Extracting cultural relationships from helicopter accidents
Culture in Aviation
Proposed Multi-Level cultural Framework
Sampled Helicopter Accidents
Results and Discussion
Analysing the Relationships in Components of the Culture Framework
Table 2: Inter-node and inter-level correlations (r)
Driving the cultural framework with data
This paper exploring the role of culture and its relationships in helicopter accidents based on the culture differences between two different countries.
Helicopter operations in Nigeria and the UK share many common features including a prominent offshore oil industry that places heavy demands on helicopter transport, but they differ in terms of the cultural context within which these operations take place. Nigeria shows themes associated with high cultural context. The UK, themes were predominately associated with low cultural context.
Air accident investigation reports were subjected to an in-depth theme-based content analysis in order to distil the role of culture in their causation. This analysis was selected as the means by which to extract cultural factors from the reports and variables influencing human error leading to accidents. Text and cluster analysis was used as an exploratory tool to observe patterns by grouping the nodes that shared similar themes.
The study shows two key findings. First, the sampled UK and Nigerian accident reports differ in respect to weak and strong uncertainty avoidance. Secondly, and more importantly, cultural factors, although not currently well represented in accident analysis were prominent in the sampled accident reports.
Greater consideration needs to be paid on accident analysis and the role of culture within the national context. This paper presents compelling results to show the extent of this gap.
The cultural framework can be used to understand the role of culture and the relationship with human factors in accident analysis.
Keywords: Human Error analysis, Accident Analysis, Culture, Systems.
Research shows that human error is implicated in around 70 to 80 percent of aviation accidents (Wiegmann & Shappell, 2001; Shappell & Wiegmann, 2004; Wiegmann & Shappell, 2003; Sarter & Alexandar, 2000; O’Hare, Wiggins & Batt, et al, 1994; Yacavone, 1993) and an important modifier of human behaviour is the culture in which it takes place. Culture has been the topic of consideration in previous study, especially in aviation domains (Orasanu, Fisher & Davison, 1997; Helmreich, 1994; Meshkati, 2002; Soeters & Boer, 2000) and others. It has been studied at the level of organisations (Helmreich & Merrit, 1998), in maintenance activities, (Soeters & Boer, 2000), design and manufacturing, (Foushee, 1984), and pilots, (Merrit & Helmreich, 1996a; 1996b; Li et al 2007; Strauch, 2010; Helmreich & Davies, 2004). Despite this, culture is still not a prominent component of accident analysis methods and neither is it studied in an integrated fashion across different layers of the system (Merrit & Helmreich, 1998). The aim of this paper is fourfold. Firstly, it is to link the extant knowledge-base on culture to the specific problem of helicopter safety and accident analysis. Secondly, to use this knowledge-base to propose a cultural framework linking elements of systems to features of culture and ultimately to human actions. Thirdly, to perform a content analysis on real helicopter accident reports from two culturally distinct regions of the world to reveal the relationship cultural factors actually in play. Fourth, to relate the discovered cultural factors back to the original framework in order to validate the components within it and their structure. The ultimate aim is to use this exercise to reveal the extent of the gap in current accident analysis approaches and provide a concrete way to gain traction on it.
Culture defines the way of life of a particular group of people, and in turn, has wide ranging influences on their values and behaviours (Hofstede, 1991). Values are part of a process of socialization and represent principles or patterns of behaviour which are held in high regard. Values influence people’s actions by helping to sustain the social and economic structures in which they co-exist (Kluckhohm, 1951; Merritt, 1993; Hofstede, 1991). Merritt (1993) defines culture as the values/norms and practices shared with others which define a group, especially in relation to other groups. Kluckhohm (1951) puts it that “culture consists in patterned ways of thinking, feeling and reacting, acquired and transmitted mainly by symbols, constituting the distinctive achievements of human groups, including their embodiments in artefacts; the essential core of culture consists of traditional ideas and especially their attached values” (Hofstede, 1991, p. 25). This can influence people to act in an individualistic or collective manner; how they relate to interpersonal power; how they will tolerate uncertainty, and the way they relate to members of the opposite gender (Hofstede, 1991 chap 4, 5). These factors have been shown to play an important role in accident causation (Merrit & Helmreich, 1998, p. 98-100).
This paper focuses on national culture, defined as the traits which describe a group of people living in a particular region (Kluckholm, 1951). Value system dimensions are used to understand and measure the influence of national culture and have been identified by several authors such as Smith & Schwartz (1997), Schwartz (1999) and Hofstede (1980; 1991). They include ‘conceptions of the desirable that guides the way social actors (organisational leaders, policy makers, individual persons) select actions, evaluate people and events and explain their actions and evaluations’ (Schwartz, 1999). It is ‘trans-situational’ in that it guides a person’s principle of living across a wide range of contexts (Schwartz, 1994; Kluckholm, 1951). The extent to which these cultural interactions offer a predictor of performance lies partly in the notion of cultural context.
In High Cultural Contexts (HCC) the individual is subordinate to the group and the emphasis is on maintaining power relationships. HCCs manifest themselves in behaviours such as extreme politeness and careful judgment, verbal messages that are implicit in their content, the needs of individuals interfering with set objectives or rules, lengthy decision-making processes, and slow pace of change (Hall, 1974; 1959; Graham, 1985; Adler, 1991). It is difficult to generalize, but HCC’s tend to be found in Africa, Arab nations, China, and Japan. Low Cultural Context (LCC), on the other hand, tends to be the opposite. LCCs tend to foreground verbal messages that are explicit in context, impersonal negotiations, a focus on economic goals, quick and efficient decision making, and rapid change. It can be found in many Western countries such as North America and Europe (Hall, 1976; Nishimura, Nevgi & Tella, 2008; Neuliep, 2012). Neither HCC or LCC is better than another, rather, it depends on the context and setting. Indeed, according to Neulip (2012) no national culture is exclusively high or low cultural context, instead falling along a continuum. The interest for the current paper is how a national culture positioned at one end of cultural context continuum interacts with organizational and professional sub-cultures positioned at the other. In aviation this is common. Aircraft and their associated operational, safety and maintenance procedures derive largely from LCC Western nations, but may find themselves embedded in HCC African or other nations. It would be surprising if there were not issues arising from this, but they are rarely addressed.
Culture in Aviation
It is not merely the aircraft that are built in non-indigenous cultures, it is the associated maintenance, operational and safety practices that go along with them. The key question is the extent to which this ‘alien’ entity of people, technology and processes translates into different cultural environments. A series of accidents have exposed the cultural dimension in aviation, including the Air Ontario Fokker F28 crash in March 1989 (CofI-MSSC, 1993), the 1992 Airbus A320 crash at Mount St-Odile (FMTT, 1992) and the Piper Navajo incident in June 1993 (AAIB, 1993). Accidents such as these, and numerous others, stimulate increased interest in the role of culture (e.g. Jing & Batteau, 2015) and the findings are revealing. Merrit & Helmreich (1996a; 1996b) studied the “cockpit management attitudes” between Asian and American flight crew and attendants using two of Hofstede’s cultural dimensions; ‘power distance’ and ‘individual-collective’ (see Appendix A). The report argues that the American crew were more independent, self-reliant (more individualistic), and exhibited a personal responsibility for their contribution to being an effective crew (low power distance). This was in contrast to the Asian crew, who for similar cultural reasons were more likely to support the authority of a superior (high power distance) and be satisfied with that role.
Despite what seems like overwhelming evidence for the importance of culture’s role in accident causation, it is still not fully accepted (Merrit & Helmreich, 1998; Maurino, 1998; Strauch, 2010). It is certainly more convenient to believe it is possible to have "culture free" work environments, yet it is clear there is high cultural diversity within the industry which directly impacts on operations (Helmreich & Davies, 2004). The problem, as Maurino (1998) forcefully states, is that “cultural factors should routinely be considered during the safety investigation process, although this might be the toughest nut in the entire lot to crack, owing to the resilient conservatism of accident investigation.’’ (pg. xxiii) The industry is therefore faced with the challenge of developing a practical model to deal with the relationship between different cultures, and further research is needed to confront this challenge.
Recent attempts have focused on examining the relationship between the Western origin of the aviation industry and the diverse non-Western cultures into which it has been implemented. This research confirms the importance of the research area, but identifies a number of knowledge gaps that need to be addressed. Li et al (2007) used the Human Factors Analysis and Classification System - HFACS (Wiegmann & Shappell, 2003) framework to compare accidents in China, United States and India to explore the role of cultural differences. The results show a significant difference in the accident outcome of the three countries in seven HFACS categories; however, certain causal factors could not be accounted for, such as cultural difference between the countries, the level of government oversight, and design/technology sophistication. This led Strauch (2010) to argue that the overall research methodology had flaws. To address gaps like this in existing methodologies what is needed is a systemic framework for cultural factors. Such a framework should take into account human failures, cultural beliefs, and the social context of the system. This framework should be developed for several reasons: firstly, it should identify the underlying cultural factors which led to the accident. Secondly, it should capture the contextual environment of the interactions. Thirdly, it should examine the cultural dimensions contributing to those interactions. Fourth, it should highlight the benefits of acknowledging the socio-cultural interaction in the accident investigation process and its outcomes. Such a framework is presented in the next section.
Proposed Multi-Level cultural Framework
The framework shown in Figure 1 is a synthesis of the wider background literature surveyed above, but refers more explicitly and directly to Hofsted’s (1991) cultural dimensions, Hall & Hall’s (1990) manifestations of culture in different contexts, Rasmussen’s risk management framework (1997) and Reason’s (2000) work on human error. Although structured differently to reflect the focus on cultural factors, the primary actors and elements from these models are all present. The multi-level framework of cultural factors consists of four levels. Between them they meet the objective of mapping out the socio-cultural context of events leading to unsafe acts. The high-level cultural dimensions reviewed above (Hofstede, 1991) are found at the top of the framework, called Level 1. How these high-level cultural dimensions manifest themselves in different contexts (Hall & Hall, 1990) is represented at Level 2. Levels 1 and 2 are linked to Level 3, which contains the system elements to be found in Rasmussen’s (1997) risk management framework. This, in turn, leads to Level 4 and the various points at which culturally influenced Human actions occur (Reason, 1990). The sum total of these influences are safe or unsafe acts, and beyond that accidents or loss events. These events, of course, feedback into the model to influence culture (Level 1), cultural dimensions (Level 2), system parts (Level 3) and human actions (Level 4). In other words, consistent with sociotechnical systems theory (Walker et al., 2008), the framework is itself part of the environment and able to influence what happens in it.