HOW WE COME TO LEAD A YOUTH PERSPECTIVE ON UNDERSTANDING MODERN LEADERSHIP

HOW WE COME TO LEAD: A YOUTH PERSPECTIVE ON UNDERSTANDING

MODERN LEADERSHIP

CHAPTER I

INTRODUCTION

Despite being 50% of the US population women make up just 36% of senior

managers, 31% of vice presidents, 27% of senior vice presidents, and 24% of C-suite

executives (McKinsey & Company, 2021). The disparity between white leaders and

leaders of color is even more significant with 85% of executives being white, only 5%

identifying as Hispanic, 3% identifying as Black, and 6% Asian (Agovino, 2022).

Women have accounted for more than fifty percent of degree holders for more than 20

years and yet struggle to find a seat at the decision-making table (Bryant, 2021; C. L.

Hoyt, 2010). Despite their lack of access, women have proven to be highly effective and

capable leaders (Eagly & Carli, 2007). Similarly, leaders of color have demonstrated

success in effective leadership despite numerous barriers in front of them (Bush et al.,

2005). However, research has repeatedly found that men, specifically white men, enter

adulthood perceiving themselves as leaders in ways that women and people of color do

not (Folberg et al., 2020; Rosch et al., 2015; Singer, 1989). This restriction in who has

access to leadership hinders not just those without power but also those with power.

Diversity in leadership access matters both philosophically and functionally as a

protector of democracy. Diversity ensures increased financial opportunities that in turn

ensure increased difference in ways of thinking, creating, and growing. Stagnancy in

leadership access prevents progress and reduces the competitive edge that organizations,

business, and communities strive for.

Much of the leadership research focused on gender and leadership addresses a

persistent link between a leader's gender and the ways they are expected to embody their

leadership (Eagly, 2007; Hentschel, 2019). Closely linked to the broader societal

expectations of gender performativity, gendered leadership research prescribes

traditionally feminine qualities (warm, helpful, collaborative, intuitive, compassionate

etc.) to female leaders and traditionally masculine qualities (bold, ambitious, assertive,

strong etc.) to male leaders (Abele et al., 2016; Henstchel et al., 2019; S. K. Johnson et

al., 2008). Bakan (1966) first described these stereotyped characteristics as either

communal or agentic, and female or male respectively (Eagly et al., 2020). Similar

dichotomies are present in how white and leaders of color are stereotyped and developed

through social construction (Chung-Herrera & Lankau, 2005; Rosette et al., 2008). While

the deep-rooted aspects of racism in leadership development are not addressed in this

dissertation the intersectionality of race and gender as it pertains to leadership

development is explored (Malcom & Malcom, 2011; Sanchez-Hucles & Davis, 2010).

Essentialists, who have dominated gendered leadership research, treat these

gender differences in leadership as a reflection of innate differences between men and

women (Cundiff & Vescio, 2016; C. Marshall & Young, 2013; Singer, 1989). In

contrast, I situate myself among social constructivists who argue instead that this

alignment between gender and leadership occurs through the production and

reproduction of societal presuppositions and patriarchal norms (Butler, 1993; Campbell

& Wasco, 2000). Social constructivists argue that concepts such as gender, sexuality,

class, race, and ability are not innate but instead constructed and defined by humans

through cultural and

historical contexts (Berkowitz et al., 2010; Gedro & Mizzi, 2014; Valocchi, 2005).

Similarly, the argument can be made that the intersection of race with gender presents a

different level of social expectations and resulting performance (Pierce, 2014; Smaje,

1997). By looking at leadership through a social constructivist lens, leadership becomes a

malleable concept influenced by both external and internal factors (Cieciuch &

Topolewska, 2017; Digman, 1997; Srivastave et al., 2003). Without this critical

assumption of plasticity, attempts to change the landscape of leadership would seemingly

be a futile pursuit. However, with this assumption important questions arise about how

and when individuals come to see themselves as leaders and the ways in which they learn

to lead.

The problem with leadership research is not just its over-reliance on essentialist

notions of gender and biases around race but also its hyperfocus on adult leaders. Indeed,

gendered and racialized leadership literature often ignores the significance of adolescence

as a pivotal time when leadership is learned (Elliott & Stead, 2008; Fairhurst & Grant,

2010; C. L. Hoyt, 2010). I contend that schools present a key, overlooked, space for

understanding how individuals come to see themselves as leaders and how they perceive

their leadership displays as being encouraged or discouraged during these developmental

years (DePalma, 2013). As youth come to understands themselves as leaders the question

becomes to what extent, if at all, are their self-ascriptions of leadership aligned with

gender and racial norms. Then there is the question of external influence and particularly

the influence of adults in their lives in perpetuating or countering gender-aligned self-

definitions. Only once we understand these can we begin to disentangle and racial

stereotypes from leadership.

Driven by a desire to understand how leaders can come to be more reflective of

the people they lead, this research attempts to tackle both the essentialist view of

gendered and racialized leadership that is perpetuating categorical stereotypes and the

failure of existing research in examining leadership development in youth. The current

practice of coupling gender and leadership or race and leadership without accounting for

the social construction of behavior in childhood and adolescence continues the restriction

of leaders. By questioning the social construction, we can also question who enters

leadership roles, the effectiveness of those leaders, and how we recreate harmful

patriarchal norms that are in line with values of white supremacy.

The continued disproportionality in leadership, I theorize is the reflection of

internalized patriarchal norms that shape individuals into gender/racially aligned

leadership roles. The expectations of how women and men are expected to lead and the

spaces in which they are permitted leadership continues to unequally benefit men who

display traditionally “masculine” leadership (Eagly, 2005; Eagly & Karau, 2002; Elliott

& Stead, 2008; Jampol & Zayas, 2021). Discussions of gender effects on leadership

would be incomplete without consideration of the intersectionality of race as well. For

instance, women of color face double jeopardy or a double bind described as barriers

faced due to both their gender and their race that interact in unique ways not faced by

white women or men of color (Malcom & Malcom, 2011; Rosch et al., 2015; Sanchez-

Hucles & Davis, 2010).

How one embodies leadership, however, is not innate and therefore the link

between gender/race and leadership ways of being can and should be critiqued. I advance

that leadership is socially constructed in ways that are unnecessarily linked to gender and

race and that this social construction occurs throughout childhood and adolescence

eventually leading to significant differences in who is seen and sees themselves as

leaders. Guided by this assumption I argue that a key step towards equity is

understanding how this coupling occurs. These internalized norms are omnipresent as

youth walk through life both in and outside of schools. However, schools provide one

potential space for disruption.

American public schools have a long history of tracking and sorting students

based on assumptions and with generational consequences (Katz, 1976). Public schools

have origins as institutions designed for assimilation and the upholding of social role

norms (D. W. Adams, 1995). This space that is meant to train youth on how to act in the

world has played a substantial role in perpetuating stereotypes seen in broader society

(Okeke et al., 2009; Romera, 2015). When considering the origins of leadership

development, exploring the roles schools potentially play in this continued perpetuation

of gendered and racial norms is vital. While public schools promise an equal playing

ground for all students regardless of identity, it is clear from post high school outcomes

that not everyone is exiting the school's system equally equipped for the world (R.

Gordon et al., 2000; Harry & Anderson, 1994; Keels, 2013). Querying the how, when,

and why these differences occur, in this case around leadership, is crucial for readjusting

who has access not just in schools but beyond them.

Teachers play a significant role in shaping students in the classroom, be it their

self-concept, their behavior, or their academic outcomes (Brophy & Good, 1970; Jussim,

1986; Rosenthal & Jacobson, 1968). Researchers have dubbed this influential power of

teachers on students as teacher expectancy. Students react to and internalize messages

about who they are meant to be and how they are meant to act. When these messages are

left unchecked, they can, and often do, recreate and maintain problematic norms around

gender and racial expectations that affect leadership development and have lasting effects

in access to and success in leadership positions (M. A. Hoyt & Kennedy, 2008). If

education can provide a space that critiques and questions gendered and racial norms

around leadership it may provide an avenue for increase diversity in not just how people

lead but what is come to be understood as “good” leadership and how is presumed to

embody “good leaders” ultimately leading to future generations that allow for an

encourage diversity in leadership roles and power access.

Purpose of the Research

The purpose of this research is twofold. First, I aim to understand how children

and adolescents conceptualize their own leadership. Second, I attempt to understand if

children and adolescents perceive their teachers encouraging or discouraging particular

leadership performances. I have chosen to examine these differences across various

grades and between students of different races and genders. As noted above, gender

differences in leadership are well documented; however, there are also significant

disparities between leadership representation across race and ethnicity. These racial

disparities and the significance of intersectionality with gender will be discussed more in

the subsequent chapter. The choice to look at students across various grades highlights

the progression of how leadership is conceptualized at various developmental levels

painting a more comprehensive picture of how youth define their own leadership.

Students exit school and enter society either a reflection of the status quo or in

opposition to it; something significantly shaped by the teachers they have. Leadership has

historically been held by a minority of white men who have engaged in ways of leading

that have become synonymous with masculinity and whiteness. But I argue these ways of

leading do not inherently algin with white males but instead are part of a cycle of

perpetuating patriarchal norms onto individuals as they learn to lead. This research

examines when these arbitrary leadership expectations emerge, and the potential role

teachers play in perpetuating them.

This mixed methods research occurred within a small rural school district focused

on the experiences of students in in elementary, middle, and high school as they reflected

on their own leadership behaviors, styles, and qualities. Through qualitative data this

research aims to address the lack of language and student voice in conceptualizing

leadership performativity. Through quantitative data this research explores to what extent

students are systematically taught to uphold gender and racial norms and maintain the

status quo in how they learn to conceptualize themselves as leaders.

Research Questions

Both the qualitative and quantitative questions below explore how children and

adolescents in schools identify their leadership in congruent or incongruent ways

according to patriarchal norms as well as the role they perceive teachers playing in the

adoption of these leadership qualities.

Qualitative Research Question

1. What vernacular do children and adolescents use to conceptualize their own

displays of leadership and the external influences they experience while

learning to lead?

Quantitative Research Questions

1. How, if at all, do students' ratings of their own leadership attributes differ

across grade (third, seventh, & 11th/12th), race (white & Hispanic), and gender

(male & female)?

a. How, if at all, do these differences align with or counter “communal” and

“agentic” leadership expectations for female and male students

respectively?

2. How, if at all, do students perceive/rate teachers responding differently to

leadership behaviors that are coded as gender congruent and gender

incongruent?

a. How, if at all, do these perceptions differ by student grade (third, seventh,

& 11th/12th), race (white & Hispanic), gender (male & female)?

The decision to use the terms Hispanic and white reflect the use of those terms in

the school demographics that were available to me while conducting this research. I do,

however, acknowledge that the identifier of students as Hispanic is an imperfect and at

times problematic choice (Taylor, 2012). Similarly, the terms male and female are used

because of their tie to the binary understanding of gender in leadership literature and in

available demographic data. However, as will be further discussed in Chapter II, the

gender binary is a social construction and because of this, participants had the option to

self-identify as male, female, nonbinary, and the option to provide an alternative to those

ternaries.

Significance

To understand the significance of the questions above requires an understanding

of the issues of inequity surrounding who is situated in positions of power and the

importance of how leaders come to lead. These questions examine both how individuals

internalize expectations in their leadership and how these expectations are reinforced.

With this understanding those supporting would-be leaders can decouple and restructure

not just who sees themselves as leaders, but who is accessing leadership, and who is

being seen by others as leaders. This research allows for both an increased understanding

of how gendered and racial leadership comes to be internalized and perpetuated in

schools while simultaneously adding young voices that have been previously ignored

(Cook-Sather, 2006).

While many agree that youth leadership is important (Coffey & Lavery, 2018;

Mortensen et al., 2014) and leadership is a learned process of being (Engelbert &

Wallgren, 2016), few have asked the students themselves about their leadership

(Karagianni & Montgomery, 2018). Indeed, leadership literature is ripe with language

that is overly complex and out of touch with the understandings of identity that youth of

the 21st century embody. Giving youth a voice in describing their own leadership

strengthens the collective understanding of how leadership is developed and the qualities

that are nurtured and lost in the developmental process (Coffey & Lavery, 2018; S.

Gordon, 1994; Mortensen et al., 2014). In seeking a path for more diverse leadership this

research attempts to center the experiences of youth as a means for understanding the

influence of educators in shaping leaders of the future (Freire, 2000).

Definitions

Agentic leadership. Defined as leadership that conveys assertion and control.

Leaders who are “especially aggressive, ambitious, dominant, self-confident, and

forceful as well as self-reliant and individualistic” (Eagly & Carli, 2007, p. 86). Agentic

leaders are “achievement oriented and thus are viewed as competent, aggressive,

independent, decisive, and forceful” (Rosette & Tost, 2010, p. 222).

Communal leadership. Defined as leadership that conveys concern with the

compassionate treatment of others. Leaders who are “especially affectionate, helpful,

friendly, kind, and sympathetic as well as interpersonally sensitive, gentle, and soft-

spoken” (Eagly & Carli, 2007, p. 86). Communal leaders are “relationship-oriented and

thus as kind, helpful, concerned, and sympathetic to others' needs” (Rosette & Tost,

2010, p. 222).

Social construction. Defined as the assumption that taken for granted norms and

meanings and are formed by social processes that “once crystalized [are] maintained,

modified, or even reshaped by social relations” (Berger & Luckmann, 1966, p. 173).

Teacher expectancy/Pygmalion effect. Defined as when “one person's

expectations of another's behavior may come to serve as a self-fulfilling prophecy”

(Rosenthal & Jacobson, 1968, p. 20). Teacher expectancy theory has claimed that a

teacher's expectations of their students can have a strong effect on student performance

(Brophy & Good, 1970; O. Johnson et al., 2019; Jussim, 1986).

Limitation of Operating Within Patriarchy

Throughout this research I walk the careful line of recognizing that merely by

addressing patriarchy I am in some ways perpetuating it. There are ineradicable

limitations related to the pervasiveness of patriarchal norms that uphold white

supremacy. Because of the social constructivist lens through which I conduct my

research I acknowledge that it is near impossible to disentangle learned ways of being

that stem from societal norms from true individual differences. Additionally, the research

itself presents a potential space for perpetuation of the very norms it seeks to question.

By grounding the interpretation of leadership through a gendered lens the research in

some ways validates these differences. Throughout the data collection and analysis I take

a feminist and queering approach, discussed further in Chapter II, to both how questions

are asked and how responses are interpreted.

Delimitations

The delimitations of this study center around the questions of why this setting,

these participants, and these data? The answer to the first is the importance of research in

generally overlooked small rural school districts (Lavalley, 2018). The second question

surrounding participants is slightly more complicated. These participants were chosen

from Grades 3, 7, and 11/12 to allow for insight into different developmental levels of

identity development. Rather than focusing on an individual developmental level the

choice to data collect from different developmental stages provides an interesting and

important glimpse at how leadership forms alongside other developmental milestones

(Erikson, 1959; Sokol, 2009).

The answer to the third, why these data, really questions the decision to utilize

mixed methods and specifically interviews and Likert scales as primary means of data

collection. While a narrative or grounded theory qualitative approach would provide

depth of experience not accessible through limited interviews, depth of individual

experiences is not the aim of this research. The individual experiences of adolescents as

they develop and discover their own leadership is undoubtably important and worthy of

examination. However, the aim of this research is to seek an understanding of collective

ways that broader patriarchal influences impact youth self-definitions of leadership.

Initially, conducting qualitative research and utilizing data produced from those

interviews in the development of the quantitative portion allowed two important things:

firstly, the qualitative component enabled this research to raise child and adolescent

voices and value the knowledge they hold. Secondly, the quantitative component allowed

their voices to be validated within a larger subsect of data (Hesse-Biber, 2010). Because

both gendered and racialized leadership research has mostly ignored children and

adolescents, very little language exists around the ways they describe their leadership.

Without their own words, conducting quantitative research relies on descriptors and

concepts that seem out of touch with the experiences of 21st-century youth. This mixed

methods approach redefines gendered leadership concepts in accessible language and

then validates those concepts in a larger sample.

Researcher’s Why

Perhaps best said by Reason and Marshall (1987) “all good research is for me, for

us, and for them” (p. 112). This chapter opened with the them, the need for future leaders

to be more reflective of broader society and less reflective of the past. Now briefly I want

to address the me and the us. As a student I had teachers who made me feel confident and

capable and others who made me feel inadequate, inarticulate, and unintelligent. The

power educators have in shaping how we see ourselves and the choices we make for the

rest of our lives is astronomical. I loved and hated school as a student. Now as an

educator, I continue to love and hate schools in new ways. Public education in the United

States is doing exactly what it was designed to do—benefit the privileged and funnel

children into various forms of labor for the betterment of capitalism. Yet I know that

there is potential for greatness in our schools and greatness in what our schools can do

for our children. As a white, able-bodied, queer, woman I benefit from several positions

of privilege while also continuing the centuries-long struggle to work against patriarchal

heteronormative expectations. My obligation, and the hopes for this research, is to

acknowledge the role we as educators play in perpetuating a structure that ultimately

hurts our students. It is a call to action for those of us that work with students to begin the

process of systematically dismantling our own biases around gender and leadership so as

educators we can raise a generation of children who lead in the ways they view best and

are valued not for their congruence to patriarchal norms but for who they are and the

potential they hold.

CHAPTER II

LITERATURE REVIEW

Theoretical Assumptions

Before addressing the issue of leadership and how its current conception is

perpetuating inequity in access, I feel it is necessary to briefly explore patriarchy and its

relation to how gender and power are intertwined and upheld. Through the theoretical

perspectives elicited from feminist and queer theorists I utilize patriarchy as a way of

understanding and revisiting definitions of gender and more importantly the implicit

coupling of power along binary gender lines. In tandem with this I briefly explore the

intersection of racism as it pertains specifically to the power dynamics and performative

nature of individuals within white supremacy. These theorists set a foundation for

reexamining how gender aligned behavior, specifically in the context of leadership, is

socially constructed and therefore can be socially deconstructed.

Patriarchy—Power & Gender

Hunnicutt (2009) defined patriarchy as “social arrangements that privilege males,

where men as a group dominate women as a group, both structurally and ideologically

hierarchical arrangements that manifest in varieties across history and social space” (p.

557). This notion of male domination and inequity between the binary genders is nothing

new. Indeed, Aristotle claimed in the 4th century BC that as “regards the sexes, the male

is by nature superior and the female inferior, the male ruler and the female subject”

(Smith, 1983, p. 467). Embedded in both of these definitions are not only assumptions of

a gender binary but also an acceptance of inequitable power divided between the binary.

Patriarchal norms place power and authority in the hands of men and reinforce

hierarchical social systems, social arrangements, and power that perpetuate injustice

(Horowitz, 1976; Hunnicutt, 2009).

This power differential is justified through the belief that there are differences in

the fundamental composition of women compared to men and that these physical and

biological differences result in gender-aligned attributes or ways of being positioning

women and men differently (Horowitz, 1976; Smith, 1983). Built into patriarchy are

expectations of predetermined social roles and social performance displayed through acts

of daily living (Smith, 1983). Gedro and Mizzi (2014) note that the majority of social

structures are marked by “heteromasculinity, where heterosexuality and masculinity are

dominant and acceptable features of day-to-day functions and interactions” (p. 445).

These systems of male privilege become normalized and obscure the ways in which

history has been manipulated to keep the status quo intact and the power solidified with

men (Newton, 1984). Likewise, patriarchal ideology has acted to minimize women's

agency and systematically reduce expectations for female power (Newton, 1984).

Patriarchal power functions within Eurocentric ways of being including, most

significantly, supporting white supremacy and perpetuating not just sexism but racism,

homophobia, and ableism as well (Hemming, 2021). Hunnicutt (2009) has examined the

importance of recognizing patriarchal power as not exclusively top-down but rather

complicated and intersectional. Ultimately, patriarchy as it functions in western society

today, perpetuates the gender-based distribution of power and the normalization of the

male majority in positions of authority.

Feminist Theorists’ Take on Power

Embedded within patriarchy is the notion of power and privilege aligned within a

gendered world. Feminist theorists have consistently fought for the purposeful critiquing

of the power dynamics at play in a world where how a person is perceived along the

gender binary equates to their opportunity (Butler, 1993; Campbell & Wasco, 2000;

Gedro & Mizzi, 2014; Hemming, 2021). Feminist theory is interested in uncovering and

dismantling oppressive practices derived from unexamined norms while questioning

“assumptions about privilege and about access” (Gedro & Mizzi, 2014, p. 447). These

theorists don't just empower women and their ways of being but also acknowledge the

limitations of restricting men towards rigid masculinity and traditional male ways of

being. Ultimately, perpetuation of sexism and patriarchy limit all individuals as it forces

them to remain within the confines of their presumed gender (Campbell & Wasco, 2000).

Feminists call for many significant structural changes including equal access to

resources, dismantling of economic and class structures that are inherently problematic,

and examining the intersectionality of race and class in women's oppression within a

culture of patriarchy and white supremacy (Campbell & Wasco, 2000; Newton, 1984).

As Hemming (2021) put it—true feminist work requires “disidentifying with patriarchal

logics in order to forge a feminist—and therefore antiracist—praxis of living” (pp. 306–

307). Rooted in this work of examining power is also an implicit questioning and

problematizing of the assumption that men and women are distinctly different entities

whose actions, behaviors, personalities, and desires are defined by their physical or

biological differences (Gedro & Mizzi, 2014; Hammers & Brown, 2004; Newton, 1984).

Adopting feminist theory requires examining power dynamics impacted by gender both

historic and current with an aim towards justice and equality. In the context of leadership,

this means acknowledging the gender gap in leadership access is not an issue of

individual merit or skill but a structural issue of patriarchal power.

Queer Theorists’ Take on Gender

Like feminist theorists, queer theorists also ask us to question power dynamics.

Specifically, they call us to question the existing ideology that states that sex and gender

are linked and result in innate differences. They critique the economic, political, and

institutional powers that benefit from upholding patriarchal norms of sex-based gender

roles (Butler, 1993; Valocchi, 2005). Queer theorists argue for not just a deep questioning

of norms, but also a movement away from the binary definitions of the distinct categories

of sex, gender, and sexual orientation (Gedro & Mizzi, 2014). They concur with

sociologists who argue that gender is performative and a reflection of actions that are

explicitly and implicitly taught, reinforced, and internalized rather than innate (Berkowitz

et al., 2010). But they move beyond that to also argue for a less clear distinction in the

binary understanding of gender presentation, positioning against strictly masculine or

feminine-coded displays of gender (Gedro & Mizzi, 2014; Valocchi, 2005). As Valocchi

(2005) stated, “these binaries incompletely or imperfectly represent a broad range of

complicated social processes surrounding the meaning of bodies and the social cues,

practices, and subjectivities associated with gender and sexuality” (p. 753).

Learning and enacting these gendered social roles comes about through a series of

external power displays that are so far-reaching they can be difficult to identify

(Hammers & Brown, 2004). Not only is gender performative and derived from the

observation and cueing of others but it is also reinforced by the actions of the individual

themselves who come to see their behaviors as intrinsic rather than socially manipulated

(Butler, 1993). This process of self-identifying inadvertently validates the categorization

of self and the adoption of what is expected from the social role one has selected for

oneself, which is then systematized and normalized (Berkowitz et al., 2010; Butler,

1993).

If we conceive of both gender and leadership as performative and derived from

acts then those acts can be systematically decoupled. As common leadership expectations

now stand, we continue to perpetuate “male” leadership and uphold patriarchal norms.

But by questioning the legitimacy of innate gender differences, queer theorists provide a

lens through which to question the legitimacy of innate gendered leadership differences.

Patriarchy and Race and Performativity

While patriarchy is mostly commonly understood as the unequal power and

privilege between women and men it is also a part of a bigger whole that includes racism,

homophobia, ablism, and classism (Bailey & Zita, 2007; A. G. Johnson, 2004). In many

ways the manhood and masculinity that is touted in patriarchy is also whiteness and

eurocentrism (A. G. Johnson, 2004). Bahlieda (2015) described the far-reaching nature of

patriarchy when they said, “attacking patriarchy and the status quo is also often like

shadow boxing because patriarchy has no form, no substance, and no reality beyond that

evoked through human interaction” (p. 19). Access, power, and societal acceptability

within patriarchy is measured by the distance from masculinity and whiteness (Torres &

Pace, 2005).

Race and social construction have complicated history (Pierce, 2014). Race as we

understand it in terms of the interactions of individuals with the world around them and

the world's interactions with them is understood better as ethnicity and culture. Smedley

and Smedley (2005) explain that “ethnicity and culture are related phenomena and bear

no intrinsic connection to human biological variations or race” (p. 17). Just as with

gender presentation and its association with biological sex, ethnicity and cultural

presentations have come to have associations with race. While the intricacies of

unpacking race from ethnicity and culture are important in general they are somewhat

irrelevant to understanding the biases and power dynamics evident in the execution of

patriarchy across racial groups (Andreasen, 2000; Smaje, 1997). This dissertation does

not call for a dismantling of cultural differences; however, it does call into question how

norms entrenched in patriarchy are held against people in terms of power and privilege

(Smaje, 1997). This difference in interactions is more closely examined through

intersectional theorists.

Intersectional Theorists and Power Dynamics

Intersectional theorists describe the ways in which “human lives are multi-

dimensional and complex, and people's lived realities are shaped by different factors and

social dynamics operating together” (Wyatt et al., 2022, p. 863). Intersectionality has its

origins in the experiences of Black women specifically (Crenshaw, 1989). While

intersectionality theory has in some circles shifted away specifically from the experiences

of women of color and specifically Black women it is important to remember the origins

of this theory (Kerner, 2017). In this research I utilize intersectionality to critique gender

and race but do not explicitly look at other intersecting elements of identity that can be

and often should be housed under true intersectional research. I also acknowledge that

power is dictated by more than just race and gender, but intersectional theorists provide a

framework for understanding how multiple components of self, and perceived self, must

be considered when considering power dynamics.

Intersectionality works in tandem with social identity theory which says that

identity is derived from perceived alignment with a group (Wijeyesinghe & Jones, 2014).

However, intersectionality allows for these multiple group memberships to coexist and

work in tandem with each other to influence the experiences and displays of self

(Samuels & Ross-Sheriff, 2008; Wijeyesinghe & Jones, 2014). As Edstrom and

colleagues (2016) put it, intersectionality allows us to “recognize identities as co-shaped

by interactions in intersecting processes and structures of power” (p. 63). Gender

performativity is impacted by racial background and ethnicity and cultural performativity

is likewise influenced by socially perceived gender. Power and access are then dictated

socially by the extent to which one falls within our outside of patriarchal and white

supremacy norms (Kerner, 2017; Patil, 2013). Intersectionality is not only relevant to the

liberation of women and nonbinary people but also to the liberation of men from

patriarchy and understanding the different experiences across the gender spectrum

(Edstrom et al., 2016). Similarly, intersectionality is not just relevant to understandings

the limitations placed on people of color, but it is relevant to understanding unearned

privilege and the role that white supremacy plays in the lives of all peoples who exist

within it.

Unifying Theory

Through an understanding of the existing patriarchal ideologies that undergird our

daily lives examined through feminist, queer, and intersectional theory perspectives, it

becomes evident that gender and gender aligned behaviors are not static components of

personality but instead are malleable and constructed through performative acts. It also

becomes evident that race plays a significant intersecting role in how power is perceived

and experienced along gender lines. Through these theoretical perspectives comes a

recognition of the intense inequity of power that has become linked to ways of being

coded as male and female and as white and of other underrepresented racial groups.

These theorists acknowledge the immense power of the individual in constructing their

own realities, norms, and identities. If leveraged in new ways the individual could be

fundamental in dismantling their own restricted place in the world (Hammers & Brown,

2004). Both feminist and queer theorists suggest that through critiquing and dismantling

gender and gendered power there is a path that is less restricted and more inclusive.

While intersectional theorists remind us that gender does not operate outside of race

together, they call into question how power is held, what prevents women and people of

color from accessing power, and how power redistribution can happen that dismantles

rather than reinforces patriarchal norms.

By their very definition, the holders and wielders of power are leaders, whether

implicitly or explicitly labeled so. Therefore, when considering how leaders come to be

leaders and more importantly how they come to define their own leadership it is

important to consider the role of gender, race, and gender/race-aligned power, reinforced

through our patriarchal society. In the following sections, I ground myself in critiquing

power and patriarchy while attempting to answer the questions: How do leaders come to

be? Is leadership fixed or malleable? How does gender impact leadership? How does race

impact leadership? And what space is there for rupture in the development of leadership

identity specifically within the context of education?

Leadership

Defining Leadership

Leadership, though a seemingly simple concept, has been defined and described

in various ways throughout the literature. For the purpose of this research I understand

leadership as a series of acts that result in a performative construct that can be described

through attributes or traits embodied and enacted by individuals. This understanding of

leadership begins within the realm of personality development and adjusts to

acknowledge the role of social pressure in constructing an imperfect reflection of

identity. Fiske (1949) provided the foundation for labeling and categorizing displays of

behavior as broad traits that are first developed in youth and continue into adulthood.

Fiske (1949) described five traits: social adaptability, emotional control, conformity,

inquiring intellect, and confident self-expression, colloquially known as “the big five,”

that other researchers have adjusted and adapted over time (Digman, 1997; Norman,

1963). Researchers like these acknowledge that the external perception of an individual is

a series of observed behavioral actions that eventually define who a person is. What they

fail to acknowledge are the societal forces that create the conditions in which behavior

and identity are enacted.

When looking more specifically at leadership, researchers strayed from the

simplicity of the big five to seek more descriptive terminology but often fell into the

same fallacy laden thinking that leadership development occurs within a vacuum (Judge

& Bono, 2000; Judge et al., 2002). Lord and colleagues (1986), for example, identified 20

leadership qualities and six broad categories: intelligence, masculinity/femininity,

adjustment, dominance, extroversion/introversion, and conservatism. Offerman and

colleagues (1994) as another example proposed that all leaders display some combination

of sensitivity, dedication, tyranny, charisma, attractiveness, masculinity, intelligence,

and/or strength. While the list of characteristics may vary by the researcher, there is a

solid foundation to understanding leadership as the performance of a collection of

observable qualities. Regardless of how these qualities are quantified, what an individual

adopts as their leadership, and how that leadership is perceived are all impacted by the

broader societal expectations and demands. These demands create performative

leadership that looks different between individuals but that follow patterns predetermined

by social roles.

Leadership as Socially Constructed

Fairhurst and Grant (2010) describe social construction as the acknowledgment

that “taken-for-granted realities are produced from interactions between and among social

agents” (p. 174). Understanding leadership as socially constructed rather than fixed or

universal has origins in Kuhn's (1962) work on paradigms where he proposed an

understanding of the human experience that has no singular truth or innate quality but

instead is understood via interactions, experiences, and social norms.

Social interactions, reception to knowledge and learning, dialogue, and direct

intervention are all means through which individuals and their leadership are developed

and defined socially rather than intrinsically (Fairhurst & Grant, 2010). Rusch and

colleagues (1991) argued that leadership is a process of learning and adapting to feedback

created only in connection with others through narratives set within patriarchal ideology.

Grint (2005) argued not just that leaders are socially construction but also that there is

social construction of how leaders define the problems or situations in which they are

called to lead.

All of these processes by which leadership is developed occur within broader

societal norms and expectations that simultaneously encourage and prohibit the

alignment of some leaders with some displays of leadership. This social construction of

leadership aligns with queer and feminist theorists' arguments around the social

construction of gender and gender roles (Butler, 1993). It also aligns with theory around

the social construction of culture and ethnicity presentation. I argue that leadership is

constructed by internal and external expectations of gender, race, and power that have

resulted in differentiated expectations of enacted leadership based on arbitrary

differences in individual social roles.

I argue that leadership is not innate, but rather developed over time, and given

my positioning within feminist, queer, and intersectional theory I assert that leaders are

not forged through random social construction but through purposeful reinforcement of

patriarchal and racialized norms (Grint, 2005; Rusch et al., 1991).

Leadership Categories

Assuming that leadership is socially constructed allows for the critique of the

categorization of leaders found throughout the leadership literature (Komives et al., 2005;

Pearce et al., 2002). Researchers have often begun with examining the traditional

accepted form of leadership that seeks leaders who demonstrate “individualism, control,

assertiveness, and skills of advocacy and domination” (Fletcher, 2004, p. 650). These

leaders are often transactional in style relying on systems of reinforcement and reward

using expectancy and exchange theory to gain desired outcomes from their followers

(Eagly & Johannesen-Schmidt, 2001; Khan et al., 2016; Pearce et al., 2002; Robinson et

al., 2008).

Alternative leadership is then presented as a means to contrast leaders against the

traditional expectations. Servant leadership, for example, suggests a leadership identity

that is people-centered and prioritizes the leader acting in communion and service of their

followers (Dierendonck, 2011). Post-heroic leadership, similarly, shifts away from

hierarchical authoritarian attributes to focus on empathy, community, vulnerability, and

skills of inquiry and collaboration (Elliott & Stead, 2008; Fletcher, 2004).

These forms of leadership tend broadly to be more transformational in nature with

leaders focused on the higher-order needs of their followers with the intention for

purposeful change (Khan et al., 2016; Robinson et al., 2008). Transformational leaders

utilize their charisma to mediate this change and move their organizations directionally

(Eagly & Johannesen-Schmidt, 2001; Pearce et al., 2002). Embedded within all of these

categories is the implicit expectation that some forms of leadership are more effective or

more accepted than others. Regardless of how leadership is defined, displays of

leadership that align with the status quo and patriarchal white supremacy, remain

privileged.

Figure 1

Social Construction of Leadership Within Patriarchy

Note. Figure 1 shows how leadership is formed within the context of broader patriarchal

social constructions of identity.

Gendered and Racial Leadership

Gendered Leadership Identities

While the categories of leadership described above do not explicitly align gender

roles with leadership identities, the following section looks at researchers who have

intentionally coupled the two. Gendered leadership literature assumes that gender norms

and expectations affect both displays of leadership and mechanisms of leadership.

Gendered leadership has become an area of increasing focus among researchers who are

interested in the interaction between an individuals' gender identity, societal gender

expectations, and the enacting of leadership (Abele et al., 2016; Eagly et al., 2020;

Henstchel et al., 2019; S. K. Johnson et al., 2008).

Bakan (1966) was one of the first researchers to describe a binary division in

leadership along gender lines. They labeled these poles of leadership as agentic or

agency-related masculine leadership behaviors and communal or communion-related

feminine leadership behaviors (Bakan, 1966). Agentic leadership has been defined by

researchers in various ways including: strength, masculinity, tyranny (S. K. Johnson et

al., 2008), assertiveness, competence (Abele et al., 2016), boldness, ambition,

decisiveness, direction, logic-based, etc. (Henstchel et al., 2019). In contrast, communal

leadership has been defined to include: sensitivity (S. K. Johnson et al., 2008), warmth,

morality (Abele et al., 2016), affection, collaboration, helpfulness, intuition, etc.

(Henstchel et al., 2019).

Despite the variation in exact definitions the alignment between agentic and

communal leadership with male and female leaders respectively has been consistently

validated for over 70 years (Eagly et al., 2020). Though there is some nuance within the

binary, research has generally found that the fundamental dimensions of agency and

communion are stable across cultures, groups, and environments (Abele et al., 2016).

Hentschel and colleagues (2019) theorized that the stereotypes men and women hold

about themselves and others perpetuate the alignment of agentic qualities with men and

communal qualities with women. Importantly, the gendered nature of society and the

power of existing ideologies reinforce patriarchal norms and continue to differentiate

women from men, specifically with regard to their respective leadership (Elliott & Stead,

2008).

Female Leadership

Gendered leadership describes the stereotypic ways in which men and women are

proposed to lead but research has also been increasingly interested in the ways in which

women have come to find their own leadership beyond stereotype and often in spite of it.

Research has highlighted the ways in which women enter leadership with different

motivations than men. Billing and Alvesson (2000) emphasize how:

Women often have a different attitude to power compared to men. Unlike men, women

tend to see power not so much as domination and the ability to control but rather as a

capacity, and particularly as a capacity stemming from and directed towards the entire

community. (p. 147)

These different motivations result in women feeling the need to be creative in their

leadership—navigating and balancing the dual expectations of their gender roles and

positional roles. Female leaders see themselves as needing to have total commitment to

their organizations vision in order to succeed despite societal expectations for them to

also prioritize homemaking and child rearing (Folta et al., 2012).

Women have also been tasked with strategically using their stereotyped leadership

qualities to navigate leadership positions where male forms of leadership are still given

preference. For instance, Cliffe (2011) discussed the strategic use of emotional

intelligence in female school leaders noting that when emotional intelligence was not

used strategically it had detrimental effects despite broadly being viewed as a positive

leadership attribute. Research has demonstrated that women leaders are hyperaware of

the spaces in which they lead and select different ways of leading depending on the

position and demands of their leadership (Billing & Alvesson, 2000). Particularly when

navigating male dominated spaces, research has shown that women display a myriad of

differences in their primary expressions of leadership even if they generally continue to

follow a more transformational rather transactional style (Nash et al., 2017). Women in

leadership have to simultaneously be self-aware, consistent, hold reasonable expectations,

have vision, keep their values intact, act authentically, make connections, find agency,

achieve wholeness, self-reflect, network effectively, empower others, mentor and be

mentored, negotiate, communicate clearly, and be knowledgeable about context and

culture all while also conducting the day-to-day expectations of their jobs and lives

(Scheckelhoff, 2007). This near impossibly high bar is set not just by men for women but

by women for one another and for themselves.

But like all things, leadership continues to change and the ways in which women

lead continues to change. With new leadership opportunities women have had to adapt

their leadership. Some researchers suggest that leadership is changing as gender

stereotypes change. Garcia-Retamero & Lopez-Zafra (2009) argue that “a less traditional

view of women might reduce the perception of incongruity between the leadership role

and the feminine gender role” (p. 494). Additionally, there are arguments that cultural

differences might dictate differences in expectations for women broadly that may

translate to differences in expectations for women as leaders (Garcia-Retamero & Lopez-

Zafra, 2009). While there remains a consistent division of both observed and expected

leadership that differs across gender, there is also reason to acknowledge nuance in

within different leaders and the extent to which their leadership is a reflection of

expectation versus a reflection of effectiveness (Billing & Alvesson, 2000).

The Problem With Gendered Leadership

Despite progress the recreation of gendered leadership through expectations, bias,

and power remains problematic to the aims of diversifying what individuals lead and the

ways in which they lead. Gendered leadership protects white patriarchal expectations for

leaders; these “stereotypes support and maintain social hierarchies by conveying causal

attributions that justify and rationalize why certain groups occupy certain positions”

(Cundiff & Vescio, 2016, p. 126). A significant barrier to women's success in gaining

elite leadership roles is the broad gender bias of preference for male leadership based on

gender stereotype expectations (C. L. Hoyt, 2010). Because of these stereotyped

expectations leaders are selected, promoted, and evaluated based more on their gender

than on their effectiveness (Garcia-Retamero & Lopez-Zafra, 2009).

Agentic leadership is seen as more desirable for its alignment with patriarchal

expectations (Eagly et al., 2020; Pireddu et al., 2021). However, not all agentic leaders

are perceived the same. Men—by being seen as congruent with agentic leadership—

experience a variety of privileges including disproportionate access, higher approval,

higher wages, increased hireability, and higher ratings of effectiveness (Eagly & Karau,

2002; Van Vugt & Spisok, 2008). In contrast when women display agentic leadership

they are perceived as being incongruent with their gender presentation and face

consequences including being seen as inauthentic, being devalued, being given less valid

feedback, and being rated less effective than their male counterparts regardless of

produced outcomes (Eagly, 2005; Eagly & Karau, 2002; Elliott & Stead, 2008; C. L.

Hoyt, 2010; Jampol & Zayas, 2021). Congruence and incongruence do not just relate to

the presumed gender of a leader but on all the ways leaders align with or counter the

male, white, able-bodied, heterosexual norms of white supremacy.

Racialized Leadership Identities

There is less literature that outlines specifically racialized leadership identities;

however, there is substantial evidence that how the leadership of an individual is

perceived is mediated by their race (Chung-Herrera & Lankau, 2005; Rosette et al.,

2008). Broadly, whiteness is still centered in leadership norms. Not only is being white

inherently seen as being more of leader, but there are also traits most commonly

endorsed and replicated in white individuals that have become inextricably linked to

traditional leadership (Chung-Herrera & Lankau, 2005). Perhaps the best example of

racialized leadership is within the binary of individualistic and collectivist leadership

(Chin et al., 2007; Wendt et al., 2009). While not universally true, research indicates that

collectivist leadership appears more often in the leadership style of people of color while

individualistic leadership appears more often in white leaders (Campion & Wang, 2019;

Dantley, 2003; Lomotey, 2019; Yammarino et al., 2012). Collectivist leaders, according

to Oyserman and colleagues (2002), “define their needs more in terms of relations within

groups and sacrifice personal needs for the sake of the group” while individualistic

leaders centralize “personal goals, personal uniqueness, and personal control” (p. 5).

Despite the lack of clear contrasted pairing of leadership across individuals of different

races there is clear evidence that racial presentation makes a difference in not just

leadership style but in leadership opportunities.

The Problem With Racialized Leadership

Leadership being tied to race acts as a barrier to understanding nuanced

leadership that is reflective of needs rather than reflective of people (Chin, 2013). When

assumptions are made about leaders based on their race we prevent diversity in

leadership roles (Gooden & Dantley, 2012). Similarly, when people are expected to

follow racial norms around how they lead we run into similar issues of incongruence

seen in gendered leadership (Begeny et al., 2021; Ziegert & Hanges, 2005). So long as

we operate within a patriarchal white supremacy, leadership values, actions, and traits

aligned with whiteness will be privileged and will lead to racist discrimination in access

to leadership and ways of leading (Dovidio & Gaertner, 2000; M. K. Ryan et al., 2020).

Gooden (2012) warns that the aim is not colorblind leadership because until we have

dismantled racism in society as a whole we cannot presume to eliminate racism from

leadership interpretation and evaluation. Instead, Gooden (2012) argues for critical

examination of both self and others through a recognition of implicit expectations around

race and leadership.

Intersectionality in Leadership

Gender and race are intertwined in significant ways that affect leadership.

Sanchez-Hucles and Davis (2010) found that the barriers women face including gendered

presumptions of leadership, stereotypes and stereotype threat, lack of existing

representation, and discrimination are exponentially more impactful for women of color.

Women of color face a double bind where they experience barriers for not just their

gender but also their race (Malcom & Malcom, 2011). Women of color face both sexist

and racist expectations that lead to further restrictions on accepted displays of leadership

(Eagly, 2007). For instance, Black women whose leadership includes authoritative or

dominant traits elicit a more significant negative reaction than their white female

counterparts displaying the same behaviors (Sanchez-Hucles & Davis, 2010). There is

also evidence that these leaders often hold values that are inconsistent with the dominant

ideology which in turn can limit their relational authenticity with their followers (Eagly,

2005).

While intersectionality, as it specifically relates to gender-aligned leadership,

remains severely under-researched, intersectionality more broadly acknowledges that

women of color, women with disabilities, and queer women all face additional barriers to

their success and access to domains traditionally held by White, straight, able-bodied

men. That is not to discount the important work they do in providing alternatives ways of

leading and being when compared to white, male forms of leadership. Indeed, these

leaders provide important paths forward and demonstrate success not just despite but

because of who they are (Holden & Bruce, 2022; Jang & Alexander, 2022; Lomotey &

Lowery, 2013; Manongsong & Ghosh, 2023; Peaters & Miles, 2021). Intersectionality

and the empowerment of not just women but women of color, men of color, and those

with disabilities runs counter to the patriarchal ideologies that continue to hierarchically

prioritize those who are most representative of white supremacy. Ultimately, both

identity formation and leadership formation are as racialized as they are gendered, and

the interactions between race and gender are highly complex; it is beyond the scope of

this dissertation to fully describe that complexity.

Figure 2

Social Construction of Gendered & Racialized Leadership

Note. Figure 2 indicates that gendered and racialized leadership is contrived from the

broader social construction of patriarchal gender norms and the social construction of

leadership.

Developing Gendered Leadership

Leadership Identity Development Model

Understanding how to decouple patriarchy from leadership requires understanding

how displays of leadership are influenced externally during an individual's broader

identity development. One means through which to understand leadership development is

Komives and colleagues' (2005, 2006) Leadership Identity Development model. This

model focuses on student leadership development making it a particularly useful

contextual piece for this research. Komives and colleagues (2005, 2006) argue for five

categories of influence on student leadership identity formation: “broadening view of

leadership, developing self, group influences, developmental influences, and the

changing view of self with others” (Komives et al., 2006, p. 403). The model also argues

for six stages of identity formation: awareness, exploration/engagement, leader

identified, leadership differentiated, generativity, and integration/synthesis (Komives et

al., 2005, 2006). The Leadership Identity Development model has been substantiated in

further research (Wagner, 2011). The model has been found to hold true with female

leaders (McKenzie, 2018) as well as across the experiences of LGBTQ+ students (Renn

& Bilodeau, 2006). It has also been used to evaluate the effectiveness of structured

interdisciplinary instruction on leadership formation (Sorensen et al., 2016).

For the purpose of this study I am primarily focused on the developmental

influences component of the model and specifically the role of teachers as influencers of

student development. Komives and colleagues (2005) define developmental influences as

the “supports in the environmental context in which the development of leadership

identity was occurring” (p. 608). Developmental influences include adult influences, peer

influences, meaningful involvement, and reflective learning (Komives et al., 2005). In the

model, developmental influences surround the other categories of influence, and Komives

and colleagues (2006) argued that adults in particular “were integral in making these

students feel special, sponsoring them into interesting challenges and opportunities, and

serving as role models'' in the early stages of identity development (p. 406). The domain

of developmental influences has also specifically been identified in further research as

having a significant impact on leadership development when considering adults in

students' lives as mentoring figures (Enlow, 2014; Priest et al., 2018). Developmental

influences were also shown to be impactful regardless of race (Acosta & Guthrie, 2021)

or gender (McKenzie, 2018).

Connecting leadership identity development to the broader issues of inequity, my

research is interested in the process by which teachers, specifically, advertently or

inadvertently perpetuate patriarchal expectations in their students' leadership formation.

Because of this narrow area of focus, I acknowledge that this research presents just one

of many facets of potential influence for students and mostly ignores others including the

influence of family norms, community significance, and individual reflection (Komives

et al., 2005, 2006; Wagner, 2011). However, this is an important angle of research

because of significant amount of time teachers spend with youth and the mass messaging

that schools can provide.

Student Leadership

Before addressing mechanisms through which educators may act as

developmental influences on students and particularly on the formation of student

leadership, I want to explore briefly the existing literature on student leadership and

gendered and racialized student leadership. Much of this research focuses on the

outcomes and displays of student leaders which continues to uphold the stereotypes of

female student leaders being more “collaborative, transformational, or democratic” and

male student leaders being more “hierarchical, transactional, or authoritative” (Haber-

Curran & Tillapaugh, 2017, p. 14). Similarly racialized leadership literature indicates

white leaders tend to be solution and progress oriented while leaders of color tend to be

collectivist oriented (Gooden, 2012). Most research on student leadership that has

examined the role of race has focused on college aged leadership (Dugan et al., 2012;

Whitney & Collins, 2021). Research has often focused on how to nurture student leaders

through various specialized programs and teaching practices (S. Gordon, 1994; Hine,

2014; Rosch et al., 2015). Coffey and Lavery (2018), for example, argued for teachers

who model leadership in the classroom, implement formal and informal structures for

student leadership, and focus on holistic student development. Research into student

leadership overwhelmingly remains adult centered with assumptions about what qualifies

as leadership and how to develop leadership remaining solidly within the control of

adults (Haber-Curran & Tillapaugh, 2017; Whitehead, 2009).

Several researchers have acknowledged the absence of student and youth voice in

how they conceptualize their own leadership (Archard, 2013b; Dempster et al., 2010;

Whitehead, 2009). When research has centered youth, the youth tend to describe

leadership that leans towards collective action—behavior that by adulthood has been

coded as antithetical to traditional hierarchical leadership (Dempster et al., 2010).

Researchers centering student voice have also found that students are increasingly

accepting of less traditional understandings of leadership and hold less rigid connections

between gender roles and leadership roles than adults (Mortensen et al., 2014). Indeed,

youth describe leadership as accessible to all, focused around change, collective in

nature, and requiring both individual and shared knowledge (Mortensen et al., 2014).

Despite youth generally leaning towards an interpretation of leadership that is

more inclusive of those who do not fit the traditional white male mold, research has

shown that many students of color, women, and LGBTQ+ students continue to

disassociate “with the term leader and [resist] the historical and societal connotations of

who a leader is or should be” (Haber-Curran & Tillapaugh, 2017, p. 13). Just as with

adult leadership, student and youth leadership is shaped by the intersection of identities

including the students' race, gender, and sexual orientation (Tillapaugh et al., 2017). Yet

research has shown that young women and young people of color can and do see

themselves as leaders. Archard's (2013b) study found that young women see their

attributes and qualities as aligning with leadership despite pushback from adults.

Similarly, students of color have been shown to position themselves towards leadership

even when these attempts at leadership are thwarted by the adults within their schools

(Bertrand, 2018).

Teacher Expectancy

To return to the developmental influence of teachers, specifically, on students'

leadership development, I turn to research on teacher expectancy. Research on the role of

teachers in shaping students has been extensive and consistently demonstrates that

differentiated expectations result in differentiated students. Research into how teacher

expectancy, Pygmalion effects, and self-fulfilling prophecies impact students have been

a mainstay of educational research since the 1970s. Some of the earliest research in this

domain comes from Rosenthal and Jacobson (1968) who published research on the

existence of a Pygmalion effect in classrooms, arguing that how teachers treated students

had a significant impact on how students performed. Brophy and Good (1970) first

substantiated a model of teacher expectancy which was then adapted by Jussim (1986)

and again by O. Johnson and colleagues (2019). A recent and simplified version of the

model includes four stages: “teacher expectations on students, teachers' development of

expectations, teachers' differential treatment of students, and students' reactions to

teacher expectations” (O. Johnson et al., 2019, p. 44). While not all researchers are fully

convinced that these models explain the differentiated outcomes, Wineburg (1987) for

instance argues these models are an oversimplification of the process, there is broad

acceptance that “students can change as a result of how their teachers see them” (O.

Johnson et al., 2019, p. 46). Widely, teacher expectancy research suggests that teachers

have the ability to alter the way students perform be that academically, behaviorally, or,

I argue, in learning to lead.

Mediating influences on teacher expectancy. The characteristics of students

including their gender, race, and relationship with their teacher are all factors that affect

differentiated expectations that interfere in the process of student leadership formation. It

has been widely theorized that these expectations are based on both stereotypes and

existing information that teachers have about their students steeped in systemic racism

and sexism (Fletcher, 2004; Good, 1987; C. L. Hoyt, 2010; S. K. Johnson et al., 2008;

Osad'an, 2012). While teachers often claim neutrality in their expectations they operate

within a patriarchal society of bias and stereotypes and intentionally or not are often

themselves a reflection of those biases and stereotypes (Brookfield, 2005; Freire, 2000;

hooks, 2010).

Student gender. Students' gender, teachers' gender, and patriarchal expectations

around gender conformity all come into play when considering the differing effects of

teacher expectancy on how students perform. Osad'an (2012) looked at how teachers

report treating students differently based on their gender presentation and found that

14.5% of teachers admitted to a different approach to children depending on their gender,

47.5 % thought that girls are more submissive than boys, and 50% agreed that boys are

more individual than girls. Somquist (2007) reviewed over a decade of research on

teacher expectancy as it relates to gender socialization and found that teacher behavior

and instructional practices reinforced differences between students with different gender

presentations. In Garrahy's (2001) research, teachers reported taking a “gender-blind”

approach to instruction, and yet findings suggested that teachers often allowed and even

encouraged different behavior for students based on their presenting gender.

The differentiated treatment by teachers has repeatedly been found to result in

differentiated outcomes in students. Differential treatment of boys has been found to

relate to higher levels of violence, more opportunities to explore ideas, harsher

interventions for disruptive behavior, and lower reading abilities (Glock, 2016;

Retelsdorf et al., 2015; Watson, 2007, C. S. Marshall & Reinhartz, 1997). Conversely,

differential treatment of girls has led to lower math ability, fewer opportunities to be

animated, and a lower likelihood of pursuing science-related fields (Brotman & Moore,

2008; C. S. Marshall & Reinhartz, 1997; McKnown & Weinstein, 2002). Similarly,

research has found a significant impact on students' self-expression when teachers

reinforce gender binary characteristics among LGBTQ+ students (DePalma, 2013). These

gender divisions in teacher expectations, rooted in gender stereotypes, reinforce

questionable binary definitions of how students with different gender presentations

should act and learn.

Student race. Similar to gender, and often in tandem with it, student race and

ethnicity have been found to have a significant mediating impact on teacher expectancy

and student outcomes. Several researchers have found that underestimates of ability had

a more detrimental effect and overestimates of ability have a less beneficial effect on

African American students' academic gains than on their white peers (Marcus et al.,

1991; McKnown and Weinstein, 2002). There has also been extensive research on

disproportionate discipline and teacher reaction to student behavior in the context of

students' race. Students of color, specifically African American students, are more likely

to be suspended, expelled, and referred to the office than their white peers (Carter et al.,

2014; Girvan et al., 2016; Smolkowski et al., 2016). Within the classroom teachers are

more likely to intervene and redirect disruptive behavior from students of color than

white students (Fenning & Rose, 2007; Glock, 2016). Researchers have found that these

differentiated expectations, grounded in reliance on stereotypes and pre-existing schema,

result in a variety of negative consequences including stigma, isolation, higher risk of

dropout, and lower self-worth (Osher et al., 2003; Skiba et al., 2011; Vallas, 2009).

Relational power. Several researchers have also looked at the power of

student/teacher relationships as a mediator for the effects of teacher expectancy. One

aspect of these relationships is the closeness felt by students towards their teachers.

McFarland and colleagues (2016) found that closeness predicted positive self-concept in

girls but not boys and conflict predicted negative self-concept in both boys and girls. The

relationship between teachers and students is also affected by students' feelings about

teacher bias. C. S. Brown and Bigler (2004) found that female students were significantly

more likely to see their teachers as discriminating against other girls which deteriorated

trust between students and teachers. These feelings of discrimination can affect student

performance and self-image (Jacobes, 2006).

Looking at specific student characteristics, like gender performance, highlights

the strength of relational power in forming student identity. A case study by Reay (2001)

found that girls in particular are inclined to perform gender based on the input from their

surroundings and to build an identity that they feel is acceptable even if it is non-

preferred. Similarly, Castellanos and colleagues (2019) explored the pressure students

felt to conform to gender norms in classrooms and found that that pressure was a product

of both adult and peer interaction (Ringrose & Renold, 2010). These expectations,

reinforced in relation with others, come to be internalized by students beyond the

confines of their classrooms (Good, 1987).

Teacher Expectancy and Mediating Factors

Figure 3

Note. Figure 3 shows the interactions between student race and gender, teacher reactions

and behavior, and student perception and responses, in mediating teacher expectancy

within the context of broader patriarchal social constructions of gendered and racialized

leadership.

Looking Back and Looking Ahead

In this literature review I began by exploring how patriarchal norms rooted in

white supremacy uphold gender and racial norms that have been purposefully linked to

leadership. Guided by feminist, queer, and intersectional theorists I assert that gender and

racial differences that relate to unequal power and access are intentionally created and

can be intentionally dismantled. What individuals perceive to be innate differences

between people are instead a reflection of the witnessing and participation in pervasive

performative gendered and racial norms. I then examined leadership as a one such

performative act. Gendered leadership presented a binary interpretation of leadership that

is stereotypically feminine and masculine, communal and agentic respectively. Racial

Teacher Expectancy and Mediating Factors

leadership research presented similar dichotomies though with less clear binaries. I

argued that these understandings of leadership are rooted in problematic assumptions of

who should or should not hold power and reflect in unequal leadership broadly seen in

society. I then explored how leadership is developed during youth. I focused on the role

of teacher expectancy as a means for shaping leadership in students. Within teacher

expectancy I highlighted the role of student gender, race, and student-teacher

relationships as mediating factors. The figure above shows the addition of teacher

expectancy and the relational role between expectations, perception, and presentation that

shape students within the greater context of societal norms and leadership assumptions.

Ultimately, my conceptual framework that follows brings together all of these concepts to

examine the problem of how self-conceptions of leadership are developed in youth and

encouraged in ways that unnecessarily perpetuate problematic patriarchal norms.

Conceptual Framework

Connecting Expectancy With Leadership Formation and Performance

Given the strong potential influence of teachers, it is important to understand the

extent to which the formation of student leadership may be vulnerable to differentiated

expectations. There has been increasing focus on the potential effects of teacher

expectancy on student self-concept, self-efficacy, and self-belief (O. Johnson et al.,

2019). Jussim (1986) highlighted the impact of these self-fulfilling prophecies on student

self-schemas and self-esteem. While Jacobs (2006) found that teachers played a

significant role in students' self-perceptions of themselves as “smart” or capable.

Whether leadership is understood through the lens of identity development (Berzonsky,

1989; Erikson, 1950) or socially constructed gendered expectations (Berkowitz et al.,

2010; Campbell & Wasco, 2000; Gedro & Mizzi, 2014) both point to leadership as a

malleable concept.

Importantly, the process of figuring out how one is going to engage in and with

the world has been shown to occur well into adulthood, suggesting that those seeking to

support and nurture an individual's leadership have many years of influence (Cieciuch &

Topolewska, 2017; Digman, 1997; Srivastave et al., 2003). I argue that leadership is the

cyclical process of internalizing the external expectations and norms that result in enacted

leadership that feels innate but is far from it (Crocetti et al., 2008; Marcia, 1966). Given

the strong potential influence of teachers through differentiated expectations—

classrooms become an increasingly important space for nurturing leadership identity.

Framework for Student Leadership Identity Formation

Despite the far-reaching influence of the patriarchy and its ideology in dictating

leaders, there is hope for disruption. By questioning the coupling of power with gender,

and gender with the binary, feminist and queer theorists point to opportunities for

change. Archard (2013a) argued that the question is not whether women are capable of

administering successful leadership but rather how to facilitate their acquisition of these

roles. By interrupting expectations, providing girls with opportunities to practice

leadership, modeling leadership that is inclusive and nontraditional, and exposing

students to a variety of leaders, teachers can develop female students and students of

color for future leadership (Archard, 2013a; Hannah et al., 2009; M. A. Hoyt & Kennedy,

2008; Elliot & Stead, 2008). The social construction of leadership occurs in various

settings and within various dynamics. For the purpose of this research, I highlight the

schools as a space for leadership construction and reinforcement but acknowledge that

schools do not operate in vacuums. Within the framework this construction remains

broad to account for the many interlocking systems of co-creation and recreation. The

social construction of leadership that does occur in schools, is influenced by teachers,

and is impacted by students' perceptions and characteristics; at any of these points, there

is room to reexamine what messages students receive and internalize about who they are

meant to be and the power they are meant to hold. Changing who sees themselves as

leaders and what types of leadership they can embody is an important step to dethroning

the patriarchal grip on leadership and diversifying who leads.

Figure 4

Conceptual Framework

Note. Figure 4 presents a conceptual framework that proposes student self-conception of

gender-congruent leadership occurs through the process of teacher expectancy within the

context of socially constructed gendered and racialized leadership perpetuated by

patriarchal norms and control.

CHAPTER III

METHODOLOGY

Sequential Exploratory Mixed Methods

Given the complexity of leadership and the lack of youth voice in understanding

the conception of young leaders the gathering of both qualitative and quantitative data

seemed necessary. Mixed methods allowed for a well-rounded understanding of the

research problem including going deeper and wider in understanding the experiences of

an often-overlooked population in gendered leadership literature—K–12 students

(Galliott & Linda, 2014). Broadly speaking mixed methods research is characterized by

the use of both quantitative and qualitative approaches in a single research project to

gather or analyze data (Cameron, 2009; Creswell & Plano Clark, 2017). There are several

different types of mixed methods research. The type used in this research was sequential,

meaning one type of data provided a basis for the collection of another type of data and

exploratory meaning that qualitative data were the data collected first followed by the

collection of quantitative data (Cameron, 2009).

Revisiting Research Questions

The qualitative portion of this research centered around the question: What

vernacular do children and adolescents use to conceptualize their own displays leadership

and the external influences they experience while learning to lead? Because of the lack of

language directly from youth in the leadership literature, the collection and consolidation

of vernacular allowed for data that could be immediately explored in quantitative form.

While the potential questions surrounding youth leadership are endless this concrete

focus of both vernacular and experience provided the beginnings of a new framework for

understanding what leadership means to children and adolescence.

The quantitative portion of this research was divided into two distinct sections the

first centering around the questions: How, if at all, do students' ratings of their own

leadership attributes differ across grade (third, seventh, & 11th), race (white & Hispanic),

and gender (male & female)? And how, if at all, do these differences align with or

counter “communal” and “agentic” leadership expectations for female and male students

respectively? By having participants rate various characteristics of their own leadership

comparisons about how leadership is understood by different demographics could be

explored. The literature suggests that by adulthood different groupings of individuals

present differing forms of leadership (Eagly, 2007; Martin et al., 2017; Wolniak et al.,

2023). These questions aimed to explore if those eventual presentations stem from early

self-identification.

The second quantitative section sought to answer the questions: How, if at all, do

students perceive/rate teachers responding differently to leadership behaviors that are

coded as gender congruent and gender incongruent? And how, if at all, do these

perceptions differ by student grade (third, seventh, & 11th), race (white & Hispanic),

gender (male & female), and students' reported feelings of closeness with their

teacher(s)? The literature suggests that teachers may respond differently to male and

female students based on their gender presentation and their congruence in behavior with

socially generated gender norms (Collins, 2009; Eder & Parker, 1987; Kunjumon, 2012).

Similarly, the literature suggests similar bias around congruence with cultural norms

(Kohli et al., 2017; Quinn & Stewart, 2019). This portion of the research aimed to

understand if youth perceived any differences in how they were encouraged or

discouraged in the leadership and furthermore to explore how that encouragement or

discouragement followed constructed norms around race and gender.

Setting

The setting for this research was a small rural school district located in a city of

roughly 3,500 people in the Pacific Northwest. The school district population according

to state records from 2021–2022 had a total enrollment of 628 students spanning

kindergarten through 12th grade. The student demographics were as follows: 46%

Hispanic/Latino, 48% White, 4% Multiracial, 1% Native Hawaiian/Pacific Islander, and

1% Asian. The student demographics mirrored the city's demographics.

Positionality

It is important to note my role in the district where this research took place. I had

been the school psychologist in this district since the fall of 2016. Many of the

participants in this research were familiar with me through my work in the district. While

I think my insider position supported both students' comfort in working with me and

supported my ability to actively use the results of this study for the benefit of this

community, I recognize that it did provide some potential risks. When participant

selection was occurring, I preemptively eliminated participants whom I have evaluated

for special education from being eligible for the interview portion of the research to

avoid any potential coercion that maybe have occurred due to the power dynamics of my

role.

Throughout the research I acknowledged my multiple roles both in consent forms and in

direct interactions with participants. Despite my attempts to mitigate any undo impact

from my insider position in the district there was undoubtably an effect that will be

further discussed in Chapter V.

Methods and Methodology

Because of the sequential exploratory mixed methods design there were two

distinct phases of data collection happened independently. The qualitative component

occurred first followed by a period of data analysis and synthesis and then the

quantitative component occurred followed by another period of data analysis and

synthesis. The qualitative portion involved interviews with a total of 12 participants.

Interviews were selected because the aim of the qualitative portion was to get individual

student voice and ensure clear delineation of what vernacular was being used by different

demographics. A focus group would have created too much opportunity for groupthink

(Boateng, 2012; Smithson, 2000). Written responses would have limited accessibility for

all participants—particularly the third-grade participants (Kirk, 2007). Ultimately, semi-

structured interviews were chosen as they allow dialogue to “meander around the topics

on the agenda—rather than adhering… to verbatim questions as in a standardized survey

—and may delve into totally unforeseen issues” (W. C. Adams, 2015, p. 493).

Likert scales were determined to be the best method for quantitative data collection due

to their accessibility and easy analysis potential without exhausting participants with

limited stamina (Waples et al., 2010; Willits et al., 2016).

Participant Selection

Initially the aim was to recruit participants from Grades 3, 7, and 11 for both the

qualitative and quantitative portions of the research. However, due to a lack of interest

and consent from 11th-grade participants for the quantitative portion, the Likert scales

were also given to 12th-grade participants. Rather than focusing on an individual

developmental level, the choice to collect data from different developmental stages

provided an interesting and important glimpse at how leadership forms alongside other

developmental milestones (Erikson, 1959; Sokol, 2009). Data collection also centered

around representation from female and male participants as well as from Hispanic/Latino

and white participants. The choice to examine the two majority genders and the two

majority races represented in the broader school population was intentional and stems

from existing research into how race and gender effect leadership. The literature

highlights how gender and race have a significant impact on the types of leadership

individuals embody and the social norms that are recreated when these gender and

racially aligned leadership styles are unexamined (Abele et al., 2016; Eagly et al., 2020;

Henstchel et al., 2019; S. K. Johnson et al., 2008; Malcom & Malcom, 2011; Sanchez-

Hucles & Davis, 2010).

Individual Interviews

For the interviews general qualitative research practices were used to gather the

vernacular children and adolescents at various ages use when describing their

experiences as leaders. Twelve interviews were conducted with participants who were

randomly selected based on the selection criteria and demographic information available

through

student information systems: one white female, one Hispanic/Latino female, one white

male, and one Hispanic/Latino male at each grade level—third, seventh, and 11th.

Consent and assent was successfully gained from all the initially selected third- and

seventh-grade participants. However, the 11th graders proved to be much more difficult.

While initially all parents consented, the first three white males who were asked to

participate declined. The first Hispanic/Latino male and Hispanic/Latino female also

declined. The participant selected as the third-grade Hispanic/Latino female candidate

was a newcomer student who only spoke Spanish. Because her perspective was still

important and relevant to the research question accommodations were made including the

use of a translator during the interview.

In total, 14 questions were posed in a semi-structured interview (Appendix A)

where “questions are pre-planned prior to the interview but the interviewer gives the

interviewee the chance to elaborate and explain particular issues through the use of open-

ended questions” (Alsaawi, 2014, p. 151). The interviews began with broad questions

about leadership, followed by participant specific leadership questions, and teacher

response to leadership questions. It ended with longitudinal leadership questions.

Potter and Hepburn (2005) remind researchers to be cautious about their use of

interviews noting several potential problems—specifically the positionality of the

interviewer and the content of the interviews. Conducting interviews with minor

participants required an acknowledgement of some power asymmetries that are

unavoidable. However, Lohmeyer (2020) suggests that by centering the interviewee and

their story and focusing on developing a positive relationship with participants rather than

just the aims of the researcher, interviews can provide fruitful information that benefits

both interviewer and interviewee. Schelbe and colleagues (2015) reinforced that there are

some specific strategies that need to be used when “managing complexities associated

with youth participation in research, to gain the benefits of first-person perspectives

directly from youth” (p. 505). Most importantly when completing interviews with minors

Schelbe and colleagues (2015) note the importance of prioritizing safety and being

adaptable to the varying needs of the participants.

During the interviews, which lasted 14 minutes on average, participants engaged

in questions about the leadership and provided insight, anecdote, and questions that

fulfilled the research question. The aim of this phase was not to reach theoretical

saturation, defined as the point in which no new concepts emerge (Strauss & Corbin,

1998). Instead the aim was to gain enough input to develop a quantitative tool using the

language and concepts shared by participants and to have sufficient experiential

knowledge to converge or diverge from the subsequent quantitative results (O'Cathain et

al., 2010).

Vernacular Coding

Interviews were recorded and then transcribed for ease of coding. Coding is

broadly defined as the taking of data in the form of words, expressions, sentences, and/or

group of sentences and relating them to topics or themes (G. W. Ryan & Bernard, 2003).

Coding began with review of the entirety of the interviews looking for themes. While the

research questions presuppose some connection between language used by participants

and the existing literature around leadership the interviews were not coded by looking

specifically for those concepts. Instead, a comprehensive list of attributes was collected,

narrowed down, and settled upon before turning to the literature. Only after the final

concepts that would become items for phase two of the research were selected were items

analyzed for their connection to agentic or communal categories as described by the

literature (Abel et al., 2016; Hentschel et al., 2019). G. W. Ryan and Bernard (2003)

warned that “by examining the data from a more theoretical perspective, researchers must

be careful not to find only what they are looking for” (p. 94). In an effort to find the most

organic and representative vernacular careful care was after codes were mostly chosen to

review their originating source—this is fully explained in Chapter IV. I also consistently

allowed for data that did not neatly fit into existing language around leadership. Most

notably concepts like “following” were included despite being antithetical to traditional

understandings of leadership. In total four rounds of coding, recoding, and narrowing

were done before the final 24 items were selected.

Experience Coding

In addition to gathering and consolidating the vernacular used by youth the

research question also addresses the experiences that participants had while learning to

lead. To analyze these data the transcripts of the interviews were reviewed and themes

were pulled. The themes ultimately became positive internal and externa experiences,

negative internal and external experiences, direct support or criticism, and directionality

of leadership over time. Once the themes were established the interviews were reviewed

once again and instances of these themes were tallied. Trends were then explored to look

ratios of positives to negative by demographic, and ratio of support to criticism by

demographic.

Likert Scales

Language provided by participants in the interviews was utilized to form the items

on the two Likert scales for the quantitative phase of this research. Consent to participate

in the surveys was sought from the remaining students in Grades 3, 7, and 11. While

sufficient responses were received from Grades 3 and 7, only seven students from 11th

grade responded. After several attempts were made, the decision was made with IRB

approval to open the survey to also include 12th-grade students, the majority of whom

were 18 years old and could consent for themselves. In the end the survey was completed

by 24 12th graders, 7 11th graders, 19 7th graders, and 31 3rd graders.

The survey was a two-part Likert scale. These scales were given to students

during the same academic year as the original interviews. They took approximately 15

minutes for participants to complete and were be scaffolded to the extent necessary given

the developmental levels of the participants. Third-grade participants were taken in small

groups and the questions were read aloud to ensure that reading barriers did not impact

their ability to complete the surveys.

The first Likert scale listed 24 items and had participants rate the extent to which

they identified as having that attribute when they were acting as leaders. The second

Likert scale had the same 24 items in a different, randomized, order and had participants

rate the extent to which their teachers positively or negatively responded to them

displaying those attributes when they were leading. Demographic information was also

collected—specifically, the participants' gender identity and racial identity. The literature

debates the optimal number of response options for a Likert scale. The traditional scale

has five possible responses, but, there is some argument for even numbered Likert scales

that force the respondent to commit to a certain position (J. D. Brown, 2000). Ultimately,

I used a four-point scale with the following descriptors: never, sometimes, often, always.

The four-point scale forced a choice and avoided the pitfall of a middle ground of

indecision (Chyung et al., 2017).

Statistical Analyses

The first necessity was to understand if any of the questions on either Likert scale

grouped together to create dependent variables which could be compared with my three

independent variables. The independent variables, all nominal categorical, included

participant gender, participant grade, and participant race. Ultimately a factor analysis

indicated that the first Likert scale resulted in four reliable factors and the second

resulted in three reliable factors. These factors became the dependent variables through

which the majority of the statistical analysis was performed. MANOVA's were run to

look at the effects of one or more independent variable (gender, race, grade) on the

responses to those factors. Because of the assumptions formed through the literature and

the knowledge that the items roughly fell into three groups based on the literature:

agentic, communal, and ambiguous, these groups were also run for reliability and then

subsequently MANOVAs were run to look for affects from the independent variables.

Item level analysis also occurred looking for differences between responses to specific

questions that, based on literature, were likely to be affected by race, gender, and/or grade

level. These select items were analyzed through ANOVAs. Finally, analysis of overall

trends in highest- and lowest-rated items were examined across the two scales.

Because of the sheer number of analyses being done some precautions were put in

place to reduce the change of false positives. For all of the relevant statistical analysis the

Bonferroni correction was completed to adjust the p-value that would indicate

significance (Croasmum and Ostrom, 2011). Additionally, limits were put on how many

overall statistical analyses were run. While there is temptation to continue to search for

results the most robust analyses, the MANOVA, led the decision making for whether or

not significance was really present.

Additional Analysis

One of the significant benefits of mixed methods research is the ability to connect

the quantitative and qualitative data to paint a more robust understanding of experiences

and trends. As O'Cathain and colleagues (2010) noted without integrating the results

from both phases the “knowledge yield is equivalent to that from a qualitative study and a

quantitative study undertaken independently” (p. 1147). Using the process of

triangulation I looked for data from both phases of the study for points of convergence,

complement, and/or dissonance, where findings agree, overlap, or contradict (O'Cathain

et al., 2010). These points of convergence, complement, and dissonance guided me

towards areas of future potential research.

By centering the experiences of youth in understanding youth leadership this

research adds not just to an underdeveloped corner of gendered leadership research but

also supports the educators in the school district of study (and similar districts) in

nurturing and cultivating student leadership in more effective and inclusive ways.

Utilizing the wisdom of children this research provides backing for reframing and

reexamining the ways teachers encourage or discourage leadership in classrooms. Only

by asking students can we understand the potential significance of school as a space for

leadership development that has lasting consequences well into adulthood.

CHAPTER IV

FINDINGS

This chapter begins with exploring the interviews conducted through a

quantitative lens looking specifically at gathering representative vernacular. It then

explores data related to the quantitative question—how do students rate their leadership

traits and how groups of students rate groups of traits differently. I then take the same

approach to looking at how students rate traits in terms of desirability from teachers. I

also explore overall trends in what participants view as most and least self-ascribed and

teacher-desired. I end the chapter looking qualitatively at themes found in the interviews

that go beyond just the vernacular.

Quantitative Approaches to Interviews

I began this research by conducting interviews that explored the qualitative

question: what vernacular do participants use to conceptualize their own leadership? I

gathered this information over the course of 12 individual interviews with participants

representing a variety of demographics including: female and males, Hispanic/Latino and

white, and participants from elementary, middle, and high school.

The Vernacular Used

I coded all 12 interviews for common themes. The codes fell into five broad

categories: traits, reactions/perceptions, roles, actions toward others, and self-

experiences. There were a total of 380 words or brief ideas coded into these five

categories. Actions toward others had the most overall codes (n=156), followed by traits

(n=82), reactions/perceptions (n=60), self-experiences (n=50), and roles had the least

overall (n=32). Actions toward others included codes such as “give directions”, “help

people”, and “stand up for others”. Traits included things like “calm”, “respectful”, and

“smart”. Reactions/perceptions included things like “follow”, “they feel good”, and

“listen to me”. Self-experiences included things like “know what to say” and

“comfortable”. Finally, roles included things like “mom”, “line leader” and “captain”.

The total number of codes attributed by grade level increased with age; 11th

graders were responsible for 175 codes, seventh graders for 115 codes, and third graders

were responsible for 90 codes. Male participants represented slightly more codes than

female participants, 198 to 182 respectively. White participants were responsible for 239

codes compared to 141 from Hispanic/Latino participants.

The division in total number of minutes of interviewing was relatively equal

among the three grade levels; 11th-grade participants accounted for 35% of the total

minutes, seventh grade for 33%, and third grade for 32%. Female participants talked for

more time than male participants with female participants accounting for 53% of the total

minutes compared to the 46% by males. The biggest difference, however, was between

white participants and Hispanic/Latino participants where the percentages were 58 and 42

respectively.

Each group of codes was then explored for further grouping. Action towards

others went from 156 individual codes to a potential 21 broad codes, traits went from 82

individual codes to 16 broad codes, reactions/perceptions went from 60 individual codes

to 12 broad codes, and self-experience went from 50 individual codes to 9 broad codes.

Examples of some of the consolidation can be seen in Figure 5 and the complete

consolidation process can be found in Appendix B. I decided moving forward to

eliminate the codes for roles because they did not fit in with rate-able vernacular for self-

ascription; however, they did provide interesting insight into participant understanding of

leadership, which will be further discussed in Chapter V.

Figure 5

Trait Consolidation

I then reduced the 58 codes once more to 24 by eliminating repeated codes and

combining common elements between codes. I aimed to have a manageable number of

codes to use for the quantitative surveys that would not lead to participant fatigue (Lietz,

2010; Pratap et al., 2010). I reviewed these 24 codes with the original 380 and the

literature and determined that they the best overall representation of the vernacular used

by participants in the interviews. These 24 codes became the items for both surveys used

in the quantitative portion of the research; on the surveys many of the codes had multiple

related components or vernacular that made it accessible to participants from various

developmental levels. The 24 items were distributed between the four original codes with

the majority originating as traits or actions toward others (See Figure 6). Ten of the 24

fell into more than one category.

Figure 6

Originating Category for Final 24 Codes

PROPORTION OF CODES

Traits Actions Toward Others

Reactions/Perceptions Self-Experience

11%

12%

43%

34%

For the remainder of this chapter, some of the longer items have been reduced

down or abbreviated for reader ease; however, the items in their entirety can be found

below in Table 1.

Table 1

Survey Items and Corresponding Shorthand

Shorthand Full Item

Above/looked up Above others, looked up to

Assertive/tell Assertive, tell people what to do

Calm Calm

Comfortable Comfortable, feel good

Confident/knowing Confident, know what to say

Right thing Doing the right thing

Following rules/directions Following rules and/or directions

Loving/helping/standing up Loving, help others, stand up for others

Talkative/listened to More talkative, people listen to me

Myself Myself

Nice Nice, not rude

Older Older than the people I am leading

On-task/productive On task, getting things done

Popular Popular

Powerful/impactful Powerful/impactful

Professional/work hard Professional, work hard, get things done

Quiet/listen Quiet, listen

Respectful Respectful

Responsible Responsible

Serious Serious

Smart Smart

Strong Strong

Sure/knowing Sure that I know what I am doing

Working together/sharing Working together/sharing with others, sharing information

While still ensuring representation of concepts, I also aimed to be consistent in

the distribution of where the items originated to give the best chance for the vernacular to

be representative of the demographics. The codes that were ultimately used for the items

on the surveys were proportional to the total original codes provided by each

demographic (Table 2). This suggests that in the reduction phases no individual group

was given more priority than another. While the ultimate vernacular was still not evenly

distributed between the genders and races, the ending codes were nearly identical in

proportion to the originating codes and in some instances closer to an equal distribution.

The largest difference was between white males who originally endorsed 30% of the

ending codes and Hispanic/Latino females who originally endorsed only 19% of the final

codes. This inequality speaks to a larger issue of voice and representation that will be

addressed in Chapter V.

Table 2

Final Codes and Original Codes Endorsement by Race and Gender

Demographic Original Codes % Final Codes %

Male 52 52

Female 48 48

Hispanic/Latino 37 41

White 63 59

Male & Hispanic/Latino 20 21

Female & Hispanic/Latino 17 19

Male & White 32 30

Female & White 31 29

I drew upon literature on gendered leadership (Abele et al., 2016; Bakan, 1966;

Eagly et al., 2020; Henstchel et al., 2019; S. K. Johnson et al., 2008) to tentatively sort

the final 24 items into agentic, communal, and ambiguous groups with the hopes that

there would be a somewhat equal distribution between the groups. I determined that the

ambiguous category provided a blurring of the line between what the literature has

deemed agentic or communal. In some cases, these concepts were found across both

categorized in different literature; myself, for example, could be interpreted as self-

assured and therefore agentic but also comfortable and therefore communal. Other items

appeared to not fit into either—being older, for example, was not explicitly communal or

agentic according to the literate. Ultimately, I decided that the category of ambiguous

could be used to either expand our understandings of the existing binary or point to

leadership that muddies the lines of the binary. In the end there were eight agentic-coded

items, seven communal-coded items, and nine ambiguous-coded items (see Table 3).

Table 3

Survey Items and Corresponding Category From Literature

Communal Agentic Ambiguous

Working together/sharing Assertive/tell Right

Loving/helping/standing up On-task/productive Talkative/listened to

Comfortable Powerful/impactful Older

Quiet/listen Strong Popular

Calm Serious Professional/work hard

Nice Confident/knowing Smart

Following rules/directions Sure/knowing

Above/looked up

Myself

Respectful

Responsible

The Experience of Leading

The second portion of the qualitative research question examined what

experiences participants have while learning to lead. I asked direct questions around

participants' experiences, changes to their leadership, and predictions about their future

as leaders. Answers to these questions, supplemented by other anecdotes shared during

the interviews, is compiled below and points to some trends in participants' experience,

support, and perception.

Teacher support/teacher indifference. In the interviews I asked participants

directly if they felt their teachers supported their leadership. In the conversations that

followed participants shared instances of being noticed or complimented as well as

instances of indifference, ignoring, or neutrality. Table 4 shows the cumulative number

of reported instances of support or indifference by each demographic group. Female

participants reported roughly one and a half more instances of support from teachers than

indifference, in contrast to male participants who reported the same ratio of support to

indifference from their teachers. There was a similar difference between white

participants and Hispanic/Latino participants with white participants reporting higher

ratios of support to indifference. High school participants reported the highest ratio of

support to indifference with double the amount of support. Middle school and elementary

school participants both had significantly less differences between the levels of support

and indifference they reported experiencing. Female participants made the most overall

references to teacher support while Hispanic/Latino participants made the most

references to teacher indifference.

Table 4

Total References to Teacher Support/Indifference by Demographic

References to

Teacher Support

References to Teacher

Indifference

Female 11 7

Male 7 7

Hispanic/Latino 8 8

White 10 6

Elementary School 6 6

Middle School 6 5

High School 6 3

External support/indifference. While participants were not directly asked about

their experience of external support or indifference, in reviewing the content of

interviews it became clear that there were different levels of support that participants felt

around their leadership from sources outside of their teachers (Table 5). These often

came through as participants discussed leadership opportunities they had been given such

as instances of being made captain of a project or team, earning awards, or being a leader

in the home. White participants made substantially more references to external support

than Hispanic/Latino participants with white participants referencing six times as much

support to indifference while Hispanic/Latino participance only referencing one and a

half more support than indifference. White participants also had the most overall

references to external support. References to external support also increased with age.

Finally, male participants also reported more external supports than female participants

but both referenced the same number of instances of indifference.

Table 5

Total References to External Support/Indifference by Demographic

References to External

Support

References to External

Indifference

Female 6 2

Male 9 2

Hispanic/Latino 3 2

White 12 2

Elementary School 2 1

Middle School 6 2

High School 7 1

Internal support/indifference. The interviews posed questions that had

participants reflect on their own leadership as well as organic opportunities for them to

reflect on how they feel about being leaders. They shared various levels of comfort or

discomfort with their own leadership. Participants shared instances of feeling happy,

confident, stressed, and unsure when leading; these were compiled to get a sense of how

participants internally view themselves as leaders. Table 6 tallies these positive or

negative self-expressions of leadership to allow for comparison in how participants'

internal perceptions differ among demographic groups. Overall, the ratios of positive to

negative expressions of internal leadership were relatively consistent between genders but

quite different between participants of different races. High school participants shared

overall many more instances of internal experience and had nearly double as many

positive experiences as negative while Hispanic/Latino participants had an 8:6 ratio of

positive to negative internal experiences. Middle school participants were the only group

to express more negative than positive internal experiences.

Table 6

Total References to Internal Positive/Negative Associations With Leadership by

Demographic

References to Positive

Internal Experience

References to Negative Internal

Experience

Female 13 7

Male 12 8

Hispanic/Latino 8 6

White 17 9

Elementary School 7 3

Middle School 5 9

High School 13 3

Shifts in Leadership Over Time

I asked participants about how their leadership had shifted or changed as they had

aged as well as predictions about how they think their leadership might continue to

change over time (Appendix C). Half of the participants indicated that their leadership

had increased overtime, three out of four of the 11th graders, two of the seventh graders,

and one of the third graders; four Hispanic/Latino participants and two white; four female

participants and two male participants. Only one participant, the seventh-grade white

male, indicated that his leadership had decreased over time. The remaining five indicated

no change. When asked about how their leadership might change in the future six

participants predicted their leadership will increase, two were male, four were female,

three were Hispanic/Latino, three were white, and two were in 11th grade, one in seventh,

and three in third. The 11th-grade white male and the third-grade white male both

indicated they expected their leadership to remain consistent while the seventh-grade

white male predicted his leadership would decrease. Finally, four participants reported

that they expected it to become more difficult to lead as they get older, these included

two third graders and two seventh graders, three Hispanic/Latino participants, and three

male participants.

Quantitative Results—Student Self-Descriptions

The quantitative surveys conducted in this research revolved around two

questions. The first, addressed in this section, explored how participants' ratings of their

own leadership attributes may or may not have differed based on their gender, race, or

grade level. This question also explored how, if at all, those differences aligned with or

countered the communal and agentic leadership expectations found in the literature—

specifically in relation to the participant's gender. The items or attributes rated by the

participants came directly from the interviews conducted in the qualitative portion of this

research.

Student Self-Descriptions Descriptive Statistics

The five highest-rated attributes across all genders, grade levels, and races

combined were Respectful (M=3.46, SD = 0.762), Nice (M= 3.37, SD = 0.749),

Loving/helping/standing up (M = 3.33, SD = 0.775), Working together/sharing (M= 3.32,

SD = 0.849), and Following rules/directions (M = 3.26, SD = 0.803). All but Respectful

were coded from the literature as communal attributes while Respectful was coded as

ambiguous, suggesting it may align with either agentic or communal or neither. The five

attributes with the lowest overall ratings were Popular (M = 2.25, SD = 0.921),

Above/looked up (M=2.47, SD = 0.923), Older (M = 2.52, SD = 0.853), Assertive/tell (M

= 2.66, SD = 1.043), and Talkative/listened to (M = 2.74, SD = 0.919). All of these items

were coded as either agentic or ambiguous. None of the lowest nine attributes were coded

as communal. The means and standard deviations for all 24 of the items can be found in

Table 7 below.

Table 7

Descriptive Statistics for Survey 1 Items

Item N Mean Standard

Deviation

Popular 80 2.25 0.921

Above/looked up 81 2.47 0.923

Older 81 2.52 0.853

Assertive/tell 80 2.66 1.053

Talkative/listened to 81 2.74 0.919

Confident/knowing 81 2.75 0.767

Serious 81 2.85 0.868

Sure/knowing 81 2.86 0.848

Powerful/impactful 81 2.88 0.842

Quiet/listen 80 2.93 0.868

On-task/productive 80 3.01 0.849

Comfortable 81 3.02 0.884

Professional/work hard 81 3.04 0.828

Calm 80 3.04 0.892

Strong 81 3.10 0.846

Responsible 81 3.12 0.765

Myself 80 3.20 0.947

Right 80 3.20 0.770

Smart 81 3.23 0.826

Following rules/directions 81 3.26 0.803

Working together/sharing 81 3.32 0.849

Loving/helping/standing up 81 3.32 0.849

Nice 81 3.37 0.749

Respectful 80 3.46 0.762

Student Self-Descriptions Item Analysis

I used an Analysis of Variance (ANOVA) to look for statistical differences

between the means of responses based on demographic information of participants:

gender, race/ethnicity, and grade level. However, rather than running an ANOVA on all

24 questions, which would lead to likely false positives and extremely stringent

significance values to accommodate for false positives, only a select few questions were

examined for each group. I utilized existing literature and theory to select which

questions should be further explored.

Gender-based item analysis. I conducted an ANOVA comparing gender

differences in responses to the following items: Powerful/impactful, working

together/sharing, assertive/tell, and loving/helping/standing up. I selected these items

because the literature suggested these were likely to be items where male and female

participants would rate themselves significantly different (Ayman-Nolley & Ayman,

2005; Ayman & Korabik, 2010; Eagly & Carli, 2003; Snaebjornsson & Edvardsson,

2013). Given the increased number of interactions being tested I used the Bonferroni

correction to reduce potential false positives and the significance value was raised from p

<.05 to p <.00625. With this more stringent significance value only one of the attributes

demonstrated a significant difference in ratings based on the gender of the participant;

that item was Loving/helping/standing up (F(1,79) = 12.397, p <.001) see Table 8.

Gender had a small but significant effect on how participants rated Loving/helping/

standing up (ηp2=0.275) with females rating this attribute significantly higher (M=3.57,

SD = 0.651) than males (M=3.00, SD = 0.816).

Table 8

ANOVA Results for Potentially Gendered Items

Measure F Value P Value

Powerful/impactful 2.449 .122

Loving/helping/standing up 12.397 <.001

Working together/sharing 4.603 .035

Assertive/tell .215 .644

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.00625.

Race-/ethnicity-based item analysis. Based on the research on racialized

leadership I selected the items nice/not rude, working together/sharing, strong, and

powerful/impactful for further exploration via and ANOVA (Bordas, 2001; Guajardo et

al., 2019; Dovidio et al., 2010). This time I was looking for significant differences in

ratings based on participant race. For this the racial category of “other” was excluded due

to the limited number of participants (n=4). As with gender the Bonferroni correction was

used to reduce potential false positives and the significance value was raised from p <.05

to p <.00625. With this more stringent significance value only working together/sharing

(F(1,75) = 8.494, p =.005) demonstrated a significant difference in rating dependent on

participant race/ethnicity (see Table 9). The participants’ race/ethnicity had small effect

(ηp2=0.239); white participants rated this trait significantly higher (M=3.51, SD = 0.731)

than Hispanic/Latino participants (M=2.96 SD = 0.871).

Table 9

ANOVA Results for Potentially Racialized Items

Measure F Value P Value

Strong 5.840 .018

Powerful/impactful .144 .706

Nice, not rude .902 .345

Working together/sharing 8.494 .005

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.00625.

Grade-level-based item analysis. As with gender and race I conducted an

ANOVA comparing differences between participants of different grade levels. The

literature suggested that comfortable, sure/knowing, assertive/tell, and myself may be

rated significantly different depending on the developmental level and grade of the

participants (Ayman-Nolley & Ayman, 2005; Lane et al., 2006; Lane et al., 2003). For

grade level the Bonferroni correction resulted in a p<.0042 level of significance. With

this more stringent p-value, only assertive/tell was rated significantly different between

participants of different grade levels (F(2,77) = 6.233, p =.003) see Table 10.

Specifically, middle school participants (M = 3.26, SD = 0.872) rated assertive/tell

significantly higher than elementary school participants (M = 2.26, SD = 1.064); high

school participants did not rate this attribute significantly different than the other groups

(M= 2.70, SD = 0.952).

Table 10

ANOVA Results for Potentially Age-Related Items

Comfortable

Measure F Value

2.137

P Value

.125

Sure/knowing 1.015 .367

Assertive/tell 6.233 .003

Myself 3.848 .026

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.0042.

Student Self-Descriptions Factor Analysis

While the section above explored individual traits, the literature suggests that

often leadership is understood as categorical with collections of common traits making

up types of leadership. In this section, and subsequent sections, I explore potential

categories of leadership and examine them for holding differences across gender, race,

and/or age of participants. The first step was to run a factor analysis to see what, if any,

questions hung together. A score of 0.857 on the Kaiser-Meyer-Olkin Measure of

sampling adequacy showed that there was enough shared variance to conduct the factor

analysis. The significance of the Bartlett's sphericity of p<0.001 also indicated that there

were some relationships between the variables. All values on the anti-image correlation

table fell below 0.5 which indicated enough variation and correlation to other variables to

include all items in the factor analysis. With five components 66.115% of the variance

was explained. The five factors included all items except for Responsible which did not

adequately load onto any of the five factors. All items had a rotated component Matrix

score of greater than 0.5. Serious loaded onto both Factor 3 (0.590) and Factor 4 (0.644)

—because it loaded comparably onto both factors it was eliminated from both factors.

The factors were renamed and items included are shown in Table 11 below.

Table 11

Factors for Student Self-Description Responses

Factor 1

Communal

Factor 2

Agentic

Outgoing

Factor 3

Agentic 2

Factor 4

Agentic

Authority

Factor 5

Self-

Assuredness

Quiet/ listen Professional/

work hard Strong Older Myself

Respectful On-task/

productive Smart Assertive/ tell Above/ looked

Following rules/

directions Confident/

knowing

Powerful/

impactful

Sure/ knowing Comfortable

Right Popular

Loving/ helping/

standing up Talkative/

listened to

Nice

Working

together/ sharing

Calm

Reliability tests indicated that all factors with the exception of Factor 4 were

reliable with Cronbach's Alpha above 0.7 for all factors (Factor 1 a = 0.886, Factor 2 a =

0.812, Factor 3 a = 0.801, Factor 5 a = 0.742). Factor 4's Cronbach's Alpha was 0.686

and therefore that factor was eliminated from further analysis.

Student Self-Descriptions Factor-Based MANOVA

I then ran a Multivariate Analysis of Variance (MANOVA) on the four reliable

factors to look at the effects of one or more independent variable, gender, race, or grade,

on the responses to those factors (see Table 12). Note that when looking at race the

category of “other” was eliminated due to its small sample size (n=4). The only

significant finding was the effect of gender and race combined on the rating of Factor 5

(F(2) = 5.784, p = 0.005, ηp2 = .151). Factor 5 which included myself, above/looked up,

and comfortable was rated highest by elementary female participants (M= 3.304, SD =

0.502) and lowest by elementary male participants (M= 2.095, SD = 0.8325). Middle

school participants rated this factor roughly the same regardless of gender; male

participants (M= 2.875, SD = 0.775) and female participants (M= 2.889, SD = 0.799).

High school male participants (M= 2.968, SD = 0.799) rated this factor higher than high

school female participants (M= 2.667, SD = 0.471). Further discussion on the implication

of gender and grade affecting the self-assuredness factor will be discussed in Chapter V.

Table 12

MANOVA Results for Four Factors

Measure Degrees of

Freedom

F Value P Value ηp2

Grade 8, 124 2.233 .029 .126

Gender 4, 62 1.154 .340 .069

Race 4, 62 1.639 .176 .096

Race & Gender 4, 62 0.691 .601 .043

Race & Grade 8, 124 1.649 .118 .096

Gender & Grade 8, 124 2.373 .022 .129

Race & Gender & Grade 8, 124 1.975 .055 .113

Note. This table shows results of the MANOVA using Wilks Lambda as the measure of

significance.

Student Self-Descriptions Literature-Based Analysis

When originally selecting items for the survey I utilized literature around

gendered leadership and specifically agentic and communal leadership. While the factors

above may present new understandings of leadership, I was interesting how ratings of

these originating categories might have been impacted by participant race, gender, and

age. The items that fell into each of these categories can be found in Appendix D.

Student Self-Descriptions Literature-Based MANOVA

In addition to the factors derived from the factor analysis I also decided to look at

responses through the leadership categories described in the literature: communal,

agentic, and ambiguous. Before running any statistical analyses on items that were

defined as communal, agentic, and ambiguous, I needed to test the reliability of those

groups. Reliability tests indicated that all three groups were reliable with Cronbach's

Alpha above 0.7 (agentic a = 0.801, communal a = 0.847, and ambiguous a = 0.826). I

then ran a MANOVA on the three literature-based groups to look at the effects of one or

more independent variable (gender, race, grade) on the responses to those groups—note

that when looking at race, the category of “other” was eliminated due to its small sample

size (n=4).

Results of the MANOVA can be found in Table 13. There were significant

differences in how participants from different grades rated the communal attributes (F(2) =

4.352, p = 0.017). The partial eta2 of 0.118 indicates a small effect size. Elementary

school participants (M=3.365, SD = 0.513) rated these items significantly higher than

their middle school (M=3.092, SD = 0.589) and high school (M=3.072, SD = 0.564)

counterparts. There was not a significant difference in how they rated agentic attributes or

ambiguous attributes.

Table 13

MANOVA Results for Literature-Based Groups

Measure Degrees of

Freedom

F Value P Value ηp2

Grade 6, 126 2.456 .028 .108

Gender 3, 63 1.136 .341 .051

Race 3, 63 1.073 .367 .049

Race & Gender 3, 63 0.507 .679 .024

Race & Grade 6, 126 3.829 .002 .154

Gender & Grade 6, 126 0.234 .965 .011

Race & Gender & Grade 6, 126 1.730 .119 .076

Note. This table shows results of the MANOVA using Wilks Lambda as the measure of

significance.

There was also a significant difference in how participants rated agentic,

communal, and ambiguous items based on a combination of their race and grade.

There was a small effect size in how agentic items (F(2) = 3.146, p = 0.050, ηp2 =

0.088) were rated based on race and grade. They were rated highest by white middle

school participants (M=3.088, SD = 0.283) and lowest by white elementary school

participants (M = 2.822, SD= 0.534) the other groups rated them in descending order as

follows: high school and white (M= 2.976, SD = 0.526), elementary school and

Hispanic/Latino (M=2.818, SD = 0.671), middle school and Hispanic/Latino (M=2.553,

SD = 0.680), and high school and Hispanic/Latino (M = 2.538, SD = 0.583).

Communal items (F(2) = 12.387, p < .001) were rated highest by Hispanic/Latino

elementary school participants (M=3.584 SD=0.329) and lowest by Hispanic/Latino high

school participants (M = 2.539, SD=0.352) the other groups rated them in descending

order as follows: middle school and white (M=3.314, SD = 0.430), high school and white

(M= 3.266, SD = 0.499), elementary school and white (M = 3.238, SD= 0.563), and

middle school and Hispanic/Latino (M=2.776, SD=0.669). There was a small effect on

how grade and race impacted the ratings of communal items (ηp2 = 0.276).

There was also a small effect on how grade and race determined ratings of

ambiguous items (F(2) = 5.164, p = .008, ηp2 = 0.137). Ambiguous items were rated

highest by white high school participants (M= 3.169, SD = 0.437) and lowest by

Hispanic/Latino high school participants (M = 2.679, SD=0.443); the other groups rated

them in descending order as follows: middle school and white (M=3.144, SD = 0.359),

elementary school and Hispanic/Latino (M=2.990 SD=0.431), elementary school and

white (M = 2.930, SD= 0.597), and middle school and Hispanic/Latino (M=2.683,

SD=0.586).

Quantitative Results—Perceived Teacher-Desired Attributes

In the second quantitative portion of the research I used the same 24 attributes but

this time asked participants to rate the extent to which they felt their teachers wanted

them to embody those attributes. This survey aimed to measure how participants

perceived their teachers encouraging or discouraging various leadership behaviors.

Through this survey I wanted to explore how, if it all, student may perceive their teachers

responding differently to leadership behaviors based on the race, gender, or grade level of

the participants. Specifically, this question examined the perception of teacher preference

for congruent or incongruent leadership attributes based on the literature.

Perceived Teacher-Desired Attributes Descriptive Statistics

The five highest-rated items that participants felt teachers would want them most

to display across all genders, grade levels, and races combined were following

rules/directions (M=3.74, SD = 0.587), right thing (M= 3.68, SD = 0.588), nice (M =

3.64, SD = 0.677), respectful (M= 3.64, SD = 0.661), and responsible (M = 3.61, SD =

0.665). All five were coded as either ambiguous or communal items. The five items with

the lowest overall ratings, suggesting participants felt teachers least desired them, were

popular (M = 2.20, SD = 1.089), older (M=2.23, SD = 1.102), assertive/tell (M = 2.53,

SD = 1.085), above/looked up (M = 2.69, SD = 1.051), and talkative/listened to (M =

2.81, SD = 1.001). All of these items were coded as either agentic or ambiguous.

Interestingly, the five lowest items in this survey were the same as the lowest in the first

survey though in a slightly different order. The means and standard deviations for all 24

of the questions can be found in Table 14 below.

Table 14

Descriptive Statistics for Perceived Teacher-Desired Attributes Items

Item N Mean Standard Deviation

Popular 81 2.20 1.089

Older 80 2.23 1.102

Assertive/tell 81 2.53 1.085

Above/looked up 80 2.69 1.051

Talkative/listened to 79 2.81 1.001

Serious 81 2.96 0.914

Confident/knowing 79 3.32 0.825

Calm 81 3.32 0.804

Sure/knowing 80 3.33 0.823

Powerful/impactful 80 3.33 0.897

Smart 81 3.37 0.782

Strong 81 3.42 0.804

Comfortable 81 3.42 0.906

Myself 81 3.43 0.836

Quiet/listen 81 3.46 0.791

Loving/helping/standing up 80 3.49 0.763

Professional/work hard 80 3.50 0.729

Working together/sharing 80 3.60 0.686

On-task/productive 81 3.60 0.646

Responsible 80 3.61 0.665

Respectful 80 3.64 0.661

Nice 81 3.64 0.677

Right 81 3.68 0.588

Following rules/directions 81 3.74 0.587

Perceived Teacher-Desired Attributes Item Analysis

As with the self-ascribed items above, I again ran several ANOVAs to determine

if there were any statistical differences between the means of responses based on

demographic information of participants: gender, race/ethnicity, and grade level. These

ANOVAs were not run on all 24 questions. To accommodate for false positives, I

selected only the items most likely, based on the literature, to have statistical differences

between groups.

Gender-based item analysis. I conducted an ANOVA comparing gender

differences in responses to the items: powerful/impactful, working together/sharing,

assertive/tell, and loving/helping/standing up. These items, the same as were chosen for

the self-ascribed research question, were selected because of their grounding in literature

(Ayman-Nolley & Ayman, 2005; Ayman & Korabik, 2010; Eagly & Carli, 2003;

Snaebjornsson & Edvardsson, 2013). Despite what the literature may have suggested

none of these attributes were rated significantly differently based on the participants'

gender (see Table 15).

Table 15

ANOVA Results for Potentially Gendered Items

Measure F Value P Value

Powerful/impactful 2.132 .148

Loving/helping/standing up 3.383 .070

Working together/sharing 2.133 .148

Assertive/tell .000 .992

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.00625.

Race-/ethnicity-based item analysis. To look at potential racial differences I ran

an ANOVA on the items: nice/not rude, working together/sharing, strong, and

powerful/impactful. Again, these are the same items examined closer for the self-ascribed

section above based on their grounding in the literature (Bordas, 2001; Guajardo et al.,

2019; Dovidio et al., 2010). They are attributes that may be inadvertently enforced

differently between students of different races/ethnicities and therefore participants may

perceive their teachers wanting these attributes to different degrees based on the

participant's race or ethnicity. For this analysis the racial group “other” was excluded due

to its sample size of only four. Given the increased number of interactions being tested

the Bonferroni correction was used to reduce potential false positives and the value of

significance was raised from p <.05 to p <.00625. The analysis found none of these items

differed significantly based on the participants' race/ethnicity (see Table 16).

Table 16

ANOVA Results for Potentially Racialized Items

Measure F Value P Value

Strong .021 .886

Powerful/impactful 1.331 .252

Nice, not rude 1.095 .299

Working together/sharing .961 .330

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.00625.

Grade-level-based item analysis. Finally, I ran an ANOVA looking for grade

level differences on the items: comfortable, sure/knowing, assertive/tell, and myself.

These items were selected because of the assumption that they may be enforced

differently depending on the grade level of the participants and therefore may be

perceived as desired to different degrees depending on the grade level of participants

(Ayman-Nolley & Ayman, 2005; Lane et al., 2006; Lane et al., 2003). These were the

same items more closely examined for grade level differences in the self-ascribed portion

of the research above. The Bonferroni correction was used again and the value of

significance was raised from p <.05 to p <.0042. With this more stringent significance

value three of the four items were found to have significant differences (see Table 17).

Elementary school participants (M= 3.84, SD= 0.374) rated myself (F(2,78) =

10.642, p < .001) significantly higher than high school participants (M= 2.97, SD=

0.948). Both elementary school participants (M= 3.81, SD=0.601) and middle school

participants (M= 3.79, SD= 0.535) rated comfortable (F(2,78) = 15.732, p < .001)

significantly higher than high school participants (M= 2.81, SD=1.014). Similarly, both

elementary school participants (M= ,3.84 SD= 0.374) and middle school participants

(M= 3.47, SD= 0.697) rated sure/knowing (F(2,78) = 23.542, p < .001) significantly

higher than high school participants (M= 2.70, SD= 0.837). Interestingly, assertive/tell,

the only item was found to be significantly different in self-ascribed attributes, was not

significantly different in perceived teacher-desired attributes.

Table 17

ANOVA Results for Potentially Age-Related Items

Measure F Value P Value

Comfortable 15.732 <.001

Sure/knowing 23.542 <.001

Assertive/tell 5.345 .007

Myself 10.642 <.001

Note. With accounting for the Bonferroni correction, a significant p value was increased

to greater than 0.0042.

Perceived Teacher-Desired Attributes Factor Analysis

As with the self-ascribed traits I was interested in not just individual traits but

whether or not groupings of traits came to light and whether those groupings may reflect

differences in perceptions across grade, race, and/or gender. To that end I ran a factor

analysis to see what, if any, questions hung together. A score of 0.863 on the Kaiser-

Meyer-Olkin Measure of sampling adequacy indicated enough shared variance to conduct

the factor analysis. The significance of the Bartlett's sphericity of p<0.001 also indicated

that there were relationships between the variables. All values on the anti-image

correlation table fell above 0.5 which indicated enough variation and correlation to other

variables to include all items in the factor analysis. The items quiet/listen and respectful

did not load significantly onto any factor with greater than 0.5 so they were not included.

The item comfortable was excluded from Factor 1 because it also loaded significantly

onto Factor 2. Similarly, the item serious was eliminated from Factor 3 because it also

loaded significantly onto Factor 2. Finally, Factors 4 and 5 were eliminated because they

only held one trait each making them not a “group” so much as an individual significant

concept. Factor 4 contained only the item above/looked up while Factor 5 contained only

the item assertive/tell; these items were explored earlier when looking at individual

items. With the remaining three components 59.907% of the variance was explained.

Reliability tests indicated that all three remaining factors were reliable with Cronbach's

Alpha above 0.7 for all factors (Factor 1 a = 0.918, Factor 2 a = 0.889, Factor 3 a =

0.713). Table 18 shows the items that loaded onto each renamed factor.

Table 18

Factors for Survey Two Responses

Factor 1

Well-Rounded

Factor 2

A+ Student

Factor 3

Demonstrating Maturity

Loving/helping/standing up On-task/productive Popular

Myself Smart Older

Following rules/directions Professional/work hard Talkative/listened to

Working together/sharing Nice

Powerful/impactful Responsible

Right

Confident/knowing

Sure/knowing

Calm

Strong

Perceived Teacher-Desired Attributes Factor-Based MANOVA

Now that the three factors were determined I ran a MANOVA to look at the

effects of one or more independent variable (gender, race, grade) on the responses to

those factors—note that “other” for race was eliminated due to its small sample size

(n=4). Grade level had a significant impact on the participants’ rating of teacher-desired

attributes (see Table 19). This significance was found in both participants’ ratings of

Factor 1 items (F(2) = 17.188, p < .001) and Factor 2 items (F(2) = 16.824, p < .001). High

school participants (M = 3.113, SD = 0.706) rated Factor 1 significantly lower than both

elementary school participants (M = 3.785, SD = 0.231) and middle school participants

(M = 3.648, SD = 0.282). High school participants (M = 3.200, SD = 0.621) also rated

Factor 2 significantly lower than both elementary school participants (M = 3.820, SD =

0.474) and middle school participants (M = 3.727, SD = 0.233). No other independent

variables, or combination of variables significantly affected how participants rated their

perception of teacher-desired behaviors.

Table 19

MANOVA Results for Three Factors

Measure Degrees of

Freedom

F Value P Value ηp2

Grade 6, 126 7.285 <.001 .225

Gender 3, 63 0.684 .565 .032

Race 3, 63 1.557 .209 .069

Race & Gender 3, 63 0.211 .888 .010

Race & Grade 6, 126 1.936 .080 .084

Gender & Grade 6, 126 1.088 .373 .049

Race & Gender & Grade 6, 126 1.258 .282 .057

Note. This table shows results of the MANOVA using Wilks Lambda as the measure of

significance.

Perceived Teacher-Desired Attributes Literature-Based Analysis

Just as with the self-ascribed responses I decided to look at literature-based

categories of traits and how the perception of what teachers desire might differ between

these pre-existing groups. Through the literature I had predetermined communal, agentic,

and ambiguously coded groups of traits found in Appendix D. The first step in examining

those groups was to run a reliability test to see if they were sound groupings. Reliability

tests indicated that all three groups were reliable with Cronbach's Alpha above 0.7

(agentic a = 0.798, communal a = 0.886, and ambiguous a = 0.758). Note that the

category of “other” was eliminated when considering race due to its small sample size

(n=4).

Perceived Teacher-Desired Attributes Literature-Based MANOVA

I ran a MANOVA to see how different groups of participants may have rated their

teachers desiring different literature-based types of leadership. Results of that MANOVA

are displayed in Table 20.

Table 20

MANOVA Results for Literature-Based Groups

Measure Degrees of

Freedom

F Value P Value ηp2

Grade 6, 126 2.456 .028 .105

Gender 3, 63 1.136 .341 .051

Race 3, 63 1.073 .367 .049

Race & Gender 3, 63 0.507 .679 .024

Race & Grade 6, 126 3.829 .002 .154

Gender & Grade 6, 126 0.234 .965 .011

Race & Gender & Grade 6, 126 1.730 .119 .076

Note. This table shows results of the MANOVA using Wilks Lambda as the measure of

significance.

Participant's grade level alone had a significant impact on the ratings of

communal items (F(2) = 4.352, p =.017). Elementary school participants (M = 3.365, SD

= 0.513) rated communal items significantly higher than both high school participants (M

= 3.072, SD = 0. 564) and middle school participants (M = 3.092, SD = 0.589).

When combining participant race and grade level there were significant

differences across all literature-based groups—communal, agentic, and ambiguous—in

how participants perceived their teachers desiring behaviors (Table 21).

Table 21

Means and Standard Deviations for Differences Across Grade and Race

Race & Grade Agentic Communal Ambiguous

M SD M SD M SD

Hispanic/Latino

Elementary

2.818 0.672 3.584 0.329 2.990 0.432

Hispanic/Latino

Middle

2.554 0.680 2.776 0.669 2.683 0.586

Hispanic/Latino

High

2.538 0.322 2.539 0.352 2.679 0.443

White Elementary 2.822 0.534 3.238 0.563 2.930 0.597

White Middle 3.088 0.283 3.314 0.430 3.144 0.359

White High 2.976 0.530 3.266 0.499 3.169 0.437

Note. This table shows the means and standard deviations for each subgroup across the

three literature-based categories.

Agentic items (F(2) = 3.146, p = .050, η2 = 0.88) were rated highest by white

middle school participants (M = 3.088, SD = 0.283) and lowest by Hispanic/Latino high

school participants (M = 2.538, SD = 0.322). All three grade levels of white participants

rated agentic items higher than all three Hispanic/Latino grade levels.

Communal items (F(2) = 12.387, p < .001, η2 = .276) were rated highest by

Hispanic/Latino elementary school participants (M = 3.584, SD = 0.329) and lowest by

Hispanic/Latino high school participants (M = 2.539, SD = 0.352). There was very little

difference between how the three grade levels of white students rated communal items

but a significant difference between how different grade levels of Hispanic/Latino

participants rated communal items. Hispanic/Latino participants rated communal items as

perceived as less desired the older the students got.

Ambiguous items (F(2) = 5.164, p = .008, η2 = .137) were rated highest by white

high school participants (M = 3.169, SD = 0.437) and lowest by Hispanic/Latino high

school participants (M = 2.679, SD = 0.443). Ambiguous items were less endorsed by

Hispanic/Latino participants as they got older while they were more endorsed by white

participants as they got older.

Comparing Self-Ascribed & Perceived Teacher-Desired Attributes Results

When looking deeper at connection and differences between the two surveys I

determined it was relevant to look at the highest and lowest items both self-ascribed by

participants and selected as teacher-desired by participants.

Highest-Rated Items—Self-Ascribed and Perceived Teacher-Desired

Below, Tables 22, 23, and 24 look at the differences between what participants of

different groups rated as their highest self-ascribed items and the highest-rated items

they believe their teachers desire from them. The highest-rated quartile, or six traits, are

explored.

Both female and male participants rated many of the same items when

considering both their own leadership and the leadership they presume their teachers

desire. However, there were some notable differences. Only female participants rated

myself, loving/helping/standing up, and following rules/directions high when considering

their own leadership while only males self-identified with the items right thing, smart,

and strong. Similarly, there was substantial overlap between teacher-desired items by

gender but only females identified teachers desiring working together/sharing and only

males identified respectful. There were also some differences between what participants

self-identified and what they felt teachers desired; on task was identified as teacher-

desired but not self-ascribed and myself, smart, and strong were identified as self-

ascribed but not teacher-desired.

Table 22

Gender Differences in Highest-Rated Self-Ascribed Items and Teacher-Desired Items

Rating Self-Ascribed Teacher-Desired

Female (N=47) Male (N=34) Female (N=47) Male (N=34)

1 Respectful Respectful Right Respectful

2 Loving/helping/

standing up Nice Following rules/

directions

Following rules/

directions

3 Working together/

sharing Smart Nice Nice

4 Nice Right Responsible On-task/ productive

5 Following rules/

directions Working together/

sharing

Working together/

sharing

Right

6 Myself Strong On-task/ productive Responsible

Note. This table shows the six highest-rated items by participant gender.

There were also commonalities and differences between white and

Hispanic/Latino self-scribed and teacher-desired items (Table 23). Myself and right were

self-ascribed high by Hispanic/Latino participants but not by white participants. In

contrast working together/sharing and smart were self-ascribed high by white

participants but not by Hispanic/Latino participants. When it came to teacher-desired

traits only Hispanic/Latino participants rated respectful as highly desired while only

white participants rated nice as highly desired. Smart was identified as self-ascribed but

not teacher-desired; conversely on-task/productive was rated as highly teacher-desired

but not self-ascribed.

Table 23

Racial Differences in Highest-Rated Self-Ascribed Items and Teacher-Desired Items

Rating Self-Ascribed Teacher-Desired

Hispanic/Latino

(N=26) White

(N=51)

Hispanic/Latino

(N=26)

White

(N=51)

1 Respectful Working together/

sharing Following rules/

directions

Right

2 Nice Respectful Respectful Following rules/

directions

3 Loving/helping/

standing up Nice On-task/productive Nice

4 Myself Loving/helping/

standing up Right Responsible

5 Following rules/

directions Smart Working together/

sharing

Working

together/sharing

6 Right Following

rules/directions Responsible On-

task/productive

Note. This table shows the six highest-rated items by participant race.

Finally, when considering the highest-rated items based on grade level, there were

again some significant differences (Table 23). Myself was self-ascribed high by only

elementary school participants and responsible by only high school participants. Middle

school participants identified no unique traits as highly self-ascribed. Quiet/listen was

rated as highly teacher-desired by only elementary school participants. Working

together/sharing was rated high by only middle school participants, and high school

participants has no unique traits rated high. Across all three grade levels only

loving/helping/standing up, myself, and smart were self-ascribed but not rated as highly

teacher-desired and only quiet/listen and on-task/productive were rated as highly teacher-

desired but not self-ascribed.

Table 24

Grade-Level Differences in Highest-Rated Self-Ascribed Items and Teacher-Desired

Items

Rating Self-Ascribed Teacher-Desired

Elementary

(N=31) Middle

(N=19)

High (N=31) Elementary

(N=31)

Middle

(N=19)

High

(N=31)

1 Nice Respectful Respectful Quiet/listen Working

together/

sharing

Following

rules/

directions

2 Loving/

helping/

standing up

Following

rules/

directions

Nice Following rules/

directions

Nice Right

3 Working

together/

sharing

Right Following

rules/

directions

Nice Right Responsible

4 Respectful Nice Responsible Right Following

rules/

directions

Respectful

5 Myself Loving/

helping/

standing up

Right Responsible On-task/

productive

On-task/

productive

6 Smart Working

together/

sharing

Smart Respectful Respectful Nice

Note. This table shows the six highest-rated items by participant grade level.

Lowest-Rated Items—Self-Ascribed and Perceived Teacher-Desired

Below, Tables 25, 26, and 27 look at the differences between what participants of

different groups rated as their lowest self-ascribed items and the items they perceived

teachers wanting the least. The lowest-rated quartile, or six traits, are explored.

There were a handful of differences between the items self-ascribed as lowest by

females compared to males (Table 25). Only females rated assertive as least self-ascribed

while only males rated powerful/impactful as low self-ascribed. Only females rated

serious as least teacher-desired while only males rated sure/knowing as least teacher-

desired. The only items that were self-ascribed low by either gender that was not rated

low as teacher-desired were comfortable and quiet/listen; conversely, the only items rated

as not teacher-desired by either gender but not rated low as self-ascribed were calm and

smart.

Table 25

Gender Differences in Lowest-Rated Self-Ascribed Items and Teacher-Desired Items

Rating Self-Ascribed Teacher-Desired

Female (N=47) Male (N = 34) Female (N=47) Male (N=34)

24 Popular Above/looked up Popular Popular

23 Older Popular Older Above/looked up

22 Above/looked up Older Assertive/tell Assertive/tell

21 Assertive/tell Talkative/listened to Talkative/listened to Older

20 Confident/knowing Confident/knowing Serious Talkative/listened to

19 Talkative/listened to Powerful/impactful Above/looked up Sure/knowing

Note. This table shows the six lowest-rated items by participant gender.

Across race there were also differences in which items were rated low (Table 26).

Hispanic/Latino participants but not white participants rated confident/knowing as a low

self-ascribed item conversely, white but not Hispanic/Latino participants rated

sure/knowing as a low self-ascribed item. There were no differences in lowest perceived

teacher-desired traits between the racial groups. However, confident/knowing and

sure/knowing were rated as low self-ascribed traits but not as low teacher-desired.

Serious was rated as low teacher-desired but not low self-ascribed.

Table 26

Racial Differences in Lowest-Rated Self-Ascribed Items and Teacher-Desired Items

Rating Self-Ascribed Teacher-Desired

Hispanic/Latino

(N=26) White

(N=51)

Hispanic/Latino

(N=26)

White

(N=51)

24 Popular Popular Popular Popular

23 Above/ looked up Older Assertive/tell Older

22 Older Above/looked up Older Assertive/tell

21 Assertive/tell Assertive/tell Above/looked up Above/looked up

20 Confident/knowing Talkative/listened to Talkative/listened to Talkative/listened to

19 Talkative/listened to Sure/knowing Serious Serious

18 Serious Confident/knowing Powerful/impactful Confident/knowing

17 On-task/productive;

Strong; Professional/

work hard a

Powerful/impactful Calm Sure/knowing

Note. This table shows the six lowest-rated items between participants of different races.

Finally, participant grade level resulted in some differences in what items were

rated lowest (Table 27). Only elementary participants rated talkative/listened to and

sure/knowing, only middle school participants rated serious and quiet/listen, and only

high school participants rated comfortable and powerful/impactful as lowest self-ascribed

items. In terms of traits perceived as least desired by teacher on high school participants

had unique traits. They rated sure/knowing and comfortable as undesired by teachers—

traits that were not identified by their middle school and elementary school peers. Across

all three grade levels serious, quiet/listen, and powerful impactful were self-ascribed low

but not rated as least teacher-desired and no traits were rated as least teacher-desired that

were not also self-ascribed low.

Table 27

Grade-Level Differences in Lowest-Rated Self-Ascribed Items and Teacher-Desired Items

Rating Self-Ascribed Teacher-Desired

Elementary

(N=31) Middle

(N=19)

High

(N=31)

Elementary

(N=31)

Middle

(N=19)

High

(N=31)

24 Popular Older Popular Popular Older Popular

23 Assertive/

tell Above/

looked up

Powerful/

impactful

Assertive/

tell

Popular Older

22 Above/

looked up Popular Above/

looked up

Older Above/

looked up

Above/

looked up

21 Older Serious Assertive/

tell Talkative/

listened to

Talkative/

listened to

Assertive/

tell

20 Talkative/

listened to Confident/

knowing

Confident/

knowing

Serious Assertive/

tell

Sure/

knowing

19 Sure/

knowing Quiet/

listen

Comfortable Above/

looked up

Serious Comfortable

Note. This table shows the six lowest-rated items between grade levels.

Consideration of Intersections

As with the MANOVAs I felt it was important to examine the intersection of all

three variables (race, gender, and grade level) to ensure that no one variable was strongly

skewing the others. Despite the disproportionate number of white respondents compared

to Hispanic/Latino respondents there was no evidence that white respondents sewed

results aligned with gender or grade level. Similarly, despite there being the least number

of middle school participants, their ratings were in alignment with the outcomes in gender

and race. In fact, when comparing responses via race, gender, and grade and accounting

for the other two the majority of highest and lowest ratings had an 80% or higher

alignment. Meaning that, for example, the highest-rated items by female participants had

80% or higher alignment with the highest-rated by white, Hispanic/Latino, elementary,

middle, and high school participants. This was true with both self-ascribed and teacher-

desired responses. This is not to say that there was no interaction between the

intersectionality of participants, but simply to say that no one group was dictating the

highest and lowest responses in another demographic area.

Incorporating Qualitative and Quantitative Findings

In this chapter I have focused on the quantitative analysis of both the interviews

and the survey responses. In the following section I explore the qualitative themes that

came to light through the inter views and were supported by quantitative data.

Leadership as Evolving

Across both the quantitative and qualitative components of this research there was

evidence that leadership is something that evolves over time and is influenced externally.

Organically participants referenced examples of external support as well as instances of

external indifference when describing the development of their leadership. They also

referenced internal narratives about their own leadership; one, for example, said, “I feel

like I'm not really a natural leader” (11WM). Others easily noted that when asked to

think of a leader they first think of themselves. These external and internal forces support

the theory that leadership changes over time. One participant said leadership is

developed from “who they like, who they also were led by, I mean, lots of people learn

from other people” (11HM) while another said leadership “comes at different times and

in different places” (11WF) for different people. Language in the interviews around

leadership also increased significantly with age suggesting an increased understanding of

leadership over time. By 11th grade, interviewees were providing nearly double the

number of leadership related codes as third-grade participants. There was also evidence

that overall confidence in leadership increased with age. In both elementary and middle

school there was an 11% gap between how much participants perceived teachers

endorsing leadership traits and how much participants endorsed those traits as part of

their own leadership. By high school this gap was almost non-existent with only a 0.4%

difference. This suggests that elementary and middle school participants saw themselves

has having less leadership qualities overall than high school participants and suggests a

development of those qualities over time.

Teachers as Mechanism of Influence

The influence of teachers, specifically, on the evolution of student leadership was

a core component of this research. In interviews participants validated both teacher

support and teacher indifference to their leadership suggesting that they are aware of

teacher feelings about leadership. Beyond just being aware there was significant overlap

between what participants in surveys indicated were their least and most self-ascribed

traits and what they viewed as teacher-desired traits. In fact, 78% of the lowest teacher-

desired traits were also rated as least self-ascribed; and in contrast 70% of the highest

teacher-desired traits were also highly self-ascribed. The highest level of influence, as

measured by internal consistency between teacher-desired and self-ascribed, was by the

high school participants at 87.5% overlap. The lowest level, at just 53% consistency, was

by the Hispanic/Latino participants. There were some notable items where the perception

of teacher desire did not appear to translate to participant self-identification. Across all

groups two traits were rated as highly desired by teachers but not self-ascribed:

Comfortable and Sure/Knowing. Interestingly assertive was highly endorsed as part

participants’ own leadership but not indicated as highly desired by teachers.

Favoritism and relationship between participants and their teachers also appeared to

be an impactful element in student leadership. One participant noted that the teacher that

they felt saw them as a leader was “definitely like a lot nicer and more empathetic” (11

HF) to the participant. Another participant shared that “there’s obviously always the

teacher’s pets that are there and the people that are the teacher’s favorites, but it’s not

like what everyone thinks of as the teachers’ favorites” (7WM). The overwhelming

commonality between what participants think of as their own leadership, their perception

of what teachers want, and how teachers treat them indicate that teachers are influential

in the development and definition of leadership in students.

Leadership as a Desired Attribute

A final but significant finding across both the quantitative and qualitative data

was the collective agreement that leadership is a desirable trait. One interviewee noted

that being a leader gave them “something to get up for” (7WF) while another stated “I've

always wanted to be, you know, some kind of leader” (11HM). One interviewee reported

that “being a leader is like a better feeling than being a follower” (7WF). This desire to be

leaders and therefore display leadership qualities was also seen in quantitative data.

Indeed, across all demographics all items on the self-ascribed survey averaged over a

two out of four rating. This suggests that all items were at least sometimes part of the

participants' leadership. More than half the items fell above a three out of four rating,

indicating they were often a part of the participants' leadership. Participants also

indicated that they felt that teachers strongly desired the majority of the leadership

behaviors with 18 of the items falling above a three (often) and seven of those falling

above a three-point-five out of four. Again, on the teacher-desired survey, no item's

average fell below a two, indicating that all leadership behaviors were seen as desired at

least sometimes.

CHAPTER V

DISCUSSION

What Does It All Mean?

Gender Differences in Leadership

When this research began one of its core tenets was to understand how gendered

leadership comes to be. The short answer is that in this study, it did not come to be, at

least not in the ways I initially assumed. The literature outlined the significant of social

construction in shaping gender. That social construction suggested that there should have

been significant differences across the leadership self-descriptions of participants based

on their gender. What I found instead were some minor but not altogether insignificant

differences. Gendered leadership, based in the literature, suggested that female students

would strongly endorse communal leadership while their male counterparts would

endorse agentic leadership. This research found that communal items in general were

rated higher by female participants than male, but the reverse was not found with agentic

items. There were some specific traits that would be presumed to hold stereotyped

gendered connotations that were found to algin with expectations: loving and following

for females and strong and right thing for males. However, the majority of the

stereotyped traits assertive, confident, quiet, powerful, working together, nice, etc. were

found to have no significance between participants of different genders. Maybe most

significantly, as a counter to literature expectations, male participants not only did not

rate powerful/impactful higher than females; they did not rate it as a desirable trait at all.

The literature also suggested that there would be significant differences in how

teachers treated students of different genders which would also play a role in the social

conformity of gendered norms and result in internalized gendered ways of being. While

actual teacher behavior was not measured in this research, student perceptions of their

teachers' desires was. There were no significant differences across genders in traits

regardless of their alignment with gendered expectations. There were only minor

differences in what female participants perceived as more and less desired by their

teachers compared to male participants. Perhaps the most gender stereotyped examples of

this were the female participants perceptions that their teachers wanted them to be

working together/sharing and not serious while male participants perceived their teachers

wanting them to be more sure or knowing.

The assumption may be that these gendered stereotypes failed to show up as

hypothesized because the social construction of gendered leadership is a thing of the past

or perhaps was never as significant as initially assumed. However, given the historical

evidence that socially constructed understandings of both gender and leadership are long

standing and relatively stable it seems unlikely that this research would have occurred

just as the tides change. Similarly, while societal understandings of gender and gender

norms are changing, the hope that the deeply ingrained gendered norms are no longer

impacting this generation seems overly optimistic. Instead, it is likely that gender

differences failed to show up for a variety of smaller but still consequential reasons.

Firstly, by using student language as the basis of the survey questions many of the items

encapsulated several components that may have made them more gender neutral than the

gendered traits from the literature. Second, it is likely that given the hierarchical dynamic

between the researcher and the participants, there was some people-pleasing behavior

that led to a neutralizing of strong opinions on surveys—particularly strongly negative

opinions. Indeed, there was an overall trend of positivity for all leadership traits across

both genders. Most significantly though, I hypothesize that there was another more

important factor at play that superseded the gender influence when participants

considered their leadership—their positionality as students. This will be further explored

below.

Racial Differences in Leadership

Similarly, to gender, there were expectations for racialized differences across self-

ascribed and perceived teacher-desired leadership traits. Based on the literature there

were assumptions that white participants may be inclined towards more individualistic

forms of leadership and Hispanic/Latino participants may be inclined towards more

community-oriented forms of leadership. There were some differences that aligned with

these stereotypes including white participants indicating they strongly endorse being

smart and do not endorse being sure/knowing. But in contrast to expectations white

participants did not view being powerful/impactful as a self-desired trait. In alignment

with expectations myself was self-ascribed high by Hispanic/Latino participants. While

communal and agentic categories do not directly algin with collectivist and

individualistic leadership the literature suggests some overlap. Therefore, there were

expectations that Hispanic/Latino participants would gravitate towards communal-coded

traits and white participants to agentic-coded traits. Communal traits were rated highest

by young Hispanic/Latino participants, but this tendency ended as participants aged and

high school Hispanic/Latino participants showed the strongest rejection of this societal

expectation. White participants also broke with societal expectations by consistently

endorsing communal traits over agentic traits at all ages.

More was in alignment with literature-based stereotypes when it came to the

perception of teacher-desired traits. All three grade levels of white participants rated

agentic items higher than all three Hispanic/Latino grade levels. This suggests that white

students perceive the feedback that agentic behaviors are more acceptable and wanted

than their Hispanic/Latino peers. This mirrors the self-identified ratings where

Hispanic/Latino participants became less agentic over time. However, it differs from the

self-identified traits in that white elementary participants were least likely to self-identify

with agentic traits but still perceived them as more highly desired than Hispanic/Latino

participants when it came to what teachers want. There was very little difference between

how white students rated their teachers' desire for communal items but a significant

difference between how Hispanic/Latino participants of different ages rated these items.

Hispanic/Latino participants rated communal items as less desired by their teachers the

older the students got, which again mirrored their self-ascribed trends where being

communal became less a part of themselves as they got older. This contrasted to white

participants, where it stayed consistent both in their self-ascription and in their

perceptions of what teachers want. White participants specifically felt encouraged to be

nice. Hispanic/Latino participants on the other hand specifically felt encouraged to be

respectful. While these differences do not directly algin with the literature's assumptions

about how racial bias plays into the nurturing of youth it does speak to some combination

of different messaging and internalizing that is occurring across racial lines.

While racial stereotypes and bias in leadership formation is less researched it was

still unexpected to find so little significant difference between the two groups self-

ascriptions and perceived teacher endorsement. As with gender it is likely that there are

several potential causes of these results. The assumption of homogeneity between all

white participants and all Hispanic/Latino participants is a known fallacy with several

other factors impacting their potential self-image and experience. Notably their gender,

age, background, economic status, upbringing, familial norms, and position within the

school all likely played a role along with their racial identity. While there should be

celebration for a lack of severe and impactful stereotyping on the perception of self as

leader in this research there should be hesitancy to say that race does not impact the

development of leadership identity and continue to impact access to leadership beyond

K–12 education.

Developmental Differences in Leadership

The hopes for exploring leadership at three different developmental levels,

elementary, middle, and high school, was to get a glimpse at how leadership identities

develop and change as students move through the K–12 education system. In this regard

the research provided some interesting insight. For example, elementary participants were

the only ones to highly rate themselves as myself when leading. In the top quartile there

were few differences in self-ascribed traits; however, in the lower quartile there were

some notable differences. Only middle schoolers for instance disagreed with being quiet

or serious whereas only high schoolers felt that being comfortable was unimportant.

Holistically though the most significant difference in how leadership was self-ascribed

came from the endorsement of communal attributes across ages. When considering their

own leadership, elementary students rated these attributes significantly higher than their

middle school and high school peers suggesting they see their leadership as more about

community support and care. The other significant difference was that high school

participants generally rated themselves as lower in their endorsement of leadership across

the board when compared to elementary and middle school participants. This would

appear counter-intuitive considering that by high school the expectations would be that

these students would inherently be stronger leaders simply by nature of hierarchy than

their younger peers.

Perhaps more relevant though to understanding the shifting tides of leadership

over time is the perception of what students think teachers want from their leadership.

Teacher expectancy theory suggests that teachers inadvertently act differently based on

assumptions about the students in front of them, and age is no exception to this. While

raising strong leaders may be the universal aim of teachers regardless of student age,

there were significant differences in the messages being received by students of different

grade levels. Elementary school participants indicated they felt encouraged to be quiet

more so than their older peers. Broadly they also saw their teachers desiring them to be

more communal than their older peers which aligns with their self-ascriptions from

above. Middle school participants felt their teachers wanted them to work together.

Interestingly, again, high school participants rated the overall endorsement of leadership

by their teachers as lower than their younger peers. High schoolers rated following

rules/directions as the highest teacher-desired trait overall. This authority compliant

mindedness as a component of leadership is interesting in the context of both the

increased independence and higher academic demands of high school. However, the

general trajectory of less leadership endorsement from high school teachers compared to

elementary and middle school teachers is worrisome as it suggests a decline just as

students are entering the world where their leadership is most valuable and necessary.

Broadly speaking, leadership—as it was identified and encouraged across participants

regardless of age, race, and gender—looked different from the leadership initially

expected prior to this research. How students define leadership in this research was non-

traditional and at times counterintuitive.

Re-Examination and Reflection

Student Leadership Redefined as Collaboration

Regardless of gender, race, community, or tradition there are some universalities

of what leaders look, sound, and act like. Leaders, by their very dictionary definition, are

in charge, in the forefront, making decisions, and guiding others. But what this research

shows is a different understanding of what it means to be a leader. Across the board

students self-identified with leadership that involved traditional values like being

respectful, nice, and smart with traits like being helpful, working together, and following

rules or directions. Indeed, traditional stereotyped leadership ideals such as telling people

what to do, being listened to, and being assertive were least likely to be self-ascribed. In

both the quantitative and qualitative phases, it became clear that student perceived

leadership was more relational and defined by action towards others rather than by

feelings of authority and power. While initially seeming somewhat counter intuitive there

is some support in the literature for these findings. Specifically, the literature on youth

activism which suggests a way of leading that is more organic, driven by issue rather than

by role, and responsive to the needs of the collective (Marri & Walker, 2008; Chang &

Gamez, 2022). Similarly, there is some support for the notion that young leaders are more

inclusive and less norm oriented than adults in their collective understandings (Bertrand,

2018; Mortensen et al., 2014).

Yet despite some research supporting the redefining of leadership as

collaborative, these findings are arguably less about youth rebellion and more about the

significant influence of teacher expectations. Coffey and Lavery (2018) have suggested

that how teachers model leadership in the classroom has a significant impact on how

students come to understand leadership. More significantly Haber-Curran and Tillapaugh

(2017) noted that the setting of expectations by the teacher who holds the power in the

classroom is a strong predictor of how the students, who hold less power, come to

understand their own behaviors. The traits students across all ages, races, and genders

identified as most desired by their teachers included following rules and directions, being

nice, being on task, and working together. In contrast they felt their teachers did not want

them to be assertive, talkative, popular, or knowing. While this may not align with

colloquial understandings of leadership these traits do align with typical expectations of

“good” students.

Does this difference in student definitions of leadership from traditional

understandings of leadership simply point to a lack of understanding of what leadership

means? Classrooms and schools are filled with messaging about being a leader,

developing leadership, and raising leaders so it is unlikely that students are simply

unfamiliar with leadership as a concept. Yet some interaction between the messaging of

being a leader and the felt expectations of being a cooperative follower has led to this

differing definition. The assumption laid out in Chapter II of this paper was that there was

an interaction between societal expectations, socially construction of self, and teacher

expectancy that created leadership congruent with an individual's gender and racial

identity. This research suggests that gender and race, and to a lesser extent grade, are less

impactful in how one comes to understand their leadership in the context of being a

student in a K–12 setting. Yet there is messaging happening around what it means to be a

leader. Revisiting the conceptual framework highlights a significant revision that needs to

occur in understanding the self-conception of leadership.

Revisiting the Conceptual Framework

The conceptual framework outlined in Chapter II and presented again below in

Figure 7 suggested an interaction between large societal pressures and norms that

dwindle down to a cyclical pattern of behavioral interactions that shape student

leadership. Specifically, the conceptual framework explores the potential societal

pressures that create and perpetuate stereotyped socially constructed gendered and racial

ways of leading. Ultimately, leadership that was self-identified in this research did not

fall as neatly as anticipated along gendered and racial lines. Similarly, the leadership that

was presumed to be desired by teachers did not appear to bias individuals towards

stereotyped congruence. While communal traits were significantly different between

genders the same was not found with agentic traits. The conclusion therefore must but

that this conceptual framework is lacking some additional point of influence.

Figure 7

Conceptual Framework Revisited

Despite its flaws, the framework does provide an insight into understanding how

factors work together to shape leadership. Working from the outside in there is still

evidence that patriarchy influenced participants in their conception of self as leader.

Primarily in interviews there were significant differences in how participants of different

genders came to understand themselves. There was also clear evidence that leadership

was socially constructed and not stagnant over time. Differences across grade levels

demonstrated a growth and re-prioritization of leadership traits. Some of these pressures

were alluded to in interviews where participants talked about feeling increased

responsibility and expectations placed on them. While agentic traits failed to fall clearly

along gendered lines, communal traits were seen to be congruent with stereotyped

expectations. Teacher expectancy also was found to be influential in the significant

alignment between self-identified and teacher-desired traits. However, I now hypothesize

that there is a more influential factor in how participants came to define their leadership

than their race or gender presentation. Instead of leadership that was stereotypically

feminine or masculine, white or Hispanic/Latino, the leadership presented by participants

in this research was that of cooperative followers. Leadership that was ultimately

congruent with being a good student more so than a good leader. I suggest that the

positioning of individuals in particular spaces and with expectations while they develop

their leadership is the untapped key to understanding leadership development.

Novel Theory

The Missing Piece—Student Positionality

To better understand the potential influence of positionality I returned to the

literature. One of the originating problems in developing the conceptual framework

presented in Chapter II was the severe lack of existing research on how youth define and

develop their own leadership. Indeed, Dempster and Lissio (2007) noted that “there is an

identifiable gap in our knowledge of students’ understanding of leadership and how they

see, experience and interpret it in different situations” (p. 279). One space that has

utilized student voice has been in the domain of youth activism. Research on youth

activism tells us that schools have traditionally taken a top-down and hierarchical

approach to leadership leaving teachers and administrators in positions of outright

authority over students (Chang & Gamez, 2022). Social movements and youth activism

presents opportunities for youth to demonstrate a different kind of leadership than that

seen in this research (Marri & Walker, 2008). In these spaces of activism students have

voiced leadership that looks and acts differently than that of the participants in this

research. Chang and Gamez (2022) argue that this activism leadership looks different

because of differences in how the adults and youth interact. They present an alternative

type of adult leadership, accompaniment leadership, that allows youth to branch beyond

the follower-manager relationship and into one where students have authority (Chang &

Gamez, 2022)

Because of the traditional hierarchical expectations of participants in this study,

with no explicit messaging around activism, no formal leadership opportunities within

their classes or schools, and limited opportunities for them to hold power and influence;

participants were inadvertently positioned as followers while simultaneously being told

they were leaders. Kezar (2002) looked at how several facets of an individual can interact

to impact their leadership and specifically found that the “role within the institution was

related to the way individuals construct leadership” (p. 100). The role of students as

hierarchically lower than their teachers impacted their internal understandings of

leadership. There is evidence that the connection between meeting teacher expectations

and considering oneself a good leader is not entirely new (Coffey & Lavery, 2018;

Wallin, 2003). But without the opportunity to hold true leadership as defined by power

and influence, participants were limited in their ability to construct their own leadership

identities inside of their position (Kezar & Lester, 2010). With this additional

component, a new conceptual framework is presented below (Figure 8) that highlights

the significant influence of where an individual is positioned in affected several layers of

their leadership development. Arguably the results of this study are not just about how

students are positioned but about who holds power in the student/teacher dynamic. Power

plays a significant role in the dictation of what is acceptable and what is not. Power in

this instance is held almost exclusively with the teachers and over the students. Through

implicit and explicit instruction teachers can sanction and elicit certain types of

leadership displays that they deem acceptable.

Figure 8

Updated Conceptual Framework

Added and adjusted elements, highlighted in red, indicate that the position an

individual is placed in influences all of their experience along with their racial and gender

identity. It impacts their participations in and experience of patriarchy and white

supremacy societal norms. It impacts how they come to socially construct leadership. It

interacts with how they are perceived and treated by others, in this case by teachers.

Additionally, this new understanding of how leadership is developed, argues more

universally to the experiences of how all individuals come to develop, understand, and

perform their leadership in the context of their positioning.

When discussing positionality is it important to clarify the many ways in which

position is used both in this colloquial dialogue and in the literature. How one is

positioned or their positioning speaks to the implicit or explicit ways in which they are

perceived by society and then funneled into particular roles, habits, and ways of

presenting. An individual's positionality in contrast speaks to how their real or perceived

identity will impact their access, power, perception, and assumptions. Both of these are

impacted by an individual's position. Our position speaks to the identity we embody and

are perceived to embody. In the case of leadership access, an individual's position as a

woman or person of color impacts their positioning as they are funneled not into

leadership but into subordinate roles and therefore impacts their positionality where they

are less likely to have access to leadership and authority. In a sense this theory speaks to

all levels of positioning. Students bring their own position, are then positioned, and then

embody a positionality. This convoluted connection is the reality of examining

individuals in the context of others and in the context of societal norms and expectations.

As queer theorists would argue, all things are complicated and simplicity should be

reexamined.

Broadening Positionality in Understanding Leadership

This dissertation began with the question of why does leadership continue to look

and be homogeneous despite evidence that diversity in leaders and leadership is

beneficial. Additionally, it asked when and how do serotypes around gendered and

racialized leadership come to be developed and what factors are perpetuating a link

between certain types of leadership with certain individuals. Approaching this work from

a social constructivist perspective suggested that how leadership comes to be embodied is

not innate but instead an accumulation of learned, practiced, and reinforced actions.

The findings of this research suggest that we have been looking at leadership

invertedly. We have looked at the end result of leadership performance and surmised

commonalities between groups of people based on their socially constructed

presentations of self—things like their gender presentation or racial identity. Instead we

need to look at leadership though the lens of what is being asked of the individual based

on where they are positioned. In this method it is not just how the individual presented

externally but how society has them positioned in hierarchy, power, and space. There is a

cyclical relationship between how they are perceived by society, where they are therefore

positioned, how that positioning impacts their actions, and how those actions are then

perceived by society. For example, it may not be that student leaders are naturally

followers—as found in this research—but instead they have been positioned as

followers, told they are leaders, and come to define their leadership as that of followers.

Similarly, it is not that women tend to be more communal in their leadership and

therefore are often in communal leadership roles but instead women have historically

been placed in communal roles, told they are leaders, and then embody communal

leadership.

If we understand positionality as having played a significant role in the

development of leadership identity than we must also see positionality as a method for

adjusting leadership. It would suggest that to change who is in leadership positions we

must change what “positions” people are in. When they find themselves in a new

position, they can see themselves as a new kind of leader. In this case if we queer the role

of student and provide meaningful opportunities for leadership by students we will see

student leaders who are more representative of socially accepted leadership and less

representative of good followers.

Perhaps the best measure of a good theory is its ability to logically explain what

has already been explored. The theory above, that the positioning of individuals has a

significant impact on their leadership identity, works harmoniously with the existing

understanding of how gender and race play into leadership identity. By looking at the

positioning of people, assumptions about how leadership comes to be socially

constructed includes gender and race but also is broad enough to include other power

dynamics. In the case of students and teachers this hierarchical dynamic played into

leadership taking on a follower form. When considering the prevalence of white male

leaders in authoritative positions and displaying agentic and authoritative leadership it is

not just that their race and gender have conditioned them to engage in this type of

leadership but also that they have been disproportionately positioned in these type of

leadership roles.

I suggest that rather than train women to lead in agentic ways, or train hirers to

value new forms of leadership, one way to diversify leadership is simply to ensure that

people are placed, in positions of leadership. I suggest that we readjust who we are

privileging, of course, with the careful caveat that the individuals place in those roles are

both seeking them and adequately supported once there. It is the obligation of change

makers to ensure they do not set people up to fail if they are unprepared or unwelcomed

for the roles in which they are placed. One way to provide support for those newer to

these spaces would be to hire groups or cohorts of female leaders of leaders of color so

that they are not alone, known sometimes as cluster hiring. While shifting deeply

engrained racial and gender stereotypes may be beyond the scope of individuals or our

current political climate the simple task of putting people in positions, they have not

historically been in may be a fundamental way of disentangling leadership from

essentialist binaries.

Despite this we cannot discount the fact that the positioning of people is dictated

at least in part by their race and gender. As I call for change, I simultaneously

acknowledge that where people are positioned is a byproduct of bias and patriarchal

norms. While different positioning may present the opportunity for more diverse

leadership, it will likely also face the same barriers as dismantling race and gender biases

from any institutional practice or societal norm.

Limitations

Demographic Limitations

While the theory presented above holds both internal validity and common-sense

validity it is important to note the several restraints of this study and resulting limitations

of any results pulled from this study. The most notable limitation is that of participants.

With only 12 interviews and only one participant per demographic group the qualitative

portion of the research does not provide substantial data to make judgments of different

experiences across groups of individuals. Similarly, the limitations in quantity of

participants in the quantitative portion of this research limit the reliability. While the total

sample size of 81 is reasonable, when looking at individual groups that intersected

multiple demographic points, the sample sizes ranged from two at the smallest

(Hispanic/Latino, male, third graders) to 14 at the largest (white, female, third graders).

When talking about participants it is also important to note significant differences in

representation of groups. White females were most represented with 35% of all survey

participants, while Hispanic/Latino males were least represented at only 16% of survey

participants. Middle schoolers were also roughly 15% less represented than elementary

and high school participants.

There were also consistent issues with representation of racially diverse voices.

Hispanic/Latino participants and white participants were equally represented in the

interviews in terms of number of participants, but Hispanic/Latino participants were

underrepresented in the ultimate amount of talk time and language/vernacular that was

found and used in the quantitative portion of the research. In the surveys there was a

significant difference in representation of voice with roughly two thirds of the

respondents being white and only one third being Hispanic/Latino. An even greater

limitation is the absence of data from other racial groups. While this was a byproduct of

conducting research where I would be most effective in using data for the betterment of

student experience it is an unfortunate gap that limits the generalizability of the findings.

Similarly, the research taking place in a small rural school district creates some

limitations to generalizability. However, small rural districts are consistently under

researched despite being the learning environment of roughly 20% of students all across

the United States and in some states as high has 54% (Lavalley, 2018). The particular

culture and customs of the town. The school district itself, and the families that were

open to the research likely had an impact on some of both the results and the previously

addressed issues of inequity of voice.

Despite these limitations of overall participants, representation, and location, this

research does hold value in its ability to demonstrate what some students from a variety

of demographics think of their leadership and of their teachers' desires around leadership.

While there are, of course, differences between rural and urban schools and youth from

different racial backgrounds there is likely a significant amount of overlap of the

experience of being a youth in the public school system in America in 2023. These

shared experiences made more shared by an ever globalizing world, are represented in

this research and add to a very small body of literature of student voice on the

development of leadership.

Methodological Limitations

There are also some important learnings to be found in critically examining the

methodology used in this research. The survey given as part of the quantitative portion

used a four-point Likert scale using the labels: never, sometimes, often, and always.

While these are commonly used labels, they leave a significant amount of room for

individual interpretation.

The interview portion of the research relied on questions generated by me based on a

combination of literature, identified gaps, professional experience, and the aims of the

research questions. While these questions ultimately were able to give robust data they

were not previously tested for validity or reliability. It is likely that my own positionality

and experiences led to these questions and may have biased the way participants

answered them.

My own positioning also likely played a role in the responses received on both

portions of the research. While attempts were made to ensure that participants knew their

answers would remain anonymous and to reduce the power hierarchy between myself and

them, it is likely that there was social desirability bias (Lohmeyer, 2020; Potter &

Hepburn, 2005). Participants were likely trying to give the “correct” answers for both

what they thought I would want and what they thought was most socially accepted. While

authenticity was the aim of this research there is also value in understanding what youth

think adults want to hear about leadership and what they believe their leadership should

be if following adult expectations. Ultimately the line between what individuals think is

the right answer and what they come to believe to actually be the right answer is blurred

and in terms of actions often irrelevant.

The Ambiguity of Leadership

Throughout the research from its conception to the literature to the research itself

one thing has become abundantly clear: while everyone seems to know what leadership

means everyone also defines it slightly differently. The very concept of leadership is a

limitation to this study. Leadership varies from situation to situation, from demand to

demand, from context to context. Therefore, as participants were asked to reflect on their

own leadership or what their teachers seemingly wanted from their leadership, they were

forced to conjure up some imagined scenario of leadership. Their scenarios likely

differed person to person and reflected in their responses. This is both a limitation and

part of the beauty of a topic that will never cease needing exploration. As challenging as

it is to define leadership it is worthy of its many iterations and definitions.

Future Research and Concepts to Explore

Reliability of Current Research

Given the limitation of this study as well as the presentation of a new theory there

is a clear need for further research. The first is the obvious need to validate the reliability

of the vernacular presented by students in this study by students from other demographics

and settings. This research may include further interviews that confirm or counter the

vernacular presented by participants in this study. It may also involve repetition of the

surveys with larger number of students and students who are representative of more

demographic groups. Importantly, for research interested in gender and race this research

failed to explore nonbinary individuals and their experiences as well as individuals from

racial groups outside of white and Hispanic/Latino.

In my initial conceptual framework presented in Chapter II, I suggested that

relational connection between students and their teachers may contribute to the teacher

expectancy bias that sustains or counters social expectations of leadership. In this

research I did not ask participants about their perceived negative or positive relationship

to their teachers in either the survey or interview. In any future research it would be

important to examine the level of relationship more closely between those in power and

those learning to lead—in this case teachers and their students.

Further Exploration of Positioning

There is also a need to expand beyond this research and better understand the role

of positioning. When considering how we prepare a generation to be leaders without

falling into the trap of leadership dictated by racial and gendered stereotypes the answer

may lie in further exploring what happens to leadership when positioning is altered.

Research that explores the shift in leadership definitions when students are empowered to

be genuine leaders within their classes and schools. Additionally, there is a need to

explore positioning outside of schools and students. Any research that explores the

intersection of positioning and leadership will need to navigate the back-and-forth

relationship of positioning and social construction. This research will also need to take

care to evaluate the ongoing impact of stereotypes and white supremacy on all levels of

leadership development. How do we change positions before changing leaders but in an

effort to change leaders? How do we acknowledge that an individual's race, gender,

power often dictate their positioning, while also knowing that racial, gender, and power

dynamics can only change if they are positioned differently.

Recommendations for School Administrators and Teachers

Ultimately the aim of research was to unpack how schools are playing a role in

the perpetuation of harmful norms around gendered and racial leadership. Additionally, it

aimed to find a way forward that reduced the continuation of long standing socially

constructed norms of how individuals should lead and replace it with leadership that is

diverse and authentic to the individual. Finally, there was the hope that if leadership itself

could be detangled from stereotyping than who becomes the “best fit” for certain types of

leadership in adulthood could similarly become decoupled. I will grant that these were

lofty ambitions far beyond the reach of any one dissertation. However, there are some

practical adjustments that I believe can be made that encourage leadership exploration

and critical examination in schools. These are by no means a comprehensive list of all

possible changes that could be or should be made to education but they are a jumping off

point of difficult but doable tasks that I believe all educators could and should fight for.

This research began as a call to action for those of us supporting students in their

leadership development in the hopes of diversifying not only who leads but also how they

lead. To that end I would like to close this dissertation with recommendations for

educators. My first recommendation is around the importance of creating structures of

repeated evaluation of stereotyping by teachers. While this research did not find

significant differences in the quantitative portion regarding teacher bias there was clear

differences in level of internal and external support felt by different groups in the

interviews. Hispanic/Latino participants referenced less than half as many instances of

feeling their leadership organically supported compared to their white peers. Examples of

these supports include:

Ongoing professional development on bias and microaggressions—encourage

reflection and self-critique.

Bias and favoritism tracking as part of administrator walk-throughs with

pointed and meaningful feedback provided swiftly to educators.

Equity and cultural competency built into student learner growth goals or

other professional goals and accountability to meeting or not meeting those

goals.

Consistent review of school and distract data disaggregated by race/ethnicity,

gender, socio economic status, and disability.

District policy around inclusive classrooms including self-assessments,

checklists, and community accountability.

District purchasing and ongoing training of culturally competent and inclusive

curriculum.

Safe and respectful classrooms that include reduction in sensory input,

predictable schedules, and restorative consequences.

Additionally, I call educators to create structures for student leadership and

decision making in district and school policy. I task them with finding ways to go beyond

feedback or input and into legitimate cocreation. Examples of these include:

Student participation in school board meetings.

Student participation in district level teams including leadership teams,

adoption committees, and data review teams.

Student participation in the development of continuous improvement plans

reported to the state.

Opportunities for student liaisons between adult decision makers and student

body for a more collective voice.

One existing tool for supporting students in meaningful engagement with problem

solving and social justice work is through Youth Participatory Action Research, YPAR.

Through YPAR teachers and students work in collaboration to reflect critically on how

political and social forces impact their lives, identify areas of need, study those needs

through system research, and then engage in actionable steps that raise awareness and

change structures. This root-cause analysis couples critical reflection with actionable

practices (Buttimer, 2018). In the classroom the use of YPAR allows students to explore

and work to dismantle systems of oppression that are created rather than innate. In daily

instruction teachers can work with student to understand how to study problems they

encounter and provide opportunities for them to work through solutions (Cammarota &

Fine, 2017). Beyond the classroom YPAR can be used when teachers select critical or

core standards as a way to ensure that what is being taught emphasizes not just content

but content in context (Kornbluh et al., 2015). Teachers and administrators can also

support YPAR-led programs; local examples are available through the Oregon

Department Health Authority or the Washington SNAP-Ed Providers.

Similarly, teachers need to create opportunities for meaningful participation in

decision making in classrooms including leading both peers and adults in instructional

practices. Examples of these include:

Student co-creation of classroom activities and learning priorities. For

instance, student interests dictating reading materials or the use of math

problems that reflect real life concerns of students.

Shared norm setting where classroom norms are derived from students and

both adults and students are held accountable to those norms.

Similarly, classroom incentives and consequences are mutually agreed upon

and determined by students—supporting the shared commitment to ownership

over the space.

Opportunities for student wisdom to lead class discussions—those most

knowledgeable rather than simply the most authoritative.

Districts and schools also need to engage in intentional diversifying of leaders

within their buildings. By diversifying who sits at tables of decision making they can be a

part of the change of positioning people differently; hire women and people of color in

higher up roles, hire men in IA and support staff roles, hire queer and nonbinary folks and

create spaces that are safe for expression.

Review of hiring practices for implicit bias around concepts like “fit”.

Constant calling out of disproportionality in staffing in schools—especially

compared to student population.

Intentional recruiting from spaces that are reflective of student population

including the community and neighboring professions.

Mentoring and adequate support for both retaining diverse individuals in

education and in supporting their movement through the ranks—avoiding the

stereotype of majority white men becoming administrators and district leaders.

Finally, I call on educators to learn about and openly support discussions of social

constructivism and bias and how it can be perpetuated inadvertently in education.

Educators themselves need to be lifelong learners who routinely access high quality

professional learning not just about how to educate but about how society operates and

impacts their ability to educate effectively and equitably.

Recommendations for Students and Youth

I believe it is the responsibility of the older generations to pave a clearer path for

the younger ones but that does not mean that the youth of today are powerless in

changing the world they will inherit. To students and young people, I have the following

recommendations:

Question the expectations you feel are being placed on you. If you are being

shaped in a way that does not feel organic to who you are or who you want to

be, know that it may be the perpetuation of misguided norms not the wisdom

of your elders.

Find and create spaces for your voice to be meaningfully heard around

decisions, particularly those that will affect your life. Call out ways in which

you are being excluded from conversations and ask why.

Try things that may go against expectations or social norms. The more of you

that explore new possibilities the more you can change those norms.

Find a version of leadership within yourself that is just uncomfortable enough

that you know you are growing. Find a space, any space, where you want to

lead and then find a way to take ownership of that space. You are all capable

of leading and each of your voices is important for reshaping this world.

Do the difficult work of recognizing your spaces of privilege and being part of

shifting that privilege. Understand that we are not a meritocracy and not all of

your leadership is truly earned—that is not to say you do not have a voice

worthy of being heard but it may mean you need to stop, listen, and step back

so others can step forward.

Above all I would say to the youth both involved in my research and not that the

norms of society will change regardless of what you do but you have the power to nudge

it in the direction you see fit. Many of the issues around access and inequity can feel

overwhelmingly large and insurmountable but like all things they were made by us and

can be unmade, slowly, by us as well. I will do what I can to help clear the road, but I

also look forward to the many new, yet to be considered, ways in which you clear your

own road towards equity.

A Parting Message to the Readers

To anyone who has made it to the end of this dissertation I hope you leave with a

sense of curiosity about yourself, your actions, and how you engage or disrupt the status

quo. I hope you question why you lead the way you lead and reflect on when your

leadership took its shape. I ask you to reflect on how you are supporting the leadership of

others in ways that supports individuals rather than societal expectations. I hope that you

remain open to changing your own ways of being. I ask you to continue to grow in your

own leadership and how you nurture leadership in those around you. If you are in a

decision-making position, I ask that you be conscientious of norms you are perpetuating.

Ask givers and holders of power, including yourself, why you feel the way you feel.

Remember that difference is valuable and diversity in how we are positioned is diversity

in who we get to be. Finally, consider stepping into a previously avoided position and be

part of a diversifying world—allow yourself to change and avoid holding on to your

socially constructed self in leu of a self that is far more interesting and beautiful.

HOW WE COME TO LEAD A YOUTH PERSPECTIVE ON UNDERSTANDING MODERN LEADERSHIP

Students also viewed