Organising mindfullyexpanding mindful organising’s nomological network through testing predictors and outcomes in high risk environments

Resumen

This chapter discusses the overall summary of the doctoral thesis titled “organizing mindfully: expanding mindful organizing’s nomological network through testing predictors and outcomes in high risk environments”. This summary starts off by outlining the background and rationale behind our research. We then outline our thesis objectives. Thereafter, we present each of the original research studies conducted as part of this dissertation by explaining the methodology, the findings and main conclusions of each study. We then finish with an overall conclusion about what was learnt from this research. BACKGROUND The sophisticated systems and technologies that exist today allow us to continue to prosper and advance humanity to live better connected and higher-quality lives. From nuclear power plants, aviation, to information system technology, humans have created wonderous complex systems that have had beneficial effects on our quality of life. However, we continue to grapple with devastating consequences and lost lives resulting from failures of these systems and accidents within these environments. For decades, scholars and practitioners from a plethora of disciplines have attempted to understand how to better manage safety and prevent accidents in countless high-risk industries worldwide. Through the insight gained from management and research over the years, we have learned a great deal about the socio-technical factors at play for safety management. These socio-technical factors are multifaceted and various authors have created frameworks and theories to give leaders and safety management practitioners tools to better manage safety. Despite the increasing sophistication in our understanding of risk and safety management, it is estimated that workplace fatalities are on the rise, with a recent report based on statistics from the International Labour Organization estimating that 2.78 million people die every year from occupational accidents and work-related illnesses (Takala et al., 2017). We are still left with the challenge of better understanding increasingly complex systems and creating better frameworks and tools to avoid accidents and keep people and communities out of harm’s way. As our systems continue to become more and more complex, our ability to manage safety does too, because the number of interactions between the social system (workers) and the technical system (technologies) in an organization continue to rise, making it impossible to map out every interaction that could happen. This rising complexity makes organizations more and more vulnerable to unexpected events that could lead to accidents and errors. Therefore, safety management theorists have started to shift their focus away from analysing the causes of errors and accidents (as this is becoming futile with rising complexity) and towards understanding how to build resilience in organizational systems so that unexpected events do not destabilise the system (Hollnagel et al., 2015). This lead scholars to study organizations that had very high levels of resilience, to try to extrapolate lessons of how to build resilience. Nuclear power plants and air traffic control centres immediately stood out as outliers in terms of resilience and reliability, as these organizations managed to operate within an exceptionally high-risk environment, with exceptionally low accident rates (Rochlin et al., 1987). They found that these organizations did experience unexpected events and errors, but these unexpected events and errors did not destabilise them (Weick et al., 1999; Schulman, 2004). Through analyses of how these organizations managed to achieve such high resilience and reliability, they saw that these organizations designed for safety on a systems level and had a very intricate understanding of their operations with highly mapped our procedures and protocols (Schulman, 2004). Beyond that, they exhibited the social and relational infrastructure that allowed them to expertly manage unexpected events (Weick & Roberts, 1993). When analysing this social and relational infrastructure, researchers discovered that these highly resilient organizations had specialised team dynamics characterised by mindful actions and interactions (Weick & Roberts, 1993). These mindful actions and interactions allowed teams to be able to anticipate when something was about to go wrong, and quickly contain this potential problem before it caused more serious harm (Weick et al., 1999). They called these team dynamics “mindful organizing”, and since its inception, it has been studied in numerous high-risk environments and has shown to have beneficial effects on performance (Sutcliffe et al., 2016). Mindful organizing offers a promising framework to use to help other high risk organizations better manage safety. However, its current utility in research and practice is stunted due to a lack of clear conceptualisation of the construct, a limited understanding of the conditions needed to support and sustain it in organizations as well as limited insight into how it affects individual behaviour and attitudes. Mindful organizing research has been criticized for being too narrow in focus, too linear in its level of analysis and not socially embedded enough, making it difficult to create and sustain in practice (Martínez-Córcoles & Vogus, 2020). Mindful organizing was discovered in the unique organizational context of high reliability organizations, with various conditions and practices that allow for this team level capability to be constantly enacted and re-enacted in teams. To advance our understanding of mindful organizing, we need to attempt to build a map of key factors and organizational practices that are important for creating and sustaining mindful organizing. In high-risk environments concerned with safety, we have empirical evidence that specific leadership characteristics and certain practices such as training can have a positive impact on mindful organizing (Sutcliffe et al., 2016). However, there has hardly been any investigations into the team level conditions and norms that are important for mindful organizing, even though it is described as a bottom-up, fragile emergent property in teams (Vogus & Sutcliffe, 2011). We also know that mindful organizing leads to better safety performance by looking at objective safety indicators and safety reports in high-risk settings (e.g. Oliver et al., 2019; Ausserhofer et al. 2013), but we still have a very limited understanding of how it impacts individual safety behaviours and attitudes. Examining these relationships empirically will allow us to understand whether the safer practices found in teams that engage in mindful organizing come just from mindful organizing, or whether mindful organizing promotes better safety practices on an individual-level too. OBJECTIVES Our research has three overarching main objectives: (1) To clarify the concept of mindful organizing by theoretically and empirically differentiating it from related team level constructs, (2) to understand which team level climates and norms to do with safety and communication are important for mindful organizing to develop and (3) to understand the impact of mindful organizing on individual safety behaviour and commitment while investigating motivational and affective mechanisms that may act as mediators of these relationships. In this dissertation, we start by explaining the evolution of safety management theories, with special focus on high reliability organization theory (Chapter I), we then move onto a theoretical review of the concept of mindful organizing (Chapter II). Chapter III describes the objectives of this thesis and the methodology used to carry out each empirical study presented within this thesis. The four studies carried out for this thesis are found in Chapters IV, V and VI, and we present a general discussion of our findings in Chapter VII. Finally, we finish with the most relevant conclusions drawn from our work in Chapter VIII. As discussed in chapter II, there has recently been an increasing number of quantitative studies into mindful organizing, but, our understanding of mindful organizing comes largely from case studies and qualitative investigations. This leaves those interested in mindful organizing with many questions about its conceptualisation as there has been limited testing of the psychometric properties of mindful organizing scales and it has not been empirically differentiated from related team safety variables. Earlier research into mindful organizing described it as “the team level enactment of safety culture” (Vogus & Sutcliffe, 2007), as these authors claimed that mindful organizing represents the team level behaviours that drive safety culture in organizations”. However, the conceptualisation of safety culture (e.g. Zohar, 2008) differs substantially from mindful organizing. Later Sutcliffe et al. (2016) recognised that mindful organizing needs to be differentiated empirically from constructs that measure the team or organizational prioritisation of safety such as safety culture and safety climate. These authors also highlighted that mindful organizing needs to be differentiated from other related team level behaviours such as team learning too. Therefore, the first overall research question we wanted to answer is: is mindful organizing a distinct construct from other important team safety variables? We attempt to answer this question in study 1 where we validate a measure of mindful organizing (in Spanish), exploring its factor structure, whether it can be reliably aggregated to a team level as well as its distinctiveness from (1) team safety climate, (2) organizational safety climate and (3) team learning. We also looked at the incremental validity of mindful organizing in predicting safety outcomes (safety compliance and participation) over and above these three important safety variables. Study 2, 3 and 4 investigate antecedents and outcomes of mindful organizing, using multidimensional models examining team level antecedents and individual outcomes of mindful organizing. Mindful organizing is said to be observed in the actions and interactions of team members, with conversations between team members described as an important driver each of the five processes (Sutcliffe et al., 2016). However, little is known about which communication conditions and norms are important for mindful organizing to develop. This led us to ask our second overall research question: which team level participatory communication conditions are important for mindful organizing to develop? The impact that mindful organizing has on employee’s wellbeing and affective responses at work is a subject of some controversy. It is expected that in intensive, high-risk environments (nursing units in hospitals), mindful organizing gives teams’ resources to cope which positively impacts their well-being, however, it negatively impacts wellbeing in less intensive environments as it is taxing making it difficult to maintain (Vogus et al., 2014). This raises important questions about the sustainability of mindful organizing in high-risk environments that experience less consistent adversity than some medical settings as it could be too taxing for employees to consistently engage in. This led us to ask a third overall research question: what impact does mindful organizing have on team’s subjective experience at work and individual’s propensity to leave their organization? Study 2 attempts to answer our second and third research question by drawing on the current theory about engagement, voice and psychological safety to propose two specific participatory communication predictors of mindful organizing: participation climate and perceived safety for upward dissent. We also draw on the job demands-resources model (Bakker & Demerouti, 2006) and traditional needs theory (Maslow, 1981) to extend theory and test the impact of mindful organizing on team’s job satisfaction and individual’s turnover intention. We do so by testing a time-lagged multilevel structural equation model using data from 47 teams working in a nuclear power plant. Study 1 showed us that mindful organizing is related to, but distinct from, team safety climate. It is argued that mindful organizing may have a reciprocal relationship with safety climate (Sutcliffe et al., 2016). However, the nature and direction of the relationship between team safety climate and mindful organizing have virtually been unexplored. This led us to ask our fourth overall research question: Is team safety climate a necessary prerequisite for mindful organizing to develop? Various studies have shown that mindful organizing improves objective indicators of safety in high-risk environments (e.g. Vogus & Sutcliffe, 2007; Ausserhofer et al. 2013). Although useful in showing us the objective value of mindful organizing in enhancing safety, these models do not test which individual safety behaviours are stimulated by team level mindful organizing leading to increased reliability and fewer accidents. Models using objective indicators of safety (e.g. medication errors) are also specific to certain environments and industries, not offering much insight to other organizations about how mindful organizing may effect more generalisable, individual behaviours. We also do not know which cognitive-motivational mechanisms mindful organizing may impact leading to individual safety behaviours. This leads us to have an incomplete understanding of the impact of team mindful organizing on individual performance in high-risk industries. Therefore, our fifth and sixth overall research questions are: Does mindful organizing increase individual in-role and extra-role safety behaviours? And if so, do capability motivational drivers mediate the relationship between mindful organizing and individual safety behaviours? We attempt to answer the fourth, fifth and sixth research questions by two studies (study 3 and 4) within a sample of chemical workers in Russia and Ukraine. Study 3 will assess whether mindful organizing mediates the relationship between group safety climate and individual safety citizenship behaviours. Study 4 will explore whether mindful organizing affects individual capability motivational state (role breadth self-efficacy), leading to increased individual safety citizenship behaviour. STUDY MEHODOLOGY Descriptive Analyses For each of the four studies enclosed within this thesis, various preliminary descriptive analyses were conducted. Descriptive statistics were calculated and reported for the data within each study (e.g. means and standard deviations). The reliability of each of the scales (as reported above) were also calculated using Chronbach’s alpha coefficient (1951). For study 1, which entailed the validation of a Spanish version of the mindful organizing scale, various other reliability analyses were carried out, over and above the Cronbachs alpha coefficient. The average variance extracted (AVE) value and composite reliability value (rho) were also examined to ascertain the internal consistency of the scale. For AVE, values of .50 or greater indicate satisfactory reliability as the variance of the construct is greater than the error variance (Fornell & Larcker, 1981). For composite reliability (rho values), a score of .70 or greater indicates good reliability (Raykov, 2001). Aggregation indices Mindful organizing is a team construct, therefore for every study in this thesis individual’s scores were aggregated to form a team mindful organizing score as is common practice in measuring this variable (e.g. Ausserhofer et al. 2013; Vogus & Sutcliffe, 2007b). Likewise, team safety climate (measured in study 1 and study 3), as well as participation climate and perceived safety for upward dissent (measured in study 2), were also analysed at the team level by aggregating individual scores. Before aggregating the scores of these four variables (mindful organizing, team safety climate, participation climate and perceived safety for upward dissent), we had to calculate the aggregation indices for each of these variables. Calculating aggregation indices are essential to demonstrate that each member’s score was similar enough to those within their team, and different enough from those not in their team, to justify aggregating these scores. Therefore, in every study, we calculated the average deviation index (ADI; Burke, Finkelstein, & Dusig, 1999), the intra-class correlation coefficients (ICC; Bliese, 2000) and the analysis of variance (ANOVA) scores for each of the variables that were aggregated to the team level. Average Deviation Indices (ADIs) and intra-class correlation coefficients (ICCs) were computed and analysed to ensure within-team agreement. Analysis of Variance (ANOVA) was computed to ascertain the between-team distinctiveness of scores. Confirmatory Factor Analysis In all four studies, the confirmatory factor analyses were run using the programme Mplus (Muthén & Muthén, 1998–2010). Each CFA differed depending on the study model and are therefore described separately, the criteria used to evaluate the CFA’s was the same for all models. Study Specific Analyses Study 1 The first study included a sample of 573 Spanish nuclear power plant workers, who made up 47 teams. In this study, we aimed to validate a measure of mindful organizing. To do so, mindful organizing was correlated with safety culture, team safety climate and team learning. This analysis was done with the aggregated group level scores. Thereafter, a CFA was run with Mplus to show the distinctiveness of mindful organizing compared with these three variables. We compared two alternative models: a one-factor model (with all items loading in a single factor), and a four-factor model (with items loading in their corresponding scale). For running the CFA individual-level scores were used, as a large number of items prevented us from using scores aggregated to the team level. Model fit indices were analysed according to the criteria mentioned at the beginning of the CFA section. Study 2 Study 2 utilised a time-lagged design, so there were two data collection points (in 2014 and 2016). In 2014 (Time 1), 58 teams comprising of 615 employees participated in the study, yielding a response rate of 76.3%. In 2016 (Time 2), 54 teams comprising of 607 employees participated in the study, yielding a response rate of 72.5%. The final sample included 47 teams (comprising 425 employees), which were those that answered both in 2014 (N = 427) and in 2016 (N = 425) and had at least 2 subjects each time (Kozlowski & Bell, 2003). In this study, we aimed to test a multilevel structural equation model of predictors and outcomes of mindful organizing with two time points. The model chosen is a moderated mediation model. The proposed model wanted to test whether the interaction of perceived safety for upward dissent and participation climate leads to mindful organizing and whether job satisfaction mediates the relationship between mindful organizing and turnover intention. All variables were measured at the team level, except turnover intention, which was measured at the individual-level. To test the first hypothesis, the statistical significance of a3 (the coefficient estimating the moderator effect of perceived safety for upward dissent in the relationship between participation climate and mindful organizing) was tested. To further probe the interaction effect we used the Process macro for SPSS (Hayes, 2018) to estimate the slopes of the relationship between participation climate and mindful organizing at high and low values (one standard deviation above and below the sample mean) of perceived safety for upward dissent and to plot the corresponding regression lines. To test the significance of the indirect effect stated in the second hypothesis, we used bias-corrected (BC) bootstrap confidence interval (CI) method (MacKinnon, Lockwood, & Williams, 2004) as implemented in Mplus. A bootstrap sample size of 5000 was used. The b1c1 indirect effect was calculated, where b1 is the coefficient estimating the relationship between mindful organizing and job satisfaction, and c1 is the coefficient estimating the relationship between job satisfaction and turnover intention. Mediation is supported when the BC bootstrap confidence interval for the indirect effect does not include the zero value. Finally, to test the conditional indirect effect stated in Hypothesis 3 we also used BC bootstrap confidence interval method as implemented in Mplus. The (a1+a3W)b1c1 conditional indirect effect was calculated, where W is the moderator variable (perceived safety for upward dissent), a1 is the coefficient estimating the relationship between participation climate and mindful organizing, and a3, b1 and c1 are the coefficients estimating the relationships previously stated. The conditional indirect effect is supported when the BC bootstrap confidence interval for the difference in the indirect effect (diff_IE) among different levels of the moderator does not contain zero (Preacher et al., 2007), which implies that the strength of the indirect effect (a1b1) depends on the level of the moderator variable (W). Study 3 The data used in study 3 was collected within a sample of Russian-based chemical plant workers (N = 1112) comprising of 98 teams. To test our proposed model in study 3, we ran a multilevel structural equation model. This model wanted to ascertain whether mindful organizing mediated the relationship between team safety climate and five individual safety behaviours. Group safety climate and mindful organizing were analysed on the team level and the five individual safety behaviours (safety compliance, routine violation, voice, initiative and helping) were analysed on the individual-level. To confirm the proposed model, we first analyzed the model fit indices (using the CFA criteria stipulated in the previous section). If the model fit was satisfactory, we considered the pathway estimates by looking at whether they were significant and the strength of each pathway. Thereafter, Monte Carlo (MC) confidence intervals were used for testing the significance of the indirect effects, as it is argued to be a more viable and robust method for calculating confidence intervals for complex and simple indirect effects when working with a multilevel model (Preacher and Selig, 2012). Study 4 The data for study 4 came from a sample of Ukraine-based chemical plant workers (N = 443) comprising of 50 teams, from the same parent company as study 3. For this study, we also ran a multilevel structural equation model to assess our proposed mediation model and the pathways between our variables. This model wanted to test whether individual role breadth self-efficacy mediated the relationship between mindful organizing and the five individual safety behaviours (safety compliance, routine violation, voice, initiative and helping). For this analysis, mindful organizing was aggregated to the team level and the individual safety behaviours were analysed at the individual-level. To confirm the proposed model, we first analyzed the model fit indices (using the CFA criteria stipulated in the previous section). If the model fit was satisfactory, we considered the pathway estimates by looking at whether they were significant and the strength of each pathway. Monte Carlo (MC) confidence intervals were used for testing the significance of the indirect effects, as it is argued to be a more viable and robust method for calculating confidence intervals for complex and simple indirect effects when working with a multilevel model (Preacher and Selig, 2012). FINDINGS AND CONCLUSIONS In study 1, we showed that the widely used mindful organizing scale (by Vogus & Sutcliffe, 2007) does measure one underlying construct, and it does so reliably. The findings confirmed that mindful organizing is positively related to team learning, team safety climate and organizational safety climate, but these variables are distinct from one another when looking at their factor structure. We were also able to provide evidence that mindful organizing affects safety behaviours (participation and compliance) over and above team learning, team safety climate and organizational safety climate. This shows empirically that engaging in mindful organizing is a distinct contributor to safer actions as it helps employees to engage in safety behaviours over and above team learning, team safety climate and organizational safety climate. After clarifying mindful organizing as a distinct construct from related team variables and testing the psychometric properties of our mindful organizing measure, we wanted to expand our understanding of the needed team level communication conditions and climate for mindful organizing to be created and sustained in study 2. This was following speculations in the literature that conversations were the main driver of mindful organizing (Sutcliffe et al., 2016) and that despite mindful organizing being created and maintained through communication channels, there is hardly any enquiry into the communication conditions that lead to higher mindful organizing or may stifle mindful organizing (Ford, 2018). In study 2, we draw on literature about psychological safety and voice and hypothesized a model where two participatory communication conditions interact to predict mindful organizing. We found that perceived safety for upward dissent significantly moderated the relationship between participation climate and mindful organizing . This showed that for mindful organizing to develop, teams not only needed to perceive that their participation (in sharing ideas and suggestions) within the organization was valued, but they also needed to feel safe to voice their concerns and disagreements with their superiors without fear of backlash. This perceived safety to disagree with supervisors turned out to be critically important, as our findings showed that without it, the relationship between a participation climate and mindful organizing became non-significant. These results are promising since the data are longitudinal, showing that the interaction of perceived safety for upward dissent and participation climate at time one leads to mindful organizing at time two, giving some evidence of a possible dynamic relationship between these variables. Study 2 also attempted to investigate the relationship between mindful organizing and team’s subjective experience at work by investigating whether mindful organizing has a positive impact on job satisfaction given the controversy around the relationship between mindful organizing and employees affective responses at work (e.g. Rerup, 2006; Vogus et al., 2014). It has been argued that mindful organizing is particularly draining for employees operating in high-risk environments with few adverse events, negatively impacting wellbeing (Vogus et al., 2014). If this is true, it suggests that mindful organizing may not be sustainable in high risk environments that only sporadically experience adverse events, as employee’s wellbeing would suffer and they would not want to remain in their current, taxing positions. However, mindful organizing has been observed in various case studies in HROs where employees engage in demanding, risky work but do not necessarily experience constant, ongoing adverse events (e.g. nuclear power plants) and these teams manage to sustain mindful organizing as an ongoing practice. We therefore tested (within our model) the impact of mindful organizing on individual turnover intention, with team job satisfaction mediating the relationship. The results of the pathway between mindful organizing and team job satisfaction showed a strong positive and significant relationship. Team job satisfaction mediated the relationship between mindful organizing and individual turnover intention, showing that through increasing a team’s job satisfaction, mindful organizing resulted in lower turnover intentions. This suggests that in a nuclear power plant, it is far better for team’s satisfaction at work to engage in mindful organizing than to not engage in mindful organizing, even though being collectively mindful can be taxing. This suggests that mindful organizing offers teams much needed resources to cope with the demands of their working environment, even if there are not continuous adverse events, leading to a higher desire to stay in their organization. These results provide evidence of the sustainability of mindful organizing as it not only improves reliable and safe performance but it also positively impacts job satisfaction and commitment. The holistic model in study 2 also showed us the critical importance of teams feeling safe to disagree with supervisors. The findings showed that perceived safety for upward dissent moderated the negative indirect effect of participation climate on turnover intention through mindful organizing and job satisfaction. This means teams feeling safe to disagree with supervisors facilitated the relationships between our study variables, and without it, a high participation climate would not lead to higher mindful organizing, stifling team job satisfaction resulting in higher turnover intention for individuals. Study 1 showed us that mindful organizing is positively related to team and organizational safety climate but it conceptually distinct from these variables. Safety climate refers to the perceived priority a team or organization give to safety above other competing demands. The nature and direction of the relationship between mindful organizing and safety climate are unclear. It is speculated that teams engaging in mindful organizing may enhance the team’s safety climate (Sutcliffe et al., 2016), but in study 3, we wanted to ask: is team safety climate a necessary prerequisite for mindful organizing to develop? Team safety climate is a well understood, implemented and measured concept in organizations. Understanding the impact of team safety climate on mindful organizing could help scholars and practitioners with enhancing this collective team capability into organizations. This is because focusing on team safety climate as a supporting prerequisite could help to sustain and “socially embed” mindful organizing in practice. Study 3 also wanted to examine the impact mindful organizing may have on individual safety behaviours, as these individual behaviours may be contributing to the higher safety and reliability seen in organizations that have high mindful organizing in their teams. We tested a holistic mediation model which examined whether team mindful organizing mediated the relationship between team safety climate and various individual safety behaviours. Our research findings confirmed this mediation model. This not only showed that team safety climate is an important prerequisite for mindful organizing (with a strong positive and significant pathways between these two variables), but mindful organizing did lead to higher in-role and extra-role safety behaviours. In other words, the results show that when teams perceive that safety is prioritised above all else (high safety climate), mindful organizing is stimulated and increases individuals’ propensity to not only comply with prescribed safety standards but also to engage in extra-role safety behaviours (helping, voice and initiative) that are not expected of them by the formal rules and procedures. After confirming that mindful organizing does stimulate individual safety citizenship behaviours and safety compliance, we wanted to further examine the possible motivational drivers that could be affecting these relationships in study 4. We were particularly interested in the impact that mindful organizing may have on individual role breadth self-efficacy. We speculated that the increased propensity of individuals belonging to teams that organize mindfully to engage in extra-role safety behaviours came from mindful organizing increasing their self-efficacy to do so. In other words, individuals that engaged in the five processes of mindful organizing in their team, feel more confident in their ability to engage in safety-enhancing tasks beyond their formal job description. To our knowledge, no study has looked at how mindful organizing affects individual capability perceptions. Even though it could explain some of the important individual motivation and behaviour changes that come about from mindful organizing, aiding in the higher safety and reliability seen in various studies. Therefore, we ran a multilevel mediation model where individual role breadth self-efficacy mediated the relationship between team mindful organizing and individual safety citizenship behaviours and safety compliance. This mediation model was confirmed and the pathways showed that team mindful organizing has a positive statistically significant impact on individual role breadth self-efficacy, and this individual role breadth self-efficacy positively predicted individual in-role and extra-role safety behaviours. These findings suggest that being in a team that engages in mindful organizing helps individuals who may not originally have the self-efficacy to engage in a wide range of extra-role tasks and actions to do with safety, grow their confidence in performing these tasks. This new-found confidence leads to individuals going above and beyond what is required of them in the pursuit of safer practices. OVERALL CONCLUSION In conclusion, the insight gained from the four studies conducted allowed us to significantly extend mindful organizing’s nomological network. In doing so, we managed to reach all three of our research objectives. First, by empirically distinguishing mindful organizing from related team variables. Second, by uncovering three team level climates and norms to do with safety and communication that are important for fostering mindful organizing. Third, by showing how mindful organizing impacts team satisfaction, individual self-efficacy and individual in-role and extra-role safety behaviours.