Answer

1. Methodology Chapter: HFACS-OGI Model

    

   Research design and Flow Chart of HFACS-OGI Model Based Data Mining

   The proposed research will deploy a quantitative methodology. In specific, a quantitative research method implements objective measurements as well as of statistics and numerical analysis of data collected via the use of computational techniques. The rationale for using quantitative methodology in this research builds from various advantages and benefits. First, since the method relies heavily on primary data, it affords a given research more reliable and credible research findings. Second, as Muijs (2010) points out, findings of research studies that follow quantitative methodology are authentic and valid because the findings are informed largely by numerical and unbiased data. However, the data may be susceptible to manipulation during analysis to align with the interests of the analyst. To avoid such cases, collected data will be analyzed by the research team.

    Being a quantitative research, this study will utilize case study approach as the research design with particular focus on the HSE database. Simply stated, HSE database will be used as the case study for this research on HFACS-OGI model. The case study design will aim at identifying the various human factors using HFACS-OGI based on the case and provide focused solutions and or interventions. As Yin (2017) aptly puts it, the rationale for using a case study design is that it allows for a more in-depth and comprehensive collection of information as it only focuses on one case. For this research study, the focus will be collecting quantitative data from the database maintained at HSE.

   Unlike other models, the use of HFACS model in data mining follows a simple flow-chart as shown in figure 1 below.

    

    

    

    

    

    

    

    

    

    

    

    

    

    

    

    

    

   Figure 1: Flow of the HFACS Model Based Data Mining for Oil and Gas (X et al. 2010).

   Reliability of the HFACS-OGI

   The ability of the HFACS model to visualize potential occurrence of accidents and disasters at four levels; namely, at the organizational failure, unsafe supervision, preconditions for unsafe acts, and unsafe acts, has afforded it a higher reliability in mitigating such accidents and their related risks (Theophilus et al. 2017). Additionally, the HFACS-OGI has the ability to classify accidents based on their causes, effects, intensity, magnitude among other criteria, and hence reliable for analyzing accidents in the oil and gas industry. Furthermore, the framework has emerged as the most reliable tool for analysis of oil and gas accidents because it goes beyond focusing on and accusing the stakeholders involved in the accidents. Instead, it employs a systems approach to identify the mishaps and deficiencies that contributed to the accident as well as missed opportunities and go-around measures that might have prevented the accident.        

   Design Procedure

   Design of the HFACS-OGI was informed by a set of quantitative data obtained through examination of the various pieces of primary information contained in the HSE database. The collected data contained information related to the multiple exposures to risks in the oil and gas industry at different levels including at the organizational, supervision, operation and maintenance levels. Preliminary findings suggest that risk for accident occurrence are highest at the operation level followed by at the supervision and organizational levels, consecutively. This is partly because the crew usually ignore early warning signs and other pre-accident conditions due to overreliance on system automation. As a result, this introduces active errors (i.e. errors caused by the crew during operation), thereby, leading to mismanagement of oil and gas vessels by the onboard crew. On the other hand, lack of proper supervision both during equipment construction, commissioning, usage, and maintenance introduces to latent errors that align with active errors to bring about devastating crises (Reasons 1990). Figure 2 below shows a proposed HFACS-OGI framework based on four levels of risks for crisis: organizational, unsafe supervision, preconditions for unsafe acts and unsafe acts (Theophilus et al. 2017).  

    

    

    

    

   `

    

    

    

    

    

   `

    

    

    

    

    

    

   `

    

    

    

    

    

    

    

    

    

    

    

    

    

   `

    

    

                           Figure 2: Proposed HFACS Framework for Oil and Gas Industry based on four                                                        levels of risks for crisis. 

   Research Paradigm

   This research study adopts a positivist paradigm. The rationale for following this philosophy is that the research study draws largely from quantitative approaches and employs empirical and statistical methods to test stated hypotheses and confirm existing constructs as Krauss (2005) notes. Thus, the positivist paradigm is suitable for this research as it will help underpin description and explanations about the effectiveness of the proposed HFACS framework in abating accidents and disasters in oil and gas, marine, aviation and other industries with higher risks for catastrophes. For instance, since the research study follows a quantitative methodology and employs case study research design, positivist philosophy will be necessary to analyze the collected numerical data, and more importantly, interpret the resulting findings, and present them in a more meaningful manner (Johnson and Onwuegbuzie 2004).        

   Data analysis

   Data obtained from the HSE database on human factors that contribute to catastrophes in transportation industry was analyzed using the SPSS software developed and provided by IBM. Data on each of the accident contributing categories for both HFACS and HFACS-OGI was recorded and analyzed through the Chi-squire test and spearman’s rank correlation to understand the differences. Since research was a secondary research, data on accident contributing categories was collated from Theophilus et al. (2017) and other sources and tabulated in table 1 below.

   Table 1: Accident Contributing Categories for both HFACS and HFACS-OGI Frameworks (Theophilus et al. 2017).

   Level/Category	HFACS		HFACS-OGI
   	Number of Incidents Identified	Percentage	Number of Incidents Identified	Percentage
   Organizational influences	22	198	34	306
   Unsafe supervision	11	99	11	99
   Preconditions for unsafe acts	11	99	16	144
   Unsafe acts	10	90	10	90

    

   Table 2: Refined Data for Analysis

   Level/Category	Number of Incidents Identified per contributing category
   	HFACS	HFACS-OGI
   Organizational influences	22	34
   Unsafe supervision	11	11
   Preconditions for unsafe acts	11	16
   Unsafe acts	10	10

    

    

   Table 3: Data analyzed using Spearman’s Rank Correlation

   Number of Incidents Identified per contributing category Level/Category HFACS HFACS-OGI Organizational influences 22 34 Unsafe supervision 11 11 Preconditions for unsafe acts 11 16 Unsafe acts 10 10   HFACS HFACS-OGI HFACS 1 HFACS-OGI 0.980808324 1

    

   Analysis of collected data using Spearman’s rank correlation shows that HFACS and HFACS-OGI have a correlation co-efficient of 0.981. This suggests existence of a stronger positive relationship between the two frameworks, thereby, indicating how HFACS-OGI can be an effective tool in identifying and classifying factors that contribute to accidents and catastrophes in the oil and gas industry. .