An assessment of business intelligence tools’ fitness in supporting decisions making processes in South Africa’s universities of technology: a case of the CUT

Abstract

The pervasive integration of Business Intelligence (BI) systems within academic institutions underscores their pivotal role in fostering data-driven decision-making. Nevertheless, grasping the intricate web of factors influencing BI fitness for decision-making in academia remains daunting. This dissertation contributes towards bridging this knowledge gap by harnessing Structural Equation Modelling (SEM) to unravel the multifaceted elements underpinning BI systems effectiveness in facilitating decision-making processes, especially by middle-level managers. An extensive case study was conducted at the Central University of Technology, Free State (CUT) where the institution depends on the Integrated Tertiary Software (ITS) as its BI system. The research methodology was thoughtfully crafted to guarantee the attainment of strong and scientifically sound outcomes. A quantitative research approach was used to collect data. This entailed the distribution of an online survey questionnaire through the SurveyMonkey. The survey achieved a 70% response rate, with 42 out of the targeted 60 academic teaching and support staff participating. These respondents predominantly held middle management positions and utilised ITS to inform their decision-making roles. The collected dataset underwent a thorough analysis, commencing with descriptive data analysis conducted using IBM Statistical Software for Social Sciences version 27. Additional data analysis in the form of SEM, using SmartPLS 4 was used to undertake an extensive analysis of the measurement model (outer model) and the structural model (inner model). Through SEM, a comprehensive evaluation of the fundamental components of this study is conducted. These components encompass (1) the current positions of the respondents, (2) their utilisation of systems, (3) the attributes defining the tasks they engage in, (4) the compatibility of these tasks with the technology in use (or Task-Technology-Fitness, otherwise known as TTF) and (5) the inherent characteristics of the technology itself. In addition, an assessment of the reliability and discriminant validity of these constructs was performed in order to establish a strong and dependable basis upon which a subsequent analysis was built. This emphasises the crucial role played by the assurance of both the consistency and distinctiveness of these foundational elements in ensuring the overall strength and validity of the research findings. The structural model findings shed light on the factors influencing the appropriateness of BI for decision-making. TTF emerges as a pivotal factor, demonstrating a strong and positive impact on the effectiveness of BI in supporting decision-making in academic contexts. Conversely, system usage has a comparatively weaker influence in this context. The dissertation extends to explore effect size () and path coefficients, providing a glimpse into the strength and significance of these interrelationships. Specific hypotheses are scrutinised, with results either confirming or challenging initial hypotheses. The research validates the significant influence of current position on system usage while negating the substantial impact of task characteristics and system usage on TTF. To summarise concisely, after conducting extensive tests on the Inner Model, which included assessments like R-Square, F-Square, and Q-Square, we found that while the standardized Root Mean Square Residual (SRMR = 0.120) exceeds the acceptable threshold of 0.08, and the Normed Fit Index (NFI = 0.430) falls short of the desired benchmark of 0.90, when considering all these results together, they suggest a positive fit between the model and the data. The key finding is therefore the critical significant impact of TTF and technology characteristics on the effectiveness of BI for decision-making tasks. This dissertation offers a substantial contribution to comprehending BI fitness within academic contexts. The findings accentuate the pivotal role of TTF and its far-reaching implications for fortifying BI systems in academia. These insights can potentially steer strategic decisions, optimise resource allocation, and cultivate data-driven academic ecosystems. The research highlights the necessity of using strong methodologies and SEM techniques to uncover the complex dynamics that govern BI in the academic sphere.