1
When politicians talk counter-stereotypically:
The impact of gender and gendered linguistic
communication styles on the public’s candidate
evaluations
Master’s Thesis for the Master’s Program Political Communication
By Lynn Leeneman (11062460)
Supervised by Linda Bos
Date of completion: June 28, 2019
Word count: 8434
Abstract
This study investigates the relationship between language, gender and trait evaluations. In an experiment (N = 185), the impact of politicians talking in counter-stereotypical ways on the public’s candidate evaluations were studied. The analyses showed that the interaction of a politician’s gender and the linguistic communication style he/she uses, which can be feminine or masculine, has no significant effect on the evaluations of any of the five trait dimensions, being Competence, Empathy, Integrity, Leadership and Warmth. However, the gender of a politician on itself does seem to have an impact: female politicians are perceived as more warm and possessing stronger leadership traits than their male counterparts. It might be the case that sex stereotypes have eased and don’t have the negative impact on the evaluations of female politicians anymore that they are often assumed to have.
2
Introduction
The number of female political leaders is growing (Campus, 2013a). Even though the political arena is still for a large part dominated by men (Jones, 2017), the amount of women that are stepping up to run for higher political office and subsequently get elected is slowly increasing: as of November 2018, 24 per cent of the representatives in all national parliaments was female, while in 1995 only 11.3 per cent was female (UN Women, 2019). At the same time, personal characteristics of politicians are becoming more and more important because of the current personalization of politics, which has led to candidate evaluations affecting the public’s vote choice (Dolan, 2014; Gatterman & De Vreese, 2017). The individual candidate evaluations of the public thus influence whether political candidates successfully get or stay in higher office. Therefore, insights on the origins of these candidate evaluations are
important, especially for the increasing amount of female politicians entering the political arena.
Because the political sphere is still dominated by men and consequently permeated by a male communication model (Jones, 2017), female politicians occupy a minority status. In the existing literature it is often stated that this minority status results in pressures on female politicians to conform to a normative, masculine style of leadership and
communication (Jones, 2017; Karpowitz & Mendelberg, 2014). The current model of
leadership is closely associated with and based on traditionally masculine qualities, such as strength, decisiveness, determination and self-confidence (Campus, 2013b; Jones, 2017). As a consequence, it is often argued that female politicians are only able to get their points across and succeed in politics when they adopt a gendered, masculine communication style (Jones, 2017; Grebelsky-Lichtman & Bdolach, 2017). After all, the institutional norms of behaviour embody and thus reward masculine styles of communication (Jones, 2017). A classic example of this is the case of Margaret Thatcher, who was encouraged to speak in masculine ways and trained to lower the pitch of her voice in order to “communicate with more authority” (Jones, 2017; Grebelsky-Lichtman & Bdolach, 2017). Female politicians that
3 fail to conform to a masculine style of communication run the risk of being criticized by the public, for instance by being considered as too weak (Campus, 2013a; Jones, 2017)
However, in reality the situation is more complex. Women in leadership roles have to deal with dual expectations from the public (Grebelsky-Lichtman & Bdolach, 2017), also dubbed the double bind. Next to the leadership model requiring politicians to communicate in masculine ways, there are stereotypical expectations surrounding the politician’s gender. Therefore, women in the political arena are likely to face backlash when violating gender norms. For instance, it is not uncommon that female politicians are criticized for being too aggressive when they fail to display qualities that are traditionally regarded as feminine (Campus, 2013a; Jones, 2017). Thus, female politicians are challenged by complex and conflicting expectations: on the one hand, they have to exhibit masculine communication patterns because of their leadership positions, but on the other hand they also have to exhibit feminine communication patterns because they are women (Grebelsky-Lichtman & Bdolach, 2017). In order to be successful, female politicians have to develop an effective and
appropriate self-presentation, of which communication is an important element (Jones, 2017).
In comparison to research focussing on female politicians and their use of gendered communication styles, little research has focussed on the effects of the use of gendered communication styles by male politicians. Based on the masculine leadership model and the expectations surrounding a politician’s gender, it can be expected that a male politician is most successful when he uses a masculine communication style. However, Jones (2017) found that the current U.S. president Donald Trump uses a linguistic communication style that is far more feminine than the style of any other U.S. presidential candidate since 2004 (Sedivy, 2016). Research has shown that the public considers politicians that speak in feminine ways as warmer, more trustworthy and more likeable (Jones, 2017). Therefore, in some cases politicians using a feminine style could have an advantage over those using a
4 masculine style. According to Jones (2017), this benefit of using a feminine linguistic
communication style is greater for male politicians than for females (Sedivy, 2016).
The research question that is central to the thesis is as follows: “What is the impact of the gendered linguistic communication style used by a male/female politician on the public’s evaluations towards this politician?” (Feminine and masculine) communication styles are already analysed in several studies. However, most studies try to describe the origins, behaviours and motives of the communication styles (Wood, 1995), or try to investigate how the gendered communication styles are used by females (or males), more specifically female (or male) politicians (Dow & Tonn, 1993; Jones, 2016). Studies in the field of communication science also often focus on how the media portray female and male politicians, and how these portrayals influence the public’s evaluations (Campus, 2013c; Carlin & Winfrey, 2009; Dunaway et al., 2013; Hayes & Lawless, 2015). It is particularly interesting to focus on the politicians’ linguistic communication styles. First of all, because the usage of language can easily be adjusted by politicians in order to present themselves in strategic ways that impact the public’s candidate evaluations in favourable ways. Next to this, it is interesting to study linguistic communication styles in particular since expectations about behaviours surrounding gender are (re)constructed trough language (Jones, 2017). The linguistic communication style does not refer to the substance or content of a speech, but to the way in which a person communicates and conveys content (ibid.).
The influence of the use of a certain communication style by a politician on the public’s evaluations of this politicians and attitudes towards him/her is not as often investigated. Up until today, it is not clear whether individuals have (implicit) knowledge about the way in which men and women make use of language. Also, it still is unclear if gendered linguistic communication styles even have an effect on the evaluations of the public towards political candidates (ibid.). Aalberg & Jenssen (2007) did investigate whether the gender of politicians affects the way citizens evaluate the qualities of a political speech, but the different communication styles were not taken into account. One of the few studies
5 that dives into the impact of linguistic communication styles on the public’s evaluations
towards politicians is that of Jones (2017). However, Jones focuses mainly on the
evaluations of the personality traits competence and warmth. In this research, the impact on more diverse personality traits are taken into account.
In this research, the relationship between gender, language and the public’s
evaluations of politicians is investigated. To answer the research question, a 2 x 2 factorial experimental design testing the impact of gendered linguistic communication style on the public’s evaluations of five personality traits of the politician (competence, empathy, integrity, leadership, warmth), which is moderated by the gender of the political candidate, was
created.
Theoretical Framework
2.1 Gender and gendered communication
Gender is a social construct that both describes and prescribes behaviours of men and women (Jones, 2017). It is defined by Duerst-Lathi & Kelly (1995) as “a coherent set of beliefs about what constitutes masculine or feminine.” According to Judith Butler (1999), gender can be understood as a performance: gender is something we do, instead of something we are (Deutsch, 2007). It is a set of actions that is closely related to societal expectations about typically male and female features and behaviour. This set of actions is learned through narratives, language, socialization, and other performative acts (Butler, 1999). By performing the acts, gender is continually reconstructed in society (Deutsch, 2007). Gender is particularly performed in social interactions (West & Zimmerman, 2009).
One major way in which gender is performed is through the usage of language (Jones, 2017). Numerous studies in which speech samples of both men and women are analysed have found reliable and consistent gendered variations in linguistic communication style (e.g. Mulac et al., 2001; Newman et al., 2008; Schwartz et al., 2013). The gendered
6 language differences are a result of the different ways in which men and women are
socialized. This leads to them employing language for different reasons (Newman, 2008; Pennebaker, 2011). According to current gender stereotypes, females should display communal characteristics, e.g., warmth, friendliness and interdependence. Their use of language is a way to establish relationships and consequently their words reflect a greater concern with relationships and understanding people (Jones, 2017; Menegatti & Rubini, 2017; Pennebaker, 2011). For women, language is used for social purposes and thus serves as an end in itself (Colley et al., 2004). Men on the other hand should display agentic
characteristics, e.g., competence, assertiveness and independence. For them, language is a way to accomplish concrete goals, preserve independence, exert control, entertain and enhance status (Jones, 2017; Menegatti & Rubini, 2017; Wood, 2009). In comparison to women, men use language more with the instrumental goal of conveying information (Colley et al., 2004). These gender stereotypes are reflected and subtly reproduced in our day to day communication (Jones, 2017; Menegatti & Rubini, 2017).
2.2 Analysing language: the use of function words
Studies about gender differences in language have analysed language in various ways and at multiple levels, ranging from analysing the overall structure of the narrative (e.g. Herring, 1993), to analysing specific phrases (e.g. Thomson & Murachver, 2001) and words (e.g. Mulac et al., 2001). But what is the right way to capture the variations in which men and women communicate?
When analysing language, a distinction can be made between words that reflect content (‘What you say’) and words that reflect style (‘How you say it’) (Pennebaker, 2011). Content words (nouns, regular and action verbs, adverbs and adjectives) are words that contain a relative specific, shared meaning in labelling an action or object, and that are necessary to convey a certain idea to someone. Function words include pronouns (both personal and impersonal), conjunctions, prepositions, determiners (such as articles,
7 demonstratives and quantifiers), auxiliary verbs, modal verbs, negations, common adverbs (also called qualifiers or intensifiers) and question words. These kind of words are used to ‘glue’ the content words together in order to create sentences. Function words fulfil a grammatical function: they connect, shape and organize the surrounding content words (Pennebaker, 2011). Verbal communication styles rely heavily on the use of function words (Jones, 2017). They are used at very high rates and can be considered as the basic
elements of a linguistic style (ibid.).
Much content analysis research tends to remove or ignore the function words, since they appear as not containing substantive meaning in itself in comparison to content words (Jones, 2017; Newman et al., 2008; Pennebaker, 2011). However, recent research has uncovered that function words in fact contain value, as we convey the thoughts we have into forms of communication by using function words (Jones, 2017; Pennebaker, 2011). The words we use mirror our thoughts about events and objects in our lives in subtle ways, and thus variations in the usage of function words reflect variations in the ways in which
individuals think and feel, and how they view and organize the world (Jones, 2017; Newman, 2008; Pennebaker, 2011). By analysing function words, we learn about the people that use them (Pennebaker, 2011). For example, when one uses the pronoun ‘we’, a different view on the relationship between the speaker and listener is reflected than when one uses the pronouns ‘you and I’.
2.3 Feminine and masculine linguistic communication styles
Since the research of Gleser et al. (1959) about differences in language of men and women was published, a lot of empirical research on this topic has been executed. Proof has been found for gender-linked differences in diverse communication forms: from monologues (e.g. Mulac & Lundell, 1982) to interactions (e.g. Poole, 1979), and from written texts (e.g. Schwartz et al., 2013) to spoken ones (e.g. Singh, 2001). Since this study focusses on linguistic communication style instead of content, and since function words carry important
8 meaning, this study puts emphasis on function words to discern between male and female linguistic communication styles. In order to create a clear picture of which function words can be understood as feminine and masculine, the findings of previous studies have been
aggregated into a linguistic communication style model. An overview of this model is shown in Table 1. The model is in particular based on insights from Argamon et al. (2007), Jones (2017), Mulac et al. (2001), Newman et al. (2008), Pennebaker (2011) and Schwartz et al. (2013). The linguistic communication styles and the related function words are further explained below.
2.3.1 Feminine linguistic communication style
The feminine linguistic communication style is more personalized and socially-oriented than the masculine linguistic communication style. This is the case because women are on average more self-aware and self-focussed than men and tend to think and talk more about
9 other people (Jones, 2017; Newman et al., 2008; Pennebaker, 2011; Schwartz et al., 2013), which results into using personal pronouns at high rates. Indeed, many scholars agree that women on average use more personal pronouns than men, especially first person singular words (also called ‘I-words’) (Argamon et al., 2003 & 2007; Jones, 2017; Mulac & Lundell, 1986; Newman et al., 2008; Pennebaker, 2011; Poole, 1979; Schwartz et al., 2013). It should be noted that most studies did not find significance differences in the usage of first person plural words, e.g., we, us, our (Newman et al., 2008; Pennebaker, 2011; Schwartz et al., 2013).
Surprisingly, in his review of studies about language differences, Mulac et al. (2001) state the opposite: men tend to use more I-references. However, this statement is based on two studies (Mulac et al., 1990; Mulac & Lundell, 1994) that were executed over 25 years ago. This could be problematic as considerations about appropriate gender behaviour, and thus language usage, change over time (Deutsch, 2007). Even within short time spans, change are possible (ibid.). Next to this, the size of the studies where Mulac’s et al. (2001) statement is based on is rather small: they consisted of respectively 32 (Mulac et al., 1990) and 40 (Mulac & Lundell, 1994) text samples that were analysed. As more recently
conducted studies are of a much larger scale and make use of greater samples (e.g. Newman et al. (2008) analysed 14.324 text samples, and Schwartz et al. (2013) analysed 15.4 million Facebook text samples written by almost 75.000 volunteers), their results appear to be more reliable.
Next to this, women often use words that emphasize cognitive mechanisms together with I-words (Jones, 2017; Mulac & Lundell, 1994; Newman et al., 2008; Pennebaker, 2011). Pennebaker (2011) calls these sentences that typically start with phrases like ‘I think that’, or ‘I don’t know, but’ hedge phrases. Hedge phrases soften a statement and add a degree of uncertainty (Turner, 2017). They point at the sensitivity of feminine language to a range of perspectives and contexts (Jones, 2017).
10 Auxiliary verbs are also used significantly more by females (Argamon et al., 2003 & 2007; Jones, 2017; Pennebaker, 2011; Schwartz et al., 2013). The feminine linguistic
communication style is, apart from being more personalized than the masculine style, also more dynamic, meaning that the female language tends to focus on how people, topics and events change (Pennebaker, 2011; Jones, 2017). These discussions of change require making use of more verbs, including auxiliary verbs, and thus females’ language is more likely to include many (auxiliary) verbs than men’s (Argamon et al., 2003 & 2007; Jones, 2011; Newman et al., 2008; Pennebaker, 2011; Schwartz et al., 2013). High rates of auxiliary verbs usually indicate a more passive speaking style (Pennebaker, 2011).
Research also finds that females tend to use significantly more intensifiers than men do (Hancock & Rubin, 2014; McMillan et al., 1977; Mulac, 2001; Mulac & Lundell, 1986; Newman et al., 2008). Because of intensifiers, listeners are directed to the emotional instead of the cognitive meaning of the message stated (Hancock & Rubin, 2014; McMillan et al., 1977). The use of intensifiers points at the emotional involvement of women with the stated ideas. Using intensifiers is a way for women to communicate and generate their involvement (McMillan et al., 1977).
Studies also find that females tend to use significantly more conjunctions (Argamon et al., 2007; Gleser et al., 1959; McMillan et al, 1977; Schwartz et al., 2013). This use of
conjunctions mirrors the tentativeness of the feminine communication style, just like their use of intensifiers and auxiliary verbs does (Biber et al., 1998).
Several studies report that women make more use of negations (Gleser et al., 1959; Lester & Leenaars, 2016; Mulac et al., 2001; Mulac & Lundell, 1986; Pennebaker, 2011; Newman et al., 2008). However, the findings of the two large scale studies of Schwartz et al. (2013) and Newman et al. (2008) did differ: Schwartz et al. (2013) found no significant difference in the use of negations between men and women, while Newman et al. (2008) found a difference that was relatively strong. The differences might be a result of the kind of texts that were analysed: Schwartz et al. (2013) exclusively analysed written Facebook
11 updates, and Newman et al. (2008) analysed different kinds of texts, both written ones and ones that were transcribed from speech. Since this study focusses on transcribed political speeches, it was chosen to follow the findings of Newman et al. (2008), which are in line with most studies that report about gendered use of negations.
2.3.2 Masculine linguistic communication style
Since the masculine linguistic communication style tends to be centred around external objects and things categorized in specific ways, men usually tend to use more nouns, as object references, than women. This frequent use of nouns results in men using more articles (Argamon et al., 2003 & 2007; Jones, 2017; Newman et al., 2008; Pennebaker, 2011; Schwartz et al., 2013), demonstratives (Mulac et al., 2001; Schwartz et al., 2013; Argamon et al., 2003) and prepositions (Argamon et al., 2007; Gleser et al, 1959; Jones, 2017; Newman et al., 2008; Pennebaker, 2011; Poole, 1979; Schwartz et al., 2013) than women. The prepositions serve as temporal and spatial hierarchies (e.g. ‘at 3:30’, ‘at the corner’) (Jones, 2017).
However, men more frequently using articles and demonstratives in comparison to women is not merely a consequence of their greater use of nouns. It is also a result from men being more likely than women to specify or indicate particular things in the world (Argamon et al., 2003). They tend to express themselves and view the world by ‘counting, naming and organizing objects’ (Pennebaker, 2011). Research has shown that because of this, men also tend to use more numbers and quantifiers (Argamon et al., 2003; Newman et al., 2008; Pennebaker, 2011; Schwartz et al., 2013; Mulac et al., 2001; Newman et al., 2008).
2.4 Effects of gender and gendered linguistic communication on the public’s candidate evaluations
We all use stereotypes in our day-to-day life (Aalberg & Jenssen, 2007). These stereotypes function as information shortcuts: because we simply cannot have access to or knowledge
12 about the full range of information, we often rely on conventional conceptions. Also for
making political decisions, voters rely on the informational shortcuts (Hardy, 2017; Lodge & Stroh, 1993; Lupia & McCubbins, 1998; Popkin, 1994). Gender stereotypes can affect vote choice both directly and indirectly by influencing the public’s candidate evaluations (Dolan, 2014). Since voters have established conscious and unconscious views on gender roles, the gender of a politician can be used as an information shortcut in order to directly estimate his or her traits, beliefs, preferences and issue positions and to evaluate these (Aalberg & Jenssen, 2007). For example, several studies have found that politicians of a different gender are perceived as having different competence areas: male politicians are considered as being more knowledgeable on traditional ‘male issues’, like economy, defense and foreign policy, and female politicians are usually affiliated with traditional ‘female issues’, like
education and healthcare (Dolan, 2014; Shapiro & Mahajan, 1986)). Voters prefer one gender over the other based on the stereotypes related to competence areas (Aalberg & Jenssen, 2007; Huddy & Terkildsen, 1993; Matland, 1994).
A more indirect way in which gender stereotypes could influence the public’s evaluations of politicians is through perceived personality traits. The perceived personality traits of politicians function as an appealing information shortcut to evaluate politicians on, as they are relatively easy to assess in comparison to e.g., policy positions: you do not have to put much time or energy into uncovering issue stances or following public affairs in order to be able to say something about a politician’s traits (Funk, 1996; Kinder, 1986). The public thus uses perceived personality traits of politicians as an inexpensive way of gaining
information about political candidates and to simplify their vote decisions (Funk 1999; Kinder et al., 1980; Miller et al., 1986; Popkin, 1994; Rahn et al., 1990).
But what traits do citizens use to evaluate politicians on? In a study of Kinder et al. (1980), insights on the traits that are used to describe the profile of an ideal political leader were gathered by asking participants to describe an anti-ideal and an ideal president. A list of traits arose from this and the participants had to selects the six that they found most
13 important. Out of these results, Kinder (1986) developed four interrelated dimensions of traits that are used by voters to evaluate politicians and base their vote choice on, being
Compentence, Empathy, Integrity and Leadership (Funk, 1999; Pierce, 1993; Schneider & Bos, 2016). These four dimensions are also used in the American National Election Studies. Jones (2017) adds that Warmth also could be an influential candidate trait.
Literature suggests that gender influences how these personality traits are perceived: men and women are understood as possessing different personalities (Best & Williams, 1990; Broverman et al., 1972; McKee & Sheriffs, 1957). Especially when information about a politician is sparse, gender stereotypes play a role in initial candidate evaluations (Alexander & Andersen, 1993). Traditionally, gender stereotypes result into a greater likelihood of men being perceived as possessing competence-related traits than women. Women on the other had are more likely to be seen as lacking competence, but instead possessing traits that are associated with warmth (Cuddy et al., 2009; Fiske et al., 2002; Glick et al., 2004; Hayes, 2011).
Indeed, several empirical and experimental studies have found evidence that females score higher than males on traits associated with the integrity dimension (Alexander & Andersen, 1993; Leeper, 1991; Schneider & Bos, 2016), the empathy dimension (Alexander & Andersen, 1993; Huddy & Terkildsen, 1993; King & Matland, 2003; Schneider & Bos, 2016) and the warmth dimension (Huddy & Terkildsen, 1993; Jones, 2017). Men on the other hand are presumed to be more competent and thus score higher on traits associated with the competence dimension than women (Aalberg & Jenssen, 2007; Cuddy et al. 2009; Fiske et al. 2002; Glick et al. 2004; Moss-Racusin, 2012). However, Alexander & Andersen (1993) found that the trait hardworking, which is part of the competence dimension, is more often associated with women than men. But as this is only one part of the several traits the competence dimension exists of, it could still be that men score higher on the competence dimension when taken as a whole.
14 Surprisingly, empirical and experimental research has also shown that a political candidate’s gender does not seem to matter for evaluations of traits associated with leadership (Aalberg & Jenssen, 2007; King & Matland, 2003). Again, Alexander & Andersen (1993) did find that females are often perceived as ‘getting things done’, which is part of the leadership
dimension, but as this is only a little part of it, it could still be that evaluations about leadership do not differ for male and female politicians when the leadership dimension is taken as a whole.
Apart from the actual gender of the political candidate, evaluations are also shaped by the language that is used (Jones, 2017). Jones (2017) found that, regardless of a
candidate’s gender, evaluations of warmth were significantly affected by the candidate’s style of communication: the participants in the study consistently rated feminine statements higher on the trait warmth than the masculine ones. Jones’ research also uncovered there was no significant impact of gendered communications style on the trait competence (ibid). Apart from the work of Jones, not much research has focused on the impact of gendered communication styles on the perceived personality traits of politicians.
2.5 Interaction of gender and gendered communication on the public’s candidate evaluations
However, in this study the focus does not lay on the impact of gender and gendered
communication style separately. Rather, it is focused on the interplay of the two, which might result into different candidate evaluations than when analyzed separately.
Research of Cuddy, Fiske & Glick (2008) suggests that men are rated higher on the trait warmth when they act counter-stereotypically, while maintaining the same rating on the trait competence. The findings of Jones (2017) are in line with this: relative to male
candidates with masculine statements, she found that male candidates with feminine statements score higher on the warmth dimension. In addition, competence ratings did not change when male politicians were talking in feminine ways. This means that they do not face backlash when violating gender norms. Two first hypotheses can be derived from this:
15 H1: A male politician making use of a feminine linguistic communication style will be evaluated higher on the trait dimension warmth (H1a), but the same on competence (H1b), compared to a male politician making use of a masculine linguistic
communication style.
As explained before, men score not as high as women on traits associated with the integrity and empathy dimension (Alexander & Andersen, 1993; Huddy & Terkildsen, 1993; King & Matland, 2003; Leeper, 1991; Schneider & Bos, 2016), and score the same as women on traits associated with leadership (Aalberg & Jenssen, 2007; King & Matland, 2003). Following this, it can be expected that a masculine communication style has effects on the public’s candidate evaluations that are comparable. However, since no literature was found on the impact of communication style and a politician’s gender taken together on the three trait dimensions, a targeted research question was formulated for those traits:
RQ1: Is a male politician making use of a feminine linguistic communications style evaluated higher on the trait dimensions empathy (RQ1a) and integrity (RQ1b), but the same on leadership (RQ1c), compared to a male politician making use of a masculine linguistic communication style?
Jones (2017) also finds that female candidates with feminine statements received the highest overall ratings on the warmth dimensions. Next to this, women speaking with a masculine linguistic communication style are expected to be evaluated higher on specific, more traditionally masculine traits, like competence, while scoring lower on personality traits that are traditionally feminine, like empathy and warmth, compared to women using a
feminine linguistic communication style. This is in line with Jones (2017) findings, suggesting that female candidates using masculine statements receive higher ratings on the
competence dimension, but lower ratings on the warmth dimension, in comparison to female candidates with feminine statements. This has to do with the double bind, that results in a backlash effect when women act in ways that violate the traditional gender stereotypes
16 (Heilman et al., 2004; Rudman & Fairchild, 2004; Rudman & Glick, 1999). The third
hypothesis is as follows:
H2: A female politician making use of a feminine linguistic communication style will be evaluated higher on the trait dimensions empathy (H2a) and warmth (H2b), but lower on competence (H2c), than a female politician making use of a masculine linguistic communication style.
As stated, women are usually perceived as more integer than man. Because of this, it can be expected that the use of a feminine linguistic communication style also leads to higher perceived integrity than when a masculine communication style is used. Since no insights are found on the impact of the interaction of gender and linguistic communication style on the perceived integrity and leadership traits of a politician, a second targeted research question was formulated:
RQ2: Is a female politician making use of a feminine linguistic communications style evaluated higher on the trait dimension integrity (RQ2a), but the same on leadership (RQ2b), compared to a female politician making use of a masculine linguistic
communication style?
Methods
In order to investigate the impact of linguistic communication style on the public’s candidate evaluations and the role gender plays in this, an online survey experiment was conducted among a sample of Dutch speaking citizens. In a 2 x 2 factorial design, the participants were randomly assigned to one of four stimuli, being (1) a speech sample with a feminine linguistic communication style performed by a female politician, (2) a speech sample with a feminine linguistic communication style performed by a male politician, (3) a speech sample with a masculine linguistic communication style performed by a female politician and (4) a
17 speech sample with a masculine linguistic communication style performed by a male
politician.
3.1 Participants
For this study, participants were recruited through Facebook and personal contact. The total research sample consisted of 185 participants, of which 174 were entered into the analysis since they passed the manipulation check (see below). Of these 174 participants, 32.2 per cent were male and 66.1 per cent were female. The participants were aged 25 on average, with 18 being the minimum and 67 being the maximum. The great majority (79.5 per cent) were aged in the range of 18 to 25. When it comes to political left-right orientation, the sample was quite evenly distributed and was skewed only a little bit to the left (M = 4.42, with 0 meaning the most left and 10 meaning the most right).
3.2 Procedure
The participants were asked to carefully read one of two manipulated speech samples that were presented as having been performed by a fictional Dutch male or female politician. In order to make sure they actually did read the speech sample, participants were obliged to stay on the speech-page for at least 25 seconds. After the exposure, the participants moved
18 on to the next page that repeated the politician’s name (Linda van Dijk, Paul van Dijk) and they completed a questionnaire about how they evaluated the politician that performed the speech on several personality traits. The participant also answered questions that were used for the manipulation check. At last, the participants filled in several background questions, like their age, sex, education level and political left-right orientation.
3.3 Stimulus Material
In this study, the stimulus material consisted of two manipulated speech samples that were presented as performed by a fictional Dutch female or male politician. Both fictional
politicians were given a common Dutch first name for women/men, namely Linda and Paul, and the last name that was chosen (Van Dijk) is a frequent surname in the Dutch parliament. The fact that the names were fictional was not revealed to the participants until after the experiment was completed.
The manipulated speech samples were based on an actual speech that was
performed by the politician Chris Stoffer, who is a member of the Reformed Political Party (in Dutch: SGP) during a general consultation in the Dutch parliament. This speech was found in a publicly available committee report of a general consultation about sustainable aviation that took place on April 4th 2019. The topic of the consultation, and thus of the chosen
speech, was sustainable aviation.
The topic of sustainable aviation was picked carefully, since persistent gender
stereotypes have led to assumptions about the policy areas that male and female politicians have expertise over and are interested in (Aalberg & Jenssen, 2007; Dolan, 2014; Hayes, 2011; Huddy & Terkildsen, 1993; Shapiro & Mahajan, 1986). It is believed that women and men stand for different policies and bring different issues into politics (Aalberg & Jenssen, 2007). For example, many people think that men are more able to deal with issues like crime, defense, military, economic development, trade, taxes, agriculture, and foreign policy issues. Other issues, such as child care, education, health care, poverty, women’s issues
19 and the environment, are labeled by Shapiro and Mahajan (1986) as compassion issues and are perceived by the public as handled better by women (Aalberg & Jenssen, 2007; Dolan, 2014; Huddy & Terkildsen, 1993). These assumptions about gendered competence areas could affect the evaluations of the gendered speeches. The topic of sustainable aviation was chosen because sustainable development is a cross-cutting issue, which means it contains many elements from different dimensions, namely the environmental, economic and social dimension (EEA Grants, n.d.). Indeed, the chosen speech sample focusses on
environmental issues, an area that is associated with female politicians, in combination with taxes, which is an issue associated with male politicians (Alexander & Andersen, 1993). As the issue of sustainable aviation is part of both traditional masculine (economic) and
feminine (environment) policy dimensions, the effects of the assumptions about gendered competence areas on the evaluations should be minimized.
The base speech about sustainable aviation of Chris Stoffer was shortened and manipulated by changing and adding certain function words to the speech. In this way, one speech sample with a distinct masculine linguistic communication style and one speech sample with a distinct feminine linguistic communication style was created. The manipulated speech sample with a masculine linguistic communication style contained 75 per cent more articles (20 vs. 35), 94 per cent more prepositions (18 vs. 35), 133 per cent more quantifiers and numbers (3 vs. 7), and 200 per cent more demonstratives (2 vs. 6) than the speech with a feminine linguistic communication style, which is in accordance with the literature on gendered linguistic speech styles. The feminine speech on the other hand contained 466 per cent more personal pronouns (17 vs. 3) when 1st person plural words are excluded1, and it
contained specifically more 1st person singular pronouns (12 vs. 0) than the manipulated
masculine speech sample. Next to this, the manipulated feminine speech sample contained
1 Since research has found no significant differences between men and women in their use of 1st person plural
pronouns, these words were not taken into account. When they are, the manipulated speech sample with the feminine style still contained 54.5 percent (17 vs. 11) more personal pronouns than the speech sample with the masculine style.
20 150 per cent more conjunctions (25 vs. 10), 26.67 percent more auxiliary verbs (19 vs. 15), 133 per cent more negations (7 vs. 3), and way more intensifiers (8 vs. 0) and hedge
phrases (5 vs. 0) than the masculine speech sample. The two manipulated speech samples counted 239 and 244 words. Both the original speech and the two manipulated speech samples can be found in Appendix A.
Before the experiment was executed, a small scale pilot study was done among 10 participants to test if the speech sample was successfully manipulated. The participants were asked to read both the manipulated speeches. After reading each speech, they rate it on the features that make a communication style masculine or feminine and they answered whether they found the style of the speech masculine or feminine. This was done by using 17 items, measured on 7-point Likert scales. Paired samples t-tests were conducted to compare the scores masculine and feminine speech sample condition. The results show that the
participants perceived the manipulated speech with the masculine style as being more masculine (M = 4.30, SD = 1.06) in comparison to the feminine style condition (M = 3.80, SD = 1.03). The manipulated speech sample with the feminine style was perceived as being more feminine (M = 4.20, SD = 1.23) than the masculine style condition (M = 3.70, SD = 0.82). However, these differences aren’t significant (t(9) = -0.80, p = .440; and t(9) = 0.92, p = .381), which might be a result of the rather small sample size of the pilot. Since the differences between the two stimuli, being a speech sample with a feminine style and one with a masculine style, were in the intended direction, it was expected that the speeches were successfully manipulated.
3.4 Measures
Dependent variables: Competence, Empathy, Integrity, Leadership and Warmth. The five dependent variables assessed the perceived Competence, Empathy, Integrity, Leadership and Warmth of the politician. The choice for these personality traits as dependent variables was based on Kinder’s (1986) four interrelated dimensions of traits (competence, empathy,
21 integrity, leadership) that are often are used by voters to evaluate politicians on. In recent work, Jones (2017) has added that Warmth also could be an influential trait dimension.
The five variables were measured by using 5 different scales (Funk, 1999; Pierce, 1993; Schneider & Bos, 2016), ranging from 3 to 5 items. The scale for Competence was composed of items measuring whether the politician was perceived as hard-working, intelligent and knowledgeable, the Empathy-scale was made up of items that indicate whether the politician was perceived as compassionate, kind, whether he/she cares about people like me, and was in touch with the people, Integrity was composed of items
measuring whether the politician was perceived as decent, moral, a good example, ethical and honest, the scale for Leadership was constructed out of items assessing whether the public found that the politician commands respect, is inspiring, provides strong leadership and gets things done, and lastly the Warmth-scale was composed of items measuring whether the politician was perceived as trustworthy, sincere and warm (Funk, 1999; Jones, 2017; Pierce, 1993; Schneider & Bos, 2016). All the items were measured by using a 7-point Likert scale, with a minimum score of 1 indicating ‘strongly disagree’ and a maximum score of 7 meaning ‘strongly agree’.
Five factor analyses indicated that all the proposed scales form unidimensional
constructs, since there was only 1 component with an Eigenvalue above 1.00 for every scale. These components explain from 62.80 per cent up to 73.89 per cent of the variance in each factor. Reliability analyses proved that all 5 scales are reliable, as all the Chronbach’s Alpha’s were above the necessary score of 0.60.
The 5 scale variables Competence (M = 4.86, SD = 1.05, Min. = 1.33, Max. = 7.00, α = .819), Empathy (M = 4.55, SD = 0.94, Min. = 2.00, Max. = 6.25, α = .817), Integrity (M = 5.06, SD = 0.93, Min. = 2.00, Max. = 7.00, α = .874), Leadership (M = 4.53, SD = 0.94, Min. = 1.25, Max. = 6.50, α = .798) and Warmth (M = 4.70, SD = 0.89, Min. = 2.00, Max. = 6.67, α = .761) were created by computing means out of the different items. A high score on the scales
22 indicates that the respondents rate the politician high on the specific personality trait, and vice versa.
Independent variable: (Linguistic) Communication Style. A dichotomous variable indicating the Linguistic Communication Style of the manipulated speech samples was created. Whether the speech samples were labeled as having a feminine or masculine linguistic communication style was dependent upon the function words that were used.
Independent variable (moderator): Gender. The Gender of the fictional politician performing the speech was indicated by creating a dichotomous variable. The speeches performed by Linda were labeled as performed by a woman and the speeches performed by Paul van Dijk were labeled as performed by a man.
3.5 Manipulation checks
In the experiment, two manipulations were made: first, the function words of the original speech were deleted and changed, and some function words were added, in order to create two versions of the speech: one with a distinct masculine style and another with a distinct feminine style. Secondly, the name of the politician that gave the original speech was
changed. Depending on the condition, the manipulated speeches were ascribed to a fictional male or female politician.
When manipulations are made, it is necessary to execute manipulation checks. To check whether the participants recognized the gender of the fictional politician performing the speech, they were asked to fill in the gender and the name of the politician at the end of the survey. Since 11 people filled in the wrong gender or stated they did not know, it was decided to filter these 11 participants out of the analysis.
Another manipulation check was executed in order to see whether the different speeches were actually perceived as manly and feminine. At the end of the survey, the participants were asked whether they agreed with the statement that the style of the speech
23 was manly and the statement that the style was feminine.2 The answer 1 indicated ‘totally
disagree’, and a 7 indicated ‘totally agree’. Independent samples T-tests were conducted to compare ratings in the feminine and masculine linguistic communication style condition. The results show that the participants in the masculine communication style conditions indeed thought of their speech sample as more manly (M = 4.11, SD = 1.30) than participants in the feminine communication style conditions (M = 3.69, SD = 1.42), which was a significant difference (t(172) = 2.05; p = .042). Also, the participants in the feminine linguistic
communication style conditions thought of their speech sample as more feminine (M = 3.76, SD = 1.51) than the participants in the masculine style conditions (M = 3.56, SD = 1.19). However, this was not a significant result (t(155.30) = .339, p = .465).
Results
Since the dependent variables Competence, Integrity, Empathy, Leadership and Warmth are all measured by using scales, and since interaction effects need to be tested, several two-way ANOVA tests were conducted. The two-two-way ANOVA analyses indicated that the interaction of the variables Politician’s Gender and Linguistic Communication Style did not have a significant effect on any of the dependent variables (Competence: F(1,170) = 0.00, p = .948; Integrity: F(1,170) = 0.36, p = .550; Empathy: F(1,170) = 0.00, p = .954; Leadership: F(1,170) = 0.02, p = .890; Warmth: F(1,170) = 0.20, p = .659). More results can be found in Appendix B.
This means that hypothesis H1a, stating that male politicians making use of a feminine linguistic communication style score higher on warmth than one that uses a masculine communication style, can be rejected. Also the first research questions can be
2 Before answering these two questions, the participants were presented with 15 statements about the
occurrence of certain function words in the speech that make a linguistic communication style feminine or masculine (e.g. ‘The speech contained many articles’, ‘The speech contained many quantifiers’) and the implications they have (e.g. ‘The speech was personal’, ‘The speech was hesitant’). The participants had to answer how much they agreed with the statements. In this way, they were primed to focus on the style of the speech, and less on the content.
24 answered: a male politician making use of a feminine linguistic communications style is not evaluated higher on the trait dimensions empathy (RQ1a) and integrity (RQ1b) than a male politician making use of a masculine linguistic communication style. Next to this, since no significant differences were found, the results do lend partial support for hypothesis H1b that states that the scores for perceived competence do not differ for male politicians with a feminine linguistic communication style and a masculine linguistic communication style. Research question 1c van be answered by stating that indeed, it could be that the public’s evaluations of male politicians with a feminine linguistic communications on the trait dimension leadership do not differ from that of male politicians with a masculine linguistic communication style (RQ1c). However, the results only lend partial support.
Moving on to the female politician, the whole second hypothesis, that stated that a female politician making use of a feminine linguistic communication style would be evaluated higher on the trait dimensions empathy (H2a) and warmth (H2b), but lower on competence (H2c), than a female politician making use of a masculine linguistic communication style, can be rejected. The second research question can be answered as follows: a female politician making use of a feminine linguistic communications style is not evaluated higher on the trait dimension integrity than a female politician making use of a masculine linguistic
communication style (RQ2a). However, the results do partially support the statement that they are evaluated the same on the trait dimension leadership, since no significant differences were found (RQ2b).
The analyses also showed that the independent variable Linguistic Communication Style did not have a significant effect on any of the dependent variables (Competence: F(1,170) = 2.38, p = .125; Integrity: F(1,170) = 0.13, p = .717; Empathy: F(1,170) = 0.96, p = .328; Leadership: F(1,170) = 1.11, p = .293; Warmth: F(1,170) = 1.34, p = .248).
However, the two-way ANOVA analyses did indicate that the independent variable Gender in itself does have a significant effect on the perceived Leadership (F(1,170) = 4.63, p = .033) and Warmth (F(1,170) = 4.69, p = .032) of the politician. The analyses showed that
25 participants that were shown a speech performed by a male politician (Leadership: M = 4.37, SD = 0.83; Warmth: M = 4.55, SD = 0.88) did differ significantly from the participants that read a speech performed by a female politician (Leadership: M = 4.68, SD = 1.02; Warmth: M = 4.84, SD = 0.88) on how they rated the politician on the traits Warmth and Leadership: the female politician scored, irrespectively of the communication style she used, higher on both the dimensions Leadership and Warmth than the male politician. For the other scales Competence (F(1,170) = 1.20, p = .239), Empathy (F(1,170) = 1.30, p = .256) and Integrity (F(1,170) = 1.62, p = .205), there was no significant effect.
Conclusion and Discussion
The main finding of this study is that, opposed to what was expected, female and male politicians were not evaluated differently on the trait dimensions Competence, Empathy, Integrity, Leadership and Warmth when they talked in counter-stereotypical ways. Also, linguistic communication style on itself does not have an impact on any of the five traits. However, the gender of the politician does have an impact: the female politician was perceived as significantly more warm and as possessing significantly stronger leadership traits than her male counterpart. These differences were solely a result of the gender of the politician.
In this thesis, written speech samples were used to test the impact of linguistic communication style on the public’s evaluations of several traits. However, audio might be a more effective tool in conveying a speaking style, especially a masculine one (Sparks, Areni & Chirs Cox, 1998). By using written texts that participants have to read, the gender of the politician is made considerably less salient than when one hears a gendered voice (Utych, 2019). However, using audio fragments creates difficulties, as now many other features might influence the results. For example, voice pitch influences how speakers are perceived (Klofstad, 2016). Still, in future research the impact of gendered linguistic communication
26 style and the role of the politician’s gender plays could be carefully assessed by using audio fragments.
The results of this study do differ from the findings of Jones (2017), who performed the only study that was found on the interaction of gender and linguistic communication style and its impact on perceived warmth and competence. In contrast to this study, Jones (2017) did find significant interaction effects. A possible explanation for the difference in impact on the trait dimension competence might be the way in which Jones (2017) measured this trait. Whereas Jones used two items that measured, on a 5-point Likert scale, how competent and capable the participants did perceive the politician to be, this study based the composition of the competence scale on Kinder’s (1986) insights and thus used three items measuring whether the politician was perceived as hard-working, intelligent and knowledgeable. However, this does not explain the different impact on the trait warmth, as this scale was measured using the same items as Jones (2017).
The different findings could also be a consequence of the sample sizes. It might be the case that the differences between the conditions are rather small and thus can only be found by using a large sample. In this study, 174 participants were analysed. Jones (2017) on the other hand gathered 557 participants, which is considerably more.
Another explanation for the differences is that culture is linked with and shapes gender stereotypes (Cuddy et al., 2010). Dependent upon whatever personality traits are culturally valued, men and women will be perceived in different ways. Language does also play a role in this: cultural frames that are created through language influence the contents of the gender stereotypes that play a role (ibid). It might be the case that the different cultures and languages of the United States and the Netherlands create different gender stereotypes, which leads to different results. Future research on the topic could investigate this claim further by performing a cross-national study.
27 A last possible reason for the different findings could be the issues of the speeches. As explained in the method section, gender stereotypes have led to assumptions about the policy areas that male and female politicians have expertise over and are interested in (Aalberg & Jenssen, 2007; Dolan, 2014; Hayes, 2011; Huddy & Terkildsen, 1993; Shapiro & Mahajan, 1986). In Jones’ (2017) experiment, each participant was asked to read 8 different speeches that were about a variety of topics. In this study however, participants were asked to read only one speech that was about sustainable aviation. Although sustainable
development is a cross-cutting issue, the speech was mostly about taxes, which is an issue that is traditionally part of the masculine policy dimension. This might have influenced the outcomes, as it could be hard to display personality traits that are often regarded as feminine when speaking about a masculine issue.
The results of this study might also simply show that nowadays sex stereotypes have eased and have not as much of an impact anymore as often thought. Alexander & Andersen (1993) showed that voters are more likely to evaluate female politicians, and possibly a feminine communication style, favourably when their attitudes about gender roles are more egalitarian. Literature has also shown that, over the past decades, stereotypes are easing and egalitarian attitudes towards female politicians are increasing (Burrell, 2008; Dolan, 2010; Eagly & Carli, 2007; Fridkin & Kenney, 2009). Already in 2008, a study performed by Pew Research Center demonstrated that women were evaluated by the American public as being superior to men on the leadership trait dimensions. This is in line with the finding in this study that female politicians are perceived as possessing stronger leadership traits, merely because of their gender. So, it might be the case that, in the Dutch context, sex stereotypes do not have the negative impact on the evaluations of politicians anymore that they are often assumed to have.
28
References
Aalberg, T. & Jenssen, A. T. (2007). Gender stereotyping of political candidates: an experimental study of political communication. Nordicom Review, 28(1), pp. 17-32.
Alexander, D. & Andersen, K. (1993). Gender as a Factor in the Attribution of Leadership Traits. Political Research Quarterly, 46(3), pp. 527-545.
4Argamon, S., Koppel, M., Fine, J. & Shimoni, A. R. (2003). Gender, Genre, and Writing Style in Formal Written Texts. Text, 23, pp. 321-346.
Argamon, S., Koppel, M., Pennebaker, J. W. & Schler, J. (2007). Mining the Blogosphere: Age, gender and the varieties of self-expression. First Monday, 12(9).
Best, D. L., & Williams, J. E. (1990). Measuring Sex Stereotypes: A Thirty-Nation Study. Beverly Hills: Sage.
Biber, D., Conrad, S. & Reppen, R. (1998). Corpus linguistics: Investigating langage structure and use. Cambride, England: Cambridge University Press.
Broverman, I. K., Vogel, S. R., Broverman, D. M., Clarkson, F. E. & Rosenkrantz, P. S. (1972). Sex-Role Stereotypes: A Current Appraisal. Journal of Social Issues, 28(2), pp. 59-78.
Burrell, B. (2008). Likeable? Effective Commander in Chief? Polling on Candidate Traits in the ‘Year of the Presidential Woman’. Political Science and Politics, 61, pp. 747-752.
Butler, J. (1999). Gender Trouble: Feminism and the Subversion of Identity. New York, USA and London, United Kingdom: Routlegde.
Campus, D. (2013a). Introduction. In D. Campus (Ed.), Women Political Leaders and the Media (pp. 1-9). London, United Kingdom: Palgrave Macmillan.
29 Campus, D. (2013b). Chapter 1: Gender and Models of Leadership. In D. Campus (Ed.),
Women Political Leaders and the Media (pp. 10-21). London, United Kingdom: Palgrave Macmillan.
Campus, D. (2013c). Chapter 3: Media Coverage of Woman Leaders. In D. Campus (Ed.), Woman Political Leaders and the Media. London, United Kingdom: Palgrave Macmillan.
Carlin, D. B. & Winfrey, K. L (2009). Have You Come a Long Way, Baby? Hillary Clinton, Sarah Palin, and Sexism in 2008 Campaign Coverage. Communication Studies, 60(4), pp. 326-343.
Colley, A., Todd, Z., Bland, M., Holmes, M., Khanom, M. & Pike, H. (2004). Style and content in emails and letters to male and female friends. Journal of Language and Social Psychology, 23, pp. 369-378.
Cuddy, A. J. C., Crotty, S., Jihye, C. & Norton, M. I. (2010). Men as Cultural Ideals: How Culture Shapes Gender Stereotypes. Working Paper at the Hardvard Business School. Retrieved on June 28, from
https://www.hbs.edu/faculty/Publication%20Files/10-097_0b862da9-45e9-40f2-b59d-796f7fc88276.pdf.
Cuddy, A. J. C., Fiske, S. T. & Glick, P. (2008). Warmth and Competence as Universal Dimensions of Social Perception: The Stereotype Content Model and the Bias Map. Advances in Experimental Social Psychology, 40, pp. 61-149.
Cuddy, A. J. C., Fiske, S. T., Kwan, V. S. Y., Glick, P., Demoulin, S., Leyens, J. P., Bond, M. H., et al. (2009). Stereotype Content Model Across Cultures: Towards Universal Similarities and Some Differences. British Journal of Social Psychology, 48(1), pp. 1– 33.
30 Deutsch, F. M. (2007). Undoing Gender. Gender & Society, 21(1), pp. 106-127.
Dolan, K. (2010). The Impact of Gender Stereotyped Evaluations on Support for Women Candidates. Political Behavior, 32, pp. 69-88.
Dolan, K. (2014). Gender Stereotypes, Candidate Evaluations, and Voting for Women Candidates: What Really Matters? Political Research Quarterly, 67(1), pp. 96-107.
Dow, B. J. & Tonn, M. B. (1993) “Feminine Style” and Political Judgement in the Rhetoric of Ann Richards. Quarterly Journal of Speech, 79, pp. 286-302.
Duerst-Lahti, G. & Kelly, R. M. (1995). The Study of Gender Power and Its Link to Governance and Leadership. In Duerst-Lahti & Kelly (Eds.), Gender Power,
Leadership and Governance (pp. 39-63). Ann Arbor: University of Michigan Press.
Dunaway, J., Lawrence, R. G., Rose, M. & Weber, C. R. (2013). Traits versus issues: how female candidates shape coverage of senate and gubernatorial races. Political Research Quarterly, 66(3), pp. 715-726.
Eagly, A. & Carli, L. (2007). Through the Labyrinth: The Truth about How Women Become Leaders. Cambridge: Harvard Business Press.
EEA Grants. (n.d.). Cross-cutting issues. Retrieved on May 6, from https://eeagrants.org/What-we-do/Cross-cutting-issues.
Fridkin, K. & Kenney, P. (2009). The Role of Candidate Sex in Voter Decision-Making. Political Psychology, 19, pp. 405-419.
Fiske, S. T., Cuddy, A. J. C., Glick, P. & Xu, J. (2002). A Model of (Often Mixed) Stereotype Content: Competence and Warmth Respectively Follow From Perceived Status and Competition. Journal of Personality and Social Psychology, 82(6), pp. 878-902.
31 Funk, C. L. (1996). Understanding trait inferences in candidate images. In M. X. Delli Carpini, L. Huddy & R. Y Shapiro (Eds.), Research in micropolitics: Rethinking rationality, Vol. 5 (pp. 97-123). Greenwich: JAI.
Funk, C. L. (1999). Bringing the Candidate into Models of Candidate Evaluation. The Journal of Politics, 61(3), pp. 700-720.
Gatterman, K. & De Vreese, C. H. (2017). The role of candidate evaluations in the 2014 European Parliament elections: Towards the personalization of voting behaviour? European Union Politics, 18(3), pp. 447-468.
Gleser, G. C., Gottschalk, L. A., John, W. (1959). The relationship of sex and intelligence to choice of words: A normative study of verbal behaviour. Journal of Clinical
Psychology, 15, pp. 182-191.
Glick, P., Lameiras, M., Fiske, S. T., Eckes, T., Masser, B., Manganelli, A. M., Pek, J. C. X., Huang, L. L., Sakalli-Ugurlu, N., Rodriguez Castro, Y., D’Avila Pereira, M. L.
Willemsen, T. M., Brunner, A., Six-Materna, I. & Wells, R. (2004). Bad But Bold: Ambivalent Attitudes Toward Men Predict Gender Inequality in 16 Nations. Journal of Personality and Social Psychology, 86(5), pp. 713-728.
Grebelsky-Lichtman, T. & Bdolach, L. (2017). Talk like a man, walk like a woman: an advanced political communication framework for female politicians. The Journal of Legislative Studies, 23(3), pp. 275-300.
Hancock, A. B. & Rubin, B. A. (2014). Influence of Communication Partner’s Gender on Language. Journal of Language and Social Psychology, 34(1), pp. 46-64.
Hardy, B. W. (2017). Candidate Traits and Political Choice. The Oxford Handbook of Political Communication. Retrieved on June 21st, from
32 https://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199793471.001.0001/ oxfordhb-9780199793471-e-017.
Hayes, D. (2011). When Gender and Party Collide: Stereotyping in Candidate Trait Attribution. Politics & Gender, 7, pp. 133-165.
Hayes, D. & Lawless, J. L. (2015). A Non-Gendered Lens? Media, Voters and Female Candidates in Contemporary Congressional Elections. Perspective on Politics, 13(1), pp. 95- 118.
Heilman, M. E., Wallen, A. S., Fuchs, D., & Tamkins, M. M. (2004). Penalties for Success: Reactions to Women Who Succeed at Male Gender-Typed Tasks. Journal of Applied Psychology, 89(3), pp. 416-427.
Herring, S. C. (1993). Gender and Democracy in Computer-Mediated Communication. The Electronic Jounral of Communication, 3(2). Retrieved on April 4, from
www.cios.org/EJCPUBLIC/003/2/00328.HTML.
Huddy, L. & Terkildsen, N. (1993). Gender Stereotypes and the Perception of Male and Female Candidates. American Journal of Political Science, 37(1), pp. 119-147.
Jones, J. J. (2016). Talk “Lika a Man”: The Linguistic Styles of Hillary Clinton, 1992-2013. American Political Science Association, 14(3), pp. 625-642.
Jones, J. J. (2017). Talk ‘like a man’: feminine style in the pursuit of political power. University of California, Irvine.
Karpowitz, C. F. & Mendelberg, T. (2014). The Silent Sex: Gender, Deliberation and Institutions. Princeton: Princeton University Press.
Kinder, D. R., Peters, M. D., Abelson, R. P. & Fiske, S. T. (1980). Presidential Prototypes. Political Behaviour, 2, pp. 315-337.
33 Kinder, D. R. (1986). Presidential Character Revisited. In R. R. Lau & D. O. Sears (Eds.),
Political Cognition (pp. 233-255). Hillsdale: Erlbaum.
King, D. C. & Matland, R. E. (2003). Sex and the Grand Old Party: An Experimental
Investigation of the Effect of Candidate Sex on Support for a Republican Candidate. American Politics Research, 31(6), pp. 595-612.
Klofstad, C. A. (2016). Candidate Voice Pitch Influences Election Outcomes. Political Psychology, 37(5), pp. 725-738.
Leeper, M. S. (1991). The Impact of Prejudice on Female Candidates: An Experimental Look at Voter Inference. American Politics Research, 19(1), pp. 248-261.
Lester, D. & Leenaars, A. (2016). A comparison of suicide notes written by men and women. Death Studies, 40(3), pp. 201-203.
Lodge, M. & Stroh, P. (1993). Inside the mental voting booth: an impression-driven process model of candidate evaluation. In S. Iyengard & W. McGuire (Eds.), Explorations in political psychology (pp. 225-263). Durham: Duke University Press.
Lupia, A. & McCubbins, M. D. (1998). The Democratic Dilemma: Can Citizens Learn What They Need to Know? Cambridge: Cambridge University Press.
Matland, R. E. (1994). Putting Scandinavian Equality to the Test: An Experimental Evaluation of Gender Stereotyping of Political Candidates in a Sample of Norwegian Voters. British Journal of Political Science, 24(2), pp. 273-292.
McKee, J. P., & Sheriffs, A. C. (1957). The Differential Evaluation of Males and Females. Journal of Personality, 25(3), pp. 356 -371.
34 McMillan, J. R., Clifton, A. K., McGrath, D. & Gale, W. S. (1977). Women’s Language:
Uncertainty or Interpersonal Sensitivity and Emotionality? Sex Roles, 3(6), pp. 545-559.
Menegatti, M. & Rubini, M. (2017). Gender Bias and Sexism in Language. Oxford Research Encyclopedia of Communication. Retrieved on April 13, from
https://oxfordre.com/communication/view/10.1093/acrefore/9780190228613.001.0001 /acrefore-9780190228613-e-470.
Miller, A. H., Wattenberg, M. P., & Malanchuk, O. (1986). Schematic assessment of presidential candidates. American Political Science Review, 80, pp. 521–540.
Moss-Racusin, C. A. (2012). Male backlash: penalties for men who violate gender
stereotypes. In R. J. Burke & D. A. Major (Eds.), Gender in Organizations: Are Men Allies or Adversaries to Women’s Career Advancement? (pp. 247-269). Cheltenham & Northampton: Edward Elgar Publishing.
Mulac, A., Bradac, J. J. & Gibbons, P. (2001). Empirical Support for the Gender-as-Culture Hypothesis: An Intercultural Analysis of Male/Female Language Differences. Human Communication Research, 27(1), pp. 121-152.
Mulac, A. & Lundell, T. L. (1982). Linguistic determinants of the gender-linked language effect. Paper presented at the annual meeting of the International Communication Association, Boston.
Mulac, A. & Lundell, T. L. (1986). Linguistic Contributors to the Gender-Linked Language Effect. Journal of Language and Social Psychology, 5(2), pp. 81-101.
Mulac, A. & Lundell, T. L. (1994). Effects of gender-linked language differences in adult’s written discourse: Multivariate tests of language effects. Language & Communication, 14(3), pp. 299-309.
35 Mulac, A., Studley, L. B. & Blau, S. (1990). The gender-linked language effect in primary and
secondary students’ impromptu essays. Sex Roles, 23(9), pp. 439-469.
Newman, M. L., Groom, C. J., Handelman, L. D. & Pennebaker, J. W. (2008). Gender Differences in Language Use: An Analysis of 14.000 Text Samples. Discourse Processes, 45(3), pp. 211-236.
Pennebaker, J. W. (2011) The Secret Life of Pronouns: What our Words Say About Us. New York, Berlin, London and Sydney: Bloomsbury Press.
Pew Research Center. (2008). Men or Women: Who’s the Better Leader? A Paradox in Public Attitudes. Washington: Pew Research Center.
Pierce, P. A. (1993). Political Sophistication and the Use of Candidate Traits in Candidate Evaluation. Political Psychology, 14(1), pp. 21-35.
Poole, M. E. (1979). Social class, sex, and linguistic coding. Language and Speech, 22(1), pp. 49-67.
Popkin, S. L. (1994). The reasoning voter: Communication and persuasion in presidential campaigns. Chicago: University of Chicago Press.
Rahn, W., Aldrich, J. H., Borgida, E., & Sullivan, J. L. (1990). A social-cognitive model of candidate appraisal. In J. Ferejohn & J. Kuklinski (Eds.), Information and democratic processes (pp. 136–159). Champaign: University of Illinois Press.
Rudman, L. A. & Glick, P. (1999). Feminized management and backlash towards agentic women: The hidden costs to women of a kinder, gentler image of middle managers. Journal of Personality and Social Psychology, 77(5), pp. 1004-1010.
36 Rudman, L. A. & Fairchild, K. (2004). Reactions to Counterstereotypic Behaviour: The Role
of Backlash in Cultural Stereotype Maintenance. Journal of Personality and Social Psychology, 87(2), pp. 157-176.
Schneider, M. C. & Bos, A. L. (2016). The Interpay of Candidate Party and Gender in Evaluations of Political Candidates. Journal of Women, Politics & Policy, 37(3), pp. 274-294.
Schwartz, H. A., Eichstaedt, J. C., Kern, M. L., Dziurzynski, L., Ramones, S. M., Argrawal, M., Shah, A., Kosinski, M., Stillwell, D., Seligman, M. E. P. & Ungar, L. H. (2013). Personality, Gender, and Age in the Language of Social Media: The Open-Vocabulary Approach. Plos One, 8(9), pp. 1-16.
Sedivy, J. (2016, October 25). Donald Trump Talks Like a Woman. Politico Magazine. Retrieved on February 26, from
https://www.politico.com/magazine/story/2016/10/trump-feminine-speaking-style-214391.
Shapiro, R. Y. & Mahajan, H. (1986). Gender Difference in Policy Preferences: A Summary of Trends from the 1960s to the 1980s. Public Opinion Quarterly, 50, pp. 42-61.
Singh, S. (2001). A Pilot Study on Gender Differences in Conversational Speech on Lexical Richness Measures. Literary and Linguistic Computing, 16(3), pp. 251-264.
Sparks, J. R., Areni, C. S. & Chris Cox, K. (1998). An Investigation of the Effects of Language Style and Communication Modality on Persuasion. Communication Monographs, 65(2), pp. 118–136.
Turner, C. (2017, December 6). Masculine-Feminine Difference: How We Talk. Huffpost. Retrieved on May 6, from https://www.huffpost.com/entry/masculinefeminine-differe_b_5559127?guccounter=1.
37 Thomson, R. & Murachver, T. (2001). Predicting gender from electronic discourse. British
Journal of Social Psychology, 40, pp. 193-208.
Utych, S. M. (2019). Speaking style and candidate evaluations. Politics, Groups, and Identities, 7(2), pp. 1-19
UN Women. (2019, January). Facts and figures: leadership and political participation. Retrieved on February 2, from www.unwomen.org/en/what-we-do/leadership-and-political-participation/facts-and-figures.
West, C. & Zimmerman, D. H. (2009). Accounting for Doing Gender. Gender & Society, 23(1), pp. 112-122.
Winter, N. J. G. (2010). Masculine Republicans and Feminine Democrats: Gender and Americans’ Explicit and Implicit Images of the Political Parties. Political Behavior, 32(4), pp. 587-618.
Wood, J. T. (1995). Gendered Interaction: Masculine and Feminine Styles of Verbal Communication. In K. S. Verderber (Ed.), Voices: A Selection of Multicultural Readings (pp. 18-29). Belmoni, California, United States of America: Wadvorth Publishing Company.
Wood, J. T. (2009). Gendered Lives: Communication, Gender, and Culture. Wadsworth, Boston, United States of America: Wadsworth Cengage Learning.
