Rater characteristics : the moderating effect of rater attractiveness

(1)

Rater characteristics:

The moderating effect of rater attractiveness

Nikki Ouborg

10648011

Msc. Business Administration

Leadership & Management Track

Supervisor: F.D. Belschak

Date of submission: 29-06-2015

(2)

Statement of originality

This document is written by Student Nikki Ouborg 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.

The Faculty of Economics and Business is responsible

solely for the supervision of completion of the work, not for

(3)

Acknowledgments

This Master Thesis forms the end of my Leadership & Management Track. As my hart and interest truly lies in ‘management’ and in ‘people’, this study has provided me with the opportunity to gain more insight into the interaction of people. A message that I learned is that people and their decisions are not as rational as we believe. There is more to it than we think, and therefore, it is key to continuously increase our knowledge on this matter.

Delivering this final version is a significant moment in my life as it announces the end of my studies. I would like to express my deepest gratitude to all the people who inspired me and provided me with knowledge, ideas, and support throughout my studies the past 4 years.

I would like to thank my supervisor, F.D. Belschak, for his guidance, knowledge,

recommendations, and supervision. His patience and advice were essential throughout the process of writing my thesis. I would also like to thank my family and friends who have supported me, believed in me, encouraged me to reach my full potential, and who have put faith in me through all the hard work.

Naturally, I am grateful to the respondents who took the time to participate in my survey and provided me with the much-needed data.

(4)

List of tables and figures

Tables

Table 1 Means, Standard Deviations, Correlations and Reliabilities Table 2 Results of Perceived leader attractiveness as predictor of Leader

selection.

Table 3 Regression results for i)Perceived leader communion, ii) Leader trust, and iii) Leader competence as a mediator of the relationship between Perceived leader attractiveness and Leader selection.

Table 4 Bootstrap results for i)Perceived leader communion, ii) Leader trust, and iii) Leader competence as a mediator of the relationship between Perceived leader attractiveness and Leader selection.

Table 5 Results of Rater attractiveness as moderator of the relationship between Perceived leader attractiveness and Leader selection. Table 6 Regression results for Rater attractiveness as moderator of the

mediation of Perceived leader communion between Perceived leader attractiveness and Leader selection.

Table 7 Regression results for Rater attractiveness as moderator of the mediation of Perceived leader trust between Perceived leader attractiveness and Leader selection.

Table 8 Regression results for Rater attractiveness as moderator of the mediation of Perceived leader competence between Perceived leader attractiveness and Leader selection.

Figures

Figure 1 Conceptual model

Figure 2 Path model of the moderated mediation.

Figure 3 Interaction effect of Leader attractiveness*Rater attractiveness on Leader competence.

(7)

Abstract

This study extends research on the process of how people make attributions, by exploring if physical attractiveness leads to a higher selection rate for a leadership role. More specifically, it fills a research gap by focusing on the role of the rater in these processes. The central question this study addresses is if the self-perceived attractiveness of the rater influences the relationship between the perceived attractiveness of the leader and leader selection. A scenario study is conducted among 313 working Dutch respondents. The respondents evaluated a facial image (unattractive or attractive) on valued leadership characteristics (communion, trustworthiness and competence) and for the selection of a leadership role. Finally, the rater evaluated their self-perceived physical attractiveness. The data is analyzed by using the PROCESS tool for SPSS to conduct mediations, moderation and moderated mediation analyses. The results show that a person with perceived leader characteristics receive a higher selection rate. Mixed results were found of the effect of physical

attractiveness. It was negatively related to leader selection and to the mediators communion and trustworthiness, and positively related to the mediator competence. The role of the rater only showed a significant effect for competence, such that the mediated relationship is stronger under high rater attractiveness than under low rater attractiveness.

As the role of the rater has received limited attention in research on leadership, this study shows that someone self-image can influence what and how attributions are made. In the context of research on leadership and the leader selection process, the new variable of self-perceived attractiveness of the beholder introduces new knowledge on which future research can build.

Key words: physical attractiveness, leader selection, communion, trustworthiness, competence, leadership

(8)

1. Introduction

Do you remember when you were just a child? An important lesson that your parents tried to teach you was: ‘Don’t judge a book by its cover’. However, referring to the globalized world in the twentieth century, these ‘covers’ are emphasized everywhere around us. The visual has become the norm and face-to-face interactions with both familiar and unfamiliar persons are easy. This influences how people make impressions and judgments of others. People seem to repeatedly ignore the lesson learned by their parents, as they can’t seem to escape the growing use of the visual. As what is visible is easily observed, the invisible effects are much more difficult to understand.

This study links this to the context of leadership and the leader selection process. Leaders are a very important factor of the success of a company, organizations and even societies (Poutvaara, 2014). To select the ‘right’ leader it is therefore important to gain insight into this process of leader selection. However, the evaluation of another person may be

influenced by the visual, their appearance, more than we think. The concept ‘thin slicing’ is in line with the above, as it states that people make accurate attributions and impressions from someone’s appearance (Nana & Jackson, 2010). More specifically, the facial physical

attractiveness is important while making these judgments. Literature explains the functioning of physical attractiveness as a ‘halo effect’ (Thorndike, 1920). The halo effect is defined as the failure of a rater to discriminate among different attributes of another person, which results in a higher covariance of attributes (Leuthesser, Kohli & Harich, 1995). In literature this is referred to as ‘the what is beautiful is good’ stereotype (Dion, Berscheid and Walster, 1972; Zebrowitz & Franklin, 2014), and implies physical attractiveness to positively influence the judgments of other attributes.

In the context of leadership, this study aims to gain more insight into the leader selection process by exploring the attractiveness halo effect on three generally valued

(9)

leadership characteristics: perceived communion, trustworthiness and competence. Therefore, it builds on the implicit leadership theory (ILT) that suggests that people have a mental image that contains characteristics of leaders and informs their perception of who is a leader and who is not (Schyns & Mendl, 2005).

In literature on leadership and leader selection there mainly seems to be a focus on the leader, but a limited focus on the beholder or rater. The rater is also a participant in the leader selection process. It does not matter if this rater is a follower, or if it concerns a supervisor selecting an individual for a promotion, its role should be acknowledged. As there is no existing research that focuses on the physical attractiveness of the rater in the leader selection process, this paper aims to fill this gap. Consequently, it aims to answer the following

research question: Does the self-perceived attractiveness of the rater influence the relationship between the perceived attractiveness of the leader and leader selection?

To answer this question the use of the ‘similarity-attraction theory’ will be extended to the similarity of physical attractiveness between a rater and a leader. Thereby, it will be used next to the attractiveness halo effect. The theory states that individuals have the tendency to engage in more positive social interactions with other that are similar to them (Byrne, 1971; Huang & Iun, 2006). It is assumed that similarity will strengthen the positive halo effect of attractiveness.

This will be put into practice by conducting a scenario study. The respondents are asked to imagine a scenario in which they work in an organization that will get restructured. In the following survey they will evaluate a candidate on the basis of a facial image (attractive or unattractive) for a leader position. The survey measures the perceived leader

characteristics, and the effect of physical attractiveness of the leader and their self-perceived physical attractiveness.

(10)

Hereby, this paper aims to advance the understanding on leadership and the process of evaluating and selecting a person for a leadership role. Moreover, it aims to increase

knowledge on the process of how people make attributions from someone’s appearance and the role of physical attractiveness in that process. It looks deeper into the attractiveness-selection relationship by including explaining mechanisms (thin slicing, the halo-effect, self-fulfilling prophecy, ILT) and valued leader characteristics as possible mediators (perceived communion, trustworthiness and competence). Furthermore, this study aims to fill a research gap by focusing on the rater in the process of leader selection. The similarity-attraction theory will be applied in a new context as the role of the self-perceived attractiveness of the rater will be explored in the attractiveness-selection relationship. Therefore, this paper could contribute to the understanding of both the production and consumption processes of the human face in leadership and leader selection.

This knowledge could be used in real life situations. For example, when making decisions during elections, when supervisors are deciding whom to promote, but also during the selection of a new employee in a job application. Even in our day-to-day life we may use this knowledge. For example, when meeting a new person and first impressions are made. When a person is more aware of the influences that someone’s appearance, facial attributions, physical attractiveness and/or their self-image may have on their perceptions and judgments of others, more rational judgments and decisions can be made.

First, the paper will provide a clear overview of the key concepts and theoretical background that allows it to build hypotheses regarding the research question (§2). The following section will describe the used research methodology (§3). Subsequently, the data analysis and results will be reported and interpreted (§4). The discussion section will reveal the significance of the findings and what this means for the research question of this paper (§5). Also it contains the limitations, the directions for future research on this topic, and the

(11)

theoretical and practical implications. Finally, conclusions are drawn regarding the findings of this study (§6).

2. Theoretical background

The following section provides a detailed theoretical background containing the key concepts of this paper and previous research that contributes to building hypotheses and answering the research question.

2.1 Physical attractiveness and leader selection

Overall, research on physical attractiveness has the tendency to treat physical attractiveness as an independent variable (e.g. Braun et al., 2012). Furthermore, research presumes that there are objectively visible traits, such as facial asymmetry and averageness that are universally perceived as attractive or unattractive (e.g. Fink & Penton-Voak, 2002; Gangestad & Scheyd, 2005). Recent work, including a meta-analysis on facial attractiveness preferences, has found high consistency between judgments of facial attractiveness within and across cultures (Coetzee, Greeff, Stephen & Perret, 2014; Langlois et al., 2000). In other words, they

concluded that there is an agreement across raters about who is and who isn’t attractive, both within and across cultures. Therefore, this paper follows the assumption that physical

attractiveness is an objective characteristic whose perception is universally shared.

Referring to Gordon Allport’s observation, this paper assumes that there is something in the nature of individuals that makes raters or observes to attribute certain characteristics to individuals (Allport, 1937). Furthermore, it follows Ambady and Rosenthal’s (1992)

assumption that this ‘something’ is communicated through expressive behavior that is mostly unintended. This expressivity can lie in someone’s speech, body, tone of voice, or such as in this study; someone’s face. Facial psychology literature suggests that the human face can be used as a source of communication (Nana & Jackson, 2010). Viewing an unfamiliar face

(12)

activates visually derived semantic codes. These codes are based on the facial physical appearance of a person and provide social information to the viewer (Bruce, 1991; Nana & Jackson, 2010). This social information includes the age, gender, attractiveness, emotions, and personality of a person. Furthermore, it is evaluated as the most frequent expresser of nonverbal information (Rule et al., 2010), and is frequently used to make judgments about individuals (Nana & Jackson, 2010). This study assumes that the face is a critical source of information upon which people make attributions from, most especially before they meet someone (in photographic or film form), or when they meet the individual in person for the first time. In exploring the process of leader selection, this study will therefore include the facial appearance. Please take note that when there is referred to physical attractiveness in this study it therefore implies the physical attractiveness of a person’s face.

In trying to explain how people make judgments from someone’s face, the concept ‘thin slicing’ is introduced. Thin slicing is the ability to form judgments and impressions of strangers based on a limited source of information (Ambady & Rosenthal, 1992). People use thin slices of information, like the facial appearance, as an information shortcut that simplifies both daily and important judgments and decision-making (Poutvaara, 2014). Even more, people are able to judge what is really important from these thin slices of information (Nana & Jackson, 2010). Research suggests that quickly made decision (thin slicing) are generally as accurate, valuable, and predictive as cautious and conscious made decisions (Houser, Horan & Furler, 2007; Tidwell, Eastwick & Finkel, 2013). The accuracy and reliability of the impressions that are made during the brief information sourcing is a surprising aspect of thin slicing (Ambady & Rosenthal, 1992).

The accuracy of thin slicing has been examined in a broad range of contexts. Overall, the literature shows that when people view short periods of nonverbal behavior, they are able to make accurate judgments of personality traits, teaching ability, marital satisfaction, sexual

(13)

orientation, and social status (e.g. Albright, Kenny & Malloy, 1988; Ambady & Rosenthal, 1992; Kraus & Keltner, 2009). A leadership study found that perceptions of leader

competence, based upon a CEOs face, were able to accurately predict the financial

performance of the CEOs organization (Rule & Ambady, 2008). A similar study focused on solely female CEOs. The facial appearance of female CEOs predicted not only their

organization’s success, but also their personal financial success as executives (Rule & Ambady, 2009).

So far, the effect of the facial appearance has been discussed. More specifically, this study will focus on the role of (facial) physical attractiveness in the process of leader

selection. Linking the concept thin slicing to physical attractiveness, it seems that someone’s physical attractiveness can influence the judgments that a person makes about him or her (Ambady & Rosenthal, 1993). Overall, research shows that attractive people invite more favorable responses from other individuals (Langlois et al., 2000; Rule & Ambady, 2010; Shinada & Yamagishi, 2014). There is a tendency for people to perceive physically attractive individuals to possess desirable personal characteristics, such as intelligence, benevolence (Shinada & Yamagishi, 2014), social competence, and emotional stability (Ambady & Rule, 2010). In addition, this seems to be the case for both children and adult perceivers and targets (Langlois et al., 2000). Furthermore, this is equal for perceivers familiar and unfamiliar with the target, and regardless of the amount of information there is available for the perceiver while forming the impression about the target (Langlois et al., 2000).

This high consistency of findings suggests it can be assumed that people are more likely to attribute positive characteristics to attractive others. This ‘positive’ effect of physical attractiveness is also known as the ‘attractiveness halo effect’, whereby more positive

characteristics are ascribed to physically attractive individuals over less attractive individuals (Zebrowitz & Franklin, 2014). Thorndike (1920) was the one to name the ‘halo effect’, that he

(14)

regarded as the tendency for judgment of a novel attribute or characteristic of an individual to be influenced by the value of a known, but objectively irrelevant, attribute. Leuthesser et al. (1995, p.58) state a clear definition of the halo effect: “Since Thordike’s original

conceptualization, the halo effect has been fairly frequently defined as a rater’s failure to discriminate among conceptually distinct and potentially independent attributes, with the result that individual attribute ratings co-vary more than they otherwise would”.

In research on the halo effect, physical attractiveness began to play the role of the ‘irrelevant attribute’ that influences evaluative judgment on other aspects. Dion et al. (1972) referred to this as the ‘what is beautiful is good’ stereotype. They suggested, “a physical attractiveness stereotype exists and that its content is perfectly compatible with the ‘what is beautiful is good’ thesis” (p. 289). They found that physically attractive individuals were assumed to possess more socially desirable personalities than less attractive individuals and, even more, to have happier and more successful lives. More studies suggested this effect as attractive individuals were rated to be healthy (White, Kenrick & Neuberg, 2013), likeable (Braun et al., 2012), successful (Poutvaara, 2014), and intelligent (Shinada & Yamagishi, 2014).

Apart from the consistency of findings on attributions and judgments made from physically attractive faces, which are mostly positive, research also shows that these seem to be predictive of outcomes (Rule et al., 2010). In many cases the perceptions of faces are indirectly or even directly predicting someone’s behavior towards another person (Rule et al., 2010). For example, judgments of physical attractiveness of a person’s face affected

employers’ decisions about who to hire (Little, Roberts, Jones & DeBruine, 2012), the prediction of a CEOs’ compensation and their companies’ profits (Poutvaara, 2014; Rule et al., 2010), and the expectations of teachers about children’s intellectual abilities (Rule et al., 2010).

(15)

An explaining mechanism on this accuracy is the ‘self-fulfilling prophecy’. Merton (1948, p.195) defined this concept as: “a false definition of the situation evoking a new behavior which makes the originally false conception come true”. It is possible that

individuals, who are treated as if they possess specific personality characteristics on the basis of their physical attractiveness, will actually develop these characteristics in response to interactions with other people (Ritts, Patterson & Tubbs, 1992). Individuals may have

different types of interaction styles for people whom they perceive to be physically attractive than for those who they perceive as unattractive. These differences in interaction styles may, in turn, evoke and nurture behavior from the target according the stereotype (Snyder, Tanke & Berscheid, 1977). In other words, a physical attractive person may start to behave in a

likeable, friendly or social manner, because the behavior of other evokes and nurtures behaviors taken to be manifestations of such characteristics and not just necessarily because they possess these characteristics. As the theorem of philosopher W.I. Thomas states: “If men define situations as real, they are real in their consequences” (Merton, 1948, p.193). This may explain why judgments of someone’s appearance are often accurate and predictive of

outcomes.

So how does the effect of physical attractiveness relate to the process of leader selection? Within research on leader selection there seems to be increasing evidence that physical attractiveness and the facial appearance of a leader are important factors for a candidate’s success in real election of leadership roles (Little, 2014; Little et al., 2012; Poutvaara, 2014; White et al., 2013). For instance, research has shown that people are more likely to vote for leaders that are physically attractive (Little et al., 2012; Poutvaara, 2014; White et al., 2013), and that people can predict electoral winners based upon facial images (Poutvaara, 2014). Different explanations are available for why people may desire facial attractive leaders. Little et al. (2012) state attractiveness is a trait that is to be valued in leaders

(16)

because it signals healthiness and the possession of positive personality characteristics. For example, facial attractiveness may signal biological quality such as a strong immune system or longevity (Thornhill & Gangestad, 1999). From an evolutionary perspective, people who are perceived as attractive, and therefore healthy, are more valuable to work with to succeed for a positive future. This may explain why attractive individuals are treated better in a social perspective and why people choose an attractive leader over a less attractive leader.

Furthermore, physical attractiveness is associated with a range of positive personality attributions. If such positive personality characteristics are perceived to be beneficial to the group, this may lead individuals to value physical attractive leaders (Eagly, Ashmore, Makhijani & Longo, 1991).

The expectation follows that when an individual is asked to rate a person when viewing a facial image of an unfamiliar attractive or an unattractive leader, through thin slicing he or she will use the face as a source of information to make an accurate evaluation and decision. The halo effect suggests that facial attractiveness will influence judgments made about the person shown in the image, such that more positive attributions will be made for the attractive person compared to the unattractive person. Drawing on these assumptions, the expectation is that an individual will be more likely to choose an attractive person over an unattractive person for a leadership role. In other words, this paper assumes a positive relation between perceived leader attractiveness and leader selection. This will be tested regarding the following hypothesis:

(17)

2.2 Leader communion, trustworthiness, competence and leader selection

So far, the direct relation between physical attractiveness and leader selection has been

discussed. However, in trying to understand the leader selection process, it is key to also focus on the indirect relation between attractiveness and selection. It is important to explore what other factors may play a role in this relationship. As discussed, this paper assumes on the basis of the thin slicing theory, that the facial appearance influences judgments of individuals (Ambady & Rosenthal, 1992). In the context of leadcer selection, the facial appearance of a leader is used as a source of information that influences an observer’s perception and evaluation of that leader. Certain characteristics will be assigned on the basis of a leader’s face and these may influence their judgments and choice when selecting a leader. The (perceived) possession of leadership characteristics makes it more likely that appropriate actions will be taken and that these actions are successful (Kirkpatrick & Locke, 1991). In other words, when individuals are perceived to possess leadership characteristics, this may influence an observer’s choice of selection.

This assumption is based on the ‘implicit leadership theory’ (ILT). This is well-known theory in leadership literature and is examined in many different contexts (e.g. Geys, 2014; Offerman, Kennedy & Wirtz, 1994; Schyns & Meindl, 2005). ILT suggests that people have ‘an image’ of an effective leader, or a leader in general (Schyns & Mendl, 2005). People carry around these mental images about the characteristics and behaviors of leaders, and

consequently, these images inform their perceptions of who is a leader. In other words, it suggests that an individual may apply this image to a certain target individual in order to identify this target as a leader (or not).

To discuss the indirect effect of leader attractiveness and leader selection, leadership characteristics should be used that are appropriate to the image discussed on the basis of ILT.

(18)

Literature shows that ILT concerns certain variability in the subscription of leadership characteristics. For example, Hogan, Curpju & Hogan (1994) suggest that most people evaluate intelligence, sociability, honesty, understanding, aggressiveness, determination, verbal skills and industriousness are important characteristics of leadership, regardless of the context. Other studies mention the ‘Big Five Model’ of personality to be consistent with ILT, containing: extraversion, agreeableness, conscientiousness, emotional stability, and intellect (Goldberg, 1990). Yet another research suggests that factors such as ambition, dominance, empathy, independence, and likeability are tied to leadership perceptions (Shondrick, Dinh & Lord, 2010). As there is a diverse range of specific characteristics to be valued leadership characteristics, this study will focus on three general-valued leadership characteristics that are highly emphasized in the leadership literature and appropriate to the ILT: leader communion, trustworthiness, and competence. Communion has recently developed itself as a valued leadership characteristic, as its importance is emphasized in literature on transformational leadership (Bass & Avolio; 1994; Judge & Bono, 2000; Rosette & Tost, 2010). Findings suggest that transformational leadership and the accompanying communal behavior have been associated with leadership success (e.g. Judge & Piccolo, 2004). Trustworthiness is a well-known leadership treat described consequently in leadership literature. It involves trust in the ability, integrity and benevolence of an individual (Colquitt, Scott & Lepine, 2007; Little, 2014; Mayer, Davis & Schoorman, 1995). Generally, people are likely to desire a leader to not solely pursue their individual goals, but to act in the best interest of the group. Therefore, trustworthiness may be assumed to be an important characteristic in judging and selecting leaders. Competence is assumed to be high on the list of valued leadership characteristics, as an incompetent leader could have harmful effects for their group or organization they lead (Little, 2014). Moreover, findings suggest that competence ratings are an important predictor of real world elections (Todorov, Mandisodza, Goren & Hall, 2005).

(19)

In the following section the leadership characteristics will be described separately and more elaborate in the context of the attractiveness-selection relationship. Moreover, it will be discussed how these characteristics may contribute to a better insight into the selection process.

2.2.1 Leader communion as a mediator

Communal leaders are frequently described as female leaders that have communal

characteristics like being kind, helpful, likeable, concerned for others, and sympathetic (Braun et al., 2012; Rosette & Tost, 2010). Furthermore, it comprises the functioning in social

relations (Braun et al., 2012). People that are perceived as communal are evaluated as interpersonally oriented and concerned for the welfare of others (Collins, Burrus & Meyer, 2014). Other words that resemble this characteristic are, for example, nurturing, sympathetic, friendly (Collins et al., 2014). It can be compared to the characteristic ‘agreeableness’ of the ‘Big Five’ personality dimensions as research uses these two characteristics next to each other (e.g. Stopfer, Egloff, Nestler & Back, 2013). Agreeableness refers to the tendency to be helpful to others and to be sympathetic (Keller, 1999). They both refer to interpersonal involvement (‘getting along’) as for instance, liking another person (Stopfer et al., 2013; Wojciszke, Abele & Baryla, 2009). In trying to clarify the concept of perceived communion, it seems that this characteristic takes on the ‘social aspect’ of leadership.

Research suggests that communal characteristics are becoming increasingly valued in leadership (Rosette & Tost, 2010). Literature on transformational leadership, a leadership construct encompassing behaviors consistent with communal characteristics, reveals consistent benefits of a communal approach in leadership (Bass & Avolio, 1994; Judge & Bono, 2000; Rosette & Tost, 2010). Communion is evaluated as an important characteristic in assessing individuals and groups (Leach, Ellemers & Barreto, 2007; Wojciszke, Bazinska & Jaworksi, 1998). A more specific example shows the valence of perceived communion in

(20)

research on politics (Bellucci, Garzia & Lewis-Beck, 2013). It argues that the perceived likeability of a leader, another word related to communion, contributes to positive evaluations and may strongly orient a voter’s choice. This is in line with the ILT that suggest that people carry mental images of leadership characteristics, which are used in order to identify leaders and nonleaders (Geys, 2014; Offerman et al., 1994; Schyns & Meindl, 2005). On the basis ILT and the above-mentioned empirical findings, the assumption follows that communion is a valued leadership characteristic (Hogan et al., 1994). Therefore, it is possible to assume that perceived leader communion may contribute to the evaluation and selection of a leader.

In line with the assumptions of this paper, findings show that communion can be judged from someone’s face and relates to one’s social attraction toward a person, as for instance, liking (Braun et al., 2012). As previously discussed, many studies have shown that, on the basis of the halo effect, facial attractiveness positively influences judgments and attributions (Langlois et al., 2000; Rule & Ambady, 2010; Shinada & Yamagishi, 2014). In trying to understand the halo effect, one may think that feelings generally overcome

cognitions in the appraisal of others. The characteristic communion and its social aspect (versus cognitive aspect) may fit well in the explanation of the halo effect. Therefore, it is possible to assume that physical attractiveness may positively influence perceptions of

communion. Accordingly, research indicates that physical attractive individuals are perceived as more likeable and kind then less attractive one’s (Berscheid & Walster, 1974; Kissler & Bäuml, 2000). In an experimental study, respondents first had to rate another individual on the basis of a picture. Not only were the attractive individuals perceived to be more friendly, likeable and sociable, this also positively affected the phone conversation afterwards between these two persons (Mierke, Aretz, Wilmsen & Heinemann, 2011). This indicates that not only did the respondents perceive the attractive individuals as more positive; they also acted upon

(21)

this such that they communicated more friendly and open towards the attractive person. This is in line with the explaining mechanism of the self-fulfilling prophecy.

In sum, perceived communion has been defined and it has been discussed in the context valued leadership characteristics and ILT. Furthermore, the positive effect of physical attractiveness on communion and the underlying halo effect have been discussed. This

follows to the expectation that, in line with the ‘what is beautiful is good’ stereotype, an attractive candidate leader will be more likely to be judged as more communal, such that he is evaluated in a more positive way (being kind, more likeable) than an unattractive candidate. In line with ILT, perceived communion is assumed to be a socially valued leader

characteristic and therefore people are expected to be more likely to select the candidate leader with the highest rated communion. In other words, this paper assumes that perceived leader communion may positively influence the relationship between leader attractiveness and leader selection. This will be tested regarding the following hypothesis:

H2a. Perceived leader communion mediates the relationship between perceived leader attractiveness and leader selection, such that perceived leader attractiveness is positively related to perceived communion, which in turn, is positively related to leader selection.

2.2.2 Leader trustworthiness as a mediator

When referring to trustworthiness, trust is discussed in the context of personal characteristics that inspire positive expectations on the part of other individuals (Colquitt et al., 2007). Leader trustworthiness contains follower’s trust in the ability, integrity and benevolence of the leader (Colquitt et al., 2007; Little, 2014; Mayer et al., 1995). Trustworthiness is a driver of trust, whereby trust is defined as “the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action

(22)

important to the trustor, irrespective of the ability to monitor or control that other party” (Mayer et al., 1995, p.712). It subsumes that a leader will act in the best interest of the organization and its followers.

According to Chen, Jing & Min Lee (2014), trustworthiness is perceived as the

primary basis for a general evaluation of others. More specifically, trustworthiness is strongly emphasized in the assessment of political figures and leaders in general (Chen et al., 2014; Little et al., 2012; Van Vugt, 2006). The value of trustworthiness is perceived as a cross-cultural concept as it was highlighted as: “Contributing to outstanding leadership in all cultures were several attributes reflecting integrity. Thus, such a leader is trustworthy, just, and honest” (Den Hartog, House, Hanges, Ruiz-Quintanilla & Dorfman, 1999, p.237). This is in line with the ILT that suggests that people are seen as a leader to the degree that their characteristics match the other’s mental image what leadership characteristics are (Geys, 2014; Offerman et al., 1994; Schyns & Meindl, 2005). In other words, the characteristic trustworthiness fits the mental image of leadership (Hogan et al., 1994), and therefore, may positively influence the evaluation and selection of a leader. On the basis of the definition of trustworthiness, the empirical findings and the ILT, it is reasonable to assume that perceived trustworthiness is an important characteristic for the overall evaluation of a leader and may contribute to the selection of a candidate leader.

Linking these results to the context of facial appearance, previous research shows that trustworthiness may be judged from a person’s face and consequently influence the likelihood of selection. For example, an experimental study compares the hypothetical votes for

trustworthy and untrustworthy appearing faces (Little et al., 2012). People were more likely to select the trustworthy face to lead their country. Another study by Nana & Jackson (2010) explored the process through which individuals made attributions from a leader’s face. They found that people judged other’s perceived trustworthiness from a facial photograph.

(23)

Moreover, respondents firstly needed to establish if the leader was trustworthy (or not) before they could judge if they would follow the leader. Overall, results are in line with this paper’s assumption that trustworthiness is an important leadership characteristic. This paper expects that a candidate’s perceived trustworthiness will increase the likelihood of being selected as a leader.

This paper also emphasis physical attractiveness to have an effect on perceived trustworthiness and thereby to have an indirect influence on leader selection. A study on physical attractiveness revealed that A leader’s physical attractiveness could positively influence the ratings of leader trustworthiness (Surawski & Ossoff, 2006). Another experiment explored the ‘beauty premium’ on the basis of a trust game (Wilson & Eckel, 2006). Results showed that attractiveness acted as a signal of trustworthiness; more attractive trustees got higher rates on trustworthiness. Moreover, the results showed that attractiveness is important for the evaluation of candidates, and emphasize the indirect effect through perceived trustworthiness. This supports the notion of the halo effect that people ascribe positive attributions and characteristics to individuals who are more physical attractive (Langlois et al., 2000).

In sum, perceived trustworthiness has been defined and it has been discussed in the context of valued leadership characteristics and ILT. Furthermore, the effect of physical attractiveness on trustworthiness and the underlying halo effect have been discussed. This follows to the expectation that, in line with the halo effect, individuals will rate a physical attractive candidate as more trustworthy. As this is assumed to be a valued leader

characteristic, people will be expected to be more likely to select the candidate with the highest rated trust. In other words, this paper assumes that perceived leader trustworthiness may positively influence the relationship between leader attractiveness and leader selection. This will be tested regarding the following hypothesis:

(24)

H2b. Perceived leader trustworthiness mediates the relationship between perceived leader attractiveness and leader selection, such that perceived leader attractiveness is positively related to perceived trustworthiness, which in turn, is positively related to leader selection.

2.2.3 Leader competence as a mediator

The characteristic competence has been defined by Boyatzis (1982, p.21) as “an underlying characteristic of an individual that is causally related to effective or superior performance in a job”. Someone is perceived as competent when he or she is thought to be efficient in

performing his or her tasks (Justis, 1975). As incompetency of a leader could have harmful effects for the group or the organization, competence is assumed to be a high valued leadership treat.

The importance of perceived competence has been frequently examined in predicting important social outcomes such as real world political elections (Rule et al., 2010; Todorov, 2005). Even more, research shows perceived competence is evaluated as one of the core dimensions in judgments of others (Chen et al., 2014). For example, Chen et al. (2014) examined the role of several personality characteristics in hypothetical elections through showing participant facial images; perceived competence appeared to be the strongest predictor. Its great impact was explained by the notion that “competence is a task-related quality and thus serves as the normative basis of evaluation for public officeholders” (p. 32). This is in line with ILT, that states that people have mental images of leadership when defining an individual as a leader or not (Schyns & Meindl, 2005). In other words, the characteristic competence fits the mental image of leadership (Engle & Lord, 1997), and therefore, it is reasonable to assume that perceived leader competence may contribute to the evaluation and selection of a leader.

(25)

The above-mentioned study of Chen et al. (2014) solely used facial images during the hypothetical elections. This is in line with the assumption that people may judge the perceived competence from someone’s face and that this may influence the selection of a leader.

Another study from Todorov et al. (2005) examined this assumption. The study shows that inferences of competence, which are based on the facial appearance of political candidates and without prior knowledge of the candidates, predicted the election outcomes for the U.S. Congress.

Linking this to physical attractiveness, studies have shown that an individual’s physical attractiveness may affect judgments about the quality of a person’s work (related to perceived competence). For example, a study found that students judged an essay and its writer as more competent for an attractive writer than for an unattractive writer (Landy & Sigall, 1974). On the basis of the halo effect and the ‘what is beautiful is good’ stereotype, perceivers are expected to positively rate physical attractive individuals on perceived competence. Findings show support as physical attractiveness has showed to positively influence perceptions of characteristics linked to competence, such as intelligence and being skilled (Feingold, 1992; Zebrowitz, Hall, Murphy & Rhodes, 2002), highly qualified (Tews, Stafford & Zhu, 2009), and social competent (Eagly et al., 1991). More specific, people perceive physical attractive individuals as more intellectual competent (Eagly et al., 1991; Feingold, 1992). A meta-analytic review by Jackson, Hunter and Hodge (1995) showed that both physically attractive adults and children were perceived as more intellectual competent than unattractive peers. Moreover, an experiment showed that individuals not only perceive physical attractive persons to be more intelligent and competent in general, but also more competent in performing an unrelated and specific task such as piloting a plane (Kanazawa & Kovar, 2004).

(26)

In sum, first it has been discussed that perceived competence is assumed to be a high valued leadership characteristic that, in line with ILT, underlies the evaluation and selection process of a candidate leader. Secondly, the assumption has been discussed that competence can be judged from someone’s face. Moreover, it has been discussed that physical

attractiveness may have a positive effect on perceived competence. This study expects that, in line with the halo effect, individuals will rate the physical attractive candidate as more

competent and as this is assumed to be a valued leader characteristic, they will be expected to select the leader with the highest rated competence. Accordingly, this paper expects that perceived leader competence may positively influence the relationship between leader attractiveness and leader selection. This will be tested regarding the following hypothesis:

H2c. Perceived leader competence mediates the relationship between perceived leader attractiveness and leader selection, such that perceived leader attractiveness is positively related to perceived competence, which in turn, is positively related to leader selection.

2.3 Rater attractiveness and leader selection

Yet so far the focus has mainly been on the leader. The effects of a leader’s physical attractiveness, valued leader characteristics, evaluation and selection have been discussed. This focus misses out on another person that is participating in the process of leader selection: the rater. It is important to take note that the evaluations that are made from another person are in the eye of the beholder. For this study this means, the rater is the one who observes and determines the perceived communion, trustworthiness and competence of a leader candidate.

This paper explores the role of the rater in the process of making impressions and evaluations of another person. It will discuss the self-perceived physical attractiveness of the rater as a possible moderator of the leader attractiveness-selection relationship. Research

(27)

suggests that a person’s own attributes may influence how they perceive someone else’s attributes (Montaya, 2008). The rater’s attributions become a ‘lens’ through which other’s attributions are evaluated. This is well explained by Combs and Snygg (1959, p. 145) who state; “the self provides the frame of reference from which all else is observed. People are not really fat unless they are fatter than we”. This has been supported by findings indicating that the self may influence evaluations of another person’s weight (Ball, Crawford & Kenardy, 2004), social comparisons (Festinger, 1957), comparative appraisals (Jones & Gerard, 1967), and mental stimulations (Markman & McMullen, 2003).

This means when we make judgments of others, we compare them to ourselves. In other words, we compare how similar or not others are to our perceived selves. Similarity is a well-known topic in social psychology research. Similarity between individuals is related to interpersonal attraction and liking, and is examined on a number of dimensions, such as attitudes, personality, and demographic characteristics (Byrne, 1971; Suazo, Turnley & Mai-Dalton, 2005). This paper explores the role of the rater by focusing on a rater’s perceived similarity with another person, referring to the extent to which the rater perceives that they are similar to each other (Turban & Jones, 1988). The guiding theory this study uses to explain the effect of similarity is the ‘similarity-attraction theory’. This theory states that individuals have the tendency to engage in more positive social interactions with others similar to them (Byrne, 1971; Huang & Iun, 2006). For example, research has shown that subordinates who perceived their supervisor to be similar to themselves reported more confidence and trust in the supervisor (Turban & Jones, 1988).

To link this to the context of leadership, similarity is also a known concept in research on person-organization (PO) fit and person-supervisor (PS) fit. PO fit concerns the perceived match of individuals’ values and those of the organization (Van Vianen, Shen & Chuang, 2011). These perceptions are very important for job applicants’ organization choice,

(28)

recruiters’ hiring decisions and employees’ commitment to the organization (Van Vianen et al., 2011). PS fit refers to the perceived fit between the employee’s and a supervisor’s

characteristics (Van Vianen et al., 2011). In line with the similarity-attraction theory, research generally assumes that an individual who is perceived as similar to the beholder is perceived as more attractive, such that the evaluations and decisions regarding that person are positively biased (Byrne, 1971; Turban & Jones, 1988).

The general attraction towards similar others is often explained on the basis of the reinforcement model (Byrne, 1971). This theory states “individuals have a fundamental need for a logical and consistent view of the world” (Montoya & Horton, 2012, p.65). This means that when an individual perceives another person as similar on attitudes, opinion or

characteristics, this reinforces the logic and consistency, and thus is associated with positive feelings.

There has been no previous research on leadership or the process of leader selection that have included the physical attractiveness of the rater. The role of self-attractiveness of the rater has mainly been discussed in the context of dating experiments. Findings such

experiments show that someone’s own attributions may affect judgments of others and thereby influencing their choice of dating (Curran & Lipold, 1975; Montaya, 2008). This paper aims to explore the role of rater attractiveness in the leader selection process on the basis of the similarity-attraction theory. It questions if someone’s own attributes affect the attributions that are made of another person. In other words, does the self-perceived

attractiveness of the rater influence the relationship between the perceived attractiveness of the leader and leader selection?

The role of physical attractiveness and perceived similarity has not yet been discussed in context of leader selection, but it has been discussed in other contexts. Findings show support for the similarity-attraction theory and the assumption that rater attractiveness may

(29)

positively moderate the relation between leader attractiveness and leader selection. For example, an experimental study of Darby and Jefferson (1988) simulated a courtroom trial in which students had to evaluate attractive, moderately attractive, and unattractive defendants as to their guilt or innocence. Their results showed that attractive judges convicted attractive defendants less frequently and severely, and evaluated them more favorably across several dimensions such as likeability and trustworthiness. This suggests that for attractive judges the ‘what is beautiful is good’ effect is in harmony with positively evaluating others similar on attractiveness. However, for the unattractive judges the ‘what is beautiful is good’ effect was limited by a desire to be lenient towards defendants similar on attractiveness. They convicted the average attractive defendants the most, then the unattractive ones, and last the attractive ones. This means that the ‘attractiveness-similarity’ effect mainly holds for similarity on high attractiveness and less for similarity on low attractiveness, as this may be overruled by the attractiveness halo effect.

In sum, the role of the rater and the similarity on physical attractiveness has been discussed. Also the similarity-attraction theory has been discussed and explained next to the halo effect of attractiveness. As the self may influence how others are perceived, the self-perceived attractiveness of a rater is assumed to influence attributions of another person. Furthermore, as the similarity-attraction theory suggests people are drawn to similar others (Byrne, 1971), it is reasonable to assume that a rater who finds him or herself attractive will be drawn to a physical attractive individual. However, focusing on the physical attractiveness of persons, the halo-effect limits the similarity effect for less attractive individuals. Regarding this study, the expectation follows that when an rater is asked to evaluate a leader, viewing an attractive or an unattractive leader, the rater’s own physical attractiveness will influence the attractiveness halo effect and thereby the evaluation of the leader. An attractive rater will prefer an attractive candidate, as the perceived similarity with the leader will strengthen the

(30)

positive attractiveness halo effect. An unattractive rater will also prefer an attractive candidate, but to a lesser extent, as the perceived similarity with the unattractive candidate will lower the strength of the positive attractiveness halo effect. In other words, the

attractiveness-selection relationship will be stronger when a rater’s self-perceived attractiveness is high than low. This will be tested regarding the following hypothesis:

H3a. Self-perceived rater attractiveness will moderate the relationship between perceived leader attractiveness and leader selection, such that the relationship will be stronger under high rater attractiveness than under low rater attractiveness.

Furthermore, research on PO fit, PS fit, and in-group suggests that fit and similarity are important in organizations and for the leader-follower relationship (Van Vianen et al., 2011). Moreover, the importance of perceived similarity is also suggested by research on personnel decisions, such that perceived similarity influences who to fire, hire, promote and demote in the organization (Leonard, 1975). As previously has been discussed, according to the halo effect and the ‘what is beautiful is good’ stereotype, perceived physical attractiveness makes raters more likely to ascribe more positive characteristics to a person (Langlois et al., 2000; Rule & Ambady, 2010; Shinada & Yamagishi, 2014). Linking this to the similarity-attraction theory, this would mean that when raters perceive themselves a similar to a person on the basis of physical attractiveness, this would increase the likeliness even more of ascribing more positive characteristics to an attractive person. Accordingly, this effect is expected for the valued leadership characteristics that this paper previously discussed. In line with the similarity-attraction theory and this study’s expectation, research shows that judgments of similarity may affect perceived commitment (Van Vianen et al., 2011), trust and loyalty

(31)

(Huang & Iun, 2006; Turban & Jones, 1988; Van Vianen et al., 2011), and likeability of the leader (Byrne, 1971; Suazo, Turnley & Mai-Dalton, 2005).

A possible explanation for why people are likely to assign more positive

characteristics to others similar to them can be described by the concept ‘self-esteem’. Self-esteem is “an overall positive or negative attitude toward the self” (Rosenberg, Rosenberg & McCord, 1978, p. 280). Assigning the self with positive characteristics may contribute to the self-esteem (Tafarodi & Milne, 2002). By viewing oneself to be similar to a person with positive attributes, may increase one’s self-esteem. Linking this to the present study, we have discussed that an attractive person would be more likely to be rated more positively on leader characteristics and therefore, will be more likely to be chosen as a leader. When a rater perceives himself to be similar on attractiveness, this would enhance the self-esteem, as more positive attributions would be assigned to the self.

So far the halo effect and the ‘what is beautiful is good stereotype’, the similarity-attraction theory and the self-esteem concept have been discussed. This paper assumes that the self-perceived attractiveness of the rater will influence the attractiveness halo effect and thereby may affect the relationship between perceived physical attractiveness and perceived communion, trustworthiness, and competence. In other words, the expectation follows that when a rater perceives him or her self as attractive the similarity attraction effect will

strengthen the positive attractiveness halo effect. In other words, the attractive rater will make more positive judgments of perceived communion, trust and competence for an attractive leader than an unattractive rater will. Furthermore, as previously has been discussed, as the valued leader characteristics are important for the evaluation of a candidate leader, this will be assumed to have a positive effect on the selection of an attractive leader. This will be tested with the following hypothesis:

(32)

H3b. Self-perceived rater attractiveness will moderate the strength of the mediated relationship between perceived leader attractiveness and leader selection via perceived leader communion, such that the mediated relationship will be stronger under high rater attractiveness than under low rater attractiveness.

H3c. Self-perceived rater attractiveness will moderate the strength of the mediated relationship between perceived leader attractiveness and leader selection via perceived leader trustworthiness, such that the mediated relationship will be stronger under high rater attractiveness than under low rater attractiveness.

H3d. Self-perceived rater attractiveness will moderate the strength of the mediated relationship between perceived leader attractiveness and leader selection via perceived leader competence, such that the mediated relationship will be stronger under high rater attractiveness than under low rater attractiveness.

To clarify what has been discussed so far, the hypotheses are summarized in a conceptual model in Figure 1.

(33)

3. Method

The method section represents the empirical part of this study. First, the important

characteristics of the collected sample will be outlined. Furthermore, the procedure of the method is briefly discussed. Finally, the variables of the survey are discussed and the corresponding reliabilities are given. The complete survey (in Dutch) is presented in Appendix 1.

3.1 Sample

The sample was collected by means of an online survey. The survey administration started on April 8th 2015. The survey was closed on May 3rd 2015. The sample consisted of Dutch employees working on a part-time or fulltime basis. People who are not acquainted with a working life situation are not able to sufficiently answer questions concerning leadership in organizations and therefore were excluded from the sample. After personal and email contact with working individuals, the non-probability, convenience sample was gathered through sending the survey links.

Please take note of the following before interpreting the method and analysis. To gain sufficient respondents for this survey within a limited amount of time, the choice was made to let the respondents evaluate and rate two persons. These ratings will be treated as independent ratings in the analysis. The author is aware that this raises limitations to the study and these will be discussed in the discussion section accordingly. From the 426 employees that started the survey, 313 respondents completed the survey (Mage = 35.28, SDage = 14.30, age-range: 18-74 years). That is a response rate of 73.2 %. Out of the 312 respondents 118 were male (37.8%) and 190 were female (60.9%). The gender of 4 respondents was missing. The sample covered a broad range of educational backgrounds. From the sample 18.5% had completed a

(34)

secondary education program (VMBO/MAVO = 0.6 %, HAVO/VWO = 17.9 %), 29.5 % had completed an educational program at university of applied sciences (MBO = 4.5 %, HBO = 25 %), 50.7 % completed an education program at the research university (Bachelor’s degree = 18.6 % and Master’s degree = 32.1 %), and the remaining 0.6 % had no completed

education (and 0.6 % was missing data). From all the respondents 32.1 % reported a job tenure less then a year, 15.4 % between 1 and 2 years, 13.5 % between 2 and 4 years, 3.8 % between 4 and 5 years, and 34.0 % more than 5 years (and 1.3 % was missing data). A percentage of 36.8 % of the respondents reported to be in a leadership position.

3.2 Measurement of variables

3.2.1 Translation, back-translation procedure

All items used in the survey were derived from several English studies in which the items were validated. Since the survey was only to be completed by respondents that had Dutch as their first language, the original items were translated into Dutch. Four native English/Dutch speakers helped translating the questions. Then a fifth native English/Dutch speaker back translated the questions into English. A small number of discrepancies was detected and adjusted to the items of the final Dutch version of the survey. The scale reliability was analyzed and measured by chronbach’s alpha.

3.2.2 Procedure

In the survey two facial images were used, on the basis of which respondents were asked to answer several questions. The facial images consisted of one ‘attractive’ male and one

‘unattractive’ male. These images were found on the Internet with the search terms ‘attractive leader’ and ‘unattractive leader’. On the basis of a female and a male opinion, two images

(35)

were selected that were expected to fit an attractive and an unattractive person. Also the images were rated to be unfamiliar, to be similar in age and to have the same photographic angle. In a pre-test these presumptions were checked (please refer to the result section).

Then the scenario study was conducted. Respondents were collected through sending online links and the survey was completed online. Prior to the questions the respondents were informed about the topic and goal of the study (to gain more knowledge about the leader selection process), a time indication (5 minutes), and the anonymity of the survey. They were asked to read a short text that inquired them to envision a scenario of a work situation in which they were in the middle of reorganization and their team got restructured. They had to evaluate a candidate for a leader position on several aspects and ultimately had to indicate how likely they would select the candidate for a leader position. To relief any felt stress by the respondents, the story imported that there are no right or wrong answers.

When starting the survey, the respondents were showed a facial image of a male candidate (attractive or unattractive). They were then asked to answer several statements about the candidate. These statements measured the perceived communion, trustworthiness and competence of the candidate leader. After filling in these statements they were asked how likely they would select the candidate as their leader. Following, the perceived attractiveness of the candidate was measured. This was followed by several demographic questions. The survey concluded with asking the respondents to fill in several statements that measured their self-perceived physical attractiveness. This was purposely asked at the end of the survey, to make sure this could not bias answers to the other questions and thereby potentially

influencing outcomes. Afterwards, the respondents were thanked for their participation and informed with contact information for further questions. The full survey is presented in Appendix 1 (in Dutch).

(36)

3.2.3 Variables

Perceived leader attractiveness

To measure perceived leader attractiveness the scale of Montaya (2008) was used. The measure consists of one item “How physically attractive do you consider the person in the picture, relative to other people of his age?”. A 5-point Likert-scale ranging from 1 (strongly disagree) to 5 (strongly agree) was used. This variable serves a control function to ensure that one person is perceived as relative attractive and one person as relative unattractive.

Self-perceived rater attractiveness

To measure self-perceived rater attractiveness the scale of Jonason (2009) (Cronbach’s α = .89) was used. The measure consists of five items and two of the five items were reverse coded, meaning that a relatively low score indicates relatively high levels of self-perceived rater attractiveness. An example item is “Relative to other people of my age I find myself not good looking”. The scale was adjusted to a 7-point Likert-scale ranging from 1 (strongly disagree) to 7 (strongly agree) instead of a 5-point Likert-scale, as the pre-test showed limitations in the range of ratings.

Leader selection

To measure leader selection the scale of Little et al. (2012) consisting of one item was used “How likely would you select the person in the picture as team leader?”. A 5-point Likert-scale was used ranging from 1 (strongly disagree) to 5 (strongly agree).

Perceived leader communion

To measure perceived leader communion the scale of Heilman and Okimoto (2007) (Cronbach’s α = .85) was used and adjusted to fit the survey. The measure consists of four

(37)

items. One of the items was “I believe this person to be understanding”. Respondents were asked to answer according a 5-point Likert-scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Perceived leader trustworthiness

To measure perceived leader trustworthiness the scale of Ohanian (1990) (Cronbach’s α = .92) was used consisting of five items. One of the items was “I believe this individual to be reliable”. A 5-point Likert-scale was used ranging from 1 (strongly disagree) to 5 (strongly agree).

Perceived leader competence

To measure perceived leader competence the scale of Heilman, Battle, Keller and Lee (1998) (Cronbach’s α = .69) was used consisting of four items. One of the items was “I believe this individual to be organized”. A 5-point Likert-scale was used ranging from 1 (strongly disagree) to 5 (strongly agree). For reliability also perceived effectiveness was measured by using the scale of Heilman, Block and Lucas (1992) (Cronbach’s α = .76) consisting of two items. One of the items was “I believe this individual to be able to competently perform this job”. The same range of 5-point Likert-scale was used. To measure perceived competence these two scales were combined into a six-item scale (Cronbach’s α = .81).

Control variables

Six other variables were measured to retrieve descriptive statistics from the sample: gender, age, education, job tenure, job sector and job position. These were included in the last part of the survey.

(38)

4. Results

In the results section the outcomes of the data analysis will be given. The statistical analyses were performed using the Statistical software Package for Social Sciences (SPSS). After recoding the reverse coded items, scale reliabilities, descriptive statistics, skewness, kurtosis and normality tests were computed and checked. First, the results from the pre-test and manipulation check will be discussed. Following, the correlation matrix in Table 1 will be showed and the correlations between the variables will be given. Subsequently, the results of the linear regression will be outlined. The direct relationship between perceived leader attractiveness and leader selection are given in advance to the mediation, moderation and moderated mediation effects respectively.

4.1 Pre-test

Prior to the distribution of the final survey a pre-test was conducted to test if the scales were reliable, and to conduct a manipulation check to control for age, familiarity and perceived physical attractiveness of the two candidates before applying them to the final survey. The sample size consisted of 21 persons of whom 10 men (47.6 %) and 11 women (52.4 %). (Mage

= 26.71, SDage = 9.63, age-range: 20-55 years).

Respondents estimated the age of candidate 1 between 35 and 50 with an average of 43 (SD = 4.36) and candidate 2 between 30 and 47 with an average of 38.7 (SD = 4.50). Thereby we conclude that the age of both candidates is in the same range. None of the respondents noted to be familiar neither with candidate 1 nor with candidate 2.

A paired-samples t-test was conducted to examine whether the two candidates were evaluated different on perceived physical attractiveness, as one ‘attractive’ candidate and one ‘unattractive’ candidate was needed for the final survey. Results indicated a significant

(39)

difference in ratings of physical attractiveness for candidate 1 and 2. Candidate 2 was evaluated as significantly more attractive than candidate 1: t(20) = 12.10, p <.001.

From the manipulation check can be concluded that candidate 1 and 2 are estimated around the same average age, are not familiar to the respondents, and have a different rating in physical attractiveness. In the final survey candidate 2 will be used as ‘attractive’ and candidate 1 as ‘unattractive’.

Findings from the reliability analysis in the pre-test, showed that the scale for self-perceived rater attractiveness, measured by one item “How physically attractive do you consider yourself, relative to other people of your age”, was not sufficient. It showed limited diversity in measurements. Results showed that out of the 21 participants, 6 perceived themselves to be not unattractive/not attractive, 12 perceived themselves as attractive, 3 as very attractive and no one reported him or herself as unattractive or very unattractive. This limitation was taken into account while constructing the final survey, such that a different scale was used.

In the final survey a control check for perceived physical attractiveness was inserted. An independent sample t-test was conducted to compare the perceived physical attractiveness for candidate 1 and candidate 2 showed in the survey. Results showed a significant difference in rated attractiveness for candidate 2 (M=5.38, SD=.93) and candidate 1 (M=.2.34, SD=.93), t(310) = -28.82, p <.001, two-tailed. As such candidate 2 is perceived as significant more attractive then candidate 1. To further analyze our data a new variable ‘image’ was made to make clear which person is rated on leadership attributes: the attractive candidate (= 2) or the unattractive candidate (= 1).

(40)

4.2 Correlation analysis

In Table 1 the overview is presented of the descriptive statistics, correlations and scale reliabilities. This table shows several correlating variables. Perceived competence was positively correlated to perceived trust (r = .18, p < .01), leader selection (r = .20, p < .01), rater attractiveness (r = .12, p < .05) and image (r = .18, p <.01). The variable perceived trust was positively correlated to perceived communion (r = .81, p < .01), leader selection (r = .38, p < .01) and negatively correlated to image (r = -.53, p <.01). Perceived communion showed a positive correlation with perceived leader selection (r = .32, p < .001) and a negative

correlation image (r = -.51, p <.01). Leader selection was negatively correlated with image (r = -.14, p < .05). The correlations show surprising effects as perceived trustworthiness,

communion and leader selection were negatively correlated with image (perceived leader attractiveness). As a positive relation was expected, these negative correlations are rather surprising.

Also the descriptive variables showed correlations. Leader position was negatively correlated with job tenure (r = -.52, p < .01), and age (r = -.46, p < .01), and positively correlated with gender (r = .26, p < .01). Job tenure was positively correlated with age (r = .64, p < .01) and negatively correlated with perceived competence (r = -.18, p < .01). Education showed a positive correlation with rater attractiveness (r = .15, p < .01). Age was negatively correlated with perceived competence (r = -.14, p < .01) and rater attractiveness (r = -.23, p < .01).

Rater characteristics : the moderating effect of rater attractiveness