Elsevier

Safety Science

Volume 49, Issues 8–9, October 2011, Pages 1110-1117
Safety Science

Possibilistic regression analysis of influential factors for occupational health and safety management systems

https://doi.org/10.1016/j.ssci.2011.02.014Get rights and content

Abstract

The code of occupational health and safety (OHS) is an influential regulation to improve the on-the-job safety of employees. A number of factors influence the planning and implementation of OHS management systems (OHSMS). The evaluation of OHSMS practice is the most important component when forming a health and safety environmental policy for employees. The objective of this research is to develop an intelligent data analysis (IDA) in which possibilistic regression being endowed with a convex hull approach is used to support the analysis of essential factors that influence OHSMS. Given such subjective terms, the obtained samples can be conveniently regarded as fuzzy input/output data represented by membership functions. The study offers this vehicle of intelligent data analysis as an alternative to evaluate the influential factors in a successful implementation of OHS policies and in this way decrease an overall computational effort. The obtained results show that several related OHSMS influential factors need to be carefully considered to facilitate a successful implementation of the OHSMS procedure.

Highlights

► The realization of intelligent data analysis (IDA) with the use of possibilistic regression completed for OHSMS analysis purposes. ► Becomes an alternative solution especially for realizing possibilistic regression analysis carried out in the presence of fuzzy data in OHSMS environment. ► Perform a real-world OHSMS data analysis process; to identify the most influential factors contributing to the successful implementation of the OHSMS. ► The proposed approach for possibilistic regression could become an efficient of computer-aided ergonomics tool. ► The proposed approach becomes a potential vehicle for analyzing real-world data in real-time processing where the ambiguity or fuzziness cannot be avoided such as OHSMS performance analysis.

Introduction

In the recent years, the implementation policy of occupational health and safety (OHS) becomes the major requirement to prepare and provide the best solution to assure health and safety in working environment (Ayomoh and Oke, 2006, Sgourou et al., 2010). There are two major groups, which directly involve this important regulation; that is employees’ personnel and administrative staffs. The overall combination of OHS code ethics and activities has to be clearly defined and their results should be periodically evaluated through some OHS management system (OHSMS). In this situation, the evaluation of factors that influence an OHS provides a useful feedback to employees’ personnel and administrative staff. This process may produce higher safety consciousness as well as contribute to the well-being in a workplace (Gallagher, 2000, Hwang et al., 2009).

In addition, the most essential albeit somewhat controversial point concerns a procedure on how to obtain sound knowledge related to the OHS practices while dealing with imprecise or vague data. The inherent complexity of OHS stems from several sources. In particular, we need to stress that a simple computational approach is often inadequate and traditional data about incidents/claims have also proved to be quite unreliable (Gallagher et al., 2003, Dul and Neumann, 2009). In light of this, a genuine need emerges to develop an alternative approach, which enables us to analyze and provide high quality results for the continuous improvement of OHSMS implementation procedures (Gallagher et al., 2001).

Currently, supervisors or auditors can use a number of instruments or audit tools in their evaluation of OHSMS influential factors. The factors or criteria that affect OHS practices usually depend on company policy and regulations, where its evaluation process involves a number of approaches or/and parameters, which are often based on imprecise data (Watada et al., 1998). Besides that, some companies/organizations still use open-ended and somewhat subjective questions. These types of questions require some judgment process such as OHS experts’ interpretation to analyze and interpret the results. Without this, it is difficult to arrive at conclusive results or generate useful knowledge (Robson et al., 2005). Moreover, emerging Internet facilities offer advantages to the administration section in companies/organizations when developing a web portal for each level of employees to assess their administrative OHS practice. Using this technology in place, related parties such as auditors, companies/organizations administration and employees can envision quite tangible benefits.

In order to acquire benefits from the analysis of the results and facilitate their comprehensive interpretation related to the OHSMS implementation, an adequate analysis is badly needed. The main objective of this research is to employ possibilistic regression analysis to the OHSMS evaluation. We anticipate that the possibilistic regression analysis has the capability of producing a meaningful and noteworthy results related to the most influencing factor of successful OHSMS practice. The possibilistic regression is implemented with the hybrid approach, which involves a certain geometric concept called a convex hull approach, specifically a Beneath–Beyond algorithm (Ramli et al., 2009). This selected approach is appropriate here because of its underlying efficiency and consistency while dealing with imprecise data. In addition, the possibilistic regression analysis enables us to determine and rank the factors that significantly influence the OHSMS practice among selected companies/organizations.

This intelligent data analysis (IDA) has been selected because of its efficiency and consistency while dealing with imprecise input data such as in OHSMS evaluation practice, which closely link to the essence of human decision-making processes, especially when dealing with a large amount of fuzzy input–output data. As it will become revealed, this research leads us to an interesting and practically relevant finding that among six selected OHSMS influence factors, the three of them are highly ranked. By this means, companies/organizations could manage their OHS policy implementation procedure by considering these important factors at some stage.

The paper is organized as follows. In Section 2, a pertinent literature on OHS policy, its influential factors and the important and efficient performance of OHSMS are reviewed. A general outline of the possibilistic regression model and an overview of the convex hull based approach for possibilistic regression model are highlighted here. Section 3 presents an analysis of implementation factors that affect the planning and implementation of OHSMS, while Section 4 interprets the results of possibilistic regression models obtained in this way. Section 5 presents some concluding remarks.

Section snippets

Occupational health and safety policy: a brief overview

Cagno et al. (2010) revealed the identification of company occupational health and safety (OHS) related significant success factors and their interactions is a crucial task issue to better understand and examine on how improvement interventions impact on the OHS performance of the company. All employers are required under the OHS Act to accept the term of ‘duty of care’ for the health and safety of all people in workplace. Implementing this action means that everyone in workplace should be

Fundamental ideas of models of possibilistic regression

In statistical regression, deviations between observed and estimated values are assumed to be due to random errors. Although conventional regression has been applied to various areas, the related problems include the vague relationship existing between input and output variables that cannot be clearly justified (Ciarapica and Giacchetta, 2009). Therefore, this becomes a major reason behind its unsuccessful usage in the meaningful interpretation of the model.

Regression analysis is one of

A determination of occupational health and safety management system influential factors

Here, we present an analysis realization related with OHSMS, which focuses on factors influencing the successful planning and implementation for OHSMS. We use selected samples of fuzzy I/O data coming from a certain number of companies/organizations, which fully performed and successfully implemented the OHSMS policy. Moreover, those selected companies/organizations are coming from similar business type, working environment and culture as well as administrative and workers population (up to

Interpretation of the coefficients of possibilistic regression

He et al., in 2005 noted that positive regression coefficients, mj in possibilistic regression imply positive contribution to the prediction outcome. With regard to this study, an analysis outcome refers to the efficiency of OHSMS implementation achievement (OHSEI) based on ranked order of influential factors. The obtained results presented in Table 4 imply that OHSPP, OHSHA and OHSRC exhibit a positive impact on OHSEI. On the other hand, OHSCS, OHSTS and OHSPM point at the negative

Concluding remarks

To acquire knowledge to support the decision-making process, the proposed idea offers an effective and efficient alternative to evaluate and rank the influencing factors related to the OHSMS implementation. The paper presents the realization of IDA with the use of possibilistic regression completed for fuzzy data.

Based on the obtained results, we were able to identify the most influential factors contributing to the successful implementation of the OHSMS: developing an OHS policy and program

Acknowledgment

The first author was supported by Ministry of Higher Education Malaysia (MOHE) under Skim Latihan Akademik IPTA (SLAI-UTHM) scholarship program at Graduate School of Information, Production and Systems (IPS), Waseda University, Fukuoka, Japan.

References (40)

  • W.P. Neumann et al.

    Integrating ergonomics into production system development – the Volvo Powertrain case

    Applied Ergonomics

    (2009)
  • G. Papadopoulos et al.

    Occupational and public health and safety in a changing work environment: an integrated approach for risk assessment and prevention

    Safety Science

    (2010)
  • D. Parker et al.

    A framework for understanding the development of organisational safety culture

    Safety Science

    (2006)
  • M.-M. Robertson et al.

    Flexible workspace design and ergonomics training: impacts on the psychosocial work environment, musculoskeletal health and work effectiveness among knowledge workers

    Applied Ergonomics

    (2008)
  • E. Sgourou et al.

    Assessment of selected safety performance evaluation methods in regards to their conceptual, methodological and practical characteristics

    Safety Science

    (2010)
  • E. Cagno et al.

    Identification of OHS-related factors and interactions among those and OHS performance in SMEs

    Safety Science

    (2010)
  • D. Challis et al.

    Impact of technological, organizational and human resource investments on employee and manufacturing performance: Australian and New Zealand evidence

    International Journal of Production Research

    (2005)
  • C. Derosier et al.

    Studying work practices: a key factor in understanding accidents on the level triggered by a balance disturbance

    Journal of Ergonomics

    (2008)
  • S. Drebit et al.

    Occupational and environmental risk factors for falls among workers in the healthcare sector

    Journal of Ergonomics

    (2010)
  • Gallagher, C., 2000. New directions: innovative management plus safe place. In: Pearse, W., Gallagher, C., Bluff, L....
  • Cited by (0)

    View full text