Effective leadership has been identified as an important factor associated with the successful implementation of evidence-based practices (EBPs) in mental health services (1–3). The focus on implementing EBPs in the healthcare system started in the 1990’s (4). Over the years, it has subsequently been shown that successfully implementing EBPs may lead to better and more effective health care services, with lower costs, higher job satisfaction among clinicians, and higher patient satisfaction (5–7). Leaders are an important part of any implementation process, as they can influence the organizational climate at the workplace, cooperation between team members, and employee’s attitudes towards the EBP ( (8–11). With the growing interest in the role that leadership plays in effective EBP implementation, there is a need for establishing reliable and valid measures to assess leadership behaviors that relate to successful implementation.
In the quest to discern the leader’s role in the implementation of EBPs, there was initially a focus on identifying general leadership behaviors associated with different implementation outcomes (11, 12), using leadership concepts such as the Full-Range Leadership (FRL;13) theory. FRL is one of the most widely used leadership theories. It describes different leadership behaviors such as transformational leadership, in which leaders motivate and encourage employees; transactional leadership, where the leader rewards and punishes employees based on performance; and non-leadership, where the leader has a more “hands off” approach and avoids making decisions (14, 15). Several studies have shown a relation between transformational leadership and different implementation outcomes, such as employees’ attitudes towards EBPs (16), motivation (17), turnover intention (18), burnout (19), and overall improved performance at all levels of the workplace (20–22).
The Multifactor Leadership Questionnaire (MLQ) is based on the FLR leadership theory, and is likely the most frequently used scale to measure leadership (23). The scale has been psychometrically validated several times, showing acceptable scores (ranging from α = .78 − .94; 21–24). The MLQ has also previously been validated in a Norwegian sample (24), where each subscale showed adequate psychometric properties (α = .62 − .84). However, the validations differ as a result of researchers altering the original factor structure by either combing or excluding certain factors or items (25, 28). The original structure consists of nine subscales, where idealized influence was separated into behaviors and attributed charisma (Bass & Riggio06). However, several researchers treat idealized influence as one factor (23, 25, 28), and the scale has accordingly been broadly used as an eight-factor scale, including idealized influence, inspirational motivation, intellectual stimulation, individualized consideration, contingent reward, management-by-exception (active), management-by-exception (passive) and laissez-faire. However, although most use the eight- or nine-factor models (25, 30, 31), some have rearranged the subscales into categories different from what was originally proposed (i.e. two subscales measuring non-leadership; (26, 32)). These alterations could cause misunderstandings when using the MLQ in relation to implementation outcomes. The creators of the scale have recommended to analyze each of the eight subscales of the MLQ individually, with the exception of the subscales of transformational leadership, which can be combined (33). The current paper therefore considers the MLQ as consisting of eight subscales: four subscales measuring transformational leadership, three subscales measuring transactional leadership and one measuring non-leadership.
While the interest in the relationship between these more general leadership concepts and successful EBP implementation was growing, research on strategic leadership behaviors related to implementation was lacking. Building on the growing evidence base that general leadership behaviors relates to effective EBP implementation, scientists have turned their interest towards specific leadership behaviors that may be more proximally related to successful EBP implementation. Newer research have established that such specific, or often termed strategic leadership behaviors, may provide additional explanatory value in the investigation into how leadership relates to key implementation outcomes (8, 34, 35).
The focus on identifying strategic leadership behaviors for implementation coincides with a more general call within the implementation research field to develop simple, brief and psychometrically sound implementation measures (36, 37). In 2014, Aarons, Ehrhart and Farahnak developed the Implementation Leadership Scale (ILS), drawing from a broad base of theory and research on implementation, leadership, and organizational climate (11). Results from studies investigating the effect of strategic leadership behaviors (i.e. implementation leadership) have revealed that these promote organizational change (12). This is consistent with findings reporting that employee-ratings on the ILS are correlated with factors considered to be important during the implementation of EBPs and their sustainment (38).
The ILS was initially developed in the U.S., and has been validated several times (39–41). The ILS has shown excellent psychometric properties in investigations using both employee-ratings and leader self-ratings, and in multiple sectors (42–46). The ILS contains four subscales, including proactive leadership, knowledgeable leadership, supportive leadership and perseverant leadership, and the suggested four-factor structure has been confirmed in all studies (5, 11, 39, 46). Analysis of reliability have found internal consistency to be excellent (Cronbach’s α ranging from .92 − .98) (42, 44, 46). Convergent validity has been investigated by correlating the ILS with the MLQ, finding moderate to high correlations (Pearson’s correlation ranging from .63 to .75) between the two leadership concepts (11). Discriminant validity has been established by correlating the ILS to theoretically unrelated implementation concepts, such as the Evidence-based Practice Attitude Scale (EBPAS), finding zero to low correlations (Pearson’s correlations ranging from .05 to .4) (11).
Only two articles have investigated the psychometric properties of the ILS outside the U.S. – in China and in Greece (5, 47). Employee ratings have previously been used when investigating the psychometric properties in a U.S. context, and a similar investigation in a Norwegian context would provide further evidence for the relevance of the concept of implementation leadership.
The purpose of this study is to examine the psychometric properties of the Norwegian version of the MLQ and the ILS. First, the factor structure and internal consistencies of the two scales will be explored. Secondly, we will examine the convergent and divergent validity of the MLQ and the ILS. Based on previous findings regarding the ILS, we expect to find support for a four-factor model and high internal consistency for the total scale and all subscales. In addition, we anticipate that the ILS will have moderate to high correlations with the MLQ and subsequently low correlations with the EBPAS. In accordance with other studies, we expect to find support for an eight-factor model (48, 49), as well as similar results regarding convergent and divergent validity as hypothesized above.