Self-perception of height matters : a study on leaders’ height-dissatisfaction as a predictor of leadership style

Self-Perception of Height Matters:

A Study on Leaders’ Height-Dissatisfaction as a Predictor of

Leadership Style

Student: Wilhelmina (Dagmar) Lund

Student Number: 11087005

Email: wilhelminalund@gmail.com

Master Program: MSc Business Administration – Leadership & Management

Supervisor: Jill Knapen

Statement of Originality

This document is written by student Wilhelmina (Dagmar) Lund, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

This study is the first study known to have examined the effect of leaders’ self-perception of height in relation to their leadership style. The aim of this research was to extend the litera-ture on height and leadership by providing an understanding of the relationship between height dissatisfaction and leadership styles, with regard to both males and females. It was expected that leaders with high height dissatisfaction would report leadership styles charac-terized by low communion, and ineffective and darker sides of leadership, i.e. withdrawn, distrustful, authoritarian, and directive leadership. Two students collected data with a ques-tionnaire that was sent out to leaders in their own network. The quesques-tionnaire consisted of questions regarding demographic variables, leadership style, and height satisfaction. A total of 176 respondents participated in the study, with a majority of males (72%). Results con-firmed that height dissatisfaction influenced respondents’ leadership styles. Leaders’ with high height dissatisfaction reported the leadership styles withdrawn, distrustful, authoritari-an, and yielding. Withdrawn and yielding leadership styles were characterized by a passive and absent leadership, while distrustful and authoritarian were seen as despotic and harsh leadership styles. No substantial differences between males and females were found. Fur-thermore, leaders actual height showed no significant effect on leadership styles, which was in contrast to previous research. As a final remark this study addressed its strengths and limi-tations, and suggested areas for future research.

Keywords: leadership, leadership styles, height, height dissatisfaction, self-perception of height, The Leadership Circumplex Scan (CLS)

Table of Content

2.2 Leadership and Height ... 9!

2.3 Leadership and Self-perception of Height ... 11!

2.3.1!Self9perception!of!height.!...!11! 2.3.2!Self9perception!of!height!and!leadership!styles.!...!13! 2.4 A Gender Perspective ... 15! 3. Method ... 18! 3.1 Sample ... 18! 3.2 Measures ... 20! 3.2.1!Demographics.!...!20! 3.2.2!Circumplex!Leadership!Scan.!...!20! 3.2.3!Height!Satisfaction!Questionnaire.!...!21! 3.3 Procedure ... 22! 4. Results ... 23!

4.1 Distribution of Data and Correlations ... 23!

4.2 The Effect of Gender, Height, and HD on Leadership Styles ... 26!

4.3 Discriminant analysis ... 29!

4.4 Moderation Effect of Gender on association HD-Leadership Styles ... 30!

4.5 The Effect of Height on HD and Moderation Effect of Gender ... 31!

4.6 Summary of Results ... 32! 5. Discussion ... 34! 5.1 Height ... 34! 5.2 Height Dissatisfaction ... 35! 5.3 Gender ... 38! 5.4 Practical implications ... 40!

5.5 Strengths and Limitations ... 40!

5.6 Future Research ... 42!

6. Conclusion ... 43!

List of Figures and Tables

Figure 1. Research Model ... 5

Figure 2. The Circumplex Leadership Scan ... 7

Table 1. Definitions of the CLS Leadership Styles ... 8

Table 2. Demographic Variables ... 19

Table 3. Skewness and Kurtosis ... 24

Table 4. Correlation Table ... 25

Table 5. Informative Table – Leadership Styles per Gender ... 28

1. Introduction

The relationship between physical height and leadership is not a new concept. Physical height played a role in the success of both nonhuman and human lives long before Darwin developed his theory of evolution (Darwin, 1859; Murray & Schmitz, 2011). Over a century later, this concept remains a topic of discussion. As later argued by Tooby and Cosmides (1992) argue that, leader qualities from our evolutionary past still play an important role in determining who will succeed in today’s society. Consequently, physical stature influences an individual’s perception of who is best positioned to become a leader (Judge & Cable, 2004; Blaker, Rompa, Dessing, Vriend, & Herschberg, 2013; Re, Hunter, Coetzee, Tiddeman, Xiao, Debruine, & Perrett, 2013).

If an individual is tall, he or she can literally reach higher and see further compared to shorter counterparts. Furthermore, a tall person is admired, looked up to, and valued (Frieze, Olson, & Good, 1990; Hensley & Angoli, 1980; Judge & Cable, 2004). The additional centi-meter a tall person has influences a follower's perception of he or she as a leader. For exam-ple, a taller person is perceived as being more charismatic (Hamstra, 2013). Followers also tend to believe leaders are taller than their actual height yet they do not believe these taller individuals are more skilled or more successful (Hensley & Angoli, 1980; Marsh, Schechter, & Blair, 2009). Thus, tallness has benefits for gaining status in the workplace. For example, a tall United States presidential candidate has a greater chance of winning votes in the election than a shorter competitor (Stulp, Buunk, Verhulst, & Pollet, 2012). Likewise, a tall person who is seeking a managerial position in a firm has a greater chance of succeeding and getting a higher salary than shorter colleagues (Judge & Cable, 2004).

Tallness is beneficial for gaining status in a career and in politics. But what about short statured leaders who are unsatisfied with their height? How does this affect their leadership

styles? Alfred Adler (1956) invented the term “Napoleon complex”; suggesting that a short man’s stature leads him to act more aggressively, hostile, and confrontational in interpersonal relationships. This behavior is an act of feeling insufficient and trying to overcompensate for one’s undersized stature and lack of power (Adler, 1956). In turn, physical height is associat-ed with a dominant behavior: a tall individual takes up more space and occupies a more dom-inant position compared to a shorter individual. Consequently this domdom-inant “non-verbal” behavior makes others act more submissively and a tall individual often takes precedence when interacting socially (Tiedens & Fragale, 2003; Stulp et al., 2015). Thus, physical height is associated with dominant behavior and individuals’ height influences social confrontations. There are several ways in which physical height may influence the behavior of a leader and also others’ perceptions of the leader’s abilities and status. It is important to gain a prac-tical understanding of how a leader's physical height could positively and negatively influ-ence the work environment and how a leader's perception of oneself can influinflu-ence his or her own and others behaviors. In the current study, we theorize that self-perception of physical height can be a predictor of leadership style. Hence, self-perception, or how individuals view themselves, strongly influences individuals’ behaviors in general (Shapka & Kahn, 2012) and self-perception of positive physical attributes (e.g. tallness) correlate with individuals’ per-sonality features (Melamed, 1992). For example, a leader who is dissatisfied with one’s height may not have great self-esteem (Melamed, 1992; Müssen & Jones, 1957) and is there-fore likely to act more suspicious or distrustful (Müssen & Jones, 1957), while a taller indi-vidual may present a more assertive and dominant behavior (Tiedens & Fragale, 2003; Stulp et al., 2015; Melamed, 1992). Thus, height’s influence on leaders’ actual behavior is notable.

Previous research concerning height and leadership has focused on height’s influence on followers’ perceptions of the leader’s abilities and tall individuals’ possibilities to succeed (Hamstra, 2013; Judge & Cable, 2004). Little research has been conducted with regard to

individuals’ height and its effect on actual behavior (Tiedens & Fragale, 2003; Stulp et al., 2015). Furthermore, there is a lack of research concerning self-perception of physical height and leadership styles in the literature. Thus, this study is the first study known to examine the effect of leaders’ self-perception of height in relation to their leadership style. The aim of this research is to extend the literature on height and leadership by providing an understanding of the relationship between height dissatisfaction and leadership styles, with regard to both males and females. The research question is outlined as follows:

To what extent are leadership styles associated with leaders’ height dissatisfaction? Is this effect stronger for males than females?

In sum, this study further explores the previously proved relationship between height and leadership (Judge & Cable, 2004; Lindqvist, 2012; Blaker et al., 2013; Hamstra, 2013). We also examine the relationship between leaders’ self-perception of height and their leader-ship style. A gender perspective is added to both relationleader-ships to observe if there are differ-ences between males and females. The next section presents the complete research model and introduces the hypotheses.

1.2 Research Model

The research model (Figure 1) describes the expected relationship between leaders’ physical height and leadership style. As previously mentioned, tall individuals were more likely to occupy a leading position compared to shorter individuals since they were highly respected and perceived as more leader-like (Frieze et al., 1990; Hensley & Angoli, 1980; Judge & Cable, 2004; Blaker et al., 2013). Therefore, leadership styles between tall and short leaders were expected to differ. Tall leaders were expected to report leadership styles charac-terized by bright traits, inspirational, coaching, participative, and yielding; while short

lead-ers were expected to report leadlead-ership styles on the dark side of leadlead-ership, directive, authori-tarian, distrustful, and withdrawn, according to the Circumplex Leadership Scan definitions (Redeker, Vries, Rouckhout, Vermeren, & de Fruyt, 2014) (H1).

Actual physical height was also expected to influence an individual’s self-perceived height (H2). For example, a short statured individual showed low self-esteem due to acquir-ing less positive feedback (Müssen & Jones, 1975) or that an individual acted dominant or submissive towards others in social interactions due to the perception of one’s own height in comparison to others (Tiedens & Fragale, 2003; Stulp et al., 2015). In addition, self-perception of physical height, i.e. height dissatisfaction was expected to influence the behav-ior and actions of a leader (H3). For example, individuals were likely to act more confronta-tional if they tried to compensate for their physical stature (Adler, 1956), or they felt suspi-cious, distrustful, and afraid of others (Freeman, Evans, Lister, Antley, Dunn, & Slater, 2014). Finally, the effects of height and height dissatisfaction on leadership style were pected to be stronger for males than females (H4 & H5). The hypotheses are further ex-plained in the literature review.

2. Literature Review

The literature review is divided into four parts. The first part provides an overview of leadership and the Leadership Circumplex Scan, followed by a review of leadership and its relation to leaders’ physical height. The third part gives an introduction to leadership and self-perception of height. Finally, a gender perspective on the relationship between height and leadership is given.

2.1 Leadership

Leadership is an extensive field of research that is studied both separately and integrat-ed with other concepts, as in this case with physical height. Previous leadership research mainly concentrated on one or two traits of leadership, e.g., the contingency model (Fielder, 1964), human- and task-oriented leadership (Syroit, 1979), or on the positive dimensions of leadership, e.g., charisma (Shamir, House & Arthur, 1993) and transformational leadership (Bass, 1985; Judge & Piccolo, 2004). In later years, researchers have also turned their atten-tion to negative leadership traits, for example abusive and despotic leadership (e.g. Johnson, Venus, Lanai, Mao, Chang, & Kozlowski, 2012; De Hoogh & Den Hartog, 2008; Hogan & Hogan, 2001). Consequently, little research in the field of leadership incorporates a broader perspective that includes both positive and negative aspects of leadership. The next section will present the Circumplex Leadership Scan of Redeker et al. 2014) and the different leader-ship styles it integrates, which formed the basis of this study.

2.1.1 The Leadership Circumplex.

To integrate conceptualizations of leadership into one central model Redeker et al. (2014) constructed a leadership circumplex: the Circumplex Leadership Scan (CLS). Redeker

the CLS is based on interpersonal interaction theory and its interpersonal circumplex. The interpersonal circumplex is based on two dimensions: agency and communion. Agency im-plies dominance and control, and the condition where individuals strive for mastery and pow-er to be diffpow-erent. Communion means affiliation and love, and the condition whpow-ere individu-als strive for intimacy, union, and solidarity within a larger social unit (Wiggins, 2003; Rede-ker et al., 2014). Thus, these two dimensions also represent the vertical (agency) and horizon-tal axis (communion) in the CLS, see figure 2.

The leadership circumplex captures not only agency or communion separately, but also other styles that fall in between. This incorporation enables the CLS to embrace a variety of different leadership styles in a way that no other model has done before. In total the CLS in-cludes eight different octants that represent eight different leadership styles, as shown in Fig-ure 2. The eight styles are inspirational, coaching, participative, yielding, withdrawn, dis-trustful, authoritarian, and directive. The styles are further described in Table 1 on page 8. Figure 2. The Circumplex Leadership Scan

Table 1. Definitions of the CLS Leadership Styles

Source/note: Retrieved and modified from Redeker et al. (2014), p. 441.

The leadership styles in the CLS are related to different aspects of leadership. The left side of the circumplex, i.e. withdrawn, distrustful, authoritarian, and directive, is related to the darker side of leadership (e.g. despotic) and these styles are also low on communion. Ad-ditionally, the leadership styles distrustful and withdrawn are associated with passive and absent leadership, e.g., laissez-faire leadership. Leadership styles high in agentic behavior, directive, authoritarian, and inspirational leadership, are related to the definition of mascu-line and task-oriented leadership. High communion, human oriented leadership, and feminine leadership styles characterize the right half of the circumplex, i.e. inspirational, coaching, participative, and yielding. Especially inspirational and coaching are associated with sub-scales of transformational leadership, i.e., idealized influence, inspirational motivation, intel-lectual stimulation, and individualized consideration. The leadership style participative has an unstable position in the circumplex, as it is more closely related to the coaching leadership style than yielding and hence high in communion. Overall, the leadership styles characterized by low communion and low agency are related to ineffective leadership, while high

commun-Leadership style Definition

Inspirational Stimulate and persuade followers by exposing a clear vision. Motivates follow-ers to perform optimally when performance/organizational problems arise. Coaching Show appreciation for followers and let them know their importance. Positive

communication and listen to followers’ opinions.

Participative Including followers in all processes. Easily accept and integrate propositions from followers. Shows their support of the feelings/emotions of their followers. Yielding Flexible in interactions with followers, hesitant to provide guidance. Put

follow-ers interest high, above the organizations interest, and avoids being the center of attention.

Withdrawn Personally and professionally absent. Evade from confrontations and responsibil-ities. Late responding when problem occurs.

Distrustful Suspicious towards followers, quick and negative in judgments, and keeps dis-tance from followers.

Authoritarian Force followers to obey them. Perform harsh against followers and do not accept critique.

Directive Competitive in reaching success, actively monitor and correct followers. Be-haves strictly towards followers.

There are leadership styles that are not incorporated in the CLS, for example narcissis-tic leadership (see e.g. Rosenthal & Pittinsky, 2006) and entrepreneurial leadership (see e.g. Gupta, Macmillan, & Surie, 2004). However, for this study Redeker et al.’s (2014) circum-plex is considered to have incorporated a sufficient and comprehensive amount of leadership styles that are suitable and reliable for measuring respondents’ leadership styles with regard to their self-perception of physical height. In turn, the next section will introduce and address physical height and its relation to leadership.

2.2 Leadership and Height

Physical stature has an impact on how we look at each other. An individual who is re-spected and highly valued is often portrayed as a ‘big man’ (Murray & Schmidt, 2011; Hens-ley & Angoli, 1980). Individuals look up to and admire tall individuals (Frieze et al., 1990; Hensley & Angoli, 1980; Judge & Cable, 2004). Additionally, individuals overestimate the height of someone who is seen as dominant and successful according to Hensley and Angoli (1980) and Marsh et al. (2009). Consequently, there are several factors that influence others’ perceptions of tall individuals.

It is further argued that leader qualities from our evolutionary past still play an im-portant role in determining who will succeed in today’s society (Tooby & Cosmides, 1992). The evolutionary (leadership) theory suggests that how we act towards each other and whom we decide to follow is rooted in an evolutionary past: it makes us believe that taller individu-als are the ones to lead us (Murray & Schmitz, 2011; Van Vugt, 2006), since a great body stature “promote survivability” (Murray & Schmitz, 2011, p.1219). Accordingly, others’ per-ceptions of tall individuals play a significant role in the perception of leadership. With medi-ating effects of dominance, health, and intelligence, results from a study by Blaker et al. (2013) suggested that taller individuals were seen as more leader-like than shorter

individu-als. Their study used manipulated pictures of short and tall individuals to examine percep-tions of who was most likely to be a leader. The study revealed that taller males were more likely to be seen as leaders due to perceptions of dominance, vitality, and intelligence, while taller females were only seen as more intelligent than shorter females (Blaker et al. 2013). Furthermore, a tall leader also makes others believe to possess unique abilities in comparison to a shorter leader, for example persuasiveness (Young & French, 1996) and charisma (Ham-stra, 2013). Finally, due to the positive social perception of taller individuals’ social esteem, leader emergence, and performance, taller individuals are more likely to succeed career wise and to receive a higher salary (Judge & Cable, 2004). Thus, followers perceive tall leaders to be different from shorter leaders in several aspects.

Research has shown that individuals perceive taller individuals as more leader-like, however less is known about individuals’ actual behavior in relation to height. Being tall im-plies being larger in overall physical stature and therefore taking up more space and occupy-ing a more dominant position. When an individual presented this non-verbal dominant behav-ior it made other individuals act more submissively towards them and vice versa (Tiedens & Fragale, 2003). In a study by Stulp et al. (2015), taller individuals tended to take precedence in a tight passage, and in a hectic street environment, individuals tended to yield for the taller ones. Consequently, late maturing males (i.e. short) tended to demonstrate “rebellious atti-tudes toward parents”, (Müssen & Jones, 1957, p.255). Thus, physical height is associated with more dominant behavior and these studies showed that individuals’ height affects social confrontations.

In sum, taller individuals (and taller leaders) are seen as more leader-like and more in-telligent compared to shorter individuals (Blaker et al., 2013). Furthermore, taller individuals can take advantage of their height by making others act submissively towards them (Tiedens & Fragale, 2013; Stulp et al., 2015) and taller individuals are also believed to feel superiority

towards others when they interact socially (Tiedens & Fragale, 2003). Therefore, being tall makes others believe that someone look like a leader, that they act like one, and that they have exceptional characteristics compared to others.

Previous research has mainly focused on others’ perceptions of tall individuals, tall in-dividuals possibilities to succeed, and social perceptions of tall leaders. In addition, few stud-ies have been found that examined the actual behavior of tall and short individuals. In turn no previous research has studied height and its relation to different leadership styles. However, research reveals that taller leaders possess unique capabilities (Hamstra, 2013; Young & French, 1996), and that taller individuals act more dominant towards others in social interac-tions (Tiedens & Fragale, 2013; Stulp et al., 2015). Hence, as a taller leader is perceived as more charismatic this could imply that a taller leader also actually demonstrates a more char-ismatic leadership style compared to a shorter leader. For example the leadership styles inspi-rational and coaching are defined as charismatic leadership according to the CLS (Redeker et al., 2014). Therefore, in this study it is expected that leaders’ actual height influence their leadership style. This leads to the statement of the first hypothesis:

H1: Leaders’ actual physical heights are related to different leadership styles.

2.3 Leadership and Self-perception of Height

2.3.1 Self-perception of height.

The way others perceive a leader is not the only thing that influences a leader’s behav-ior. Self-perception, i.e. an individual’s belief about him or herself, strongly influences his or her general behavior according to Shapka & Kahn (2012). More specifically, self-perception of positive physical attributes (e.g. tallness) correlates with an individual’s personality fea-tures, for example independence and assertiveness, according to a study by Melamed (1992).

Thus it is reasonable to expect that an individual’s actual height influences self-perception of physical height. For example, the feeling of being short or tall in comparison to other indi-viduals, or self-perceived height, influences individuals to act dominant or submissively to-wards others in social interactions (Tiedens & Fragale, 2003; Stulp, 2015). Moreover, an in-dividual’s experiences of power reveal a positive relation with to self-perception of physical height; an individual who feels powerful (e.g. a leader) overestimates his own height (Duguid & Goncalo, 2012). This is consistent with previous research discussing the notion that leaders (i.e. powerful individuals) are often perceived as taller, more respected and successful (Hens-ley & Angoli, 1980). Due to this, taller individuals receive more positive feedback because of their height compared to shorter individuals (Müssen & Jones, 1957; Frieze et al., 1990; Hensley & Angoli, 1980; Judge & Cable, 2004; Melamed, 1992). In turn, shorter individuals who obtain less positive feedback feel insufficient and have lower self-esteem due to their height (Adler, 1956; Müssen & Jones, 1957; Prieto & Robbins, 1975; Melamed, 1992). Hence, there are various features that can influence an individual’s self-perception of physi-cal height.

Thus, an individual acts submissively when feeling less dominant in social interactions and an individual’s physical attributes can affect him or her to feel insufficient, which can lead to lower self-esteem. Therefore, it is expected that a leader’s actual physical height in-fluence the leader’s self-perception of physical height. In other words, a leader who is tall is expected to report a lower level of height dissatisfaction than a shorter leader because a shorter leader is expected to report a higher level of height dissatisfaction. Based on this, the second hypothesis is stated below:

2.3.2 Self-perception of height and leadership styles.

Physical height is related to one’s own and others’ perceptions of dominance (Hensley and Angoli, 1980; Marsh et al., 2009; Sorokowski, 2010). Additionally, feeling inadequate is suggested to be a consequence of an individual’s height dissatisfaction (Müssen & Jones, 1957; Prieto & Robbins, 1975). When individuals feel smaller in comparison to others, they reveal a lower level of self-esteem or act submissively (Müssen & Jones, 1957; Prieto & Robbins, 1975). Freeman et al. (2014) argued that feeling smaller increases an individual’s level of paranoia, i.e., they feel suspicious, distrustful, and afraid of others. These results are in line with Alfred Adler’s Napoleon complex (also known as short-man syndrome), as it explains the phenomenon of a short person’s stature affecting oneself to act more aggressive-ly, hostile, and confrontational in interpersonal relationships (Adler, 1956). Though, the Na-poleon complex has been dismissed and argued to be a myth: a small study found evidence that “taller men were more likely to lose their temper” (“Short men ‘not more aggressive’”, 2007, para. 2) compared to short men, when they dueled each other with sticks (“Short men ‘not more aggressive’”, 2007). However, this line of research has not been extended and studies finding significant results in line with the Napoleon complex are of greater occurrence (see e.g., Freeman et al., 2014; Müssen & Jones, 1957). Consequently, this study hypothesize that height dissatisfaction is associated with more aggressive and distrustful behaviors to-wards others.

In summary, self-perception of physical height influences individuals’ self-esteem and behavior. Additionally, an individual’s physical appearance (e.g., height and attractiveness) correlates with an individual’s personality. Tallness increases an individual’s confidence and esteem, which in turn results in an assertive, dominant, independent, uninhibited, and self-assured personality (Melamed, 1992). Thus, it is expected that an individual’s self-confidence plays a significant role in his or her behavior. Since an individual who is dissatisfied with his

or her physical height is likely to have lower self-esteem, the individual is also likely to act in a more suspicious and confrontational manner in interpersonal relations. This behavior is due to an attempt to compensate for one’s short figure and low power. Consequently, an individ-ual who is satisfied with his or her own height is more likely to show confident behavior to-wards others (Adler, 1957; Melamed, 1992; Müssen & Jones, 1957; Prieto & Robbins, 1975). Thus, an individual’s height satisfaction influences the individual’s approach towards others.

The influence of an individual’s height on their self-esteem and interpersonal relations, in regard to the CLS implies that an individual with high height dissatisfaction would report leadership styles on the left side of the circumplex. These styles are seen as low in commun-ion, ineffective, despotic and passive. Someone with high height dissatisfaction is expected to to act more aggressive towards others and experience a higher level of paranoia (Adler, 1957; Freeman et al., 2014). Therefore, he or she is expected to report an authoritarian or directive leadership style, according to the CLS definitions. Authoritarian leadership style defines leaders who force followers to obey them, perform harsh against followers and do not accept critique. The directive leadership style defines leaders who are competitive in reaching suc-cess, actively monitor and correct followers, and behave strictly towards followers (Redeker et al. 2014). We expect that authoritarian and directive leadership styles characterize an indi-vidual who is dissatisfied with one’s height and consequently tries to overcompensate for his or her lack of power and acts aggressive due to this. Hence, the dominant behavior of an in-dividual due to height dissatisfaction is related to the upper left side of the CLS.

The leadership styles distrustful and withdrawn represent the other quarter of the cir-cumplex that are also expected to be associated with leaders’ height dissatisfaction. A leader who is dissatisfied with one’s height is also characterized by low self-esteem and feels insuf-ficient. It is expected that the leader could feel distrustful and suspicious towards others, and acts submissively due to this (Müssen & Jones, 1957; Prieto & Robbins, 1975). The

leader-ship style withdrawn is defined as leaders who are personally and professionally absent, evade from confrontations and responsibilities, and are late to respond when problem occurs. In turn, being suspicious, quick and negative in judgments and keeping distance from follow-ers, characterizes the distrustful leadership style (Redeker et al., 2014). This in comparison to someone who is tall (and satisfied with one’s height) and therefore is seen as more charis-matic, persuasive, admirable, and acts more dominant (i.e. leader-like) (Hamstra, 2013; Hens-ley & Angoli, 1980; Frieze et al., 1990 Young and French, 1996). Consequently, a leader who reports low height dissatisfaction is expected to also report leadership styles on the right side of the circumplex, i.e. inspirational, coaching, participative, and yielding. While a leader who reports high height dissatisfaction is expected to report right-sided leadership styles of the circumplex, i.e. withdrawn, distrustful, authoritarian, and directive (see, page 8 for defi-nitions). In regard to this the third hypothesis expects the following:

!

H3: Leaders’ height dissatisfaction is positively related to the left side of the Circumplex Leadership Scan (CLS). So that a leader with high height dis-satisfaction is more likely to score higher on the leadership styles labeled withdrawn, distrustful, authoritarian, and directive.

!

2.4 A Gender Perspective

It is well known that far more males than females govern countries and occupy leading positions in organizations all over the world. Furthermore, leadership styles are not only characterized and defined as feminine or masculine (Redeker et al., 2014); there are also sig-nificant differences between genders in the relationship between height and leadership. For example, male careers have a more significant relationship to height than female careers (Judge & Cable, 2004). Males are also perceived as better leaders (Re et al., 2013), to possess more charisma (Hamstra, 2013), and perceived leadership capabilities are stronger for taller males than for taller females (Blaker et al. 2013). Hence, leaders’ height is proven to affect

that the relationship stated in hypothesis 1, “leaders’ actual physical heights are related to different leadership styles”, is stronger for males than females. In turn, the fourth hypothesis is stated as follows:

H4: The direct effect of a leader’s actual height on leadership styles is moderat-ed by gender so that this effect is stronger for males than for females.

!

It was previously hypothesized that leaders’ heights influence their leadership style (H1) and their height dissatisfaction (H2). Based on previous research that indicated gender differences in the relationship between height and leadership styles (H4), gender differences are also expected on the indirect relationship between leaders’ actual height and their leader-ship style through leaders’ height dissatisfaction. Thus, a male’s height is expected to influ-ence their height dissatisfaction to a greater extent than a female’s. This expectation is based on previous research, which indicated that an individual’s actual height affected their self-perception of height (Tiedens & Fragale, 2003; Stulp et al. 2015; Müssen & Jones), and re-search, which indicated differences in male and female leadership (Hamstra, 2013; Judge & Cable, 2004).

In addition, research showed that males and females over-reported their own height (Duguid & Goncalo, 2012; Spencer et al., 2002), though males overestimated their height to a greater extent than females did (Spencer et al., 2002). Furthermore, a study by Murray and Schmitz (2011) revealed that it was only significant that males with a greater body stature, not females, perceived themselves “as efficacious in terms of leadership ability, which makes them more likely to express interest in pursuing a leadership position.” (Murray & Schmitz, 2011:1221). In turn, Melamed’s (1992) ‘personality correlation study’ revealed that only male positive physical attributes correlated with independence, this was not significant for females. Therefore, this study expects that height dissatisfaction in males will affect their leadership style to a greater extent than in females. Hence, the fifth hypothesis is stated

be-H5: The indirect effect of leaders’ actual physical height on leadership styles through height dissatisfaction is moderated by gender, so that this effect is stronger for males than for females.

!

Consequently, in line with previous research this study expects to find significant dif-ferences between male and female leadership styles, both in relation to a leader’s height and their height dissatisfaction. This is expected even though previous research is proven to be rather varying: several studies have focused only on males (e.g., Lindqvist, 2012; Stulp, 2012; Young & French, 1996), other studies have found small, but significant differences (e.g. Judge & Cable, 2004), or substantial significant differences between males and females (e.g., Blaker et al, 2013; Hamstra, 2013). The expectation to find significant differences are based on the impression that most studies expected gender differences even if they were based on only one gender or only found small but significant results.

3. Method

This study was part of a larger research project and was conducted as a quantitative ex-ploratory study as it sought to evaluate the unexplored topic of leadership with regard to height dissatisfaction. Due to time limitations the study was cross-sectional and data were collected with a questionnaire. The chapter is outlined as follows: first a presentation of the study’s sample, second a description of the measures used, and finally a presentation of the research procedure is given.

3.1 Sample

The population related to this study was large (anyone with leadership experience) hence the sampling frame could not be determined. The goal was a minimum of 180 re-spondents in total. This goal was reached, however not all data were valid due to missing values, as some questionnaires were not completely filled out.

Respondents were selected through a non-probability sampling technique. The data were first collected through a convenience sample. The questionnaire was distributed to or-ganizations and leaders in our own networks: e.g., current or previous work, oror-ganizations we were active in, and to friends and family. Additionally, snowball sampling was used as re-spondents were asked to forward the questionnaire to other potential rere-spondents in their network. To increase the incentive to make respondents willing to participate the respondents were offered to receive feedback on their leadership style. This was highly appreciated and more than 150 reports were distributed to respondents.

A total of 176 respondents, 127 males and 49 females, with a current or previous lead-ership position participated in this study. Age of respondents was normally distributed and ranged from 19 to 70 years, with an average age of 46.6 years (SD=11.49). The average height for males was 183.77 cm (SD=7.08; range=165-200 cm) and for females 170.13 cm

(SD=6.81; range=153-185 cm). In total 60% of the respondents had a university education and 26 % had a higher vocational education. Most of the respondents (29%) reported “other” as their occupational sector (branch). Further, nearly 26% reported government as their sec-tor, followed by transportation/logistics (20%). Finally, since one Dutch and one Swedish student together collected data for this study, most of the respondents were located in the Netherlands (55%), Sweden (24%) and other parts of Europe (14%). Locations of respond-ents were determined by respondrespond-ents’ IP-addresses since no variable measured nationality. For further details see Table 2.

Table 2. Demographic Variables

Demographic variable N % Gender Male 127 72.16% Female 49 27.84% Total 176 Education Primary education 1 0.57%

Secondary vocational education 5 2.84% Higher secondary education 7 3.98% Pre-university education 4 2.27% Higher vocational education 46 26.14% University education 106 60.23% Other 7 3.98% Total 176 Branch Finance/Banking 15 8.57% Communication 1 0.57% Education 9 5.14% Government 45 25.71% Healthcare 9 5.14% Hospitality 1 0.57% Manufacturing 8 4.57% Retail 1 0.57% Transportation/Logistics 35 20.00% Other 51 29.14% Total 175 Country/Location Netherlands 97 55.11 % Sweden 43 24.43 % Other EU 25 14.20 % Asia 7 3.98 % Other 4 2.27 % Total 176

3.2 Measures

The questionnaire consisted of three sections: (1) Demographics, (2) Circumplex Lead-ership Scan, and (3) Height Satisfaction Questionnaire.

3.2.1 Demographics.

Control variables and other relevant background information were measured in the in-troduction of the questionnaire. The first items measured were: gender, age, height, educa-tion, occupational sector, and job description. Secondly, the respondents were asked if they had a management position. If the respondents answered yes, they were asked two additional questions: number of followers and numbers of years as a manager. If the respondents an-swered no, they were instead asked if they had or have had a leadership role (could be infor-mal such as team coach). This was asked to ensure that the respondents had a leadership ex-perience and in turn were suitable to participate in the study. The formats of the items were categorical (e.g., education, occupational sector, gender) and open (age, height, job descrip-tion, leadership experience).

3.2.2 Circumplex Leadership Scan.

Redeker et al.’s (2014) “Leadership Circumplex” was used to measure respondents’ leadership style. The Circumplex Leadership Scan (CLS) consisted of a questionnaire, which integrated statements associated to eight different leadership styles: inspirational, coaching, participative, yielding, withdrawn, distrustful, authoritarian, and directive (see Figure 2, page 7). The items, a total of 117 statements, were measured on a 5-point Likert scale (never, seldom, sometimes, often, always) and the questionnaire displayed four items per page. Eve-ry respondent were asked to “imagine themselves in a leadership role” before they answered the CLS statements. Examples of items were “listens to what the people say”, “clearly com-municates his/her role in employees' personal development”, “is invisible if there are any

problems”, and “judges employees harshly”. Reliability was shown to be good: the Cronbach’s alpha for the scales ranged from .70 to .83 according to the study of Redeker et al. (2014). For the current study, the reliabilities ranged from .69 to .86. The means and standard deviations are presented in the Table 4 (page 28).

3.2.3 Height Satisfaction Questionnaire.

To measure respondents’ height satisfaction, Knapen and Pollet’s (unpublished, 2016) Height Satisfaction Questionnaire (HSQ) was used. The HSQ was divided into three sub-scales: Height Dissatisfaction, Height Benefits, and Social Expectations of Tallness. It con-sisted of totally 20 items: seven graphics with statements and 14 statements regarding re-spondents’ self-perception of their physical height. The items were measured on a 7-point Likert Scale (e.g., 1= never/very dissatisfied and 7= always/very satisfied). Examples of items were “have you ever felt too small?”, “have you ever felt your height was advantageous to you?”, and “choose the person that you think is most respected by others”. Four items were reversed since they were counter-indicative. Further, one problematic reversed item, “would you like to be taller”, which was part of Height Dissatisfaction and Height Benefits, was re-moved. When this was done the level of reliability was acceptable: Height Dissatisfaction, α=.79 (M=12.99, SD=5.26); Height Benefits, α=.71 (M=3.33, SD=1.15); and Social Expecta-tions of Tallness, α=.85 (M=29.72, SD=4.87). According to the study of Knapen and Pollet (unpublished, 2016), the reliabilities differed between subset, subscales, and gender from α=.609 to α=.781.

As a final note, for this study all three HSQ subscales were included in the question-naire. However only the items for Height Dissatisfaction were used when analyzing the data, since the research question and the hypotheses were only stated to measured leaders’ height dissatisfaction.

3.3 Procedure

Since this study was part of a larger research project, two students collected data to-gether. The questionnaire was created with the online software Qualtrics (Qualtrics, 2016) and was available in both Dutch and English. The respondents could select the relevant lan-guage themselves.

The questionnaire was accessible through an online link and upon visiting this link each respondent received an informed consent message. The message included a brief presentation of the research project, information regarding anonymity and personal data, and contact in-formation. The respondents only received information that the study examined leadership styles and not in relation to self-perception of physical height. This information, together with the hypotheses, was disclosed after the respondents had completed the questionnaire in a final debriefing message. Upon stating their consent, respondents could access the questionnaire. After this the respondents filled out the questionnaire and in the end they could enter their email if they wanted to receive a leadership style report and/or a summary of the study’s re-sults. When the questionnaire was completed the respondents received a thank you message and were informed that the study examined leadership styles in relation to self-perception of physical height, since this was not stated before they began the questionnaire. A brief presen-tation of the study’s expecpresen-tations was also given.

The data collection took four weeks and reminders were sent out after circa one to two weeks. When enough respondents had completed the questionnaire the data was downloaded from Qualtrics (Qualtrics, 2016) and later uploaded to IBMs program SPSS 24.0. The com-plete research model and each hypothesis were tested in SPSS 24.0. The methods of analysis are further presented in the results chapter.

4. Results

The result chapter presents the findings from the hypothesis testing and briefly inter-prets the results. Since the research model included mediation and moderation effects, the hypotheses were tested with different analysis methods in SPSS version 24.0. Each analysis method is presented separately. The chapter is outlined as follows: first, a presentation of the data and the correlations within the model. Second, a test of the effects of height, height dis-satisfaction, and gender on leadership styles is presented. Third, a discriminant analysis and a further interpretation of the effect of height dissatisfaction on leadership styles are given. Fourth, the findings of the moderation effect of gender on the association height dissatisfac-tion-leadership style are presented, and fifth, the effect of height on height dissatisfaction is interpreted. Finally, a summary of the results is given.

4.1 Distribution of Data and Correlations

First, tests for normality, skewness and kurtosis were conducted. Results showed that the data was normally distributed and that the values of skewness and kurtosis were within accepted levels. The most notable variable was gender, which showed a high value of Skew-ness, .997 (SE=.183), and low value of Kurtosis, -1.017 (SE=.364). This could be explained by the fact that 72% of the participants were males and only 28% were females. For further details see Table 3, page 24.

Table 4 (page 25) presents the means, standard deviations, reliability, and the correla-tions of the variables used in this study. As shown in Table 4, gender significantly correlates with the leadership styles coaching, participative, and directive. Height showed significant correlations with withdrawn (-.174) and height dissatisfaction (-.452). Furthermore, the lead-ership style labeled withdrawn showed the strongest correlation with height dissatisfaction

(HD): Pearson’s r(171)=.347, followed by distrustful: r(171)=.328 and authoritarian: r(171)=.291.

Table 3. Skewness and Kurtosis

Variable Skewness Std.Error Kurtosis Std.Error

Gender .997 .183 -1.017 .364 Age .368 .183 -.457 .364 Height .352 .184 .390 .365 Leadership styles Inspirational .052 .186 .029 .369 Coaching -.039 .186 .080 .369 Participative -0.1 .186 .370 .369 Yielding .012 .186 -.283 .369 Withdrawn .107 .186 .068 .369 Distrustful .525 .186 1.227 .369 Authoritarian .388 .186 .669 .369 Directive .697 .186 .834 .369

HD showed a negatively significant correlation with the leadership style inspirational: r(171)=-.197. Coaching (-.144), participative (-.035), and directive (.046) leadership styles showed no significant correlation with HD. There were also correlations between the differ-ent leadership styles. This was expected since the leadership styles in the CLS showed simi-larities with each other: correlations between the CLS leadership styles and correlations with other commonly used leadership styles were also presented by Redeker et al. (2014).

The leadership styles inspirational, coaching, and participative had the highest mean scores as shown in the correlation table (Table 4). Additionally, a dominant leadership style score was calculated. This score represents the style where each respondent scored the highest, thus only one dominant style per respondent. In turn, the most dominant leadership style was coaching where 115 respondents (81 males and 34 females) scored the highest. This was fol-lowed by inspirational (42) and participative (9). Further details can be viewed in Table 5 on page 28. Finally, the mean score of HD (M=12.99; SD=5.26) showed that the respondents scored low on this variable since the maximum score was 42 (six items on a 7-point Likert Scale), as shown in Table 4.

Table 4. Correlation Table

Means, Standard Deviations, Reliabilities, and Correlations of Study Variables

Variable M SD 1 2 3 4 5 6 7 8 9 10 11 1. Gender 1.28 0,45 _ 2. Height (cm) 180.03 9.28 -.658** _ Males 183.77 6.81 Females 170.13 7.08 3. INSPIRATIONAL 3.96 .38 .022 .121 (.83) 4. COACHING 4.19 .35 .162* .043 .595** (.86) 5. PARTICIPATIVE 3.81 .33 .202** -.100 .273** .678** (.78) 6. YIELDING 2.73 .41 .026 -.150 -.484** -.088 .218** (.79) 7. WITHDRAWN 2.09 .41 .026 -.174* -.697** -.477** -.188* .609** (.83) 8. DISTRUSTFUL 2.03 .42 -.065 -.102 -.437** -.530** -.410** .299** .663** (.83) 9. AUTHORITARIAN 2.34 .45 -.129 -.032 -.122 -.350** -.465** .030 .361** .683** (.80) 10. DIRECTIVE 3.08 .42 -.181* .080 .324** -.053 -.250** -.171* -.096 .288** .555** (.69) 11. Height Dissatisfaction (HD) 12.99 5.26 -.097 -.452** -.197** -.144 -.035 .228** .347** .328** .291** .046 (.79) Note. N=171. Gender is labeled 1=Males and 2=Females. The Cronbach's alpha is presented along the diagonal in parentheses.

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

As revealed in Table 4 there are some significant correlations within the conceptual model. These correlations are in line with the hypothesized path: height dissatisfaction was positively related with leadership styles on the left side of the CLS. The left side of the cir-cumplex included the leadership styles withdrawn, distrustful, authoritarian, and directive.

4.2 The Effect of Gender, Height, and HD on Leadership Styles

To further explore these results and to test the overall research model a Multivariate Analysis of Covariance (MANCOVA) was conducted. This was done to examine the effect of the independent variables actual height, HD, and gender on the dependent variables (the eight leadership styles). The Box’s Test showed a significant result (p=.041), which implied that the assumption of homogeneity was not met. Therefore a bootstrapping procedure (2,000 resamples, 95% CI) was added to check for the robustness of the outcomes. A bootstrapping analysis seeks to make interpretations about population parameters by using information computed from the original sample, i.e. it calculates a resample (IBM Corp., 1989; Field, 2013).

The multivariate test was interpreted with Wilk’s Lambda, as it is the most common measure to evaluate the mean differences between groups on a combination of dependent variables (Everitt & Dunn, 2001; Field, 2013). Wilk’s Lambda was later used to interpret the results in the discriminant analysis as well. The multivariate test revealed that neither gender (p=.174) nor height (p=.590) showed any significant effects on leadership styles. However, there was a significant effect of HD on leadership styles: F(8, 159) =3.198, p=.002; Wilk’s Λ=.861, ηp2 = .139. According to the value of Wilk’s Lambda, 14% of the variance in lead-ership styles was explained by the differences in the level of height dissatisfaction. The ef-fects of height, HD, and gender on leadership styles were not consistent and further outcomes are explained below.

First, a leader’s height showed no significant effect on leadership style in the multivari-ate analysis. The follow-up ANOVAs with bootstrap analysis showed no significant results either. P-values ranged from .102 to .878 and bootstrap p-values from .130 to .889. This im-plies that leaders’ physical height showed no significant effect on leadership styles. Thus, hypothesis 1, which stated that physical height was positively related with differences in leadership styles, was not supported.

Second, the follow-up ANOVAs with bootstrap analysis showed that respondents’ HD had a positive robust significant effect on withdrawn (t(1)= 3.925, p< .001, ηp2=.085, 95% CI [.013— .039]; bootstrap: p< .001, 95% CI [.012 — .041]), distrustful (t(1)= 3.822, p< .001, ηp2=.081, 95% CI [.013 — .040]; bootstrap: p< .001, 95% CI [.012 — .042]), authori-tarian (t(1)= 3.627, p< .001, ηp2=.073, 95% CI [.012 — .041]; bootstrap: p=.001, 95% CI [.012 — .042]), and yielding leadership style (t(1)= 2.185, p=.030, ηp2=.028, 95% CI [.001 — .028]; bootstrap: p=.029, 95% CI [.002 — .028]). Thus, as hypothesized (H3), when re-spondents revealed a high level of height dissatisfaction they also represented leadership styles located on the left side of the circumplex. Notably, the leadership style directive (left side of CLS) showed no significant results with HD. However yielding (right side of CLS) showed significant results, although a lower significance than the three left-sided styles.

Finally, while gender showed no main effect on leadership styles in the multivariate analysis, there were some significant effects in the follow-up univariate ANOVAs with boot-strap. Gender showed a significant effect on the leadership style participative: t(1)= -2.530, p=.012; ηp2=.037, 95% CI [-.257 — .032]; bootstrap: p=.002, 95% CI [-.237 — -.0.55]. Fe-males’ mean score for participative was .144 points higher (on a 5-point Likert Scale) than males’, as shown in Table 5. For the directive leadership style the results were also signifi-cant: t(1)= 2.406, p=.017, ηp2=.034, 95% CI [-.032 — .324]; bootstrap: p=.006, 95% CI [.049 — .306]). The mean score for directive showed opposite results, males scored

signifi-cantly .178 points higher compared to females. Significant results were also found for coach-ing after bootstrap was added: t(1)= -1.967, p=.051, ηp2=.023, 95% CI [-.257 — .032]; boot-strap p=.037, 95% CI [-.226 — -.011]. The difference in mean score of males and females was nearly significant (.051); males scored .119 points lower than females. Thus gender in-fluenced some leadership styles, however not to a substantial level.

Table 5. Informative Table – Leadership Styles per Gender

Leadership Styles: Means and Standard Deviations per Gender

Overall scores Dominant Leadership

Style Mean

Difference

Leadership Style M SD Std. Error Sig. Number %

1. Inspirational 3.96 0.38 Male 3.95 0.39 -0.016 0.067 0.806 34 80,95% Female 3.96 0.36 0.016 0.067 0.806 8 19,05% Total 42 24,56% 2. Coaching 4.19 0.35 Male 4.15 0.36 -0.119 0.06 0.051 81 70,43% Female 4.27 0.30 0.119 0.06 0.051 34 29,57% Total 115 67,25% 3. Participative 3.81 0.33 Male 3.77 0.35 -0.144 0.057 0.012 5 55,56% Female 3.91 0.24 0.144 0.057 0.012 4 0,65% Total 9 5,26% 4. Yielding 2.73 0.41 Male 2.73 0.42 -0.026 0.07 0.712 1 100,00% Female 2.76 0.37 0.026 0.07 0.712 0 0,00% Total 1 0,58% 5. Withdrawn 2.09 0.41 Male 2.08 0.42 -0.01 0.068 0.881 - - Female 2.12 0.36 0.01 0.068 0.881 - - Total 6. Distrustful 2.03 0.42 Male 2.05 0.45 0.084 0.07 0.235 - - Female 1.99 0.31 -0.084 0.07 0.235 - - Total 7. Authoritarian 2.34 0.45 Male 2.38 0.47 0.144 0.075 0.055 1 100% Female 2.26 0.36 -0.144 0.075 0.055 0 0% Total 1 0,58% 8. Directive 3.08 0.42 Male 3.13 0.44 0.178 0.074 0.017 3 100% Female 2.95 0.36 -0.178 0.074 0.017 0 0% Total 3 1,75% Total 171 100%

Note. N=171. Overall score represent the mean and SD for each respondent’s score for all leadership styles. Dominant leadership style represents the style where each respondent scored the highest.

4.3 Discriminant analysis

The multivariate (MANCOVA) and bootstrap analysis were followed up by a discrimi-nant analysis to further examine the results of HD on the eight different leadership styles and to test the main hypothesis: that a leader with strong height dissatisfaction was more likely to represent a leadership style on the left side of the circumplex. A leader with high height dis-satisfaction was expected to report the leadership styles withdrawn, distrustful, authoritarian, and directive.

We divided HD into three categories: low HD score, moderate HD score, and high HD score. The score was calculated using the HD mean value of 12.99 ± one standard deviation (5.26): low HD < 7.73; moderate HD = 7.73 — 18.25; and high HD >18.25. Six respondents ended up in category 1 (low HD), 135 in category 2 (moderate HD), and 30 in category 3 (high HD). A total number of 171 observations (respondents), which represented 97,2% of the total observations, were grouped for the discriminant analysis.

Two discriminant functions were created. Function 1 discriminated the high HD score group from the low and moderate HD group. Function 2 discriminated the low HD score group from the high and moderate HD group. The first discriminant function explained 85,9% of the variance, canonical R2=0.12, and the second one explained only 14,1%, canoni-cal R2=0.02. When the two functions were combined no significance for differing the treat-ment groups were found: Λ=.862, χ²(16)=24.506, p=.079. Similar results were obtained when the first function was removed (Λ= .978, χ²(7)=3,628, p=.821). This implies that the group predictor variable of HD was not able to predict any statistically significant differences when discriminating groups.

As the discriminant analysis revealed, the relationship between leaders’ height dissatis-faction and their leadership style remains ambiguous. The discriminant function classified only 52% of the HD score groups correctly. From a total of six for group 1 (low HD), three

were correctly classified. For group 2 (moderate HD) only 67 were correctly classified from a total of 135. Finally, for group 3 (high HD) with a total of 30, 19 were correctly classified. This confirmed the non-significance of Wilk’s Lambda and indicated a rather low predictive capability of HD as a discriminant function.

4.4 Moderation Effect of Gender on association HD-Leadership Styles

Since there were no significant results found for height on leadership styles this rela-tionship was not further examined. The moderation of gender on the indirect effect of height on leadership styles was not tested either. However, since there was a significant effect of HD on leadership styles a moderation effect of gender on this relationship was tested.

To test the gender moderation effect on this relationship, Hayes’ Process Model 1 was used (Hayes, 2013). HD represented the independent variable (X), leadership style the de-pendent variable (Y), and gender the moderator (M). Each leadership style was then added to the model individually. As predicted from previous findings in the multivariate and bootstrap analysis, significant effects of gender on the leadership styles participative (p=.012; bootstrap p=.002) and directive (p=.017; bootstrap p=.006) were expected. When the test of the moder-ation effect was conducted no significant results were found for any leadership style. The same results were revealed when adding a bootstrap effect (2,000 resamples, 95% CI). P-values ranged from .060 for inspirational to .907 for directive. Participative showed a non-significant value of p=.778.

The effect of HD on leadership style was not moderated by gender. The expectation that gender had a different effect on respondents’ height dissatisfaction was not accurate. Hypothesis 4 was not supported, as there was no significant relationship between height and leadership style. In turn, the additional moderation effect of gender for the relationship be-tween HD and leadership styles was not supported either.

4.5 The Effect of Height on HD and Moderation Effect of Gender

Hypothesis 2 stated that leaders’ actual physical height was negatively related to height dissatisfaction. In addition hypothesis 5 stated that this relationship was moderated by gen-der, with the effect being stronger for males than for females. To test these hypotheses a re-gression analysis was conducted.

Results showed that there was a negatively significant effect of actual physical height on HD (b=-.513; R²=.250; F(1,173)=57,585; p< .001). The results indicated that 25% of the variance in HD was explained by leaders’ physical height. Tall leaders scored low on HD; thus tall leaders showed low dissatisfaction with their actual height. This finding was there-fore supportive of the second hypothesis of this study. To test for the moderation effect of gender on this relationship we centralized the mean scores of gender and height and then multiplied them to create an interaction variable (gender*height), as suggested by Aiken and West (1991).The first step was to enter height and gender into the model separately, which showed significant results: R²=.268; F(2,172)=31.504; p< .001. The second step was to in-clude the interaction term of gender*height into the model. Results showed R²=.275; F(3,171)=21,581; p< .001, however the moderation effect of gender only accounted for a change in R² of 007, and p=.217, which was not significant. Therefore it can be concluded that these outcomes showed no support for the fifth hypothesis.

For a final observation, a regression analysis was conducted to separately test the po-tential direct effect of gender on HD. The results were not significant, R²=.009; F(1,174)=1,668; p=.198. Thus, there was no difference between males and females in the HD score. The effect of gender on HD proved to be insignificant as a moderation effect and as a direct effect on HD.

4.6 Summary of Results

Leaders’ actual physical height revealed no significant effect on the eight leadership styles. Hypothesis 1 which stated, “leaders’ actual physical heights are related to different leadership styles”, was therefore not supported. The second hypothesis, “leaders’ actual phys-ical heights are negatively related to height dissatisfaction”, was supported. Results showed that short leaders were more dissatisfied with their height than tall leaders were.

The third hypothesis, “leaders’ height dissatisfaction is positively related to leadership styles on the left side of the circumplex”, was supported. It was expected that a leader with a high height dissatisfaction score was more likely to score higher on the leadership styles withdrawn, distrustful, authoritarian, and directive. This hypothesis was (mostly) supported. The leadership styles withdrawn, distrustful, and authoritarian, which represented the left side, showed positive statistically significant results with leaders’ height dissatisfaction. In addition, the leadership style yielding (on the right side) also showed a positive significant result. However, no significant results were found for the leadership style directive.

Hypothesis number 4, “the direct effect of leaders’ actual physical height on leadership styles is moderated by gender, so that this effect is stronger for males than for females”, was not tested, since there was no significant result found for the effect of height on leadership styles, and therefore not supported. Instead a moderation effect of gender on the relationship between height dissatisfaction and leadership style was tested, but no significant results were found for this relationship. The final test examined the fifth hypothesis; “the indirect effect of a leader’s actual physical height on leadership styles through self-perceived height dissatis-faction is moderated by gender, so that this effect is stronger for males than for females”. The moderation effect of gender showed to be insignificant and no direct effect of gender on HD was established either. However, the multivariate test showed significant results of gender on the leadership styles participative, directive, and coaching.

Conclusively, as predicted by the main hypothesis, leaders’ dissatisfaction of physical height affected their leadership style. However this relationship was not stronger for males than females. Further interpretations will follow in the discussion chapter.

5. Discussion

This chapter digs deeper into the findings presented in the results chapter. It compares and contrasts results with previous research and addresses the study’s strengths and limita-tions. The discussion is first divided into three parts: the effect of height, the effect of height dissatisfaction, and the effect of gender on leadership styles. This is followed by a presenta-tion of the findings’ practical implicapresenta-tions and a discussion of the study’s strengths and limi-tations. Finally, directions for future research are addressed.

5.1 Height

In accordance with previous research, we expected tall leaders to report leadership styles related to charisma and transformational subscales in the CLS (i.e. inspirational, coach-ing, and participative leadership style), while shorter leaders were expected to report leader-ship styles related to darker and passive traits of leaderleader-ship (i.e. directive, authoritarian, dis-trustful, withdrawn; Redeker et al., 2014). However, hypothesis 1 was rejected since no ef-fects of physical height on self-reported leadership style were found.

Consequently, these results are not in line with previous research, which stated that tall individuals were more likely to become or to be seen as leaders (Lindqvist, 2012; Judge & Cable, 2004; Blaker et al. 2012), because taller individuals were perceived as more leader-like and respectful (Frieze et al., 1990; Hensley & Angoli, 1980; Judge & Cable, 2004; Blaker et al., 2013), and perceived to possess unique capabilities associated with leadership, for example charisma and persuasiveness (Hamstra, 2013; Young & French, 1996). This study adds to previous literature as it questions the strength of the relationship between height and leadership; leaders’ actual height showed no direct effect on leaders’ behavior and atti-tudes towards others.

As further expected in hypothesis 2, tall leaders showed little dissatisfaction with their actual height. Thus, the second hypothesis was supported: there is a direct effect of a leader’s actual height on their height dissatisfaction. These results relate to previous research, suggest-ing that a short statured individual has lower self-esteem because he or she receives less posi-tive feedback and therefore feels inadequate (Müssen & Jones, 1975; Melamed, 1992). Con-sequently, leaders’ height only showed an effect their height dissatisfaction but not their leadership style.

5.2 Height Dissatisfaction

A leader’s self-perception of his or her actual height, i.e. height (dis)satisfaction seems to play a more important role than leaders’ actual height. This study’s research question and main hypothesis (H3) was to examine whether leaders’ height dissatisfaction affected their leadership style and to determine whether or not this effect was stronger for males or females. The third hypothesis was supported. Leaders who scored high on height dissatisfaction also reported the leadership styles withdrawn, distrustful, and authoritarian, which are part of the left side of the circumplex. The leadership style yielding, as part of the right side of the CLS, was also associated with leaders’ height dissatisfaction. Therefore, height dissatisfaction re-vealed to have a stronger influence on leadership styles then leaders’ actual height.

This study adds to the existing leadership literature, as it is the first to demonstrate an association between height dissatisfaction and leadership styles. Leaders who were dissatis-fied with their physical height reported leadership styles defined as forcible and harsh in in-teractions (i.e. authoritarian); suspicious, negative and quick in judgments, and keep distance from followers (i.e. distrustful); absent and avoidable of interpersonal confrontations (i.e. withdrawn); and hesitant in guiding employees and escaping center of attention (i.e. yield-ing). This is in line with the third hypothesis and these findings advocate the importance of