Vast amounts of data are collected on a daily basis, making it difficult for humans to derive at valuable information to make effective decisions. In recent years, the field of Business Intelligence (BI) and Information Visualisation (IV) have become a key driver of an organisation’s success. BI tools supporting decision making need to be accessible to a larger audience on different levels of the organisation. The problem is that non-expert users, or novice users, of BI tools do not have the technical knowledge to conduct data analysis and often rely on expert users to assist. For this reason, BI vendors are shifting their focus to self-service BI, a relatively new term where novice users can analyse data without the traditional human mediator. Despite the proliferation of self-service BI tools, limited research is available on their usability and design considerations to assist novice users with decision making and BI analysis. The contribution of this study is a conceptual framework for designing, evaluating or selecting BI tools that support non-expert users to create dashboards (the BI Framework). A dashboard is a particular IV technique that enables users to view critical information at a glance. The main research problem addressed by this study is that non-expert users often have to utilise a number of software tools to conduct data analysis and to develop visualisations, such as BI dashboards. The research problem was further investigated by following a two-step approach. The first approach was to investigate existing problems by using an in-depth literature review in the fields of BI and IV. The second approach was to conduct a field study (Field Study 1) using a development environment consisting of a number of software components of which SAP Xcelsius was the main BI tool used to create a dashboard. The aim of the field study was to compare the identified problems and requirements with those found in literature. The results of the problem analysis revealed a number of problems in terms of BI software. One of the major problems is that BI tools do not adequately guide users through a logical process to conduct data analysis. In addition, the process becomes increasingly difficult when several BI tools are involved that need to be integrated. The results showed positive aspects when the data was mapped to a visualisation, which increased the users’ understanding of data they were analysing. The results were verified in a focus group discussion and were used to establish an initial set of problems and requirements, which were then synthesised with the problems and requirements identified from literature. Once the major problems were verified, a framework was established to guide the design of BI dashboard tools for novice users. The framework includes a set of design guidelines and usability evaluation criteria for BI tools. An extant systems analysis was conducted using BI tools to compare the advantages and disadvantages. The results revealed that a number of tools could be used by non-experts, however, their usability hinders users. All the participants used in all field studies and evaluations were Computer Science (CS) and Information Systems (IS) students. Participants were specially sourced from a higher education institution such as the Nelson Mandela Metropolitan University (NMMU). A second field study (Field Study 2) was conducted with participants using another traditional BI tool identified from the extant systems analysis, PowerPivot. The objective of this field study was to verify the design guidelines and related features that served as a BI Scorecard that can be used to select BI tools. Another BI tool, Tableau, was used for the final evaluation. The final evaluation was conducted with a large participant sample consisting of IS students in their second and third year of study. The results for the two groups revealed a significant difference between participants’ education levels and the usability ratings of Tableau. Additionally, the results indicated a significant relationship between the participants’ experience level and the usability ratings of Tableau. The usability ratings of Tableau were mostly positive and the results revealed that participants found the tool easy to use, flexible and efficient. The proposed BI Framework can be used to assist organisations when evaluating BI tools for adoption. Furthermore, designers of BI tools can use the framework to improve the usability of these tools, reduce the workload for users when creating dashboards, and increase the effectiveness and efficiency of decision support.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:nmmu/vital:27136 |
| Date | January 2016 |
| Creators | Smuts, Martin Bradley |
| Publisher | Nelson Mandela Metropolitan University, Faculty of Business and Economic Sciences |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis, Masters, MCom |
| Format | xiv, 287 leaves, pdf |
| Rights | Nelson Mandela Metropolitan University |
Page generated in 0.0026 seconds