic decision-making where top management were found to have a negative relationship with PU. Clearly as indicated in many studies DSS implementation needs to be user-led and consequently, strategic decision makers need to be directly involved in the process of evaluating and implementing DSS. The integration of technical improvements and functionality of DSS in terms of (1) usability and (2) decision maker support is also an imperative if success of DSS in SDM in local authorities is to be ensured. The results also showed a negative relationship between external support and PU, suggesting dependence on internal support to guarantee productive usage of DSS. Next generation groupware and user-centric software tools enable users to overcome some issues such as lack of internal support and software that specifically meets individual needs. In the future, moves towards adaptive software applications through profiling solutions which capture behavioral patterns, has the potential for radical technology-led change in perceived ease of use and use of DSS in the strategic decision-making process.