Leadership and Semantics: Explorations of the Semantic Theory of Survey Responses
MetadataShow full item record
- Master of Science 
There are various theories arguing for one definition of leadership over another, and yet we find ourselves using circular explanations all the same (Eddy & VanDerLinden, 2006). In fact, the words “leader” and “leadership” have different, nuanced meanings depending on the different operating language (Schedlitzki, Ahonen, Wankhade, Edwards, & Gaggiotti, 2017). Though we use the word “leadership” in many settings--for example in the workplace, in schools, in sports, in politics--we rarely pause to contemplate if we truly know what exactly such a word entail. How academic research on leadership translates to a practical setting can be even more complicated. Researchers Avolio and Hannah (2008) observe that there is not one accepted theory about how to develop leaders. Neither is there a clear understanding about the organizational or individual factors that facilitate or accelerate such development. Yet millions of dollars are invested in leadership development programs in corporate and educational entities (Riggio, 2008). For example, in a study completed by Lunsford and Brown (2017) which studied 69 collegiate leadership centers, annual budgets ranged from $1,500 to $900,000, with near two thirds of the centers having a budget of $100,000 or more. Further investigation led to the conclusion that the implementation of evidence-based practices were not reflected in the leadership centers’ programs, philosophies or mission statements. One has to wonder if wasted effort and money on leadership and leadership development is due to our failure to solve the basic problem of what leadership is and what it means to people. Majority of the research on leadership and leadership styles has been conducted through the use of field surveys (Yukl, 2013). For example, the Multifactor Leadership Questionnaire (Bass & Avolio, 1990) is considered the standard instrument for assessing a range of transformational, transactional, and non leadership scales (Rowold, 2005). However, as researchers Podsakoff, MacKenzie and colleagues (2003) point out, surveys are often part of the problem when it comes to common method bias, especially due to item embeddedness. In their critical literature review, they cite research by Harrison and McLaughlin (1993), which further describes item-embeddedness as when the respondent, after analyzing clues from context, use an easily accessible set of cognitions to answer subsequent items. Could these context clues be the logic, linguistics, and meanings of the words used in the questions themselves? In line with the research of Arnulf and colleagues (2014), we think the answer could be ‘yes’. The purpose of this paper is to examine whether it is possible to design an experiment that allows us to study the difference between semantic and emotional determinants of survey responses. To accomplish this, we will first review the literature on leadership surveys and how the semantics of survey questions have potentially skewed our understanding of the concept itself. We will then discuss some of the empirical studies examining the cognitive processes that take place when subjects fill out such surveys, and whether it is possible to contribute to the theory using fMRI technique. Finally, we will conduct a behavioral experiment to test the STSR by specifically manipulating temporal aspects of surveys, which in theory should have no effect on semantic predictability of responses.
Masteroppgave(MSc) in Master of Science in Leadership and Organizational Psychology - Handelshøyskolen BI, 2018