sustainability

Article
Impact of Exposure to a Counter-Stereotypical STEM Television
Program on Children’s Gender- and Race-Based STEM
Occupational Schema
Fashina Aladé 1, * , Alexis R. Lauricella 2 , Yannik Kumar 3 and Ellen Wartella 4

1 Department of Advertising & Public Relations, Michigan State University, 404 Wilson Rd.,
East Lansing, MI 48824, USA
2 Erikson Institute, 451 N. LaSalle Street, Chicago, IL 60654, USA; [email protected]
3 CVS Health, Chicago, IL 60606, USA; [email protected]
4 Center on Media and Human Development, Northwestern University, 2240 Campus Drive,
Evanston, IL 60208, USA; [email protected]
* Correspondence: [email protected]; Tel.: +1-517-353-2938

Abstract: Gender and racial diversity in STEM has been deemed an essential need for a sustainable
future, but girls and children from underrepresented racial/ethnic backgrounds continue to show
less interest in STEM than their White and male counterparts. Media has been shown to reflect
children’s occupational schema from an early age, and therefore might be used to help broaden
children’s beliefs about who participates in STEM. In this field-based pre/post-experimental study,
children in kindergarten and first grades (N = 48, 62.5% female, Mage = 6.57) viewed episodes of a
STEM-focused educational television series that features a diverse group of protagonists two to three
times a week for eight weeks. Their occupational schema were measured before and after exposure.
Citation: Aladé, F.; Lauricella, A.R.; Results suggest there was no quantifiable change in their attitudes. However, qualitative analysis of
Kumar, Y.; Wartella, E. Impact of
their open-ended responses sheds light on how children’s beliefs about who participates in STEM are
Exposure to a Counter-Stereotypical
shaped, i.e., by both mediated and real-world exposure.
STEM Television Program on
Children’s Gender- and Race-Based
Keywords: occupational attitudes; STEM identity; gender; race; children’s television
STEM Occupational Schema.
Sustainability 2022, 14, 5631. https://
doi.org/10.3390/su14095631

Academic Editors: Anita Tabacco, 1. Introduction

Gavin Duffy, Alicia García-Holgado
Knowledge, skill, and efficacy in science, technology, engineering, and mathematics
and Rachel Riedner
(STEM) are essential for preparing children for an increasingly technological global workforce,
Received: 2 April 2022 which in turn is essential for a sustainable future [1,2]. The likelihood that school-age children
Accepted: 5 May 2022 will participate in the STEM workforce later in life is highly dependent on their early attitudes
Published: 7 May 2022 toward STEM subjects [3,4]. However, U.S. students’ interest in STEM is relatively low when
Publisher’s Note: MDPI stays neutral
compared to their international peers, which contributes to the U.S. lagging behind other
with regard to jurisdictional claims in countries in STEM achievement [5–7]. This is true not only for American children, but also
published maps and institutional affil- for children in similarly industrialized countries where quality of life is generally high and
iations. pressure to pursue specific careers is low [5]. In order to address this ‘STEM gap’ and to move
towards achieving the UN’s Sustainable Development Goals, it is essential that we better
understand how children’s early attitudes towards STEM are formed and what factors might
influence children to want to pursue a STEM career later in life.
Copyright: © 2022 by the authors. Media is one environmental factor shown to shape children’s attitudes and beliefs [8].
Licensee MDPI, Basel, Switzerland. Educational television, especially, has been found to boost children’s positive attitudes
This article is an open access article towards learning and specific school subjects. Fisch [9] explains that one of the ways
distributed under the terms and that educational television can supplement formal education is by encouraging positive
conditions of the Creative Commons
attitudes toward academic subjects, especially among populations that are less likely to
Attribution (CC BY) license (https://
pursue those subjects on their own. This idea that educational television can promote
creativecommons.org/licenses/by/
learning and attitudes towards learning is well documented [10–13]. However, it is unclear
4.0/).

Sustainability 2022, 14, 5631. https://doi.org/10.3390/su14095631 https://www.mdpi.com/journal/sustainability

Sustainability 2022, 14, 5631 2 of 16

from extant research whether educational television can have a measurable impact on
children’s career interests and occupational schema. This paper presents an experimental
study that investigates the ability of a STEM-focused educational television show to shape
children’s expectations about who participates in STEM and in STEM careers.

1.1. Early Attitudes towards STEM Careers

Historically, the association between STEM interest and participation in STEM career
pathways has been studied most rigorously in the context of postsecondary education [14,15].
However, there is recent mounting evidence that attitudes towards STEM develop early and
persist into adulthood. Maltese and Tai [16] found that eighth-grade students who believed
science would be useful in their future and who were interested in a science career were more
likely to earn degrees in STEM. Sadler, Sonnert, Hazari, and Tai [17] found that students’ career
interests when entering high school were the strongest predictors of their career interests when
leaving high school. Going even younger and looking at elementary/primary school children,
Auger, Blackhurst, and Wahl [18] found that first-graders’ self-reported career aspirations were
just as specific and realistic as those of fifth graders. Selkow [19] investigated kindergarten and
first-graders’ occupational choices and found significant differences as a result of maternal
employment status, suggesting that even as early as kindergarten, children’s occupational
interests have real, concrete connections to their real-world exposure and experience. These
findings, combined with the knowledge that achievement gaps in science and math begin
before first grade [20,21], provide strong reason to investigate the ability of educational media
to influence young children’s attitudes and beliefs about who participates in STEM.

1.2. Media and the Development of Occupational Schema

Much of the existing work on the presence of occupational portrayals in the media
has focused on adult-directed television. For example, Signorielli [22] found that women
on television, and particularly African American women, had the least diversity and the
least prestige in terms of the occupations in which they were portrayed. A 2012 report
by the Geena Davis Institute on Gender in Media specifically investigated sex roles and
occupational portrayals in children’s television and family films. The authors found that
despite women comprising 47% of the U.S. labor force, they represented only 25% of
employed characters in children’s television shows [23].
These media portrayals begin to have an effect on viewers from an early age [24].
Research suggests that young children are likely to develop occupational schemas that
reflect the images they see on television [22], and that gender- and race-based occupational
schemas are in place by age six [25,26]. Jeffries-Fox and Signorielli [27] found that middle
school children who watched more television were more likely to want more glamorous and
high-status jobs, like those most often portrayed on TV. Similarly, Signorielli [28] found that
high school students who watched more television aspired to high-status jobs that allowed
them to make a lot of money, much like the TV characters they frequently viewed. In 1974
study, Beuf [29] looked at preschoolers’ gendered occupational aspirations in relation to the
amount of television they viewed. She found that children held sharply contrasting beliefs
about which occupations should be held by men and which should be held by women. For
example, on average, children reported that doctors and telephone repairpersons could
only be men, while feeding a baby was a woman’s job. Though the specific examples used
in that study are now somewhat dated, the major finding of Beuf’s [29] study that is still
relevant today was that the more television the children watched, the more likely they were
to apply gender stereotypes to careers and household work roles.

1.3. The Need for Diverse STEM Portrayals

Not only are American children’s attitudes towards STEM concerningly low [5,30],
their ideas about who participates in STEM are also quite narrow. A meta-analysis by
Miller and colleagues [31] showed that, despite gradual improvement over the last fifty
years, when children are asked to draw a scientist, the majority of drawings still depict
Sustainability 2022, 14, 5631 3 of 16

White men. These impressions are not inaccurate as women and people of color are
largely underrepresented in STEM fields in the United States and other western industrial
nations [32]. Blickenstaff [33] writes that there is no singular reason why this is the case,
but that it is a complex and multi-faceted problem. The underrepresentation of female
characters and racially diverse characters in STEM television shows [34,35] is likely one
contributing factor.
It is not only critical that girls and children of color believe they can pursue STEM, but
also that all children believe that girls and children of color can pursue STEM. Studies show
that women, and particularly women of color, often drop out of the STEM pipeline even after
receiving STEM degrees in higher education because of hostile work environments [17,36].
In other words, it is not that they do not believe they are smart enough or capable enough;
the problem often lies in the fact that their White male colleagues do not see them as equal
and valuable. Many studies have looked at ways to increase girls’ self-efficacy in STEM,
e.g., [37–39], or racial minority students’ self-efficacy in STEM, e.g., [40,41], but in addition
to boosting self-efficacy for individuals from these underrepresented groups, we must also
work to broaden all children’s ideas about who participates in STEM. Accordingly, this study
utilizes a sample of children that roughly reflects the population of the U.S. in terms of gender
and race in order to investigate whether exposure to a STEM television show that features girls
and racially diverse characters can impact children’s ideas about who holds STEM careers.

1.4. Television Viewing and Attitude Change

1.4.1. Theoretical Frameworks
The relationship between television exposure and attitude change is complex, but
significant effects have been shown across many domains, from political attitudes and
beliefs to health attitudes and behaviors [42,43]. For children specifically, researchers
have found effects of television exposure on social attitudes [44], racial attitudes [45], sex
role attitudes [46], and attitudes towards learning [11]. Cultivation theory [47] and social
cognitive theory [48] are two useful frameworks for understanding the relationship between
exposure and attitudes. Cultivation theory suggests that heavy television viewers are more
likely to adopt attitudes that are consistent with the themes portrayed on television. Via
long-term, cumulative media exposure, we come to view the world through the perspective
that is most dominant in the media [47]. Social cognitive theory posits that people learn
not only through their own experience, but also through the observation of others’ actions,
including those of media figures, and the consequences of those actions [49]. In the context
of the current study, both of these theories would suggest that exposure to programs
that portray all different types of characters modeling STEM behaviors, and, importantly,
showing intrinsic and extrinsic rewards for STEM participation, would lead children to
believe that they also can and should participate in STEM, whether they are boys, girls,
Black, White, etc.

1.4.2. Single vs. Repeated Exposure

Specifically looking at exposure to a science show and children’s reported interest in
science, Mares, Cantor, and Steinbach [50] found that while one-time exposure did not
affect attitudes, repeated exposure over the course of eight weeks did have a significant
effect on attitudes towards science. Other studies have similarly found long-term and/or
regular exposure to be more effective at influencing attitudes than short-term or one-time
exposure. Interestingly, Bond [37] found that a single, one-time exposure was enough to
influence attitudes in a negative, stereotypical direction. However, there were no effects
of one-time exposure for children who were exposed to positive, counter-stereotypical
STEM depictions. The author points out that more research is needed to determine if long-
term and/or repeated exposure to counter-stereotypical STEM depictions might influence
children in a positive way. Building from that work, and guided by the tenets of cultivation
and social cognitive theories, the current study investigates the effects of eight weeks of
 influence attitudes in a negative, stereotypical direction. However, there were no effects
of one-time exposure for children who were exposed to positive, counter-stereotypical
STEM depictions. The author points out that more research is needed to determine if long-
term and/or repeated exposure to counter-stereotypical STEM depictions might influence
Sustainability 2022, 14, 5631 children in a positive way. Building from that work, and guided by the tenets of cultiva- 4 of 16
tion and social cognitive theories, the current study investigates the effects of eight weeks
of exposure to counter-stereotypical depictions of STEM engagement. Figure 1 presents a
graphical depiction
exposure of the theory of change
to counter-stereotypical undergirding
depictions of STEM this study.
engagement. Figure 1 presents a
graphical depiction of the theory of change undergirding this study.

Figure 1. Theoretical model of the effects of repeated exposure to STEM-focused educational televi-
Figure
sion 1. Theoretical
featuring model
diverse of the effects
characters. Thisofmodel
repeated exposure elements
incorporates to STEM-focused educational
of Bandura’s televi-
social cognitive
sion featuring
theory diverse characters.
and Gerbner’s cultivationThis model incorporates elements of Bandura’s social cognitive
theory.
theory and Gerbner’s cultivation theory.
1.5. The Current Study
1.5. TheBased
CurrentonStudy
the theoretical frameworks presented above and the limited extant literature
in Based
this area, we hypothesized
on the that repeated
theoretical frameworks exposure
presented to counter-stereotypical
above and the limited extant STEM por-
litera-
trayals might in fact positively influence young children’s attitudes and beliefs
ture in this area, we hypothesized that repeated exposure to counter-stereotypical STEM about who
participates
portrayals mightin STEM. Specifically,
in fact positively we hypothesized:
influence young children’s attitudes and beliefs about
who participates in STEM. Specifically, we hypothesized:
Hypothesis 1 (H1). Repeated exposure to a counter-stereotypical STEM program will be related
to broader ideas
Hypothesis 1 (H1).about womenexposure
Repeated and girlstoparticipating in STEM compared
a counter-stereotypical to a control
STEM program will group.
be related
to broader ideas about women and girls participating in STEM compared to a control group.
Hypothesis 2 (H2). Repeated exposure to a counter-stereotypical STEM program will be related
to broader ideas about individuals from underrepresented racial/ethnic backgrounds participating in
Hypothesis 2 (H2). Repeated exposure to a counter-stereotypical STEM program will be related
STEM compared to a control group.
to broader ideas about individuals from underrepresented racial/ethnic backgrounds participating
in STEM compared
Given to a control
the relative group.
dearth of research exploring children’s race- and gender-based
occupational attitudes, we also sought to explore how children themselves would describe
Givenexplain
and/or the relative
their dearth of these
biases in research exploring
areas. children’s
Specifically, race- and gender-based oc-
we asked:
cupational attitudes, we also sought to explore how children themselves would describe
RQ1: (a) How do children articulate their explicit attitudes and beliefs about who partic-
and/or
ipatesexplain their(b)
in STEM? biases
Whatininformation
these areas. Specifically,
to they drawwe asked:
from to form judgments about who
participates in STEM?

2. Materials and Methods

2.1. Study Design
To address our hypotheses and research questions, we conducted a pre/post-
experimental study that took the shape of a media intervention. The study was conducted
in before- and after-school childcare centers in a suburb of a large city in the Midwestern
United States. Over the course of eight weeks, participants in the treatment group regularly
viewed episodes of Cyberchase, a STEM show that features main characters who are diverse
Sustainability 2022, 14, 5631 5 of 16

in terms of race and gender and exhibit high levels of STEM engagement. Participants in
the control group went about business as usual. All participants completed assessments
before and after the eight-week exposure period.

2.2. Study Sites and Recruitment

After approval of the study by the university’s Institutional Review Board, three local
childcare programs were contacted and informed about the study and agreed to take part.
These childcare centers provide before- and/or after-school care for elementary school-aged
children. The programs are relatively unstructured; students can engage in a variety of
games and activities, do homework, and eat snacks. Center directors sent consent forms
home to parents of all kindergarten and first graders at their centers. Parents who consented
to have their child participate were also asked to complete an online survey to provide
demographic information about their family. One center had two participating locations,
resulting in four study sites. The study sites were randomly assigned to either the treatment
or control condition.

2.3. Participants
A total of 55 children were recruited into the study across the three participating
centers. Of the 55 children whose parents granted consent, 48 successfully completed all
components of the study (i.e., pretest assessment, posttest assessment, and, for treatment
group, viewed at least 50% of the episodes) and, thus, could be included in analyses. The
final sample of 48 children (n = 25 treatment, n = 23 control) were 62.5% female and ranged
in age from 5.5 to 7.53 years (M = 6.57, SD = 0.52). The sample was quite diverse in terms
of race/ethnicity and also included some socioeconomic diversity (i.e., about 20% of the
sample reported household incomes between USD 60,000 and 85,000, whereas the median
household incomes in the school districts we sampled from range from USD 85,469 to
123,618). Table 1 presents demographic information of the treatment and control groups.
t-tests revealed no significant differences between treatment and control groups on any
demographic variables.

Table 1. Participant Demographics.

Variable Control Treatment

Mean (SD) Percent Mean (SD) Percent
Child Age (in years) 6.29 (0.67) 6.83 (0.48)
Gender
Males 43.5 45.4
Females 56.5 54.5
Race
White 56.5 66.7
Black 26.1 16.7
Asian 4.3 4.0
Hispanic 4.3 16.7
Other 8.6 8.0
Household Income
USD 60,000–84,999 21.3 18.2
USD 85,000–99,999 9.0 0.0
>USD 100,000 69.7 81.8
Parental Education
Bachelor’s Degree 43.5 36.4
Master’s Degree 29.1 36.4
Professional/Doctorate Degree 27.4 27.2
Parent Age (in years) 37.30 (4.76) 39.91 (2.43)
Parent’s Relationship to Child
Mother 78.3 64.0
Father 21.7 36.0
Sustainability 2022, 14, 5631 6 of 16

2.4. Procedure
2.4.1. Pre- and Posttest
Participants were interviewed in a quiet space in their childcare classroom. Each one-
on-one interview session lasted approximately 15–20 min. In addition to the occupational
attitudes measure reported in this study, participants also completed several additional
measures reported in a separate study: program and genre familiarity, attitudes towards
STEM, program appeal, and parasocial relationships. Interviews were audio recorded for
later transcription. It took approximately two weeks to complete pretest interviews for all
participants across the three centers. After exposure, posttest interviews took an additional
two weeks.

2.4.2. Exposure
Participants in the treatment condition watched an episode of Cyberchase two or three
times a week for eight weeks (the exposure length was determined based on research by
Mares et al., 1999). Viewing sessions were conducted by the lead researcher or a trained
research assistant. For each viewing session, the researcher took the participating students
to one side of the classroom or an adjacent classroom. The episodes were shown using a
classroom projector with children sitting on a rug approximately four to six feet in front
of the screen. Viewing occurred in groups of 8–14 children at a time. The researcher took
attendance at each viewing session to ensure that all participating children watched at
least two episodes every week. In order to accomplish that goal, sometimes the researcher
would need to come in on a third day, which resulted in some children occasionally viewing
three episodes in a week. Children were instructed to sit and watch the video quietly. Most
children were excited about the opportunity to watch the show, and thus were quite attentive
throughout. There was often laughter at jokes, and occasionally children made remarks
aloud such as, “Why are they doing that”. The researchers would reply with a reminder to
keep watching, such as “I don’t know, let’s keep watching and see what happens”.
During the eight-week exposure period, children in the control condition went about
business as usual. They did not view any videos during their before- or after-school
programs. They participated in activities that were very similar to what children in the
treatment group did when they were not in viewing sessions, such as coloring, reading
books, making jewelry, playing board games, etc.

2.5. Stimulus
The stimulus program for this study was chosen using data from a prior content
analysis of children’s STEM-focused television [34]. Our goal was to select a STEM program
that (A) showed high levels of STEM engagement from main characters (i.e., modeling of
STEM behaviors such as asking questions, making observations, investigating, and problem
solving), and (B) did a particularly good job of featuring characters that were diverse in
terms of both gender and race. To accomplish this goal, we created a mathematical formula
that ranked each of the 30 STEM programs included in the content analysis on both of
those features. Since the main unit of analysis in coding was the individual character, the
value of a show was posited to be the sum of the value of each individual character within
that show. Each character’s value was calculated as the sum of their STEM engagement
weighted by the centrality of the character to the show (i.e., major characters were weighted
more heavily than minor characters), with multipliers for female and minority characters.
A show value for each program was then calculated as the sum of all character values, and
the 30 programs were ranked based on their show values. Creation of the formula was an
iterative process that included several stages of trial and error to ensure that the rankings
produced by the formula both matched expectations of face validity and took the form of a
roughly normal distribution with appropriate variance, skew, and kurtosis. The program
that came out as number one in the final ranking (as well as in several other iterations of the
formula and rankings) was Cyberchase, and thus, this was chosen as the stimulus program.
Sustainability 2022, 14, 5631 7 of 16

Cyberchase is an animated PBS show designed to teach math concepts “in a fun way
that kids can understand” (http://www.pbs.org/parents/Cyberchase/about-Cyberchase/,
accessed on 17 March 2018). The main team of protagonists includes Jackie, an African
American girl, Inez, a Latin-American girl, and Matt, an Irish-American boy. These three
main characters, often referred to as the Cybersquad, are accompanied by Digit, a “cybird”
who helps them solve problems in order to defeat the evil villain Hacker. Along the way,
they learn that math is everywhere and that it is a useful tool for solving problems. In “For
Real,” the live-action segment following each animated episode, adults show the viewers
how math can help solve life’s problems in the real world. Common Sense Media rates
Cyberchase as appropriate for children ages 5 and up, which made it perfectly appropriate for
the kindergarten and first graders in the study sample. Participants were shown a random
selection of episodes from the two most recent fully available seasons of Cyberchase (seasons
8 and 9). The episodes were shown in relative sequential order, though the narratives are
not interdependent, so missing an episode did not pose any comprehension problems for
the participants. Anecdotally speaking, participants were unfamiliar with Cyberchase before
the study, but seemed to enjoy the show quite a bit, and were always excited when it was
time to watch another episode.

2.6. Measures
2.6.1. “Me/Not Me” Self-Identification Task
In order to have an understanding of the individual contexts participants were bring-
ing into the study, it was important to know how the children self-identified, both in
terms of gender and race, and in terms of their relationship to STEM. Self-identification
was measured using the “Me/Not Me” task, adapted from Rogers and Meltzoff [51]. The
task included the following social identity labels: boy, girl, daughter, son, brother, sister,
Asian, Black, Latino/a, White, student, athlete, artist, scientist, problem solver, investigator.
Following the procedure used by Rogers and Meltzoff (2017), the order of presentation
was held constant for each participant. For each card, the child was asked, for example,
“Are you a [boy]?” and told to sort the card into the “Me” pile if the label described them
and into the “Not me” pile if it did not. After all cards were sorted in this manner, the
researcher verbally went through each pile again and gave the child an opportunity to
revise any selections. Labels placed into the “Not Me” pile were coded as 0, and labels
placed into the “Me” pile were coded as 1.

2.6.2. Occupational Attitudes

The occupational attitudes measure, used at pre- and posttest, was created specifically
for this study and was informed by previous research [25,26,29]. Due to time and space
constraints of data collection, we did not use a traditional implicit association test (IAT) for
this measure. Instead, we used pictorial response options and asked for children’s quick
‘gut reactions’ as a quasi-implicit measure (adapted from [25,26]), and then asked them to
verbally elaborate on their response in order to capture their explicit attitudes (adapted
from [29]).
The measure consisted of 12 items, six of which were designed to capture children’s
adherence to gender stereotypes about STEM occupations, and the other six of which were
designed to capture children’s adherence to racial stereotypes about STEM occupations.
For the first four items, participants were shown an illustrated picture of a boy and a girl.
The boy and girl were dressed very similarly in a t-shirt and jeans and both were carrying a
book and wearing a backpack (i.e., intended to look like ordinary schoolchildren), and both
were Caucasian with brown hair and brown eyes (i.e., holding race as a constant while
manipulating gender). Participants were asked to select “Which of these kids is more likely
to grow up and become a [scientist, mathematician, engineer, computer programmer]?”
Participants could answer verbally or point to the picture of the boy or girl to select their
answer. After each question, the researcher asked the participant to explain why they
chose the answer they gave, and responses were recorded verbatim for later analyses.
Sustainability 2022, 14, 5631 8 of 16

The next four questions were identical to the first four, except that rather than a picture
of a boy and girl, the picture showed two boys, one White and one Hispanic, who were
otherwise depicted very similarly, both with a backpack and book and similar clothing
(i.e., manipulating race while holding gender as a constant). Again, participants were
asked to select “Which of these kids is more likely to grow up and become a [scientist,
mathematician, engineer, computer programmer]?” and were asked to explain why they
thought that was true.
The final four questions consisted of four different drawings: a Black male scientist
next to a Black female scientist, a White female computer programmer next to a White male
computer programmer, a Black male airline pilot next to a White male airline pilot, and a
Black female doctor next to a White female doctor (the former two manipulating gender
and the latter two manipulating race). For each drawing, participants were asked “Which
of these two people do you think really works as a [scientist, computer programmer, airline
pilot, doctor]?” and “Why do you think that?”. In each drawing, the two people were
dressed in the same clothes and carrying the same materials and were drawn to look as
identical as possible aside from the gender or race. All drawings were created specifically
Sustainability 2022, 14, x FOR PEER REVIEW 9 of 17
for this study by a locally commissioned artist to allow for a clean manipulation of only the
race or gender while keeping everything else about the pictures constant. Figure 2 shows
examples of the images used in this measure.

(a) (b)

Figure 2. Examples of image prompts from the occupational attitudes measure. Participants were asked
Figure 2. Examples of image prompts from the occupational attitudes measure. Partici-
“Which of these two people do you think really works as a [(a) computer programmer] [(b) doctor]?”.
pants were asked “Which of these two people do you think really works as a [(a) com-
There were twelve items total: (a) six items manipulated gender while holding race constant; (b) six
puter programmer] [(b) doctor]?”. There were twelve items total: (a) six items manipu-
items manipulating race while holding gender constant.
lated gender while holding race constant; (b) six items manipulating race while holding
gender Toconstant.
create a quantitative measure of participants’ gender- and race-based occupational
attitudes, responses
To create from themeasure
a quantitative six gender manipulationgender-
of participants’ items were
and summed
race-basedtooccupa-
create a
scale that measured participants’ adherence to traditional gender stereotypes about STEM
tional attitudes, responses from the six gender manipulation items were summed to create
occupations, and responses from the six race manipulation items were summed to create a
a scale that measured participants’ adherence to traditional gender stereotypes about
scale that measured participants’ adherence to traditional racial stereotypes about STEM
STEM occupations, and responses from the six race manipulation items were summed to
occupations. For each item, the more stereotypical responses (White or male) were scored
create a scale that measured participants’ adherence to traditional racial stereotypes about
as a 1 and the less stereotypical responses (non-White, female, or both) were scored as
STEM occupations. For each item, the more stereotypical responses (White or male) were
a 0, such that higher scores on the summed 0–6 scales represented greater adherence to
scored as a 1 and the less stereotypical responses (non-White, female, or both) were scored
stereotypes.
as a 0, such that higher scores on the summed 0–6 scales represented greater adherence to
stereotypes.
2.6.3. Coding of Open-Ended Responses
To explore RQ1, we conducted a qualitative content analysis of participants’ responses
2.6.3. Coding of Open-Ended Responses
to the open-ended ‘why’ component of the measure. Qualitative content analysis is an
To explore
analytical RQ1, that
technique we conducted
draws on the a qualitative
strengths content
of both analysis
qualitativeof participants’ re-
and quantitative
sponses
researchtoby
thefirst
open-ended ‘why’
qualitatively component
assigning of the
codes measure.
to each Qualitative
textual content and
unit of analysis analysis
then
isanalyzing
an analytical
the technique
frequency that draws
of each on occurrence
code the strengths of both
[52]. qualitative
Following and quantitative
standard procedures
research by first qualitatively assigning codes to each textual unit of
for qualitative content analysis of interview data, responses were first transcribed analysis and then
from
analyzing the frequency of each code occurrence [52]. Following standard procedures
audio recordings and researcher notes. Responses were typically one or two sentences long. for
qualitative
Two coders, content analysis
the lead authorofand
interview data,
a research responses
assistant, were
then readfirst transcribed
through from audio
the responses and
recordings and researcher
devised separate notes.
preliminary Responses
coding were
schemes fromtypically one
this first or two
pass sentences
through long.
the data. Two
Coders
coders, the lead author and a research assistant, then read through the responses and de-
vised separate preliminary coding schemes from this first pass through the data. Coders
then discussed their preliminary coding schemes, compared codes, and reconciled any
differences to form a more refined coding scheme. The coders then double coded a sample
of the data and established strong inter-rater agreement (Cohen’s ĸ = 0.938, p < 0.001),
Sustainability 2022, 14, 5631 9 of 16

then discussed their preliminary coding schemes, compared codes, and reconciled any
differences to form a more refined coding scheme. The coders then double coded a sample
of the data and established strong inter-rater agreement (Cohen’s k = 0.938, p < 0.001),
before individually coding the remaining responses.
The final coding scheme consisted of 15 codes. The exposure code was given to
responses that cited the media they’ve consumed, their family, or their own personal
experience (e.g., “my grandpa is an engineer” or “I saw someone like that on a TV show”).
There were two codes for race-based responses and two codes for gender-based responses.
Responses that cited a character’s race or gender to justify choosing the more stereotypical
option (e.g., “boys are mostly better than girls at math”) were coded as race-stereotypical or
gender-stereotypical. Responses that cited a character’s race or gender to justify choosing
the counter-stereotypical option (e.g., “because girls are better than boys at math”) were
coded as race-counter-stereotypical or gender-counter-stereotypical. The exposure code could be
combined with these race and gender codes to label responses that expressed race/gender
stereotypical/counter-stereotypical beliefs due to exposure (e.g., “I saw a movie and the
girl knew how to use a computer”), resulting in four additional codes: race-stereotypical
exposure, gender-stereotypical exposure, race-counter-stereotypical exposure, and gender-counter-
stereotypical exposure. Responses that expressed a belief about equality and picked both
characters (e.g., “both boys and girls can do it”) were given an equality code. The visual
characteristics code was given to any response that referred to a specific visual element of the
picture (e.g., “he has a book about math” or “he is smiling”). Explanations that were about
a character’s ability to perform the occupation (e.g., looks smarter, seems stronger, knows
science/math) were given a skills code. Vague, non-descriptive responses (e.g., “they look
like it”) were coded as non-committal. When a child did not offer an explanation or said
“don’t know” when asked why, their response was coded as don’t know. Responses that
did not fit into any of the aforementioned categories were coded as other. Each response
could only receive one code, so coders were instructed to choose the code that seemed most
prominent in the response.

3. Results
3.1. Between-Groups Effects
All analyses were completed using SPSS version 24. Prior to hypothesis testing, we
used regression analyses to check for differences in all variables of interest by classroom, by
treatment group, and by researcher who conducted the interview. No significant differences
were found, and therefore these variables were not included as covariates in the analyses.
To test the first two hypotheses, which dealt with differences between treatment
and control in their change in attitudes over time, we ran Mixed Analyses of Variance
(ANOVAs). A mixed ANOVA compares the mean differences between groups that have
been split on two factors, or independent variables, where one is a “within-subjects” factor
(in this case, time) and the other factor is a “between-subjects” factor (in this case, condition).
That is, it tests whether there is a significant difference in slope of the outcome variable
from pretest to posttest between the treatment group and control group.
A 2 × 2 mixed ANOVA was used to test H1, which hypothesized that participants in
the treatment group would have a more positive change in their gender-based occupational
STEM attitudes from pretest to posttest than participants in the control group. There was no
main effect of time (F(1, 45) = 2.28, p = 0.138), no main effect of condition (F(1, 45) = 0.297,
p = 0.588), and no interaction between time and condition (F(1, 45) = 1.38, p = 0.247).
A 2 × 2 mixed ANOVA was used to test H2, which hypothesized that participants in
the treatment group would have a more positive change in their race-based occupational
attitudes from pretest to posttest than participants in the control group. There was no
main effect of time (F(1, 46) = 0.575, p = 0.452), no main effect of condition (F(1, 46) = 1.32,
p = 0.257), and no interaction between time and condition (F(1, 46) = 1.20, p = 0.280). Taking
the results of H1 and H2 together, we find that children who watched Cyberchase did not
Sustainability 2022, 14, 5631 10 of 16

experience a measurable broadening of their race- or gender-based occupational attitudes

about who participates in STEM, compared to children in the control group.

3.2. Open-Ended Occupational Responses

Responses to the ‘why’ component of the occupational attitudes measure were coded
to investigate RQ2. There was no observable difference of the frequency of any type of
response between the treatment and control groups. Of all the types of responses, don’t
know was the most common, making up 26.4% of all 1152 responses. The following results
exclude don’t know results and look at the responses of both the treatment and control
groups together.
For items that manipulated gender, responses that cited visual characteristics of the
character were most common, making up 42.9% of the 438 responses. Children would
often point to a character’s backpack or book as a reason why they were more likely to
hold a STEM occupation, despite the fact that both characters had a backpack and a book.
Gender-stereotypical responses made up 7.1% (n = 31) of the total. Children who gave these
responses either talked about how it was normally boys who did science and built things
(e.g., “because I thought only boys could study bugs or trees or leeches”, “because girls
usually aren’t interested in building things”) or how girls did not belong in the profession
(e.g., “because girls weren’t expected to do it. They were expected to be nurses. Having
a job was off-limits”). Only 2.7% (n = 12) of the total responses were gender-counter-
stereotypical. The majority of these were expressed in relation to exposure (e.g., “because
one of my great grandmas was an engineer”, “I keep watching shows about girls doing
engineer stuff . . . I also watched a movie about it. It was a Ghost Busters movie . . . one of
the girls was an engineer”). Non-committal responses made up 22.8% and skills responses
made up 12.3% of the total 438 responses. Only 1.8% (n = 8) of responses expressed a belief
about equality. 9.4% of responses did not fit into any other category and were coded as
other. See Table 2 for the breakdown of frequencies of each code.

Table 2. Qualitative Codes for Gender-Based Occupational Attitude Responses.

Code Frequency Percent

Visual characteristics 188 42.9
Non-committal 100 22.8
Skills 54 12.3
Other 41 9.4
Gender-stereotypical 28 6.4
Equality 8 1.8
Gender-counter-stereotypical exposure 7 1.6
Exposure only 5 1.1
Gender-counter-stereotypical 5 1.1
Gender-stereotypical exposure 3 0.7
Total 438 100
Note: This table excludes don’t know responses.

For items that manipulated race, responses that cited visual characteristics were also
most common, making up 33.9% of the 410 responses. As with questions about gender,
children often used books and backpacks to differentiate the characters, despite the fact
that both characters had identical accessories. Race-stereotypical responses made up 7.1%
(n = 29) of the total. There were a variety of justifications for these responses. Some
children held ideas that linked race to ability (e.g., “White skinned people take extra
care of their people [patients]”), a few children held ideas about whether a particular
job was appropriate for Black people (e.g., “Because doctors don’t accept Blacks into
medical school—Blacks don’t get very good jobs”), and a couple of children came up with
nonsensical explanations (e.g., “he had White skin, so rain won’t show on his hands” as an
explanation for picking a White pilot over a Black pilot). Only 1.5% (n = 6) of responses were
race-counter-stereotypical (e.g., one child thought that it was “mostly Black people who go
Sustainability 2022, 14, 5631 11 of 16

to become doctors”). Non-committal responses made up 29.5%, and skills responses made
up 14.4% of the total. Beliefs about equality were expressed in 2.9% (n = 12) of responses,
and 9.8% of responses did not fit into any other category and were coded as other. See
Table 3 for the breakdown of frequencies of each code.

Table 3. Qualitative Codes for Race-Based Occupational Attitude Responses.

Code Frequency Percent

Visual characteristics 135 33.9
Non-committal 121 29.5
Skills 59 14.4
Other 40 9.8
Race-stereotypical 24 5.9
Equality 12 2.9
Race-stereotypical exposure 5 1.2
Exposure 4 1.0
Race-counter-stereotypical 4 1.0
Gender-stereotypical 3 0.7
Race-counter-stereotypical exposure 2 0.5
Gender-counter-stereotypical 1 0.2
Total 410 100
Note: This table excludes don’t know responses.

Across the gender and race questions, responses that cited skills often talked about
‘knowing’ and ‘smartness’. One other oft-cited skill was strength, though this was ex-
clusively in relation to questions about being an engineer. There were no observable
differences in the frequency of any type of response between pre-test and post-test for the
treatment group or the control group.

4. Discussion
4.1. Summary and Interpretation of Results
The importance of a diverse and technologically skilled workforce for our future
existence has been well documented in the sustainability literature [2,53,54]. Thus, many
programs have been developed in recent years to engage girls and children from diverse
backgrounds in STEM from an early age. One way to reach children en masse is via the mass
media. Specifically, curriculum-backed public television has been shown to have positive
effects on children’s learning across many domains. Yet, connections between STEM-
focused educational television and young children’s career interests and aspirations have
been yet untested. To investigate the effect of repeated exposure to a counter-stereotypical
STEM show on children’s occupational attitudes about who participates in STEM and
hold STEM careers, we conducted a pre/post-, between-subjects experimental study with
48 children in kindergarten and first grades. For eight weeks, participants in the treatment
group regularly viewed episodes of Cyberchase, a children’s television program that models
high levels of STEM engagement from a diverse group of characters. Participants in the
control group went about business as usual. All participants completed assessments before
and after the eight-week exposure period.
Our two hypotheses were not supported. Analyses revealed no significant differences
between treatment and control groups in their change in occupational attitudes. There was
also no main effect of time; for both treatment and control groups, attitudes, on average,
did not change from pretest to posttest. This was true whether students started out with
high or low attitudes, was true for boys and girls, as well as for younger and older children.
However, to probe this issue a bit further and investigate children’s reasoning behind
their responses, we conducted an exploratory qualitative content analyses of children’s
open-ended responses to the occupational attitudes measure. The results across gender
and race questions were very similar. With both kinds of questions, many children pointed
to superficial visual characteristics of the images to explain their choices. Additionally,
Sustainability 2022, 14, 5631 12 of 16

many children also gave non-committal responses, which might reflect their inability or
reluctance to articulate ideas of race and gender or their compulsion to give an answer for
the sake of the researcher. Only a minority of children articulated ideas about race and
gender, but when they did, it was most-often in a stereotypical way. Of those gender- and
race-based responses, a recurrent theme was that their belief was based on some kind of
exposure; a handful of children cited media they had consumed, some referred to what
their family members did, and others mentioned what they had seen personally.
It was clear that these open-ended data presented a very different picture than the
closed-ended response data. Unlike with our quantitative measure of attitudes, which did
not show any clear patterns, explicit race and gender stereotypes were quite prevalent
in children’s open-ended responses, showing that children in this age group do express
constrained beliefs about who participates in STEM. Encouragingly, there were also some
counter-stereotypical ideas put forth by the children, such as expressions of equality and
efficacy for all types of people to participate in STEM. However, these were much less
frequent than the stereotypical responses. As with the quantitative measure, there did not
appear to be any differences between responses from participants in the treatment and
control groups. Taken together, these findings provide valuable contributions to theory and
practice, but must be considered in light of some study limitations.

4.2. Limitations
The sample for this study was limited in several ways. First, the sample size was
relatively small and may not have been sufficient for detecting small effects. Though
the target sample size was larger, the time-intensive nature of the data collection made
it impossible to recruit more participants. Additionally, though we made an effort to
recruit childcare centers who cater to diverse families, and indeed had a racially diverse
sample, there was not much socioeconomic diversity within the sample. Participants’
parents reported relatively high education levels and household incomes. The hypotheses
and research questions presented here should be further explored with a larger and more
socioeconomically diverse sample in order to make more generalizable claims and to have
more confidence in the observed results.
There were also some methodological limitations to this study. Most notably, the
measure of occupational attitudes was created specifically for this study and therefore
had not been previously validated. Since other studies have successfully used Implicit
Association Tests (IATs) with young children, future research should utilize an IAT to
investigate implicit gender- and race-based STEM occupational attitudes. There is also an
inherent limitation of using only one stimulus program rather than a variety of television
shows as stimuli. The effects of this study may be specific to the program, and thus, may
not be generalizable to educational STEM television in general.
Finally, it should be noted that discussions of race and gender are highly subjective
to the culture in which they are being discussed. The measures used in this study were
chosen and developed specifically for a U.S. context. However, the research questions
addressed are certainly globally relevant; researchers investigating these questions in
different geographic contexts should be sure to use culturally appropriate adaptations of
the measures used here.

4.3. Theoretical and Methodological Contributions

Despite these limitations, the results of this study provide valuable contributions
to our understanding of the early development of race- and gender-based occupational
attitudes. Overall, these findings suggest that a single counter-stereotype, even with
repeated exposure, may not be strong enough to override the stereotypical occupational
schema that are already in place by kindergarten. This runs counter to effects predicted
by the drench hypothesis, put forth by Greenberg [55], which posits that a few critical
portrayals of non-traditional roles can alter stereotypical attitudes. Greenberg argues that
exposure to just a few salient counter-stereotypical portrayals can cut through the buildup
Sustainability 2022, 14, 5631 13 of 16

of traditional stereotypes to produce attitude change in viewers. The drench hypothesis was
created in contrast to the drip hypothesis, which states that the dominant stereotype-laden
content present in our provides a steady drip of stereotyped media portrayals, reinforcing
and cultivating stereotype-based attitudes. There has been some empirical evidence that
supports the drench hypothesis [56–59]. However, the results of this study suggest that
the effects of the stereotypical ‘drip’ might in fact be too powerful to be overridden by a
counter-stereotypical ‘drench’.
This is much in line with what Bond [37] discovered. In his study of six- to nine-year-
old girls, videos depicting stereotypical portrayals of women were very effective at altering
participants’ attitudes to be less favorable towards STEM. However, videos depicting
counter-stereotypical portrayals were not effective at creating more favorable attitudes
towards STEM. Bond explains that this is likely due to the fact that stereotypical gender
schemata are already so strongly in place in children’s minds that new information that
maps onto those schemata is easily processed by the brain. However, without pre-existing
schemata to map onto, the counter-stereotypical images did not have as strong of an effect.
One important nuance to consider is that previous studies that found support for the
drench hypothesis looked at outcome measures such as recall of a counter-stereotypical
character portrayal [56,59], or self-reported familiarity and comfort with an outgroup [57,58].
This study, like Bond’s [37], went a step further to see if those counter-stereotypes could
actually impact attitudes. Considering these findings together, it seems likely that the drench
of a counter-stereotype may be helpful for recall and familiarity, but not as effective for truly
altering people’s worldviews.
In terms of effective exposure periods, this study adds to extant literature on television
exposure and STEM attitudes by moving beyond a single exposure and looking at repeated
exposure over the course of eight weeks. Eight weeks has been shown to be a sufficient
time period for other similar media interventions [50,60], however, in this study we did
not observe such effects. Cultivation theory does suggest that it is cumulative exposure
over years that affects the way people view the world. Thus, in the case of changing
occupational schema, it may be that even eight weeks is not a long enough exposure
period. Future research should investigate the effects of even longer-term exposure to
counter-stereotypical character portrayals to see if there might be more measurable effects.

5. Conclusions
In sum, this multi-method analysis of the effects of educational television on children’s
gender- and race-based STEM occupational attitudes provides nuanced, but important
insights. The quantitative measure of attitudes did not reveal any patterns or trends, which
may be an indication that these attitudes towards STEM careers are not yet strongly formed
for children in this age group. Prior work has shown that by age six, children do have ideas
in place about which genders and races perform certain occupations [25,26], but this had
not been examined specifically in the STEM context. It may be that children are not familiar
enough with STEM occupations to have strong ideas about who should perform them.
Analysis of participants’ open-ended responses showed that exposure is an important
precursor to attitudes; when children had gender- or race-based beliefs about who should
hold certain occupations, it was most often because they either personally knew someone
who holds that occupation or had seen a relevant portrayal on TV or in a movie. Theories
based in psychology and communication science certainly suggest that media can be a
useful tool for pushing us towards a sustainable future, particularly in terms of children’s
interest in STEM. This study serves as a useful starting point, but more research is needed to
continue to investigate this relationship between media exposure, children’s occupational
schema, and their eventual decisions about whether or not to pursue and persist in STEM.

Author Contributions: Conceptualization, F.A.; methodology, F.A.; formal analysis, F.A. and Y.K.;
writing—original draft preparation, F.A.; writing—review and editing, F.A., A.R.L., Y.K. and E.W.;
project administration, F.A. and A.R.L.; funding acquisition, E.W. All authors have read and agreed
to the published version of the manuscript.
Sustainability 2022, 14, 5631 14 of 16

Funding: This research was supported by a grant from the National Science Foundation Reese Pro-
gram, DRL-1252121 to Ellen Wartella. Any opinions, findings, and conclusions or recommendations
expressed in these materials are those of the author(s) and do not necessarily reflect the views of the
National Science Foundation.
Institutional Review Board Statement: The study was conducted in accordance with the Declaration
of Helsinki and approved by the Institutional Review Board of Northwestern University (Study ID
STU00204780).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the
study. Parents provided written consent for their children to participate, and children provided
verbal assent.
Data Availability Statement: Not applicable.
Acknowledgments: The authors are grateful to the families that participated in this research, the
childcare centers and directors who helped facilitate the data collection, and the members of the
Center on Media and Human Development who helped with data collection efforts.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Benavent, X.; de Ves, E.; Forte, A.; Botella-Mascarell, C.; López-Iñesta, E.; Rueda, S.; Roger, S.; Perez, J.; Portalés, C.; Dura, E.; et al.
Girls4STEM: Gender Diversity in STEM for a Sustainable Future. Sustainability 2020, 12, 6051. [CrossRef]
2. Silvestre, B.S.; Ţîrcă, D.M. Innovations for sustainable development: Moving toward a sustainable future. J. Clean. Prod. 2019, 208,
325–332. [CrossRef]
3. Unfried, A.; Faber, M.; Stanhope, D.S.; Wiebe, E. The development and validation of a measure of student attitudes toward
science, technology, engineering, and math (S-STEM). J. Psychoeduc. Assess. 2015, 33, 622–639. [CrossRef]
4. President’s Council of Advisors on Science and Technology. Prepare and Inspire: K-12 Education in Science, Technology, Engineering
and Math (STEM) Education for Americans’ Future; Executive Office of the President, Office of Science and Technology Policy:
Washington, DC, USA, 2009.
5. Stoet, G.; Geary, D.C. The Gender-Equality Paradox in Science, Technology, Engineering, and Mathematics Education. Psychol.
Sci. 2018, 29, 581–593. [CrossRef]
6. National Science Board. Science and Engineering Indicators. Available online: https://www.nsf.gov/statistics/2018/nsb20181
/report (accessed on 31 October 2018).
7. National Science Board. Our Nation’s Future Competitiveness Relies on Building a STEM-Capable U.S. Workforce: A Policy
Companion Statement to Science and Engineering Indicators 2018. Available online: https://www.nsf.gov/nsb/sei/companion-
brief/NSB-2018-7.pdf (accessed on 31 October 2018).
8. Jordan, A.B. The role of media in children’s development: An ecological perspective. J. Dev. Behav. Pediatrics 2004, 25, 196–206.
[CrossRef] [PubMed]
9. Fisch, S.M. Children’s Learning from Educational Television; Lawrence Erlbaum: Mahwah, NJ, USA, 2004.
10. Anderson, D.R. Educational television is not an oxymoron. Ann. Am. Acad. Political Soc. Sci. 1998, 557, 24–38. [CrossRef]
11. Fisch, S.M. Educational Television and Interactive Media for Children: Effects on Academic Knowledge, Skills, and Attitudes. In
Media Effects: Advances in Theory and Research, 3rd ed.; Bryant, J., Oliver, M.B., Eds.; Routledge: London, UK, 2009; pp. 402–435.
12. Hurwitz, L.B.; Schmitt, K.L. Raising Readers with Ready To Learn: A Six-Year Follow-up to an Early Educational Computer
Game Intervention. Comput. Hum. Behav. 2019, 104, 106176. [CrossRef]
13. Piotrowski, J.T. Is educational media an oxymoron? In 20 Questions about Youth and the Media|Revised Edition; Jennings, N.A.,
Mazzarella, S.R., Eds.; Peter Lang: New York, NY, USA, 2018. [CrossRef]
14. Chen, G.; Gully, S.M.; Whiteman, J.-A.; Kilcullen, R.N. Examination of relationships among trait-like individual differences,
state-like individual differences, and learning performance. J. Appl. Psychol. 2000, 85, 835. [CrossRef]
15. McAlexander, S.L.; McCance, K.; Blanchard, M.R.; Venditti, R.A. Investigating the Experiences, Beliefs, and Career Intentions
of Historically Underrepresented Science and Engineering Undergraduates Engaged in an Academic and Internship Program.
Sustainability 2022, 14, 1486. [CrossRef]
16. Maltese, A.V.; Tai, R.H. Pipeline persistence: Examining the association of educational experiences with earned degrees in STEM
among US students. Sci. Educ. 2011, 95, 877–907. [CrossRef]
17. Sadler, P.M.; Sonnert, G.; Hazari, Z.; Tai, R. Stability and volatility of STEM career interest in high school: A gender study. Sci.
Educ. 2012, 96, 411–427. [CrossRef]
18. Auger, R.W.; Blackhurst, A.E.; Wahl, K.H. The Development of Elementary-Aged Children’s Career Aspirations and Expectations.
Prof. Sch. Couns. 2005, 8, 322–329.
19. Selkow, P. Effects of maternal employment on kindergarten and first-grade children’s vocational aspirations. Sex. Roles 1984, 11,
677–690. [CrossRef]
Sustainability 2022, 14, 5631 15 of 16

20. Curran, F.C.; Kellogg, A.T. Understanding Science Achievement Gaps by Race/Ethnicity and Gender in Kindergarten and First
Grade. Educ. Res. 2016, 45, 273–282. [CrossRef]
21. Morgan, P.L.; Farkas, G.; Hillemeier, M.M.; Maczuga, S. Science achievement gaps begin very early, persist, and are largely
explained by modifiable factors. Educ. Res. 2016, 45, 18–35. [CrossRef]
22. Signorielli, N. Minorities representation in prime time: 2000 to 2008. Commun. Res. Rep. 2009, 26, 323–336. [CrossRef]
23. Smith, S.L.; Choueiti, M.; Prescott, A.; Pieper, K. Gender Roles & Occupations: A Look at Character Attributes and Job-Related
Aspirations in Film and Television. In Geena Davis Institute on Gender in Media; USC Annenberg School for Communication and
Journalism: Los Angeles, CA, USA, 2012.
24. Chambers, N.; Kashefpakdel, E.T.; Rehill, J.; Percy, C. Drawing the Future: Exploring the Career Aspira Ons of Primary School Children
from Around the World; Education and Employers: London, UK, 2018.
25. Bigler, R.S.; Averhart, C.J.; Liben, L.S. Race and the workforce: Occupational status, aspirations, and stereotyping among African
American children. Dev. Psychol. 2003, 39, 572. [CrossRef]
26. Liben, L.S.; Bigler, R.S.; Krogh, H.R. Pink and blue collar jobs: Children’s judgments of job status and job aspirations in relation to
sex of worker. J. Exp. Child. Psychol. 2001, 79, 346–363. [CrossRef]
27. Jeffries-Fox, S.; Signorielli, N. Television and children’s conceptions of occupations. In Proceedings of the Sixth Annual
Telecommunications Policy Research Conference, Lexington, MA, USA, November 1978; Dordick, H.S., Ed.; Lexington Books:
Lexington, MA, USA, 1979; pp. 21–38.
28. Signorielli, N. Television and adolescents’ perceptions about work. Youth Soc. 1993, 24, 314–341. [CrossRef]
29. Beuf, A. Doctor, lawyer, household drudge. J. Commun. 1974, 24, 142–145. [CrossRef]
30. Master, A.; Cheryan, S.; Meltzoff, A.N. Social group membership increases STEM engagement among preschoolers. Dev. Psychol.
2017, 53, 201–209. [CrossRef] [PubMed]
31. Miller, D.I.; Nolla, K.M.; Eagly, A.H.; Uttal, D.H. The Development of Children’s Gender-Science Stereotypes: A Meta-analysis of
5 Decades of U.S. Draw-A-Scientist Studies. Child. Dev. 2018, 89, 1943–1955. [CrossRef] [PubMed]
32. Verdugo-Castro, S.; García-Holgado, A.; Sánchez-Gómez, M.C.; García-Peñalvo, F.J. Multimedia Analysis of Spanish Female Role
Models in Science, Technology, Engineering and Mathematics. Sustainability 2021, 13, 12612. [CrossRef]
33. Blickenstaff, J.C. Women and science careers: Leaky pipeline or gender filter? Gend. Educ. 2005, 17, 369–386. [CrossRef]
34. Aladé, F.; Lauricella, A.; Kumar, Y.; Wartella, E. Who’s modeling STEM for kids? A character analysis of children’s STEM-focused
television in the US. J. Child. Media 2021, 15, 338–357. [CrossRef]
35. Steinke, J.; Long, M. A lab of her own?:Portrayals of female characters on children’s educational science programs. Sci. Commun.
1996, 18, 91–115. [CrossRef]
36. Burke, R.J.; Mattis, M.C. Women and Minorities in Science, Technology, Engineering, and Mathematics: Upping the Numbers; Edward
Elgar Publishing: Cheltenham, UK, 2007.
37. Bond, B.J. Fairy Godmothers > Robots:The Influence of Televised Gender Stereotypes and Counter-Stereotypes on Girls’ Percep-
tions of STEM. Bull. Sci. Technol. Soc. 2016, 36, 91–97. [CrossRef]
38. Post-Kammer, P.; Smith, P.L. Sex differences in math and science career self-efficacy among disadvantaged students. J. Vocat.
Behav. 1986, 29, 89–101. [CrossRef]
39. Marra, R.M.; Rodgers, K.A.; Shen, D.; Bogue, B. Women engineering students and self-efficacy: A multi-year, multi-institution
study of women engineering student self-efficacy. J. Eng. Educ. 2009, 98, 27–38. [CrossRef]
40. MacPhee, D.; Farro, S.; Canetto, S.S. Academic self-efficacy and performance of underrepresented STEM majors: Gender, ethnic,
and social class patterns. Anal. Soc. Issues Public Policy 2013, 13, 347–369. [CrossRef]
41. Hurtado, S.; Newman, C.B.; Tran, M.C.; Chang, M.J. Improving the rate of success for underrepresented racial minorities in STEM
fields: Insights from a national project. New Dir. Inst. Res. 2010, 2010, 5–15. [CrossRef]
42. Wilson, E.J.; Sherrell, D.L. Source effects in communication and persuasion research: A meta-analysis of effect size. J. Acad. Mark.
Sci. 1993, 21, 101–112. [CrossRef]
43. Shrum, L. The Psychology of Entertainment Media: Blurring the Lines between Entertainment and Persuasion; Taylor & Francis: New
York, NY, USA, 2012.
44. Castelli, L.; De Dea, C.; Nesdale, D. Learning social attitudes: Children’s sensitivity to the nonverbal behaviors of adult models
during interracial interactions. Personal. Soc. Psychol. Bull. 2008, 34, 1504–1513. [CrossRef]
45. Vittrup, B.; Holden, G.W. Exploring the impact of educational television and parent–child discussions on children’s racial
attitudes. Anal. Soc. Issues Public Policy 2011, 11, 82–104. [CrossRef]
46. Signorielli, N.; Lears, M. Children, television, and conceptions about chores: Attitudes and behaviors. Sex. Roles 1992, 27, 157–170.
[CrossRef]
47. Gerbner, G.; Gross, J.; Morgan, M.; Signorielli, N. Living with television: The dynamics of the cultivation process. In Perspectives
on Media Effects; Bryant, J., Zillmann, D., Eds.; Lawrence Erlbaum: Hillside, NJ, USA, 1986; pp. 17–40.
48. Bandura, A. Social Foundations of thought and Action: A Social Cognitive Theory; Prentice-Hall: Englewood Cliffs, NJ, USA, 1986.
49. Bandura, A. Social Learning Theory; General Learning Corporation: Morristown, NJ, USA, 1971.
50. Mares, M.-L.; Cantor, J.; Steinbach, J.B. Using Television to Foster Children’s Interest in Science. Sci. Commun. 1999, 20, 283–297.
[CrossRef]
Sustainability 2022, 14, 5631 16 of 16

51. Rogers, L.O.; Meltzoff, A.N. Is gender more important and meaningful than race? An analysis of racial and gender identity
among Black, White, and Mixed-Race children. Cult. Divers. Ethn. Minority Psychol. 2017, 23, 323–334. [CrossRef] [PubMed]
52. Mayring, P. Qualitative Content Analysis: A Step-by-Step Guide; SAGE Publications: Thousand Oaks, CA, USA, 2021.
53. Queiruga-Dios, M.Á.; López-Iñesta, E.; Diez-Ojeda, M.; Sáiz-Manzanares, M.C.; Vázquez Dorrío, J.B. Citizen Science for Scientific
Literacy and the Attainment of Sustainable Development Goals in Formal Education. Sustainability 2020, 12, 4283. [CrossRef]
54. Maj, J. Embedding Diversity in Sustainability Reporting. Sustainability 2018, 10, 2487. [CrossRef]
55. Greenberg, B.S. Some uncommon television images and the drench hypothesis. In Television as a Social Issue: Applied Social
Psychology; Oskamp, S., Ed.; Sage: Newberry Park, CA, USA, 1988.
56. O’Keeffe, G.S.; Clarke-Pearson, K. The impact of social media on children, adolescents, and families. Pediatrics 2011, 127, 800–804.
[CrossRef]
57. Graves, S.B. Television and prejudice reduction: When does television as a vicarious experience make a difference? J. Soc. Issues
1999, 55, 707–727. [CrossRef]
58. Farnall, O.; Smith, K.A. Reactions to people with disabilities: Personal contact versus viewing of specific media portrayals.
J. Mass Commun. Q. 1999, 76, 659–672. [CrossRef]
59. Calvert, S.L.; Kotler, J.A.; Zehnder, S.M.; Shockey, E.M. Gender stereotyping in children’s reports about educational and
informational television programs. Media Psychol. 2003, 5, 139–162. [CrossRef]
60. Hurwitz, L.B. Getting a Read on Ready To Learn Media: A Meta-analytic Review of Effects on Literacy. Child. Dev. 2018, 90,
1754–1771. [CrossRef]