Healthy organization: analysing its meaning based on the HERO Model / Organizaciones saludables: analizando su significado desde el Modelo HERO

The aim of this study is to analyse the meaning of healthy organization from an empirical-theoretical perspective based on the HERO Model (HEalthy & Resilient Organizations). Analyses were performed by four independent judges on 14 interviews carried out with 14 CEOs or human resources managers in 14 Spanish organizations using content analysis. Qualitative results show: (1) a partial overlap in the categories proposed by the theoretical model (based on the concordance index, Cohen’s Kappa and ICC); and (2) that the empirical definition mainly focuses on employees’ psychosocial health as a key element of the meaning of healthy organization. Finally, categorical matrixes provide evidence of subcategories emanating from the key elements that comprise a healthy organization. Results as well as theoretical and practical implications are discussed based on the HERO Model.

organization stems from the company's behaviour through a shared mission and effective leadership; this achieves a balance in the relations between the employees, the clients and the organization, which then results in its commitment to social responsibility in both its values and its results. Therefore, considering an organization healthy means taking a broad view of it, where aspects like the characteristics of the work systems, cultural values and organizational climate are taken into account (Wilson, DeJoy, Vandenberg, Richardson, & McGrath, 2004). One of the aspects that studies have pointed to as relevant when developing a healthy organization is the employees' health, in that this poses a competitive advantage for organizations and caring for employees therefore has positive consequences in its wellbeing in terms of organizational performance and the organization's financial health (Cooper & Cartwright, 1994;Luthans et al., 2007;Salanova & Schaufeli, 2009;Shuck, Rocco, & Albornoz, 2011;Wright & McMahan, 1992).
In this sense, the mixed committee of the International Labour Organization (ILO) and the World Health Organization (WHO) suggested that the goal of health at work consists of successfully promoting and maintaining the highest degree of physical, mental and social wellbeing of employees in all jobs (ILO, 2003). Tarride, Zamorano, and Varela (2008) conducted a review of the definitions of healthy organization and concluded that work organizations are a system that involves a state of physical, mental and social wellbeing that is neither additive nor linear but that instead depends on the context of the organization and the people making it up. Therefore, physical, mental and social wellbeing belong to the organization, that is, to the system as a whole and not to its parts. Thus, we understand that encouraging the health of both the employees and the organization is a core factor in promoting healthy organizations. In this way, healthy organizations can simultaneously fulfil their mission and develop and encourage their employees' learning, growth and health.
From this, studies emerged that propose a comprehensive model of healthy organizations (DeJoy, Wilson, Vandenberg, McGrath-Higgins, & Griffin-Blake, 2010;Wilson et al., 2004). These studies try to test the heuristic model of healthy organizations which integrates employees' health as well as variables referring to the organization's context (such as work demands, tools and technologies and the social setting) and performance. These studies are an initial approach to understanding how an organization's practices are related to its employees' health. However, the validation of these initial comprehensive models of healthy organizations (DeJoy et al., 2010;Wilson et al., 2004) showed several limitations: (1) the data were gathered using the same source of information (employees) with the same measurement instruments, turning the common variance into potential bias in the data; and (2) the constructs were tested on the level of individual analysis, even though the premises underlying the concept of healthy organization require these models to be examined at a collective level of analysis. Following the same lines, other studies have considered organizations that invest in the health, resilience and motivation of their employees and work teams, as well as in the structure and control of work processes, and in healthy outcomes oriented at achieving income and excellence for society, to be healthy and resilient organizations (HEalthy & Resilient Organizations, HERO;Salanova, Llorens, Cifre, & Martinez, 2012). From a psychosocial perspective, the HERO model takes a step further towards considering that a healthy organization encompasses the health of the employees not only in their work environment but also outside of work, affecting the community. Here is where the organizational resources and practices that the organization invests in become a cornerstone in the development of HEROs.
Recently Salanova et al. (2012, p. 788) have defined HEROs as 'organizations that make systematic, planned and proactive efforts to improve the processes and results of their employees and of the organization. These efforts are related to organizational resources and practices and to the characteristics of the work at three levels: (1) job level (such as redesigning jobs to improve autonomy, feedback); (2) social level (such as transformational leadership); and (3) organizational level (such as work-family balance practices)'.
Based on these theoretical and empirical premises, we believe that a healthy, resilient organization combines three key elements that interact with each other: (1) healthy organizational resources and practices (such as leadership); (2) healthy employees (such as work engagement); and (3) healthy organizational outcomes (such as high performance) (Salanova, 2009;Salanova, Cifre, Llorens, Martínez, & Lorente, 2011;Salanova et al., 2012) (see Figure 1). Since it is a heuristic model, so far specific relationships between some variables of the key components of the HERO model have been tested using quantitative and qualitative methodologies. Some examples of quantitative studies reveal the mediating role: (1) of organizational trust between organizational practices implemented from Human Resources Management and team work engagement (Acosta, Salanova, & Llorens, 2012a); (2) of team work engagement between transformational leadership and performance (Cruz-Ortiz, Salanova, & Martínez, 2013); (3) of collective engagement between personal resources and service quality (Hernández, Llorens, & Rodríguez, 2014); and (4) of team work engagement between team resources and performance as evaluated by supervisors (Torrente, Salanova, Llorens, & Schaufeli, 2012). However, to our knowledge, no studies have been performed that qualitatively examine the definition and key elements of a healthy organization. Specifically, the studies on HEROs carried out by our team using the qualitative methodology have focused on: (1) evaluating the perceptions of healthy organizations using a 10-point Likert scale which ranged from 0 ('not healthy') to 10 ('very healthy') (Salanova et al., 2011); (2) analysing healthy organizational practices and healthy organizational outcomes ; and (3) analysing the frequency of healthy organizational practices in small and medium-sized enterprises (SMEs) (Acosta, Salanova, & Llorens, 2012b). In this sense, Sorge and van Witteloostuijn (2004) and Vanderberg et al. (2002) suggest that there is a broad corpus of knowledge on theories of healthy organizations but that this knowledge is not interconnected. According to these authors, this knowledge should be integrated through evidence based on consulting, as well as empirical evidence that would provide the groundwork for newer theoretical models. Therefore, this study strives to go a step further by more deeply examining the definition and key elements of a healthy organization using a qualitative methodology using content analysis from both an empirical and theoretical approach based on the HERO model (HEalthy & Resilient Organization;Salanova et al., 2012).

Method
Participants and procedure The sample was 14 key stakeholders (80% men) belonging to 14 Spanish organizations. The interviewees had to have thorough knowledge of their  organizations. To ensure this, we considered two requirements: (1) their current position in the company, which should enable them to have a global view of the organization; and (2) their tenure in the company. We interviewed 11 (79%) CEOs and three (21%) human resources managers. The average number of years working in the company was 18 years (SD = 10). Ten (77%) of the organizations belonged to the services sector (including education, retail, entertainment and leisure, research, tourism, financial services and non-governmental organizations) and four (23%) belonged to the production sector (including construction and manufacturing).
The organizations were chosen by convenience, and participation was voluntary. The contact with the key stakeholders was initially via telephone and later in person. They were told the objectives of the study and were guaranteed the confidentiality and anonymity of the information. Once they agreed to participate, two expert researchers held interviews lasting approximately 45 minutes. To avoid biases, with the consent of the key stakeholders the interview was recorded and later transcribed verbatim.

HERO interview
We used the interview that is part of the HERO battery of instruments , which evaluates healthy and resilient organizations. Specifically, the interview script contains 27 open-ended and semi-structured questions divided into four sections: (1) history of the organization (such as achievements and organizational changes); (2) definition of a healthy organization; (3) healthy organizational practices (such as implementation of healthy organizational practices); and (4) healthy organizational outcomes (such as financial health). In this study, we focused on the second part of the interview, that is, the definition of a healthy organization. To date, no studies have been conducted that focus on defining a healthy organization based on the perception of CEOs and human resources managers in organizations.

Analysis of the interviews
The interviews were analysed using content analysis (Ahuvia, 2001). This technique is widely used to analyse categories and reach conclusions based on a previous theoretical framework (Denecke & Nejdl, 2009;Dick, 2004). Furthermore, content analysis is a flexible technique which combines categories in a proposed theoretical model with sub-categories that emanate from the data analysed (Hsieh & Shannon, 2005). This analysis is performed by trained, independent codifiers with the aim of creating a system of categories that are mutually exclusive, reliable and valid (Weick, 1985). Four judges were chosen to participate in the analysis of the information because they are experts in Positive Occupational Health Psychology. Two of them hold PhDs in psychology and two have a Master's in work and organizational psychology.
Specifically, the information was codified through two analysis strategies: (1) one focused on defining a healthy organization; and (2) another focused on the key elements making up a healthy organization. The first strategy enables us to identify categories related to the definitions of healthy organizations. Based on Cassell and Symon (2004), the four judges reached the consensus that they believed that the theoretical definition of a healthy organization contained two categories. The first of them, practices, included the following sub-categories: job practices, social practices, organizational practices and individual practices. The second category, results, included employees' health results, financial results, excellence results and results associated with the environment and community with which the organization interacts. Afterward, utilizing inter-judge assessments, we identified the identical features in the proposed definitions of healthy organizations provided by the key stakeholders. To ascertain the degree of agreement among the judges, we calculated Cohen's Kappa statistic, which evaluates whether the degree of agreement among the judges is lower or higher than what could be expected at random (Kottner, 2009); it is a useful coefficient when the pattern of all responses is comparable to an already determined standard (Muñoz-Leiva, Montoro-Ríos, & Luque-Martínez, 2006). Values between .81 and 1 can be interpreted as 'very good'; those between .61 and .80 as 'good'; those between .41 and .60 as 'moderate'; and those between .21 and 40 'low'. Values under .21 are regarded as 'poor' agreement (Altman, 1991). Following these analyses, we calculated the concordance rate (CR). In this case, we considered evaluations to be concordant when the CR values of agreements ([agreements + disagreements]) is ≥ .80 (Tversky, 1977), which is more restrictive than Cohen's Kappa values. We also considered the percentage of agreement, calculated as (number of agreements/total possible agreements)/100. Finally, the judges evaluated the degree of fit between the theoretical definition proposed by the HERO model and the definition provided by the key stakeholders (Likert scale ranging from 0 = 'No match' to 6 = 'Total match'). In this case, since the variables are continuous, we used the SPSS programme (version 19.0) to calculate the Intraclass Correlation Coefficient (ICC; Bliese, 2000) with the goal of evaluating the consistency of the information yielded. The average reliability for this calculation of all the judges using the mean ICC was calculated applying the Spearman-Brown reliability correction (Wuensch, 2007). For the data saturation (Guest, Bunce, & Johnson, 2006), we used four principles proposed by Francis et al. (2010) for content analysis, namely: (1) initial sample; (2) stopping criterion (a criterion which considers whether the data saturation has happened); (3) independent judges; and (4) data saturation (Francis et al., 2010). Regarding the practices category, no new information emerged after the tenth interview. Specifically, the interviewees mentioned: job practices (e.g., strategic planning), social practices (e.g., interpersonal relations, leadership, teamwork and communication) and organizational practices (e.g., worker development, working conditions, work-family balance policies). Regarding the results category, just as in the previous category no new information emerged after the tenth interview. Specifically, the interviewees mentioned: individual health (e.g., workplace psychosocial wellbeing, and psychosocial wellbeing outside of work), financial health (e.g., production), excellence results (e.g., performance) and environmental results (e.g., reputation). The judges decided to include four more interviews with the goal of ensuring the data saturation process. After the 14 interviews analysed, the judges found no new information coming from the data.
The second analysis strategy revolved around the key elements making up a healthy organization. It comprises the categorization and codification of information which the judges did using paper and pencil with template analysis (King, 2004) based on the three key elements proposed by the HERO model (healthy organizational resources and practices, healthy employees and healthy organizational outcomes; Salanova et al., 2012). Template analysis is a flexible technique which allows the qualitative information obtained to be organized, and which captures the codified data in an explanatory matrix. The same four expert judges categorized the 14 interviews by consensus. They used two criteria to codify the information: (1) each company was assigned a number (from 1 to 14); and (2) correlative numbers were assigned to each statement said by the key stakeholders from each company. The phrases were numbered from 1 to 50. Therefore, regarding the order of codification, the first number corresponds to the company and the second to the statement (such as 2:11). Later, the judges made a category tree by consensus in order to make a category map of the information provided by the key stakeholders regarding the elements of a healthy organization. Table 1 shows the categories and agreement among the four expert judges (number of agreements) in the categories defining a healthy organization given by the key stakeholders. To reach this, we considered Cohen's Kappa alpha, the concordance rate (CR) and the percentage of agreement. The results reveal that the judges showed 'some' degree of agreement in the definition of a healthy organization given by the key stakeholders when comparing it to the theoretical definition, with the most agreement in the sample being 'the result of health in employees'. In short, the judges stated that all the definitions of healthy organization given by the key stakeholders referred primarily to the health of the employees.

Results related to the definition of a healthy organization
The degree of agreement that the judges showed regarding whether the empirical definition (given by the key stakeholders) fit the theoretical definition of healthy organization proposed by Salanova et al. (2012), the results (reached using the SPSS programme, version 19.0) show a 'high' level of agreement between the judges on the scores given to the fit between the definition of healthy organization provided (ICC: .74, p < .01) and the theoretical definition.
Furthermore the results showed a 'medium-low' agreement in the definition of a healthy organization (M = 3.21, SD = 0.50) with a response range from 0 ('No agreement') to 6 ('Total agreement').

Results focused on the elements of the HERO Model
The results of the categorization based on the three key elements in the HERO Model (healthy organizational resources and practices, healthy employees and healthy outcomes) resulted in a category tree.
Regarding the first element, healthy organizational resources and practices, two sub-categories emerged: social resources and healthy organizational practices. The first sub-category, social resources, encompasses style of communication among the members of the organization, leadership, teamwork and interpersonal relationships. One example of this sub-category is: 'organizations with fluid, direct communication' (7:26). The second sub-category, healthy organizational practices, encompasses the channels of communication used in the organization, strategic planning, traditional human resources practices, working conditions and worker development. One example of this sub-category is 'organizations in which the worker has information on their jobs, their objectives' (7:23).
Regarding the second element, healthy employees, two sub-categories emerged: psychosocial wellbeing on the job and psychosocial wellbeing off the job. One example is: 'the kind [of company] in which people work in a healthy environment for individuals in both the physical and emotional sense' (4:13).

Discussion
The purpose of this study was to analyse the meaning of healthy organization from both an empirical and theoretical approach based on the HERO Model (HEalthy & Resilient Organization; Salanova et al., 2012) in 14 Spanish organizations. This study provides a specific view of the perceptions of key stakeholders from 14 Spanish organizations regarding how they define a healthy organization and what the elements that they believe comprise one are. These conclusions lead us to discuss different theoretical implications regarding how CEOs and human resources managers conceptualize a healthy, resilient organization, as well as practical implications for management and human resources.

Theoretical implications
The results of the content analysis focused on definitions show that there is a 'partial' fit between the definition proposed theoretically by the HERO Model  and the empirical definition provided by the key stakeholders, as the latter offered a much more restrictive definition in which employees' health is at the core of the discourse. This agreement between the judges regarding employees' health (both work and non-work) as a core aspect in the information provided by the key stakeholders fit studies in POP which state the importance of caring for employees (e.g., Luthans et al., 2007) in the performance and productivity of the organization. Unquestionably, this information is extremely important given that employees' health is a factor in achieving the organizational objectives (Shuck et al., 2011), especially in these times of change and crisis, because employees are a competitive advantage (Cifre & Salanova, 2004). However, this empirical point of view ignored other basic factors such as healthy organizational practices, which the theoretical definition does include. Healthy organizational practices are the cornerstone in developing HEROs Salanova et al., 2012). A study performed by Acosta et al. (2012a) showed that healthy organizational practices (e.g., work-family balance, mobbing prevention, psychosocial health and communication programmes) are positively related to the wellbeing of work teams (e.g., organizational trust and team work engagement).
Secondly, the results of the content analysis focused on the elements comprising a HERO expand and specify each key element of the model (healthy organizational resources and practices, healthy employees and healthy organizational outcomes). Specifically, it pinpoints the element of healthy organizational resources and practices, where social resources emerge, such as kind of communication, leadership, teamwork and interpersonal relationships. These kinds of resources are important in organizations because they serve two purposes (Schaufeli & Bakker, 2004): first, they increase psychosocial wellbeing (healthy employees) and healthy organizational outcomes, and secondly they decrease psychosocial impairment (such as burnout and stress). Furthermore, channels of communication, strategic planning, traditional human resources practices (such as hiring and recruitment) and working conditions (such as kind of contract) emerge specifically as organizational practices.
Another theoretical contribution is to extend the concept of healthy organizational outcomes, which considers three intra-organizational outcomes (e.g., intra role performance) and extra-organizational outcomes (e.g., good relations with the community). The former stress production and financial results, while the latter emphasize organizational reputation. Different studies (e.g., Cooper & Cartwright, 1994;Salanova, 2008Salanova, , 2009Wright & McMahan, 1992) have stressed the organization's financial health as a core subject; however, the fact that the organization's reputation emerged as a new element opens up a new perspective on the importance of how others (clients, community, society) perceive an organization. This aspect is not included in the theoretical definition of a healthy organization. For this reason, based on the results obtained we suggest including organizational reputation in the definition and considering it within the key component called healthy organizational outcomes.

Practical implications
From a practical point of view, the results of this study reveal the limited vision that key stakeholders (CEOs or human resources managers) have compared to how the literature defines a healthy organization and the elements that comprise it, as they largely limit their definitions to employees' health without including the factors that could cause or maintain this health. Therefore, the results of this study could be used to share the importance of basic factors like organizational resources and practices when developing a HERO. In this sense, organizations can develop themselves in a healthy manner through positive interventions , such as by implementing training in specific skills sets (e.g., leadership skills), communication practices (e.g., intranet) and healthy practices (e.g., mobbing prevention practices) (Acosta et al., 2012a;Shuck et al., 2011) which would have positive repercussions on the employees, such as by increasing the levels of work engagement and team work performance (Salanova & Schaufeli, 2009).

Limitations and future research
First, the sample is made up of 14 key stakeholders belonging to 14 Spanish organizations. However, the sample size is appropriate for performing content analysis. In fact, previous studies published in scholarly journals have considered this same number of companies when performing qualitative studies (e.g., Salanova et al., 2012). Following this idea, we should stress that the companies that participated in the study come from different economic sectors. Therefore, the perceptions of the key stakeholders are varied and provide a perspective of the concept being study from both the services and production sub-sectors.
On the other hand, the analysis focused on qualitative information. Further studies could combine qualitative and quantitative methodologies, which would enable us to triangulate the information, such as through self-reporting questionnaires or daily studies by employees, supervisors and customers of the organization. These different sources of information would provide a more integrated, comprehensive view of what is meant by a healthy organization and provide specific proposals for future interventions (such as training).

Final note
This study enabled us to analyse the meaning and key components of a healthy organization from both an empirical and a theoretical approach based on the HERO Model (HEalthy & Resilient Organization; Salanova et al., 2012) in 14 Spanish organizations. Today, in times of crisis and profound changes, knowing and developing these HEROs may be the key to emerging from this situation healthy and strengthened.