Este estudio analizó la relación entre liderazgo transformacional e indicadores de bienestar y malestar de empleados de países en desarrollo, así como la mediación de la confianza en el líder. Quinientos noventa y siete empleados de organizaciones colombianas y mexicanas respondieron un cuestionario. Los resultados indicaron que el liderazgo transformacional se relaciona de forma positiva con la satisfacción laboral y de forma negativa con los síntomas de malestar y que estas relaciones están mediadas totalmente por la confianza en el líder; no se encontró relación entre el liderazgo transformacional y el bienestar psicológico. Estos resultados confirman la relación, descrita en estudios previos, entre liderazgo transformacional, menor malestar psicológico y mayor satisfacción laboral; sin embargo, sugieren que el efecto del liderazgo transformacional sobre la salud y el bienestar del empleado se limitaría a promover aspectos afectivos del bienestar, pero no del bienestar psicológico. Este resultado invita a una revisión profunda del significado de los conceptos bienestar afectivo y psicológico y las diferencias entre ellos.

Ser un alto ejecutivo puede ser un trabajo solitario. Tiene que dar mensajes difíciles, no siempre puede ser transparente acerca de sus propios desafíos y debe mantener confidenciales las decisiones clave hasta que llegue el momento adecuado.

No hay forma de escapar de las cargas de autoridad necesarias. Y, de vez en cuando, puede desarrollar una amistad con alguien en su trabajo. Una cosa es tener una amistad de compañero a compañero en el trabajo, pero otra es tener un desequilibrio de poder con su amigo. ¿Puede ser amigo de alguien que trabaje para usted, especialmente cuando su rol requiere que guarde secretos de ellos?

Considere este ejemplo: Mariah y Einat se hicieron amigas hace más de una docena de años después de descubrir su amor compartido por las actividades al aire libre. Han hecho caminatas, viajes y largos paseos en bicicleta juntas. Han sobrevivido a algunas piedras en el camino, literal y figurativamente, en su amistad.

Sin embargo, Mariah y Einat se conocieron en una relación diferente. Mariah era vicepresidenta de su compañía y jefa de Einat. Einat era directora principal en el grupo de Mariah. Claramente, cualquier amistad entre gerente y empleado está cargada de trampas. Puede dañar la amistad o la relación de trabajo. El resto del personal puede retener comentarios valiosos sobre el empleado si sienten que son amigos. Además, podría perder la confianza con su amigo y con el resto de su personal si no tiene cuidado al caminar por la delgada línea entre la confidencialidad y la transparencia.

En su rol, Mariah a menudo conocía información que impactaría en el trabajo de Einat, incluyendo posibles despidos y ascensos. A pesar de que eran amigas cercanas, Mariah tenía que mantener confidencial esta información. Confiaba en que Einat entendería las limitaciones de la transparencia debido a sus roles en el trabajo. Cuando hablé con Einat, ella lo entendió y dijo que era algo que tenía que “aguantar”.

Sin embargo, no todos los empleados tienen esta mentalidad y pueden ofenderse o sentirse heridos por la información retenida o los comentarios negativos. Por otra parte, no todos los jefes saben cómo navegar por la delgada línea de cuánto compartir y cuándo hacerlo. Sin embargo, hay formas en que puede preparar su amistad para el éxito. Aquí hay cinco consejos sobre cómo administrar una amistad con uno de sus empleados.

Elija a sus amigos cuidadosamente. Muchos de nosotros escuchamos este consejo cuando nuestras madres nos dieron un beso de despedida en nuestro primer día de jardín de infancia, pero este consejo es especialmente importante en el trabajo. Tener un amigo que a su vez es uno de sus empleados requiere altos grados de confianza y juicio por ambas partes. No es posible con cada relación de trabajo. Mariah señala que ambas partes deben ser maduras y tener suficiente autoestima para generar confianza a lo largo del tiempo. “Realmente no se trata de ciencia espacial”, explico Mariah, “solo comunicación y límites impecables”.

Establezca expectativas al comienzo. Tendrá conocimientos y responsabilidades más allá del rol y la autorización de su amigo, y su amigo debe saberlo. Sea transparente por adelantado sobre lo que puede y no puede compartir. No seguí este consejo anteriormente en mi carrera y simplemente hice una suposición. Cuando estaba trabajando con mis compañeros en nuestro equipo de gestión en una reorganización de la empresa, sabía que mi amiga Alice se vería afectada. Sin embargo, yo no era su gerente y tenía que mantener la información confidencial. Después de que se anunciaran los cambios, Alice estaba molesta por haberle ocultado la reorganización. Había asumido que era obvio que, aunque éramos amigas, no podía compartir ninguna información a la que Alice no tuviera acceso en el curso normal de los negocios. No era una cuestión de confianza; era una cuestión de ética. Y compartir la información de la reorganización prematuramente también habría puesto a Alice en una posición incómoda con sus compañeros quienes no tenían conocimiento de esa información. En retrospectiva, establecer límites con Alice puede que no le haya facilitado la reorganización, pero habría evitado cualquier tensión en nuestra relación.

Sea claro acerca de sus roles en la conversación. Establecer normas explícitamente para la forma en que trabajará y actuará crea igualdad y ecuanimidad en su amistad. Ya sea en una conversación personal en la oficina o en un rato juntos después del trabajo, sea transparente sobre el tipo de conversación que está teniendo. Diga algo como, “Hablemos de esto en modo amigo”. O: “Aquí hay un tema de trabajo que me gustaría mencionar y quitármelo de encima”. Sin embargo, también verifique si está bien que su amigo tenga esa conversación en ese momento. Como gerente, puede decir: “Quiero saber cómo van las cosas con tu proyecto. ¿Podemos hablar de ello ahora?”. Esto le permite a su amigo tener la misma voz en los temas que se debaten.

Sea transparente con los demás. Dado que Mariah estaba dos niveles por encima de Einat, cuando estaba en reuniones con sus empleados directos, se excusaba en las discusiones sobre el sueldo de Einat. Mariah dejaba claro que ella y Einat eran amigas, por lo que no quería que eso influyera en la toma de decisiones del equipo. Otros pueden sentirse incómodos al revelar sus sentimientos acerca de su empleado, especialmente si tienen comentarios negativos. Podrían preguntarse si mantendrán sus comentarios en contra de ellos o si podrían influir indebidamente en el resultado de la discusión. Por otro lado, es posible que sepa más de lo que se supone que debe aportar al entorno profesional. Mariah señaló: “Quería tener cuidado para que mi información privilegiada no fuera sacada de contexto”.

Haga su trabajo. Sea directo y rápido en las comunicaciones, especialmente cuando se trata de comentarios negativos o noticias desagradables, como un despido. Incluso si tiene miedo de herir los sentimientos de su amigo o teme que puedan ponerse a la defensiva, hable, pero prepárese para que haya momentos difíciles o incluso pausas en su relación. Al finalizar la discusión, dígale a su compañero que quiere ser su amigo, pero dele espacio para tomar su propia decisión sobre si también desea mantener la amistad. Por ejemplo, Ben y Ravi habían sido amigos durante muchos años cuando comenzaron a trabajar juntos. Ravi fue contratado como gerente de Ben. Después de un par de años, Ravi tuvo que decirle a Ben que su trabajo estaba siendo eliminado. Ben dijo: “Fue un desafío para cada uno de nosotros. No quería tener que decirme que había perdido mi trabajo”. Después de dejar su trabajo, Ben siguió siendo amigo de Ravi porque se dio cuenta de que la decisión de Ravi era comercial, no personal, pero le tomó un poco tiempo antes de que pudiera llegar a esa mentalidad.

Las amistades se basan en la confianza mutua y la transparencia. Navegar por las amistades entre gerentes y empleados es complicado, especialmente cuando, como jefe, usted tiene acceso a información que su empleado no conoce. Las amistades de trabajo que sobreviven también se basan en la confianza y la transparencia: transparencia sobre los límites dentro de los cuales podrá comunicarse y confiar en que sus acciones son profesionales y no personales.

A principios de 2016, Google anunció que había descubierto la clave del equipo perfecto. Tras años de analizar entrevistas y datos de más de 100 equipos, para impulsar un rendimiento de equipo eficaz, se hace necesario un nivel medio deinteligencia emocional en el grupo y un alto grado de comunicación entre sus miembros. La receta de Google de ser agradable y sumarse tiene mucho sentido.

Lo que tal vez sí sorprenda sea que la investigación de Google sugiere que los tipos de personas que componen el equipo no son tan relevantes. Aunque puede que eso sea cierto en Google, famosa por seleccionar a los empleados en función de su personalidad (o “Googlinidad“), el hallazgo no encaja con otras pruebas científicas que indican que las personalidades de los individuos juegan un importante papel en determinar el rendimiento del equipo. En particular, la personalidad afecta a:

• El rol que se tiene dentro del equipo
• La interacción con el resto del equipo
• La alineación de valores (creencias centrales) con los del resto del equipo

De forma importante, los procesos descritos anteriormente tratan de factores psicológicos (en lugar de aptitudes técnicas) que afectan tanto el rendimiento individual como el del equipo. Estos factores psicológicos son los que más determinan si la gente trabajará bien junta. Si el encaje del equipo sólo estuviera asociado con las capacidades y la experiencia, Donald Trump podría invitar a Bernie Sanders a trabajar en su administración, pero es improbable que funcionasen bien juntos. Del mismo modo, suelen existir importantes diferencias de compatibilidad entre usted y sus compañeros, sin importar cuán similares sean sus experiencias y formaciones técnicas.

Por ejemplo, un estudio de 133 equipos de la industria manufacturera encontró que los niveles más altos de sensibilidad interpersonal, curiosidad y estabilidad emocional se asociaron con los equipos más cohesionados y aumentaron los comportamientos prosociales entre los miembros del equipo. Los equipos más eficaces estaban dominados por personas curiosas, altruistas y serenas. En la misma línea, un gran metaanálisis demostró que las personalidades de los miembros del equipo influyen en la cooperación, el intercambio de conocimiento y el rendimiento general del equipo. En otras palabras: quiénes somos, afecta cómo nos comportamos y cómo interactuamos con otras personas, por lo que las personalidades de los miembros del equipo actúan como diferentes funciones de un único organismo.

Considere la tripulación que algún día (¿de un futuro próximo?) viajará hasta Marte, tal vez a las órdenes de Elon Musk o una de las agencias espaciales gubernamentales. Las simulaciones de tales viajes juntan a los astronautas en espacios reducidos durante cientos de días. Estas experiencias muestran que la tripulación crea diferentes grupos en función de la similitud de valores, y que niveles más altos de afabilidad y más bajos de inestabilidad emocional predicen mejor la cohesión del equipo y su cooperación.

Una manera útil de pensar en equipos con la mezcla correcta de aptitudes y personalidades consiste en considerar los roles que cada uno juega en un grupo de trabajo: un rol funcional, basado en su posición formal y sus capacidades técnicas, y un rol psicológico, basado en el tipo de persona que es. Demasiado a menudo, las organizaciones se centran únicamente en el rol funcional y esperan que de alguna manera le siga un buen rendimiento de equipo. Por esto, incluso los equipos deportivos con mayores recursos, a menudo, no consiguen rendir de acuerdo al talento individual de cada jugador: no existe ninguna sinergia psicológica. Un enfoque más eficaz (como el ejemplo de la misión a Marte) se centra tanto en las personalidades  como en sus aptitudes individuales.

En nuestro propio trabajo, encontramos que los roles psicológicos de equipo son en gran parte un producto de las personalidades de sus miembros. Por ejemplo, considere los siguientes perfiles de miembros de un equipo:

• Orientados a los resultados. Los miembros del equipo que organizan el trabajo de manera natural y asumen el mando tienden a ser socialmente seguros de sí mismos, competitivos y enérgicos.
• Centrados en las relaciones. Los miembros del equipo que se centran de forma natural en las relaciones están compenetrados con los sentimientos de los demás y se les da bien cohesionar el grupo. Tienden a ser cálidos, diplomáticos y accesibles.
• Seguidores de procesos y reglas. Los miembros del equipo que prestan atención a los detalles, procesos y reglas tienden a ser fiables, organizados y concienzudos.
• Pensadores innovadores y disruptivos. Los miembros del equipo que se centran de manera natural en la innovación, se anticipan a los problemas y reconocen cuando el equipo necesita cambiar, tienden a ser imaginativos, curiosos y abiertos a nuevas experiencias.
• Pragmáticos. Los miembros de equipo que son prácticos y realistas desafiadores de ideas y teorías tienden a ser prudentes, emocionalmente estables y sensatos.

Observar el equilibrio de roles dentro de un equipo ofrece extraordinarios conocimientos sobre sus dinámicas. También indica la probabilidad de éxito o fracaso en una tarea asignada. Por ejemplo, trabajamos con un equipo de finanzas encargado de lanzar un producto nuevo de informes empresariales para transformar la cultura de una sobria agencia gubernamental. Pero el porcentaje de actores en cada rol demostró que el equipo estaba condenado al fracaso:

• El 17 % de los miembros del equipo estaba orientados a los resultados.
• El 100 % de los miembros del equipo eran pragmáticos.
• El 0 % de los miembros del equipo eran innovadores.
• El 50 % de los miembros del equipo estaban orientados a procesos.
• El 0 % de los miembros del equipo eran buenos desarrolladores de relaciones.

Puesto que nadie desempeñaba el rol de construcción de relaciones, el equipo carecía de cohesión interna y no logró establecer ninguna conexión con los líderesde primera línea a los que se les requirió adoptar el nuevo proceso de contabilidad del equipo. De forma parecida, con tan sólo unos pocos miembros desempeñando un rol orientado a resultados (y un líder que no era uno de ellos), al equipo le resultó difícil impulsarse hacia delante.

A la inversa, cuando demasiada gente juega el rol de desarrollo de relaciones esto puede producir un entorno bonito, casi empalagoso, con apenas desafíos o demasiada poca contención, como sucedió con el equipo de liderazgo de esta organización de trabajo social:

• El 0 % de los miembros del equipo estaban orientados a los resultados.
• El 0 % de los miembros del equipo eran pragmáticos.
• El 29 % de los miembros del equipo eran innovadores.
• El 29 % de los miembros del equipo estaban orientados a procesos.
• El 86 % de los miembros del equipo eran buenos desarrolladores de relaciones.

En este ejemplo, el equipo dedicó demasiado tiempo a generar armonía y cohesión y demasiado poco a lograr resultados. Cuando uno se centra demasiado en llevarse bien con los compañeros, probablemente no le quedará demasiado tiempo ni energía para adelantarse a otros equipos u organizaciones.

Es muy útil utilizar estos tipos de perfiles para evaluar cómo impactará el rendimiento y la dinámica del equipo en un nuevo integrante. Como dijo la célebre investigadora de equipos, Suzanne Bell, que trabaja en el proyecto de Marte de la NASA: “suponemos que los astronautas son inteligentes, que son expertos en sus áreas técnicas y que disponen de al menos alguna capacidad de trabajo en equipo; lo que resulta complicado es compenetrarse bien entre sí“.

Por tanto, hacer una evaluación individual completa puede ofrecer conocimientos fundamentales sobre cómo trabajarán juntas determinadas personas, y puede ayudar a resaltar áreas de conflicto y afinidad. Cualquier resultado de valor se produce como el por el esfuerzo en equipo, en el que la gente aparca sus intereses egoístas para lograr algo, de manera colectiva y que no podría haber logrado sola. Los equipos más exitosos aciertan con esta combinación de personalidades.

Will technology kill jobs and exacerbate inequality, or usher in a utopia of more meaningful work and healthier societies?

While it is impossible to know what tomorrow holds, research by global professional services company PwC explores four possible futures – or “worlds” – driven by the “mega trends” of technological breakthroughs, rapid urbanization, ageing populations, shifting global economic power, resource scarcity and climate change.

The Red World – innovation rules

The world becomes a perfect incubator for innovation in one PwC scenario. Digital platforms enable those with winning ideas and specialist, niche profit-makers, to flourish.

However, PwC warns, the risks are high if innovation outpaces regulation. “Today’s winning business could be tomorrow’s court case.”

Image: PwC Workforce of the Future

Projects will develop at a fast pace and specialists will only stay with them as long as they, or the business, last. There will be few in-house human resources teams, with outsourcers or automation providing the human services needed.

Companies may see little regulation that prevents them doing what they like, while workers will enjoy fewer benefits like health insurance, pensions and long-term employment.

Image: REUTERS/Charles Platiau

The Blue World – corporate is king

Corporations grow so big and influential that some become more powerful and larger than national economies.

In a frightening vision, almost worthy of Aldous Huxley’s Brave New World, PwC predicts: “Human effort [will be] maximized through … physical and medical enhancement techniques and technology and, along with automation, analytics and innovation, push performance in the workplace to its limits.”

Image: PwC Workforce of the Future

While rewards for some will be high, the price will be people’s data, which will “predict performance and anticipate people risk [predict behaviour that may damage a business financially or reputationally]”.

In both the Blue and Red Worlds, people who have strong skills – and update them – will be in demand, those who do not will be discarded.

The Green World – companies care

“This is a world where corporate responsibility isn’t just a nice-to-have, but it’s a business imperative,” PwC says.

A strong social ethos places a heavy emphasis on diversity, human rights and the non-financial impacts of business on the planet and people’s lives.

Competition for talent is intense and financial rewards are still important, while incentive packages include “three weeks’ paid leave a year to work on charity and social projects”.

Image: PwC Workforce of the Future

However, workers are expected to reflect their employers’ values at work and at home and travel is tightly controlled.

“In this world,” the writers say, “automation and technology are essential elements to protect scarce resources and minimize environmental damage … But … technology is a double-edged sword: it allows organizations to meet their ethical and environmental agenda, but at what cost to humans?”

The Yellow World – humans come first

Financial technology enables more crowd-funded capital to reach ethically “blameless” brands, while workers and companies seek greater meaning and relevance in everyday life.

Artisanal skills return, as do workers’ guilds, which protect members’ rights and train new craftspeople: “It’s a world where humanness is highly valued,” says PwC.

Non-financial rewards are given in a trade-off for less money, work is often a fluid concept and the standard 9-to-5 working week is rare, while the divisions between home and work blur.

Image: PwC Workforce of the Future

However, while the automation of tasks that are dull, damaging or impossible for humans continues, the writers say: “Conflicts remain around the use of technology, as people are less likely to take the downsides of automation without a fight.

“As more people are impacted by technical advances and see their skills become obsolete, disaffection and the push-back against policies that favour the elite grow.”

Which way to the future?

All of the four possible futures in PwC’s report share the common theme of increasing use of technology to assist, augment and replace human work.

Some foresee the dominance of global corporations, others predict the growth of smaller, more individual endeavours. All, however, depend on digital technology to link talent pools and customers, and create financially beneficial relationships, whether these are between individuals and corporations, or groups of people.

“By replacing workers doing routine, methodical tasks, machines can amplify the comparative advantage of those workers with problem-solving, leadership, emotional intelligence, empathy and creativity skills,” PwC says.

“Those workers performing tasks which automation can’t yet crack become more pivotal – and this means creativity, innovation, imagination and design skills will be prioritized by employers.”

Changing lanes

Any of these futures – or a combination of them – are possible, but how we reach 2030, and who will benefit, needs careful planning and consideration.

In the film Escape from the Planet of the Apes, a scientist compares reaching the future to a driver changing lanes: “A driver in lane ‘A’ may crash while a driver in lane ‘B’ survives. It follows that a driver, by changing lanes can change his future.”

Working out which lane will lead to a less fractured world is one of the greatest challenges facing policy makers and corporate leaders today.

Working in hospitals as a health care professional (HP) is demanding. Doctors, nurses, and assistants must deal, on a daily basis, with emotional situations, such as the suffering of patients and relatives, cognitive challenges (for example, timely decision making and analysing several indicators in order to establish a diagnostic and a treatment plan), interpersonal tensions (conflicts between different specialties or professionals, uncooperative patients, impatient relatives), physical hassles (e.g., working nights, lifting heavy patients), and logistic complexity, such as the lack of necessary resources, time consuming bureaucratic processes, and heavy workload (e.g., Ghodse & Galea, 2009). High rates of burnout are reported in health care professionals in both Europe and the U.S. (Aiken et al., 2012; Soler et al., 2008). Consequently, the well-being and health of HPs is likely to be impaired, and their ability to work effectively may also be diminished. Stress, anxiety, and burnout have consistently shown positive relationships with decreased performance in HPs and also with maladaptive coping strategies, such as substance abuse (Firth-Cozens, 1995). Nonetheless, while working under these conditions, health professionals must assure that all the patients are given the best possible quality of care (QoC). Indeed, the quality of organizations is one antecedent of its competitive advantage. Considering quality of care in health services, there are also other fundamental issues at stake, namely human life, human rights, and human dignity. Yet, the Institute of Medicine (1999) reports that tens of thousands of American patients die every year due to suboptimal care. The existence of quality of care problems is widespread and is not restricted to the United States. For example Bartlett, Blais, Tamblyn, Clermont, and MacGibbon (2008) claim that in Norway three people might die every day due to poor hospital quality. Also, according to the European Commission (2008), it is estimated that between 8% and 12% of patients admitted to hospitals will suffer from adverse effects while receiving healthcare. Therefore, patient safety was identified as a key area for action in the Commission’s Health Strategy White Paper of October 2007.

Considering the importance of quality within healthcare, the aim of the present study is to analyse the impact of job demands on quality of care and to investigate possible team and individual processes that will help to buffer the impact of high work demands on the quality of care delivered to patients, therefore ensuring their safety.

Quality of Care – A Multidimensional Concept?

According to McGowan et al. (2011), the definition and measurement of quality of care in healthcare lack consistency across studies. The Institute of Medicine defines QoC as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge” (IOM, 1999). This institute defines six main pillars fundamental for delivering a high quality of care: health care must be safe, effective, patient-centred, timely, efficient, and equitable (IOM, 2001). For another author (Donabedian, 1980), QoC is also a multifaceted concept. It encompasses health outcomes, the process of care delivery (such as information obtained and coordination) as well as the structure where it is delivered (equipment, administrative processes, etc.). Campbell, Roland, and Buetow (2000) define QoC as “whether individuals can access the health structures and processes of care which they need and whether the care received is effective” (p. 1614). For the authors, the consequences of care reflect the effectiveness of the structure and processes and are assessed by the health status of patients and by user evaluation.

Job demands in Hospital Settings

According to the Job Demands-Resources Model (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001), job demands are “those physical, social, or organizational aspects of the job that require sustained physical or mental effort and are therefore associated with certain physiological and psychological costs” (p. 501). Within job demands that can be found in a hospital setting, we can name, for example, time pressure, physical workload, shift work, or recipient contact. According to the model, when meeting those demands requires a high level of effort from the employee and an adequate recovery from that effort is not possible, then the development of job strain develops. Consequently, workers may develop a health impairment process (because of an increased autonomic and endocrine activation and of the increased subjective effort) and their task performance may deteriorate indirectly because of the need for strategic adjustments (e.g., narrowing of attention) and of fatigue after-effects, such as risky choices.

An extensive review of all the possible job demands found in hospital settings is beyond the scope of this paper. As an example, Ecklebery-Hunt et al. (2009) searched for the antecedents of residents’ burnout symptoms and found several factors: lack of control over schedule, poor relationships with colleagues, difficult and complicated patients, excessive paperwork, not enough time in the day, and perfectionism (to name a few). In Isikhan, Gomez, and Danis’ (2004) study, unfairness in promotion opportunities, imbalance between jobs and responsibilities, conflict with colleagues, lack of appreciation of efforts by superiors, responsibilities of role, long and tiring work hours, inadequacy of equipment, and problems experienced with patients and their relatives were the main factors associated with the stress experienced by health professionals working with cancer patients. Other studies focused on specific stressors and their relationship with health professionals’ well being, such as aggression towards them (e.g., Winstanley & Whittington, 2004).

These job demands impact on professionals’ well-being and performance. Shanafelt, Bradley, Wipf, and Back (2002), as well as Toral-Villanueva, Aguilar-Madrid, and Juárez-Pérez (2009) concluded that the presence of stressors and acute stress on HPs was associated with self-reported suboptimal practices. Other studies report the relationship between job demands and decreased productivity (e.g., Kazmi, Amjad, & Khan, 2008) and between job stress and increased medical errors (e.g., Fahrenkopf et al., 2008; West et al., 2006) and mental health impairment in HPs, mainly depression and anxiety (e.g., Caplan, 1994; Chambers, Wall, & Campbell, 1996; Toral-Villanueva et al., 2009; Weinberg & Creed, 2000).

Therefore, considering this evidence, it is likely that job demands at the hospital will impact on quality of care either directly (for example, lack of resources needed, too many patients to diagnose in little time, etc.) or indirectly (through the depletion of health and cognitive and emotional resources of HPs).

H1. Job demands are negatively related to quality of care.

Job Resources in Hospital Settings

Parallel to the existence of job demands, in every job there are also job resources, defined as the physical, psychological, social, or organizational aspects of the job that reduce job demands and the associated costs, that are in achieving work goals, and stimulate personal growth and development (Demerouti et al., 2001). Resources are seen as motivation boosters, either through an intrinsic or extrinsic path. Some authors (e.g., Richter & Hacker, 1998) posit the resources can be either organizational (such as feedback, job control, supervisory support, or task variety) or social (e.g., support from colleagues and peers). Xanthopoulou Bakker, Demerouti, and Schaufeli (2006) proposed also personal resources (self-efficacy, organizational-based self-esteem and optimism) as important in facing demanding work environments. In a similar reasoning, Sweetman and Luthans (2010) suggest that self-efficacy, optimism, hope, and resilience (that they name as PsyCap) are positive agentic resources that have a motivational impact on workers and leads to desired work outcomes.

Another important premise of the Job Demands-Resources model (J D-R) is that job resources will buffer the impact of job demands on job strain. More specifically, resources will have the potential to reduce the tendency of organizational properties to become or generate stressors, to alter the perceptions evoked by those stressors, or to reduce the health impairment consequences of the stress response. Finally, the model postulates that the positive impact of job resources on both diminishing the negative impact of job demands and fostering workers motivation will be higher in situations where job demands are high. This model has received empirical support across several occupations (e.g., Bakker, Demerouti, & Schaufeli 2003; Bakker Demerouti, & Verbeke, 2004; Demerouti et al, 2001; Hakannen, Bakker, & Schaufeli, 2006).

In the present study, we will focus on two possible job resources that, to our knowledge, have not been studied in the hospital context: backup behaviors of team members and individual emotions. The analysis of these specific resources is critical, for HPs often work in teams (e.g., surgical teams) and the ability of those teams to perform well in a context of high demands is critical for patients’ health outcomes and, in more general terms, for quality of care. Indeed, as Krokos, Baker, Alonso, and Day (2009) point out, “health-care workers perform interdependent tasks (e.g., removing a patient’s appendix) and function in specific roles (e.g., surgeon, anesthesiologist, surgical assistant) while sharing the common goal of providing safe care to patients” (p. 384-385). Moreover, the emotions of individual HPs accounts for the emotional dimension of work, often overlooked by scholars (Ashkanasy, 2003), should be taken into account. Since some of the demands faced by health professionals have an emotional nature, we must, then, consider emotional resources as important factors for health professionals well-being and effectiveness.

Backup Behaviors in Teams

In 2003, Porter et al. proposed the construct of backing up behaviors, defined as “the discretionary provision of resources and task-related effort to another member of one’s team that is intended to help that team member obtain the goals as defined by his or her role when it is apparent that the team member is failing to reach those goals” (p. 319-320). According to those authors, back up behaviors emerge when team members recognize that the distribution of workload is inaccurate and may cause trouble in task and social performance of their team. For example, backup behaviors happen when a team member fills in for a co-worker who is unable to meet the demands of his or her role at a specific moment. They can be either physical (helping lifting a heavy patient or dealing with complex equipment) or verbal (suggestions, cautioning advice, or feedback).

Within the literature on teams, backup behaviors have been labelled one of the “big five” in teamwork (Salas, Sims, & Burke, 2005), together with team leadership, mutual performance monitoring, adaptability, and team orientation. Indeed, it is seen as a skill “at the heart of teamwork, for it makes the team truly operate as more than the sum of its parts” (McIntyre & Salas, 1995, p. 26). According to Salas, Rosen, Burke, and Goodwin (2009), backup behaviors support team performance in three ways: allowing members to provide (1) assistance during task performance, (2) timely feedback so that performance processes can be adjusted, and (3) help to teams, so that dynamically adjust their performance strategies and processes when a detrimental imbalance of the workload is detected.

Therefore, the relevance of backup behaviors for hospital teams social and task performance is explained by the teams’ ability to reduce work overload (considered a job demand) and by its impact on team processes, resulting in a higher degree of adaptability in face of environmental and situation changes. Faced with challenging demands, hospital teams will differ on the degree of backup behaviors they provide team members. Depending on the accuracy and timeliness of these backup behaviors, the impact of the job demands on quality of care may be diminished – workload is better divided between HPs, reducing the degree of stress, and diminishing the need to rush or, in other words, reducing job demands.

H2. Backup behaviors will mediate the impact of job demands on quality of care.

Positivity

While working at the hospital, HPs have the opportunity to experience many positive and negative feelings. This is true for every job and maybe more so in a health context, where not only is there the need to deal with people but also the emotional valence of disease, ageing, and suffering. Being able to maintain a positive mood may, thus, be a challenge.

Following the trend of positive psychology (Seligman & Csikszentmihalyi, 2000), Fredrickson and Losada (2005) proposed that the positivity ratio (ratio of pleasant feelings and sentiments to unpleasant ones over time) would predict subjective well-being. The benefits of positive affect have already been documented and ranged from resilience and physical and mental health (Fredrickson, Tugade, Waugh, & Larkin, 2003) to happiness (Fredrickson & Joiner, 2002) and increased intuition (Bolte, Goschkey, & Kuhl, 2003) and creativity (Isen, Daubman, & Nowicki, 1987). Indeed, according to the broaden-and-build theory of positive emotions (Fredrickson, 2001), positive emotions will widen the array of thoughts and actions available, resulting in more behavioral flexibility, generativity, and adaptability. Moreover, overtime, the benefits of the broader repertoires of thought and action will, as a consequence, build enduring personal resources, such as coping mechanisms, social connections, and environmental knowledge.

Therefore, the positivity ratio of HPs is likely to influence how they are able to cope with job demands – the higher the positivity the more likely they are to be able to overcome difficulties – to be able to find more than one solution for a problem and find creative solutions for problems. Consequently, the quality of care they will be able to deliver to patients will also be increased in a context of a high positivity.

H3. The positivity ratio of HPs will mediate the impact of job demands on quality of care.

Backup Behaviors and Positivity – A Process Relationship

Receiving a helping behavior from a coworker in a moment where one feels that he or she is unable to successfully accomplish his or her tasks is likely to generate a positive feeling. According to Bakker and Demerouti (2007) social support is the best well-known variable that buffers against job strain. The receiver of backup behaviors will perceive social support from the colleagues and also may feel less pressure from an excessive workload – all of these are events that most likely will be perceived as positive. Therefore, the presence of backup behaviors in a team is likely to influence the positivity ratio of team members, which, in turn, will impact on the quality of care they are able to provide patients with, according to the broaden-and-build theory of positive emotions (Fredrickson, 2001). It wouldn’t be completely odd to consider a reverse causality relationship: the higher the positivity ratio the higher the well-being and happiness of an individual and his/her willingness to invest in social relationships. Nonetheless, backup behaviors are not just helping or social behaviors – by definition they happen in a context of an uneven distribution of workload. Therefore, the existence of an uneven distribution of workload is a precondition for backup behaviors to happen; only after those backup behaviors will team members increase their positivity ratio, since before that they were experiencing trouble in workload distribution.

H4. The impact of job demands on quality of care is sequentially moderated by backup behaviors and positivity.

Method

Participants and Procedure

Data were collected within the scope of ORCAB project (“Improving quality and safety in the hospital: The link between organizational culture, burnout and quality of care”), a 7^th Framework Program financed by the European Commission. A total of 9 European countries participated in this project and six countries collected the same data on one or more teaching hospitals between September and December 2011 (cf. Table 1).

After the formal agreement of the Administration Boards and Ethics Committees of the participating hospitals, questionnaires were either emailed to all of the health professionals or collected on site. From all countries, 2,890 health professionals completed the questionnaire. On average, participants were 39 years old (SD = 10.26) and had been working at their hospital for, in average, 12 years (SD = 10.73). Seventy four percent of participants were female. Considering their staff position at the hospital, 50.5% were nurses, 22.1% were physicians, 20% were residents and 7.4% had other positions (e.g., pharmaceutics, social workers, etc.).

Measures

The following scales were used in the present study:

Job demands. The Hospital Work-Experience Scale¹ was constructed as part of the ORCAB project (ORCAB, 2012). The scale comprised six items (Cronbach’s α = .57) for organizational demands (e.g., “The communication between hospital departments is problematic”), six items (Cronbach’s α = .68) for emotional demands (e.g., “I have to deal with verbally abusive patients”), seven items (Cronbach’s α = .78) for physical/environmental demands (e.g., “I have too much paperwork to do”) and five items (Cronbach’s α = .73) for cognitive demands (e.g., “I have to take decisions under time pressure”). Participants answer in a 5-point scale (1 = never, 5 = always).

ORCAB Quality of Care Scale¹. In order to capture the complexity and multidimensionality of the construct, a three-factor model of QoC was developed via the ORCAB project (Orcab Report, 2012). The ORCAB project is concerned with improving quality and safety in the hospital via the link between organizational culture, burnout, and quality of care. This multi-centre study among hospitals in 9 European countries utilized systematic reviews, focus groups, surveys, and action research to identify the key mechanisms within quality of care.

Patient Centeredness, Effectiveness, and Personal Barriers to Providing Good QoC were identified as key dimensions of QoC. Informed by the IOM’s framework, the project used a bottom-up approach to develop both a model and an instrument that measures the construct. Items were generated based on interviews and focus-groups conducted with HPs – certified physicians, residents and nurses – in seven European countries. The three-factor model was found to be consistent across HPs’ organizational position and gender.

Patient Centeredness captures the quality of the one-on-one interaction with patients, referring to the availability of HPs in terms of time, information provided to patients, pleasantness, and equity in providing medical care. Effectiveness captures the effort to provide the best medical care possible, considering the resources available, mainly in terms of expertise. Personal Barriers to Providing Good QoC captures failures to perform as expected when interacting with patients as a result of personal hindrances interfering with professional life.

The scale comprises the dimensions of patient centeredness, measured with four items (Cronbach’s α = .77) (e.g., “I was able to ensure good and pleasant communication with patients”), effectiveness, measured with three items (Cronbach’s α = .70) (e.g., “I was able to provide the patient with the best medical care available”) and personal barriers to providing good care, measured with two items (Cronbach’s α = .67) (e.g, “Sometimes my personal problems impact on the quality of care”). Participants answer in a 10-point scale (1 = never, 10 = always).

Backup behaviors. Backup behaviors were assessed using four items (Cronbach’s α = .68) from the Hospital Survey on Patient Culture, US Agency for Healthcare Research and Quality (2004) (e.g., “When one area in this unit gets really busy, others help out” ; “When a lot of work needs to be done quickly, we work together as a team to get the work done”). Participants answer in a 5-point scale (1 = never, 5 = always).

Positivity. We computed the positivity index by the ratio of positive/negative emotions. Positive and negative emotions were assessed by the short version (Thompson, 2007) of PANAS (Watson, Clark, & Tellegen, 1988), including five positive (Cronbach’s α = .73) (e.g., inspired, active) and five negative (Cronbach’s α = .70) (e.g., ashamed, afraid) emotions. Participants answer in a 5-point scale (1 = very slightly or not at all, 5 = extremely).

Statistical Analysis

Considering the low reliability (Cronbach’s α = .57) of the scale measuring organizational demands, this scale was removed from further analysis. For testing hypothesis 1, we calculated simple linear regressions in SPSS, using the method Enter (inclusion of all variables at the same time). In order to test hypothesis 2 to 4, we followed the method presented by Preacher and Hayes (2008) and used the SPSS PROCESS macro provided by Hayes (2012) to run the analysis. These authors propose a procedure for assessing the significance of the indirect effects of one variable on another through the influence of the mediator, which is in line with the objectives of the present paper. Their method allows for testing indirect effects in samples where researchers should not theoretically assume a normal distribution, by using bootstrapping, a nonparametric resample procedure. The output of the analysis yields confidence intervals for the indirect effects. Since the authors strongly recommend using the bootstrap procedure, we decided to perform our analysis in accordance, using a bootstrapping of 5,000 samples in each analysis, as recommended (Preacher & Hayes, 2008). For hypothesis 2 and 3 (simple mediations), we used the model 4 of the PROCESS macro (Hayes, 2012).

Results

In Table 2 we present the correlations and descriptive statistics for all of the variables in the study. All of the variables were significantly correlated (p < .05), and the correlations were in the expected direction. Job demands showed positive and correlations between them (values ranging from r = .48 between physical and cognitive demands to r = .44 between physical and emotional demands). All of the job demands correlated negatively with backup behaviors (values ranging from r = -.26 for physical demands to -.23 for cognitive demands) and, with smaller magnitude, with positivity (r = -.14 for emotional demands and -.13 for physical and cognitive demands). All of the dimensions of quality of care (patient centeredness, effectiveness, and personal barriers) correlated negatively with all of the job demands, and positively with backup behaviors and positivity. Correlations amongst the dimensions of quality of care were higher between patient centeredness and effectiveness (r = .55) than between personal barriers and any of the other two (r = .08 with effectiveness and r = .10 with patient centeredness).

Considering H1 (see Table 3), the existence of job demands (cognitive, emotional, and physical) significantly predicted patient centeredness (p = .0001), effectiveness (p = .000), and personal barriers (p = .000). Therefore, H1 was supported. The value of the betas shows relationships of greater magnitude between job demands and the personal barriers dimension of QoC. Considering effectiveness, physical demands are its weakest predictor and cognitive demands are the ones with lowest impact on patient centeredness.

Table 4 presents the results of testing hypotheses 2 and 3, providing the value for the indirect effect for each model and the confidence intervals (at 95%) for testing the significance of the indirect effect. Indirect effects are considered significant (signaled with **) when 0 falls out of the confidence interval (Preacher & Hayes, 2004, 2008).

The indirect effects of job demands on quality of care through backup behaviors were found to be significant, supporting our second hypothesis for all of the dimensions of quality of care. The hypothesized indirect effect of job demands on quality of care through positivity (hypothesis 3) was also significant, except considering the effect of emotional demands on personal barriers (indirect effect = -.0190; 95% CI: -.0408, .001 with 5,000 resamples).

The indirect effects are mostly higher when quality of care effectiveness dimension is the output, when emotional demands are the independent variable, and considering backup behaviors as the mediator. The greater indirect effect was found in the relationship between emotional demands and quality of care effectiveness, mediated by backup behaviors (indirect effect = -.1964; 95% CI: -.2397, -.1596 with 5,000 resamples).

The test of the final hypothesis (hypothesis 4) was also made using the PROCESS macro, but choosing model 6 (Hayes, 2012), a double mediation model. Results are presented in Table 5.

Only two of the double mediations were not significant, both considering personal barriers as the dependent variable (indirect effect = -.0017, 95% CI: -.0046, -.0000 with 5,000 resamples for cognitive demands as independent variable and indirect effect = -.0019, 95% CI: -.0062, -.0000 with 5,000 resamples for emotional demands as independent variable). Thus, hypothesis 4 was partially supported. The higher indirect effects of the double mediation were found between physical demands and the personal barriers of quality of care (indirect effects = -.0021).

Discussion

The main goal of the present paper was to contribute to our understanding of the relationship between job demands and quality of care in hospital settings, and to specifically highlight the role of backup behaviors and positivity as dampers of that negative relationship. Our results shed some light on this relationship and on hospital dynamics.

Firstly, considering the direct effects, as expected, job demands predicted quality of care: the existence of physical, emotional, and cognitive demands tends to worsen the quality of care provided. Job demands show a greater direct impact on personal barriers to providing good quality of care. Therefore, it seems that demands are central, have a stronger direct impact on health professionals’ health and well-being (e.g., tiredness) and, consequently, on their ability to focus on their tasks and patients.

Secondly, the impact of job demands on the personal barriers dimension of QoC is likely to be direct, while backup behaviors and positivity are important mediators between job demands and the patient centeredness and the effective dimensions of QoC, and more so with this last dimension.

The existence of backup behaviors may foster the positivity ratio of HPs. However, the magnitude of the double mediation effect is smaller than the effects obtained through simple mediations. It would not be completely odd to consider a reverse causality relationship: the higher the positivity ratio the higher the well-being and happiness of an individual and his/her willingness to invest in social relationships. In addition, backup behaviors have a more relevant role, indicating that the improvement of QoC is more likely to happen in a cooperative context (backup behaviors). Hence, for the time being and in terms of practical implications, it is central to consider team dynamics when thinking about improving the quality of care in the hospital. We must then go beyond a rather simplistic view of what will facilitate providing adequate quality of care – having the right equipment and facilities, adequate cognitive challenging activities, support from co-workers and supervisors, and an adequate administration of resources do not alone account for all the fluctuations in QoC. In a demanding context, where professionals usually have a high workload and work under time pressure (Carayon & Gurses, 2005; Linzer et al., 2000; Silva et al., 2013), having the ability to detect and act upon situations where co-workers have a workload that goes beyond what they can actually achieve (i.e., backup behaviors) is fundamental for the effectiveness of the professionals. This implies that HPs are not only attentive to detect those situations but also willing to reach out to their peers and team members. However, providing backup behaviors may also have negative effects. According to Barnes et al. (2008), backup behavior providers risk neglecting their own work and backup behaviors receivers may decrease their investment in taskwork in a subsequent task, particularly when they are aware that other team members can recognize their workload. This may lead to situations where workload becomes, again, unevenly distributed, generating the noxious consequences already mentioned. The monitoring of adequate amounts of backup behaviors might be an important function of the team leader (for example, the department leader/coordinator or the chief nurse in a specific service). Indeed, Teng, Lee, Chu, Chang, and Liu (2012) found that employees’ intention to help their co-workers is negatively related to their supervisor’s negative mood when the employee-supervisor relationship is weak.

Third, the personal barriers dimension of QoC presents a somewhat different pattern of relationships than the other two QoC dimensions. Indeed, one unexpected finding was the non-significance (although closely reaching significance) of the indirect effect of emotional demands on personal barriers in providing good quality of care through positivity. We would have anticipated that positivity could be a powerful mediator specifically between those two variables since all are related to affective experiences and it was not the case. The results of the double mediation are also non-significant, considering the effects of both emotional and cognitive demands on personal barriers. However, positivity as a simple mediator has a greater impact on physical demands’ relationship with this dimension of QoC, in comparison with the other two. It seems then that the prevalence of positive experiences and emotions over negative ones is not sufficient to influence the impact of emotional demands, such as the fear of doing something wrong or the spillover of work life into family life, on the tiredness of HPs, or on their personal problems and its influence on the quality of care they deliver. At the same time, an investment in decreasing physical strain may result in a bigger improvement on this particular dimension of QoC, especially when backup behaviors exist. Indeed, considering physical demands, the double mediation has a greater effect precisely with the personal barriers dimension of QoC, and not with effectiveness or patient centeredness, which show similar effect values.

To sum up, the results of this study imply that quality of care in hospital settings is closely related to job demands. Interventions aimed at improving the effectiveness and patient centeredness of QoC will have greater success if they are directed at physical demands. Also, hospital managers should not overlook the importance of cooperation within teams and should find ways to develop teamwork. For example they should foster shared cognitions (e.g., team shared mental models, team situation awareness), provide opportunities for team training (e.g., in explicit communication skills) (Salas, Cooke, & Rosen, 2008), and develop teamwork adjustment behaviors such as intra-team coaching or collaborative problem solving and task related collaborative behaviors (e.g., coordination, information exchange) (Rousseau, Aubé, & Savoie, 2006). This will impact quality of care indirectly in all of its dimensions.

Health care around several European countries is moving towards a patient-centred and consumer-focused system based on a market-oriented approach (Sofaer & Firminger, 2005) that requires a careful consideration and monitoring of its quality of care. Furthermore, some studies (Arocena & García-Prado, 2007) show an improvement in hospital performance mainly driven by an increase of quality of care. In the end, improving QoC in the hospital reflects an improvement of human life, human rights, and human dignity.

In the present study, we explored the influence of transformational leadership on group potency, a variable that is closely related to group performance. We identified two ways through which leadership may have an impact on group outcomes: group identification and group cohesion.

Our research focused on small military units in the Spanish Army. The profile of small units in military operations has progressively increased over the last few years. In armies, small units often operate in remote scenarios and under extreme conditions with no direct supervision from commanding officers. Thus, factors such as leadership, cohesion, and potency in the most basic units of military organization are becoming increasingly important for military decision makers.

^{Bass, Avolio, Jung, and Berson (2003)} proposed a model of transformational leadership (TL) in which they analyzed the influence of TL on group cohesion, group potency, and unit performance of military units. Their model inspired many studies in the past decade. In the military field, research about these aspects is especially scarce at the squad level, even though these units are of great importance, both considering the level of responsibility and the roles assigned to them to achieve success in military operations.

Group potency is a key component of group effectiveness. Several studies have pointed to confirm the existence of a positive relationship between group potency and group performance (^{Gully, Incalterra, Joshi, & Beaubien, 2002}; ^{Shea & Guzzo, 1987}).

In the military environment, where it is often difficult to measure effectiveness, group potency can be a very useful indicator of the degree of preparation of units to tackle their missions. In previous studies, group potency has significantly predicted efficacy and was also positively correlated with mental task performance, physical task performance, and commander ratings of team performance (^{Jordan, Field, & Armenakis, 2002}).

Authors such as ^{Shamir, Zakay, Brainin, and Popper (2000)} and more recently ^{Haslam, Reicher, and Platow (2011)} have underlined the importance of group identification processes in achieving group’s objectives. Such identification is materialized as a process of reciprocal influence between leaders and followers.

Throughout history, cohesion is another factor that has been considered critical for groups to be able to fulfill their missions (^{Griffith, 1988}; ^{Jung & Sosik, 2002}; ^{Tziner & Vardi, 1982}). Recent studies (^{Shamir et al., 2000}; ^{Siebold, 2007}, ²⁰¹²) have highlighted that cohesion is a determining factor of the effectiveness of military units (^{Tziner & Chernyak-Hai, 2012}), since these units often face high-risk situations (^{Hannah, Uhl-Bien, Avolio, & Cavarretta, 2009}).

^{Zaccaro and Klimoski (2002)} raised the importance of developing models that explain the collective effectiveness of teams including the variables that contribute to collective action and leadership processes. We consider that it is especially interesting to jointly explore collective effectiveness, leadership, and relevant variables for teams, such as group cohesion and group identification in military organizations where leaders are formally established.

The objective of the present research was to analyze the direct and indirect effects of the transformational leadership of squad leaders (i.e., non-commissioned officers) on group potency. Group Identification and group cohesion were considered as mediating or indirect variables. The relationship between the various group variables that we included in our research is explained below.

Group Cohesion

One of the definitions of group cohesion most frequently used in the literature is “a dynamic process that is reflected in the tendency for a group to stick together and remain united in the pursuit of its instrumental objectives and/or for the satisfaction of member affective needs” (^{Carron, Brawley, & Widmeyer, 1998, p. 213}). Regarding cohesion in military units, ^{Siebold (2012)} highlighted the existence of a social group that remains united despite external threats and is able to achieve material and psychological objectives thanks to the psychosocial reinforcement among its members.

The importance of cohesion in military units was a clear finding of the meta-analytical review conducted by ^{Oliver, Harman, Hoover, Hayes, and Pandhi (1999)}. The review revealed the importance of cohesion and its positive relationship with group and individual performance, job satisfaction, retention, well-being, and readiness. In the present study, we followed the approach of ^{Ahronson and Cameron (2007)}, who studied cohesion in the Canadian Army using the model developed by ^{Carron, Widmeyer, and Brawley (1985)}.

Several studies have analyzed the relationship between cohesion in military units and leadership (^{Arthur & Hardy, 2014}; ^{Bass et al., 2003}). ^{Tziner and Chernyak-Hai (2012)} argued that high-cohesiveness crews perform best under the leadership of a commander who exercises high involvement both in the process of task accomplishment and in the interpersonal arena.

^{Bartone, Johnsen, Eid, Brun, and Laberg (2002)} found that the influence of leadership on group cohesion can be increased by familiarity and the fact of performing challenging tasks in small military units. In accordance with the characteristics of transformational leadership style behavior, in which leaders combine an individual relationship with followers with a shared vision of the future at the squad level, we developed the following hypothesis:

H 4. The transformational leadership style of squad leaders will be directly and positively related to group cohesion in squads.

Several studies have shown a relationship between group cohesion and effectiveness (^{Beal, Cohen, Burke, & McLendon, 2003}). Cohesion can be considered as an important characteristic of groups, as it allows them to assume that they will be able to overcome the difficulties of their environment. Cohesion enables groups to have the commitment of their members, to coordinate actions, and to persevere in the performance of tasks. In the Spanish Army, the psychological potential of units is often measured with an instrument known as Cuestionario para la Estimación del Potencial Psicológico de Unidad (CEPPU-03; Questionnaire to estimate the psychological potential of units). According to this questionnaire, cohesion is one of the factors that determine the psychological potential of the unit. This instrument is used to study both companies and battalions (^{García, Gutierrez, & Núñez, 2005}). However, a review of specialized literature shows that little research has been conducted on the effects of group cohesion on group potency in small military units, particularly in Spain, where it takes two years of specific training in a military academy to become a squad leader. ^{Zaccaro, Rittman, and Marks (2001)} developed a model that can also be analyzed considering the self-perception of followers. In this model, the stages prior to the effectiveness of the teams are based on actual processes of coordination and motivation linked to cohesion. Based on the above-mentioned points, we developed the following hypothesis:

H 5. Group cohesion will be directly and positively related to group potency in squads.

In the present research, we proposed that TL develops processes of influence that contribute to a greater identification of members with the group (^{Haslam et al., 2011}; ^{Reicher et al., 2005}). In the research conducted by ^{Walumbwa, Avolio, and Zhu (2008)}, the authors explored the effect of transformational leadership on rated performance in individuals and found that it was also mediated by the interaction of identification and means efficacy. ^{Shamir, Zakay, Breinin, and Popper (1998)} highlighted leader behaviors that raise the salience of certain values and identities in followers’ self-concepts and constitute a framework for a group’s mission and the roles of followers based on such values and identities. Based on their research and the theoretical approach of ^{Haslam et al. (2011)} and ^{Reicher et al. (2005)}, according to which leadership plays a predominant role related to organizational identification in the development of group processes, we developed the following hypothesis:

H 6. The transformational leadership style of squad leaders will be indirectly related to group potency through group identification in squads.

The relationship between transformational leadership and cohesion has been explored in various studies, which have highlighted the influence of cohesion on collective-efficacy team performance and unit performance (^{Bass & Bass, 2008}). In the military context, ^{Tziner and Vardi (1982)} also conducted a study on tank crews in the Israeli army that revealed a significant effect of leadership style on performance, mediated by cohesion.

^{Sivasubramaniam et al. (2002)} argue that the behavior of transformational leaders influences group potency. According to them, military unit leaders must articulate a vision oriented toward the missions and tasks assigned to the squad; these missions and tasks must be achieved by means of team cohesion and the distribution of responsibilities.

However, there are practically no studies in military literature on the effect of group cohesion as a mediator between the transformational leadership of squad leader and group potency in squads. Thus, it would be interesting to further explore the intermediate processes that characterize military units. For this reason, we proposed the following hypothesis:

H 7. The transformational leadership style of squad leader will be indirectly related to group potency through group cohesion in squads.

Our hypotheses are summarized in the exploratory model shown in Figure 1, in which TL is considered as an antecedent of group potency. The model postulates the existence of direct and indirect relationships between TL and group potency through both group identification and group cohesion.

Figure 1 Teorical research model and hypothesis. 

We considered group size and time spent working with the evaluated leader in the team as control variables (^{Becker, 2005}). Prior studies have found that these variables could have significant group effects. The teamwork literature has suggested that the size of the team has an inverse relationship with team performance (^{Easley, Devarj, & Crant, 2003}) and time spent following the orders of the leader (^{Wheelan, 2005}). Both variables could increase barriers among group members (^{Liden, Wayne, Jaworski, & Bennett, 2004}).

RESULTS

Measurement Model

We explored the reliability of constructs applying the criterion of ^{Hair, Black, Babin, Anderson, and Tatham (2006)}, that is, reaching values of loadings over .60 and a critical value of 1.96 for p < .05. Each indicator was assessed by exploring the loadings of the indicators of each construct (λ). Indicator loadings and composite reliability are shown in Table 1. Overall, the factors showed adequate discriminant validity. Specifically, they were all above .70, the reference value, which shows that the discriminant validity of the factors was adequate for the criterion used (^{Hair et al., 2006}). This procedure yielded a composite reliability in which the different loadings of the indicators were taken into account. Composite reliability is similar to Cronbach’s α but is a better indicator of reliability (^{Henseler, Ringle, & Sinkovics, 2009}).

Table 1 Individual Loadings (λ), Composite Reliability Coefficient (CRC), t-values and Average Variance Extracted (AVE). 

To measure convergent validity, ^{Fornell and Larcker (1981)} proposed using Average Variance Extracted (AVE) and recommended that the AVE should exceed .50. Table 2 shows the construct’s factors, including means, standard deviations, correlations, and the square root of the AVE in the diagonal. One of the discriminant validity criteria was that the correlation between constructs should be lower than the indicator defined by the square root of the AVE to ensure that different phenomena are being measured (^{Fornell & Larcker, 1981}). The data suggested adequate discriminant validity. It is worth noting that significant relationships were found between all the factors and particularly between group identification and group cohesion.

Table 2 Descriptive Statistics and Correlations between the Variables Studied (N = 243 subjects and 51 squads). 

Variables	Mean	SD	1	2	3	4	5	6
1. Time under leader’s orders	1.39	0.67	1
2. Team size	6.27	1.39	.18	1
3. Transformational leadership	2.43	0.60	.03	.06	.91
4. Group potency	3.26	0.54	.17	.10	.28**	.94
5. Group identification	2.87	0.60	.20	.01	.65**	.37**	.78
6. Group cohesion	3.07	0.44	.09	-.10	.49**	.46**	.52**	.76

Note. Elements in the diagonal are the square root of the AVE between constructs and their indicators.

Time spent under the orders of the leader in years. Team size in number of members. The remaining variables are reported on a 5-point Likert scale (0-4).

Structural Model

This was done by conducting a linear regression in which the loadings could be interpreted as standardized beta coefficients. To determine the statistical significance of the overall results and calculate Student’s t for each structural effect, confidence intervals were based on a bootstrapping procedure with 500 samples (^{Henseler & Chin, 2010}).

To analyze the predictive value of the model for the dependent latent variables, we considered the criterion of ^{Falk and Miller (1992)}, according to whom the value of the proportion of variance explained (R ²) should be greater than .10.

First, we analyzed the direct relationships between variables (Figure 1) according to the established hypotheses, without considering the possible effects of indirect relationships between variables. Our analysis showed the following results: we found a positive and direct relationship between TL and Group Potency (β = .37, p < .01) that explained a proportion of variance (R ²) of .18. This point to confirm Hypothesis 1. An analysis of the relationship between TL and Group Identification without considering the direct relationship between TL and Group Potency showed a direct and positive relationship (β = .64, p < .01) that explained a proportion of variance of .47. This support Hypothesis 2. We found a significant positive and direct relationship between Group Identification and Group Potency (β = .56, p < .01) that explained a proportion of variance of .34, which points to confirm Hypothesis 3.

The analysis of the relationship between TL and Group Cohesion based on the value of the path without considering the direct relationship between TL and Group Potency showed a positive and direct relationship (β = .51, p < .01) that explained a proportion of variance of .29, which supports Hypothesis 4.

We found a positive and direct relationship between Group Cohesion and Group Potency (β = .56, p < .01) that explained a proportion of variance of .35, which points to confirm Hypothesis 5.

Exploring direct and indirect of multiple mediation we considered the hypotheses based on the postulates of ^{Preacher and Hayes (2008)}, proposing the existence of two mediators (Figure 1). Results are shown in Figure 2.

Figure 2 Search teorical model of multiple mediation 

When we explored the direct relationship between TL and Group Potency considering the overall model, we observed a considerable change in the relationship from β = .37 (p < .01) to β = – .07 (ns), with an increase in the variance explained from R ² = .18 to R ² = .40. The model suggests the existence of an indirect relationship between TL and Group Potency that is mediated by Group Identification and Group Cohesion, as we found significant values in the paths and the variance explained by such variables. Such results support hypotheses 6 and 7.

Team size and time spent following the orders of the leader variables were analyzed in the model; however, neither of these two variables presented a significant relation to the other variables of study.