Authentic Leadership Questionnaire applied to Brazilian nurses: evidence of validity

Abstract Objective: to establish the psychometric properties of the Authentic Leadership Questionnaire (ALQ) applied to Brazilian nurses. Method: cross-sectional observational study with a non-probabilistic sample. The psychometric properties of the RATER and SELF versions of the ALQ were calculated using confirmatory factor analysis with the WLSMV robust estimation method. The following indices were used to assess the goodness-of-fit of the model: chi-square by degrees of freedom (χ2/df), Tucker-Lewis Index (TLI), Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA) and Standardized Root Mean Squared Residual (SRMR). Data reliability was analyzed using the ordinal coefficient alpha and composite reliability. Results: 181 nurses participated of the study (female gender: 80.1%; mean age of 34.6 years; working time of less than five years: 76.3%). The complete ALQ RATER and ALQ SELF models did not present an adequate fit. Therefore, the refined models presented a better fit to the sample data (ALQ RATER: χ2/df=2.77; CFI=0.97; TLI=0.97; RMSEA=0.10; SRMR=0.05; ALQ SELF: χ2/df=2.74; CFI=0.94; TLI=0.92; RMSEA=0.10; SRMR=0.08). In the ALQ RATER model, items 1, 7 and 13 were excluded. Due to the high correlation between the factors Relational Transparency and Moral Perspective, a three-factor model based on the combination of the factors mentioned above was proposed. In the ALQ SELF model, items 2, 5, 9 and 10 were excluded. Likewise, a three-factor model based on the combination of two factors, now called Self-Awareness Balance, was proposed. Conclusion: the data obtained with the Authentic Leadership Questionnaire with Brazilian nurses were valid and reliable.


Introduction
The concept of "authenticity" is related to positive psychology and has its roots in Greek philosophy. It has first been applied in the fields of sociology and education (1) , but, in 2003, the construct gained greater prominence in the area of management and leadership, and authentic leaders began to be defined as those who are aware of how they think and behave and are perceived by others as aware of their own values, moral perspectives and knowledge and of the strengths of others (1) .
Authenticity requires trust, optimism, resilience and high moral character and is related to being true to yourself. The theoretical model of Authentic Leadership is characterized by transparency in relationships, sharing of information and feelings, organizational commitment, and satisfaction in performance through conduct consistent with a system of personal values and convictions.
Authentic Leadership can result in greater motivation to engage in leadership roles as it creates a healthy work environment and promotes leadership self-efficacy (2) .

Furthermore, a study suggests that Authentic Leadership
is associated with the quality of customer service (3) .
The American Association of Critical-Care Nurses considers authentic leadership as the main leadership model and claims that it is capable of creating and sustaining healthy work environments in critical care settings (4) . Based on the assumptions of this model, those who are led can find greater meaning in the activities performed and greater commitment to work, creating an environment that supports both leaders and their subordinates.
The theory that supports Authentic Leadership is based on four pillars: Self-awareness (leaders' willingness to constantly analyze their strengths and opportunities for improvement); Relational Transparency (leaders' ability to remain consistent with their values in the relationship with subordinates); Balanced Processing (unbiased decision making) and Internalized Moral Perspective (leaders' values are consistent with their moral conduct) (5) .
Aiming to assess the authenticity of leaders, evaluating their ethical, moral, behavioral and leader; and the SELF scale, to assess the participant's own leadership profile. Subsequently, validation studies of the ALQ were carried out for its use in various countries, such as Spain (7) , France (8) and Pakistan (9)(10) .
Criticism of the original 2008 study (6) pointed out that the researchers did not report the use of modification indices related to the tests or the adjustment of models.
In this sense, the authors encouraged other researchers to carry out new studies to explain and detail the analyzed performed (11) .
Recent research reported that the four factors that make up the ALQ (Self-awareness; Relational Transparency; Balanced Processing; and Moral Perspective) explain the composition of Authentic Leadership (11) .
It is important to emphasize that so far this instrument has not been validated in Portuguese nor for Brazilian people, and its psychometric properties have not been evaluated in a sample of nurses working in a hospital environment. However, as it one of the most used instruments to asses leadership style (10) , the cultural adaptation of the ALQ to Brazilian Portuguese and its evaluation by leaders working in the health area are urgent matters. This may help to design strategies that can contribute to a more effective leadership through fair and authentic management and transformative leadership.

Study design
This is a cross-sectional observational study with a non-probabilistic sample, carried out in two general hospitals located in the Center-West Region of Brazil.

Population and sample
The first institution (Hospital A) is a public teaching hospital with 210 inpatient beds in different specialties and approximately 1,600 professionals, of which 230 are nurses. The second institution (Hospital B) is a private, non-profit entity, under municipal management, with 671 beds and more than 3,600 professionals, of whom 260 are nurses (12) .
The target population of this study was composed of nurses working in these institutions. The inclusion criteria were nurses with an active bond with the institutions, and a total of 300 nurses were considered eligible.
The minimum sample size for the factor analysis was calculated considering five to 10 subjects per parameter of the factor model (13) . As the ALQ (SELF and RATER) has  Sometimes, 3: Often and 4: Very often, almost always) (6) .
The authorization to use the ALQ was granted directly by Mind Garden, Inc., the company responsible for licensing the questionnaire, representing Dr. Bruce J. Avolio.

Psychometric properties of the ALQ -evaluation of the internal structure
The psychometric sensitivity of the items, the factorial, convergent, and discriminant construct validity, and the reliability of the instrument for the sample (internal structure) were estimated.
Psychometric sensitivity was estimated using summary (mean, median and standard deviation) and shape measures (asymmetry and kurtosis) of the distribution of the responses to the items and was considered adequate when the absolute values of asymmetry and kurtosis were less than three and seven, respectively (14)(15) .
The factorial validity was tested by means of Confirmatory Factor Analysis (CFA) using the robust estimation method of mean-and variance-adjusted weighted least squares (WLSMV). The following indices were used to assess the goodness-of-fit to the sample data: chi-square by degrees of freedom (χ 2 /df), Comparative Fit Index (CFI), Tucker-Lewis Index (TLI),

Root Mean Square Error of Approximation (RMSEA) and
Standardized Root Mean Squared Residual (SRMR).

Convergent validity was evaluated by Average
Variance Extracted (AVE) and considered adequate if AVE ≥ 0.50 (18) . The method used to assess the validity of the discriminant construct was the one was the one proposed in 1981 (18) , which considered that if AVE i and AVE j ≥ ρ ij 2 , the existence of discriminant validity can be confirmed.
The reliability of the factors was evaluated using the ordinal coefficient alpha (α) and the Composite Reliability (CR). Values of α and CR greater than 0.70 were considered indicators of reliability (15,18) .
After the factor adjustment of the ALQ RATER and SELF scales, a second-order hierarchical model (SOHM) was proposed in order to verify the contribution of each factor to the general construct of leadership.

Ethical aspects
The ethical precepts on the guidelines and standards  Regarding the analysis of the psychometric properties of the ALQ for the sample, the CFA of the complete and refined models of the ALQ RATER are shown in Figure 1.

Discussion
In the ALQ RATER scale, the high correlation between the factors Relational Transparency and Moral Perspective made it possible to combine these factors.
This is justified by the direct relationship between their theoretical concepts, since showing the true self to others (relational transparency) is an essential characteristic of those who behave according to their personal values (moral perspective). In addition, it is known that people tend to perceive and evaluate transparency based on their own structures, values, emotions and cognitive limitations (21) , and that relational transparency requires genuineness and transparency of leaders to gain the trust of their followers (22) , propagate their thoughts and seek harmony in the group (9) .
It is important to emphasize that it is necessary to invest in strengthening interpersonal relationships between leaders and team members, so that communication can be more assertive and self-confidence can be developed in the team (23) . Similarly, moral perspective requires a leader that can cultivate and promote values, so that they can make fair decisions based on high ethical standards (6,22) . Thus, the Relational and Moral factors are associated with transparency and ethics, which must go hand in hand. Corroborating the above, a study carried out in Belgium shows that the authentic leadership and behavioral integrity of the leader are related to follower performance and organizational commitment, and that this relationship is maintained by controlling the ethical organizational culture (24) , correlating the factors Relational weaknesses) (7,21) . Furthermore, the items are not always correlated with a single factor, as there is some degree of association with conceptually related factors (25) . In Spain, the ALQ was applied to a sample of 623 workers from public and private organizations from different sectors and, as in Brazil, there was a need to exclude three items (7) . sizes, which found that the dimension with the greatest explanatory power for the leader's assessment was Selfawareness, followed by Balanced Processing (26) . In this sample, which compared 1019 Brazilian employees to 842 Portuguese employees, the theoretical explanation for this fact was that leaders' self-knowledge and relationship with subordinates are essential for an authentic leadership (26) .
As in the RATER scale, the exclusion of items 2, 5, that it is possible for the factors to overlap (25) , as occurred between Balanced Processing and Self-Awareness, which showed perfect correlation. For this reason, a single factor was proposed (Balance and Self-Awareness). This can be justified using the theoretical framework of the instrument itself, considering that in order to make a coherent decision it is necessary to analyze it beforehand (balanced processing) and to be aware of one's own weaknesses and strengths (self-awareness). Balanced processing refers to the leaders' ability to carefully analyze a situation before making decisions, while also being able to accept other points of view, even if they are different from their own (1) .
Similarly, self-awareness refers to the deep perception of their values and beliefs and to how they behave and are perceived by others (1,27) . It is also worth noting that the creator of the instrument himself suggests that it is not reasonable to conceptualize the four factors of the ALQ as assessing entirely separate and distinct constructs (6) .
Given the above, the psychometric evaluation of the ALQ indicated the validity of the internal structure of the three-factor models of the RATER and SELF versions.
Regarding the differences in the adjustment of the two versions (SELF and RATER), when participants assess themselves as leaders, they are able to perceive balanced processing, being able, in their opinion, to make unbiased decisions; the same occurs with self-awareness, as individuals believe they understand the impact they can have on people (5) . However, the participants present difficulties in understanding the importance of relational transparency and moral perspective between the leader and the subordinates, which are characterized by high ethical standards guiding the behavior (5)(6) .
In this investigation, the second-order hierarchical model was only possible for the RATER scale, which showed a strong correlation between the factors. This result corroborates the study by the creators of the ALQ (6) , which highlights that the four factors (Relational Transparency, Moral Perspective, Balanced Processing and Self-Awareness) are not independent and that the leadership construct, a single second-order factor, can explain this dependence.
The review of the validation by the authors of the original version of the ALQ (11) , in line with the original 2008 article (6) reinforced the importance of the four theoretical components of the Authentic Leadership model. However, the authors encourage the development of other model validation studies (11) .
Thus, the proposed ALQ validation analysis may contribute to expanding the knowledge of professionals and researchers in the areas of nursing and hospital management, as the study presents the cultural adaptation of the ALQ to Brazilian Portuguese (31) as well as evidence related to the validity of the instrument's internal structure when applied to nurses working in Brazilian hospitals.
The limitations of this study refer to the use of a non-probabilistic sample and to the sample size, aspects that prevent the generalization of the results. However, it is worth noting that the SELF and RATER versions of the ALQ were considered valid and reliable instruments to be applied to the population of nurses, which will certainly contribute to the improvement of investigations on authentic leadership in Brazilian healthcare settings.

Conclusion
The SELF and RATER scales of the Authentic Leadership Questionnaire (ALQ) applied to a sample of Brazilian nurses allowed the collection of valid and reliable information, which made it possible to assess the authentic leadership of these professionals. However, for the ALQ to be used, it was necessary to make adaptations in the instrument's internal structure.
The SELF and RATER scales can be used