For many years, science has emphasised the importance of transparency. We, the team behind 13Leaders, would like to meet this requirement. Therefore, on this page, you will find the evaluation of our current status and a description of the theoretical background to the 13Leaders check. We want to take you through the entire process, from initial idea, to the design of the questionaire. This includes: Where do we come from? Where are we now? And where will we go from here?
IMPORTANT: If you have not yet taken the test, please do so before reading on. Knowing too much about the test can influence your answers, we don’t want that, and neither do you. 😉.
Written by Lennart May
28. June 2022
For over 100 years, organisations have been searching for the right leadership style. It was argued in terms of the Great Man Theory for a long time that leadership is a trait (Carlyle, 1888). That is, leadership is in the blood and intuitive. Such trait theories do not provide much behavioural guidance for leaders since leadership occurs according to traits. This further implies that leaders must possess a similar repertoire of traits that increase the likelihood of becoming a leader. However, we have known for some time that personality traits can only explain a small portion of leadership potential (House and Shamir, 1993).
In the early 20th century, employees were still rewarded with high pay and fewer working hours. Leadership was authoritarian (Taylor, 1911). Due to the growing dissatisfaction with working conditions, authoritarian leadership style, and the increasing influence of unions, neoclassical organizational theories emerged. Famous studies, such as that of Roethlisberger and Dickson, who became known for the Hawthorne Effect, suggested that social factors and job satisfaction had a more substantial influence on behavior than monetary stimuli (Roethlisberger & Dickson, 2003). Likert concluded that groups have an enormous influence on the behavior of individuals in an organization.
For management, this meant, among other things: Flat hierarchies, overlapping work groups, motivation, and a wide span of control that allowed employee decisions (Likert, 1961).
Work was humanized and employees’ mental and physical abilities were enhanced, increasing productivity (Maslow, 1954; Herzberg, 1959). Emery and Thorsrud coined the term sociotechnical system and brought a new view of organizations, again expanding the tasks of management (Emery & Thorsrud, 1969). The maturity model with its situational approach to leadership changed the understanding of leadership again. (Hersey & Blanchard, 1977)
Now it is not only the personality traits of the leader, neither the motivation and skills of the employees, or the systemic characteristics of an organisation that leaders have to consider, but it is each situation that should be accompanied by a different leadership style (Hersey & Blanchard, 2007). The transactional and transformational leadership approaches are followed. Leaders can choose between the extrinsic motivation of employees and intrinsic motivation depending on the task. Leaders can create a vision and meaning for their employees. They can challenge them intellectually and respond to them individually (Kuhnert & Lewis, 1987).
So over the last hundred years, leadership has evolved from an autocratic authority to a democratic, delegating authority.
Successful leaders can take multiple approaches to leadership. They motivate, encourage, and challenge their employees. They set goals, coordinate, review progress, and evaluate performance (Daft, 2015).
Why do we see leadership behaviour in all human communities all over the world?
All these leadership theories interpret how people lead. We ask ourselves, why is leadership practiced? Why do we see leadership behavior in people all over the world? Evolutionary psychology also tries to answer this question. Evolutionary psychology builds on a simple logical proposition. If the bodies of all organisms, animals, and humans are the result of evolution, then the same is true for our brains. And our brain is us if you want to get philosophical. Our experience and behavior are therefore subject to the evolutionary mechanisms that are also responsible for the fact that female peacocks recognize the health of a male peacock by his plumage.
Evolution did not begin with humans. That’s why we also look at our relatives in the animal kingdom. Leadership has also evolved in the wild as a solution to the challenges of some species. Nevertheless, we recognize commonalities in qualities that we currently expect from a leader in humans as well. For example, it is the experienced matriarchs that lead orcas and African elephants in herd movements (van Vugt & Smith, 2019). Leadership, however, is not something rigid in the animal kingdom. For example, lactating zebras initiate herd movements, due to their increased energy demands (Fischhoff et al., 2007). And contrary to what is often portrayed, it is not always about eating or being eaten.
As in the animal kingdom, leadership positions provide an individual with prestige and standing. However, such leadership positions only relate to a specific situation or task. Someone good at hunting does not dictate to the gatherers how they should operate. The Navajo have a leader for war, peace, medicine, hunting, and ceremonies (Shepardson, 1963). To avoid authoritarian leaders, some mechanisms intervene as soon as someone becomes too greedy for power (Boehm & Boehm, (2009).
Many evolutionary psychologists distinguish leadership among two types. Prestige-style and dominance-style. Dominance-style leadership works through aggression and is forced on the rest of the group by physically superior males. Prestige-style leadership works by persuading and admiring. To do this, the leader must be charismatic and bring special qualities to the table (Henrich, 2016). A detailed discussion of evolutionary leadership is beyond the scope of this paper.
The findings from research on leadership raise the suspicion that there is no one-size-fits-all approach to leadership. Different situations and tasks call for different approaches to leadership. With this knowledge as an insight, we define leadership as a task that is solved by the leader and followers together. (Alznauer, M., 2016).
How leaders solve this task therefore also has something to do with their motivation, why they accepted responsibility for this task in the first place. Insofar as this decision was self-determined, the orientation can be traced back to three distinct human needs (Deci et al., 2017), which are aligned with the three dimensions of identity, intention and integration. These shape the behaviour of leaders. Leaders differ in the strengths and preferences within these dimensions.
Leaders use their identity to solve their leadership tasks. This characteristic is composed of the leader’s understanding of his or her role, which manifests itself in self-awareness as a leader. This refers to people who identify with the leadership position and feel that they are destined for it.
Numerous studies support the assumption that the self-confidence of leaders has a positive effect on employees and the success of their work. For example, employees identify more with their team the more confident the leader is (Fransen et al., 2016). Self-confident leaders, for one, may be viewed as charismatic, more attractive, and motivate employee job involvement (Bandura, 1986; Vroom, 1964). According to De Cremer and van Knippenberg (2004), self-confident leaders can be expected to tend to self-sacrifice on the job and create identification on the part of employees with the leader. Charisma may have evolved to quickly form groups and collectively solve an acute problem (Grabo, Spisak & van Vugt, 2017).
Leaders can draw on their intention to solve their leadership tasks. Intention describes the purposeful energy that leaders use to solve their tasks. Visions and goals are instrumental in driving direction. Good visions tend to be abstract and timeless. They reach far into the future and can never be completely fulfilled (Kirkpatrick & Locke, 1996).
Goals, on the other hand, can be directed towards a clear, temporally determined state (Locke & Latham, 1990). Their commonality is the desirable end state. Leadership includes goal setting (Northouse, 2018). For example, effective CEO’s were very good at creating their vision, communicating it and steering the organisation consistently and persistently in one direction (Bennis, 1982). Leaders rely on their vision to communicate strategies, beliefs and values (Conger & Kanungo, 1998).
Vision can make work more stimulating and employees more open to new situations (Bass, 1985). There are numerous studies that support the positive effects of vision-led leadership. Examples include organisational commitment, job satisfaction and financial performance (Kirkpatrick & Locke, 1996; Lowe et al. 1996; Waldman, Ramirez, House & Puranam, 2001). Leaders who set specific goals can direct their behaviour towards achieving them, and avoid useless behaviour (Locke & Latham, 2002). Specific goals increase focus and commitment (Chesney & Locke, 1987). Leaders who effectively formulate goals can make their goals more abstract or clear, depending on the complexity of the task (Berson, Halevy, Shamir & Erez, 2015).
Leaders can use the relationship with their employees to solve the leadership task. This factor is defined as integration. It includes what value the employees have for a leader. Factors such as promotion, motivation to work and well-being of the employees are decisive.
Integration goes beyond influencing employees to achieve overall organisational goals. Successful leaders build relationships with their employees based on trust and respect (Hogan & Kaiser, 2005). Studies support the importance of leader behaviour for employee motivation and behaviour (Gilbert & Kelloway, 2014). In addition, there is a growing literature on the positive impact of leadership behaviour on employee motivation (Barling et al., 2010). Leaders can be responsive to their employees, provide opportunities for decision-making and input, and inform rather than control (Baard et al. 2004; Hadre & Reeve 2009; Su and Reeve 2011). This allows leaders to increase employee engagement and promote intrinsic motivation (Deci et al., 2017).
In 2020, we began developing the test. First, we researched leadership styles, conceptualised a model, and then created questions to make that model measurable.
We hit on three categories, pretty quickly, that are commonly found in the literature: Confidence, goal setting, and employee development:
We built our model on these three basic topics. Leaders can thus solve their leadership tasks with three factors: Charisma, their goals, and their team. However, leaders do not use only one factor on its one. We assumed that leaders would use a combination. The results are 13 different possible combinations of the characteristics. Et voila: 13 different leadership types. Now we were allowed to fill the types with content based on the combination of the three factors.
In test construction, questions are called items. We created the items for our test using Logical Item Creation. This means that we created items by logical reasoning based on the literature and our theoretical model. We assumed that these items could measure identity, intention, and integration.
Each item was rated by test takers on a Likert scale from 1 to 5. Unlike currently, we did not confront the items. More on this later. If one wants to know how good the items are, there are several tests to check this. We calculated a Cronbach’s alpha for the first draft of our questions with 109 participants. The results are in the table.
How is Cronbach’s Alpha interpreted?
Conbrach’s Alpha measures internal consistency. In other words, it is a reliability measure. We use alpha to describe the extent to which the items on a test measure a construct (Tavakol & Dennick, 2011). When we develop a test that measures depression, for example, different items on a scale should measure a construct (e.g., listlessness). However, we do not ask “Are you listless?”. A researcher generates different items that all try to measure listlessness. These items should measure in the same direction, which is what Cronbach’s Alpha tests.
The table shows a simple guide for the interpretation of Cronbrachs Alpha (Gliem & Gliem 2003).
After you read the table, you see that our Cronbach’s alpha for the intention factor was unacceptable. The questions were measured crosswise and not in one direction, like those for identity and integration. So we had to formulate new items for Intention. Not all the questions had to be thrown out. It’s easy to see which items work quite well and which don’t. We kept the best ones and formulated a few new ones. A second pretest then gave us an acceptable score for Intention (α = .74).
The next challenge came quickly. All three factors correlated very strongly with each other. That is, high expressions in one factor are associated with high expressions in the other factor and vice versa. Do people want to be good at all three factors? Or are we perhaps measuring the same thing three times? We couldn’t be sure, but we knew we had to change something. So the same type of leadership would always come out.
The solution to our problem was to contrast the factors. We force our testers to choose between one of the two factors. This sounds harsh, but it is common in test construction. It’s called forced choice.
In the example, you can see how we contrast identity with intention. You contrast 15 times during one test.
Since we decided to contrast the scales, we again had to measure the internal consistency of the 13Leaders questionnaire. N = 225 test participants were included in the evaluation. You can see from the table that we calculated acceptable to good Cronbach’s alpha.
You can see that a new value has been added. We still show you the confidence interval for each Cronbach’s alpha. For multiple (infinitely many) measurements, our Cronbach’s alpha is within the confidence interval in 95% of the measurements. We know this is a strange interpretation. For complaints, you can turn to the grandchildren of Neyman and Pearson.
Having examined internal consistency, we have one more thing to check. We now know that the individual scales each measure in one direction. Do they measure something different?
Constructing tests, we don’t want to measure one construct with one question. We ask several questions that measure the same thing. That is called face validity. Of course, the questions must also be defensible according to our theoretical understanding. Ultimately, what matters is whether the pooled variables share variance.
We conducted an exploratory factor analysis, which we will explain in more detail.
In the table, you can see how the questions can be grouped into three factors. Item 11 does not fit into any factor. You can see this because the uniqueness (0.719) is too large. Factor 1 bundles items 12 through 15 and represents the preference between integration and identity. Factor 2 bundles items 1 through 5 and is interpreted by us as the preference between identity and intention. Factor 3 bundles items 6 through 10 and represents the preference between intention and integration.
A path diagram can more easily represent the relationship between the items and the factors. Here you can see how strongly the items load on a factor and whether the factors are related to each other. It shows that factors 1 and 3 are slightly correlated. We are still investigating why this might be the case.
The factor analysis was computed with an Oblique promax rotation since we assumed there could be correlations between the factors. We chose to use maximum likelihood estimation (MLE) in the calculation. MLE calculates under which factors the underlying correlation matrix is most likely. Additionally, we looked at the RMSEA value. This allows us to examine how well our three-factor model fits the data. We calculated an adequate fit (RMSEA = .081, 90% CI .065 – .097, BIC = – 185.665).
We deal openly with the fact that not everything is perfect from the beginning. Our 11th question doesn’t seem to fit into any factor. According to our theory, this item should have loaded on factor 2. If item 11 can’t be sorted into factor 2 even with more data, we will reformulate the item.
You can help us with this. Take our test and give us feedback
Our methods and this text will be updated regularly. ‘in the mean time, do get in touch, we value aour opinions.
Your 13Leaders Team.