ArticlePDF Available

"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students' Comics about Friendship

Authors:
  • Center for Children and Technology

Abstract

Effective data literacy instruction requires that learners move beyond understanding statistics to being able to humanize data through a contextual understanding of argumentation and reasoning in the real-world. In this paper, we explore the implementation of a co-designed data comic unit about adolescent friendships. The 7th grade unit involved students analyzing data graphs about adolescent friendships and crafting comic narratives to convey perspectives on that data. Findings from our analysis of 33 student comics, and interviews with two teachers and four students, show that students engaged in various forms of data reasoning and social-emotional reasoning. These findings contribute an understanding of how students make sense of data about personal, everyday experiences ; and how an arts-integrated curriculum can be designed to support their mutual engagement in both data and social-emotional reasoning.
"I happen to be one of 47.8%": Social-Emotional and Data
Reasoning in Middle School Students’ Comics about Friendship
Ralph Vacca
rvacca2@fordham.edu
Fordham University
New York, New York, USA
Kayla DesPortes
kayla.desportes@nyu.edu
New York University
New York, New York, USA
Marian Tes
mct300@nyu.edu
New York University
New York, New York, USA
Megan Silander
msilander@edc.org
Education Development Center
New York, New York, USA
Camillia Matuk
cmatuk@nyu.edu
New York University
New York, New York, USA
Anna Amato
ada437@nyu.edu
New York University
New York, New York, USA
Peter J. Woods
peterwoo@mit.edu
Massachusetts Institute of Technology
Cambridge, Massachusetts, USA
ABSTRACT
Eective data literacy instruction requires that learners move be-
yond understanding statistics to being able to humanize data through
a contextual understanding of argumentation and reasoning in the
real-world. In this paper, we explore the implementation of a co-
designed data comic unit about adolescent friendships. The 7th
grade unit involved students analyzing data graphs about adoles-
cent friendships and crafting comic narratives to convey perspec-
tives on that data. Findings from our analysis of 33 student comics,
and interviews with two teachers and four students, show that
students engaged in various forms of data reasoning and social-
emotional reasoning. These ndings contribute an understanding
of how students make sense of data about personal, everyday expe-
riences; and how an arts-integrated curriculum can be designed to
support their mutual engagement in both data and social-emotional
reasoning.
CCS CONCEPTS
Human-centered computing
Visualization;Applied com-
puting Computer-assisted instruction;Media arts.
KEYWORDS
data literacy, social-emotional learning, data comics, math educa-
tion, arts education, data reasoning
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for components of this work owned by others than the
author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specic permission
and/or a fee. Request permissions from permissions@acm.org.
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
©2022 Copyright held by the owner/author(s). Publication rights licensed to ACM.
ACM ISBN 978-1-4503-9157-3/22/04. . . $15.00
https://doi.org/10.1145/3491102.3502086
ACM Reference Format:
Ralph Vacca, Kayla DesPortes, Marian Tes, Megan Silander, Camillia Matuk,
Anna Amato, and Peter J. Woods. 2022. "I happen to be one of 47.8%": Social-
Emotional and Data Reasoning in Middle School Students’ Comics about
Friendship. In CHI Conference on Human Factors in Computing Systems (CHI
’22), April 29-May 5, 2022, New Orleans, LA, USA. ACM, New York, NY, USA,
18 pages. https://doi.org/10.1145/3491102.3502086
1 INTRODUCTION
Data literacy can be understood as, “the ability to read, work with,
analyse and argue with data as part of a larger inquiry process” [
17
].
For an individual to be data literate, they must be equipped to read,
interpret, and make decisions around data based on its occurrence
in news, politics, healthcare, nance, advertisement, and media
[
17
,
70
]. To meet these requirements, data literacy education must
expand beyond the mathematical and statistical understanding of
data to include a broader set of competencies such as reasoning
around real-world contexts, communicating stories from data, and
engaging in the critical aspects of recognizing and understanding
the ethical implications of data and data practices [70].
In contrast to framing data as objective and factual, a humanistic
lens on data literacy emphasizes the contextual and interpretive
nature of data [
45
,
46
]. By bringing context to the center of con-
versation with data, educators can integrate relevance and avoid
automistic problem-solving and inert knowledge [
6
,
46
]. In particu-
lar, the focus on context has opened up opportunities to leverage
various art forms such as narrative development in data stories [
52
]
as well as visual forms like murals [
10
] for developing data literacy
skills. In these types of experiences, learners can work with data
in new ways while still engaging in argumentation and reason-
ing. One subset of this work examines the construction and use
of data comics, in which narratives are composed of visual and
textual elements to convey stories about data [
3
]. Given their po-
tential relevance to youth’s interests and their promise for making
information more accessible (ex.[
67
], data comics oer an under ex-
plored opportunity for students to engage in reasoning about data
and social-emotional issues. In this work, the researchers examine
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
the opportunities to expand the humanistic approaches to data
through co-designing a comic unit for 7th grade students focused
on data about their friendships. We examine how students’ process
of constructing comics can support their data and social-emotional
reasoning.
This study is part of a larger project called Data Literacy Through
Art (DLTA), which brings together university researchers and cross-
subject middle school teachers to co-design curriculum and re-
sources around art and data literacy. We worked with a 7th grade art
and math teacher at the same school, to co-create a data comic unit
with the intention of providing students with a personally relevant
context in which to develop perspectives and skills with lifelong im-
plications. The researchers and teachers decided on friendship as a
topic and on comic-making as the medium. In this paper, we present
ndings from the implementation of the data comic unit with 33
seventh graders, who examined data about themselves and their
friendships. They practiced data analysis skills, like graph reading
and statistical reasoning, while creating digital comics to commu-
nicate stories about and with the data. Specically, we examined
two research questions:
RQ1. In what kinds of data reasoning do students engage
through their creation of data comics? and
RQ2. In what kinds of social-emotional reasoning do stu-
dents engage through their creation of data comics?
Through an analysis of 33 data comic artifacts, interviews with
two middle school teachers, and interviews with four student par-
ticipants, we present key ndings. First, through the process of
constructing narratives, students engaged in complex forms of data
reasoning wherein they used data and the context of their stories to
determine narrative implications. These acts of comic making led
to instances of informal inferential reasoning [
45
,
46
]. Second, by
centering the data on teen friendships, learners not only engaged
in data reasoning but also forms of social-emotional reasoning. We
demonstrate how these two competencies supported one another
within their comic narratives.
The paper contributes to the literature by providing a set of il-
lustrative examples that demonstrate the potential for data comics
to broaden the ways learners can engage in data reasoning through
narratives that force them to represent and construct social-emotional
context around the data. We argue that comic authoring tools that
focus on narrative construction, character development, and scene
composition, can expand the potential for data comics by provid-
ing greater opportunities to move towards a more humanistic and
interdisciplinary view of data literacy.
2 BACKGROUND LITERATURE
2.1 Data Literacy
The processes, competencies, and dispositions that are important for
negotiating and working with data are everyday life skills [
70
]. Data
literacy intersects and draws on multiple other literacies, including
statistical literacy, scientic literacy, computational literacy, media
literacy, and critical literacy [
9
,
70
]. Key priorities for data literacy
education in K-12 include supporting social engagement with data,
participation in a process of inquiry, and development of critical
consciousness [
9
,
62
,
70
]. These critical data literacy perspectives
view data as situated in a social context and explicitly challenge
the perception of data as neutral or objective [
9
,
11
,
18
,
53
,
62
,
69
].
As Kitchin
[39]
explains, “no data are pre-analytic, or objective and
independent” [
39
]. In this study, we adopted an inquiry-focused
denition of data literacy [
16
,
64
,
70
]. Data literacy is “the ability to
ask and answer real-world questions from large and small data sets
through an inquiry process, with consideration of ethical use of data”
[
70
]. The inquiry process includes asking questions, developing
a hypothesis, collecting data, analyzing data, and evaluating data
[
70
]. Traditional denitions of data literacy have emphasized a
narrow set of mathematical competencies. By situating data analysis
within this larger process of context-bound inquiry, we aimed to
engage students in making informal inferences about data and in
challenging neutral, objective representations of data.
Within K-12 math, there has been a shift toward teaching sta-
tistics as a way to reason about the world rather than as a set of
procedures and calculations [
8
]. The goal is to broaden students’
participation with data and prepare them for more advanced work,
such as inferential statistics. Formal inferential reasoning requires
an understanding of complex statistical techniques that are not
accessible to middle school students. But students can still engage
informally in this process. Makar and Rubin
[45]
developed a three-
part framework dening informal inference-making. Students need
to (1) explicitly identify evidence, (2) make a claim about the aggre-
gate that goes beyond the data, and (3) articulate the uncertainty
embedded in an inference. These objectives address persistent chal-
lenges that students have in working with data. For example, de-
termining what counts as appropriate and sucient evidence is
challenging for students. They may rely on personal beliefs or fail
to align data with their claims [
32
]. In data analysis, students of-
ten rely on case value comparisons, which limits their ability to
make claims that go beyond the data [
35
,
40
,
55
]. In contrast, ap-
proaching data as an aggregate reinforces the value of calculating
statistics [
7
,
40
]. Finally, recognizing and communicating uncer-
tainty within statistical claims engages students with tough topics
such as the variability of data and challenges misconceptions that
data are neutral and objective. To demonstrate informal inference-
making, students might generate hypotheses, evaluate claims, or
make predictions based on statistics [45].
Recent research in data literacy has highlighted the promise of
arts-based techniques, such as storytelling, to engage students in
constructing explanations, arguments, and inferences grounded
in real world contexts [
9
]. Researchers also note the potential of
storytelling to support data inclusion by connecting local, expe-
riential ways of knowing to data [
9
,
18
]. For example, students
may construct stories from personal experience to reason about
mathematical concepts, such as center and spread [
42
]. They may
co-construct evidence-based claims by challenging each other to
support personal stories with data [
60
]. Students may also broaden
their understanding of what counts as data and how it can be used
for personally relevant purposes [
10
,
59
]. Telling stories with per-
sonal data can help students understand themselves in new ways,
but it can also overshadow other goals such as connecting personal
data to culturally or socially-relevant issues [
59
]. In this study, we
investigate the potential of narrative-based arts, such as comics,
to help students situate their personal stories in a broader social
context through data and engage students in informal inference
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
making. We identify how their artifacts were indicative of their
engagement in social-emotional learning.
2.2 Social-Emotional Learning
Social-emotional learning (SEL) has been dened as the process
by which we develop our competencies to recognize and manage
emotions, empathize and care about others, make ethical and re-
sponsible decisions, and develop constructive relationships [
21
,
71
].
The importance of cultivating such social-emotional competencies
is particularly important in early adolescence since internalizing
diculties becomes more prevalent, alongside increases in mood
variability and intensity in negative emotional experiences [
5
]. In
addition, adolescence is largely understood as a period in which
many emotion-related words are fully comprehended, but the un-
derlying concepts have not yet reached mature levels of abstraction
[
50
,
51
]. As crises such as the COVID-19 pandemic emerge, cultivat-
ing social-emotional competencies in adolescent education spaces
has only grown in both importance and interest.
A variety of frameworks have emerged to operationalize social-
emotional learning. Lipton and Nowicki
[43]
developed the Social
Emotional Learning Framework (SELF) to identify the skills neces-
sary for people to personally relate to one another. They break
up the SELF into three components: (1) awareness – the ability to
be aware of and recognize emotions in others, (2) meaning – the
ability to understand the perspectives of others through linking
social-emotional cues to the context in which they are situated, and
(3) reasoning – the ability to use social-emotional information to
construct one’s social behavior and problem-solve [
43
]. The compe-
tencies involve both verbal and non-verbal skills as people learn to
pick up on, understand, and respond to the variety of emotional sig-
nals that humans enact [
43
]. While verbal signals include cues that
are expressed aloud, non-verbal signals include things such as fa-
cial expressions, postures, gestures, tones of voice, distance in personal
space, rhythm of interaction, apparel, and touch [
43
]. Working within
comics, provides a unique medium for us to examine how learn-
ers were able engage with verbal and non-verbal social-emotional
reasoning within their narrative artifacts.
The Collaborative for Academic, Social, and Emotional Learning
(CASEL) organization, which aims to make evidence-based SEL
an integral part of U.S. education, posits a framework with ve
key SEL competencies for children and adolescents: self-awareness,
self-management, social awareness, relationship skills, and respon-
sible decision-making [
24
]. This framework emerged from earlier
work [
29
,
47
] with studies supporting that children and adoles-
cents who have acquired these competencies typically exhibit fewer
mental, emotional, and behavioral problems later in life [
28
,
68
].
CASEL’s ve competency outcomes framework merge and balance
the cognitive, emotional, and social skills and is evidence-based
[
20
]. Self-awareness is the ability to accurately recognize one’s own
emotions, thoughts, and values and how they inuence behavior.
Social awareness is the ability to take the perspective of and em-
pathize with others, including those from diverse backgrounds and
cultures. Self-management is the ability to successfully regulate
one’s emotions, thoughts, and behaviors in dierent situations. Re-
lationship skills allow students to establish and maintain healthy
and rewarding relationships with diverse individuals and groups.
Responsible decision-making refers to the ability to make construc-
tive choices about personal behavior and social interactions based
on ethical standards, safety concerns, and social norms. SEL imple-
mentation varies widely from school-wide holistic approaches, to
competency-focused approaches that may tie into curriculum [
24
].
One challenge in implementing SEL, is aligning the core com-
petencies with other disciplinary content in order to integrate SEL
across the curriculum [
71
]. Prior work has explored the use of narra-
tives in English Language Arts and Social Studies to have students
reect on characters’ decisions and problem-solving behavior. This
enables students to build their emotional vocabularies as they ana-
lyze the emotions and feelings of the characters, action sequences,
results of the characters, and alternative paths that the characters
could have taken [
21
]. In this work, we explore how narratives,
such as comics, can be used within math and art curricula to deepen
learners’ data reasoning. We demonstrate how the use of comics
presented unique opportunities for social-emotional reasoning as
students crafted both visual and textual artifacts as part of their
work.
2.3 Using Comics to Expand Data Reasoning
Comics are a form of sequential art that communicate narratives
through the interplay of text and images, often with a mix of both
realistic and abstract symbols, to convey ideas and emotion [
19
,
48
,
54
]. The combination of pictures and text in comics, “oers range
and versatility with all the potential imagery of lm and painting
plus the intimacy of the wrien word” [
48
] (emphasis retained
from original citation). The comic art is additive in that it can use
written and drawn features to aim attention to certain details or
ideas that would not be possible in a realistic drawing. For example,
emanata are stylized lines that can be used to communicate speed of
a moving object or stench of a garbage can. Yet, it is also subtractive,
in that the cartooning of comics often removes details that would
make them look more realistic. This leaves space to draw attention
to what is left and allows for the readers’ imagination to ll in
the details [
48
]. Whether creating features within characters or
connecting the comic panels through time and space, the reader is
actively engaged with what is not there. The reader creates closure,
building up the whole of the story from the parts, based on what
is on the page and what is not [
27
,
48
,
49
]. One of the dening
features of the comic art form is that the author can manipulate
the panels and the gaps between panels to guide the spatial and
temporal aspects of the narratives in ways that are only possible
through comics [48].
The unique features of comics have led researchers to examine
the ways they can be connected to data reasoning and data-driven
storytelling through capitalizing on the spatial layout and linear
narrative to present information, charts, and data, that the reader
can traverse through the comic panels [
4
]. Bach et al
. [3]
highlight
four main components of data comics: the visualization of the data
ranging from the realistic to abstract, the ow of the comic ranging
from an explicitly directed to an open format, the narration of the
comic ranging from factual statements to a story-based format,
and the words and pictures break down of how verbal and pictorial
components are combined. Research has identied the potential
for data comics to improve understanding and engagement with
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
data over other visual presentations such as infographics [
67
] while
providing access to complex information such as explaining how
quantied-self apps handle data [
57
], or explaining results from HCI
research studies [
66
]. Creating data comics can be dicult since the
medium traverses artistic and mathematical domains. Bach et al
.
[4]
examined the design space for data comics in order to provide
design patterns that could help guide their construction. Others
have convened comic artists and data scientists in workshops to
integrate their expertise in the data comic construction process [
65
].
Technologies have also been developed to scaold the authoring of
data comics. Some of the tools are datacentric such as DataToon [
38
]
and ToonNote [
37
], which are focused on how learners can create
new visualizations through complimenting data representations
with textual information to communicate narratives about the data.
However, these are limited in that they forefront issues of data
visualization, and fail to take advantage of the complementary
opportunities in comic design, such as in the creation of characters
and settings. Although not typically focused on in data comics,
there are a variety of comic and digital storytelling tools that are
intended to support users of all abilities. These include BitStrips
(which was morphed into Bitmoji), Comic Life, MakeBeliefComix,
and Pixton (review in [2]).
In our work, we explore Pixton, a tool designed around users’
ability to tell stories rather than represent data. Pixton has been
previously used in educational research contexts. For instance, the
tool has shown to motivate and engage students while sustaining
gains in learning in English as a Foreign Language (EFL) classes in
Ecuador [
12
]. Furthermore, in their study, Smith et al
. [58]
demon-
strate how Pixton can be woven into a multimodal experience,
where learners create science ction narratives to explore a variety
of scientic concepts. Compared to other digital storytelling tools
such as Bitstrips, Comic Life, MakeBeliefComix and Cambridge
University’s Comic Builder, Pixton earned the highest scores across
12 dimensions of storytelling, with its highest ratings in Cognitive
Eort—“the required user understanding to create stories”, Con-
trol—“how much control is given to users in narrative evolution”
Immersion—“how much a user is drawn into the narrative” [
2
].
With respect to Control, Pixton oers users with exibility in set-
ting scenes in their comic panels as they are able to choose from a
variety of backgrounds and perspectives (zooming in and out) as
needed.
The exploration of digital comic tools that focus on the char-
acter and setting development is important for data comics to be
centered within the larger movement to “feel” [
44
] and experience
data in ways that can contextualize and humanize the data and data
processes [
1
]. The narrative form of comics can provide characters
and settings that can be leveraged for a humanistic understand-
ing of the data. Toh et al
. [61]
, for example, demonstrates how the
comics opened up opportunities in the math classroom to stimulate
discussions about privacy and condentiality within data practices.
Learners were able to engage in perspective taking of the comic
characters as they discussed the ethics behind the characters’ de-
cisions [
61
]. Our work taps into these opportunities within data
comics to connect them to the needed contextualization for learners
to engage in inferential reasoning. We demonstrate how the digi-
tal comics and technology supported learners in focusing on the
human side of the data that supported them in reasoning around
their data as they explored the social-emotional dimensions within
their narratives.
3 METHODS
3.1 Context and Participants
Participants included 33 7th grade students who were in both math
and art classes from a public school in an large urban city in the
Eastern United States. The school serves a diverse student popula-
tion: 33% White, 32% Black, 16% Hispanic, and 16% Asian. 46% of
the student population qualies for free or reduced lunch, 18% of
students have disabilities, and 4% are English Language Learners.
While there were more students that participated in the unit, not
all of the students in the art class were in the math class, and vice-
versa. Furthermore, shifts in scheduling due to COVID restrictions
further complicated the possibility of students attending all of the
sessions with both teachers, so some participants that initially were
in both classes were switched out. Consequently, two entire classes
(about 42 students) were dropped from the study. However, the use
of the comic artifacts in our analysis still enabled us to understand
how learners were engaging and reasoning around data and the
social-emotional contexts developed throughout their narratives.
3.2 Unit Description
One math and one art teacher in the same school taught the data
comic unit during the Spring 2021 semester. In the art classroom,
students convened once a week over a six-week period, which
amounted to about one-third of their semester’s art classes. The
math portion of the unit ran three times a week for two weeks and
was integrated within a larger unit on statistical reasoning. The art
class started prior to the math unit.
Researchers co-designed the unit with the teachers starting in
the prior semester. Researchers met biweekly with teachers either
together or separately to coordinate across and within both do-
mains of art and math. During the planning phase, the teachers
chose the topic of friendship as the focal point of the unit, because
they were interested in helping students foster life skills. The art
teacher wanted to build on a comics unit that she had previously
implemented in her class. The teachers and researchers chose this
medium because of the narrative quality of comics, the closeness
of comics to youth’s existing interests, and the available tools for
comic-making to oer an accessible medium for telling stories about
friendship. Teachers also felt strongly that students should analyze
their personal data.
To provide students with a set of personally-relevant data about
friendships in which to ground their inquiry, the research team
developed a 17-item survey, using Google Forms, on friendship
based on similar questions asked on a PEW survey on teens and
technology use [
13
]. While the PEW survey focused broadly on
teens and technology, the 17-item survey focused primarily on
beliefs about friendships, their describing their own friendships,
and certain friendship experiences (e.g., bullying experiences). As
described in Table 1, these items aligned to four categories.
The teachers provided this survey to all four 7th grade classes
in the school (N=92) during their math class. Researchers cleaned
and analyzed the survey data. Because the math teacher wanted
students to focus on graph reading and inference making in this
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
Figure 1: A timeline of the unit implementation
Table 1: Sample Survey Questions
Category Example Question # of items
Beliefs about Friendship
How much do you agree/disagree with the following statements? [It’s
okay to walk away or take a break from relationships that aren’t sup-
portive.]
3
My Friendships
How many of your CLOSE FRIENDS are... [In this school, online, female,
non-binary, male]
3
Friendship Experiences
Which of the following describes you and your friends? I can count on
my friends when things go wrong
7
Demographics Which gender do you identify with?; How often are you lonely? 4
unit, the researchers created graphs including boxplots, dot plots
and bar graphs based on the survey data. Students based their data
comics on their interpretations of these graphs.
In the art sessions, students began in the rst lesson (A1) by
exploring comics as a storytelling approach. Building o of prior
lessons around ancient manuscripts, students explored comic vo-
cabulary and elements such as panels, bubbles, transitions, captions,
by viewing and discussing comic examples. In the second lesson
(A2), students started using Pixton to create single-panel comics
focused on expressing emotions using facial expression, body posi-
tions, colors, zooming, and environmental backgrounds. In the third
lesson (A3), students explored narrative arcs using a three-act story
structure. Using Pixton, students created multiple-panel comics that
provided a context, conict, and resolution. In the fourth lesson
(L4), students focused on building narratives from graphs analyzed
in math class.
The week when students began their third lesson (A3) in art
class, they also began their initial math lessons, focusing on reading
graphs generated by the research team using the student survey
data. In the rst math lesson (M1), the students focused on making
claims using dot plots and bar graphs, while the second lesson (M2)
focused on box plots. In this third lesson (M3), the focus shifted
to samples contrasting the 7th grade survey questions against the
national sample from a Pew research dataset of teenagers’ attitudes
about social media and friendship. In the nal lesson (M4), students
created their data comics in both their math and art class. In math
class, students selected one of the graphs to explore through their
data comic, followed prompts that asked them to make claims from
the graphs they selected, identied an argument about friendship
that they felt was important to share, and nally, used a comic to
communicate this claim. In addition to creating their data comic,
students generated artist statements to reect on their message and
data. Students were made aware the unit was part of a research
study on connecting comic-creation with data literacy, and a rubric
was co-designed by the math and art teacher to provide feedback
on the data comics (see Table 2).
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Table 2: Data Comic Rubric
Level of Per-
formance
Excellent Understanding
Demonstrated
Strong Understanding
Demonstrated
Unclear Understanding
Demonstrated
Little to No Understand-
ing Demonstrated
My Claim
I made a clear and accurate
inference about the data.
I made an accurate infer-
ence but it is somewhat un-
clear.
I made an inference but
there are some errors or I
made an inference that is
somewhat supported by the
data
I did not make an inference
or I made an inference that
is unsupported by the data.
My Evidence
I correctly used at least two
statistical measures to sup-
port my inference or I com-
pared two or more mea-
sures from two or more
groups
I correctly used one one sta-
tistical measure to support
my inference
I used one statistical mea-
sure but it does not support
my inference
I did not use any evidence
to support my inference
My Artist
Statement
I can describe why I se-
lected the visuals I used and
how they relate to my claim,
my evidence, and my argu-
ment.
I can partially describe why
I selected the visuals I used
and how they relate to my
claim, my evidence, and my
argument.
I can describe why I se-
lected the visuals I used but
I did not describe how they
relate to my claim, evidence,
and/or argument.
I cannot describe why I se-
lected the visuals I used nor
did I explain how they re-
late to my claim, evidence,
and argument.
My Message
about Friend-
ship
I constructed a clear mes-
sage that helps my peers re-
ect on friendship.
I constructed a message
about friendship but it is not
entirely clear what my mes-
sage is.
I constructed a message that
is clear but not related to
friendship.
I did not construct a clear
message.
Because of the pandemic, teachers co-designed and implemented
the data comic unit remotely, mostly coordinated through the suite
of Google Apps for Education. Although this format was familiar
to students, and facilitated researchers’ and teachers’ real-time
collaboration over the instructional materials we were co-designing,
it also presented various logistical and pedagogical constraints. For
example, the art teacher expressed concern about the students’
abilities to create comics without their having access to the drawing
materials that they would have in their classroom. The art teacher
was initially hesitant to adopt a new digital tool. However, after
reviewing it, she became excited about the potential for the tool
to support students in quickly creating comics while also covering
important topics about the art form, such as scene composition,
character creation, narrative development, and communication
through text and visuals. The key to the success of the tool was that
students would not require signicant support from her to begin
using it.
3.3 Pixton Comic & Storyboard Builder
Pixton is a tool for digital storytelling through comics developed by
Clive and Diana Goodinson in 2008 [
31
]. Pixton provides users with
the ability to rapidly construct comics using a library of predened
content that the user can manipulate to build visual and textual
narratives. We chose the tool in collaboration with the teachers
who needed something that could work in a remote context, en-
abled students to create artistic artifacts without additional tools or
materials (outside of the computing devices they were already using
to connect to class), and supported a range of expertise. Within
the tool, users create comics by adding and designing each panel–
situating characters, objects, and text within settings. As illustrated
in Figure 2, the panels have three main components that can be
manipulated: the background and setting elements, the characters
including their outts, facial expressions, and body movements, as
well as the textual components.
3.3.1 Background and Seing. The tool provides hundreds of back-
grounds that can be searched for by the user. This includes a diverse
set of indoor and outdoor scenes that represent real and ctional
settings across cultures and contexts, from various house interiors,
to graveyards and concerts, to underwater ruins and planetary land-
scapes. The content also includes options that are artistic in nature,
such as a spiral or a frame full of colorful question marks, that can
be used to convey a certain aesthetic. The user can then choose the
focus of the scene which allows them to choose from preset zooms
within the chosen background. The user can tint the scene to make
it appear to be dierent times of the day, and can choose to draw
attention to a particular panel by integrating a monochromatic
tinting across the characters and background (such as turning the
characters all a shade of red and the background elements all a
shade of blue). Additionally, the user can add overlays of weather
(various types of wind, rain, and snow), symbols (such as arrows
and explosions), and predened text art such as the iconic comic
POW! in bubble letters with an explosion behind it. Lastly, there is
a feature to upload an image for your background; however, it had
several limitations and thus we were not able to use it.
3.3.2 Character Controls. The tool provides a selection of over
a thousand characters that users can choose to start from. Users
can customize the clothing, skin color, hair color and style, facial
expression, eye gaze, body position, and location of the character in
the frame. They can also choose from hundreds of clothing options
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
Figure 2: Screenshot of the Pixton tool
that represent a diverse set of contexts and cultures from space suits
and animal costumes to gladiators and togas. The tool provides a
subset of skin colors that are automatically linked to certain options
for hair and eye color. This limits options such as being able to create
a dark skinned character with blond hair. The tool does provide
options for non-natural skin and hair colors like green, blue, and
purple. For the hair styles, the user has a selection of just over a
hundred to choose from that have a range in textures.
The user can add emotive features to their characters such as
facial expressions and body positions. The facial expressions range
from those that are expressed on the character’s face to those that
can be expressed through symbolic representations such as an
explosion coming o the top of the head to indicate anger or a
question mark to signal surprise. In Pixton, one can search through
a list of facial expressions by using social-emotional vocabulary,
selecting from predened emotions or typing in your own. You
can also link multiple emotions together within your search such
as concerned and hurt which will give you a dierent subset of
results compared to concerned and angry. Additionally, the user can
manipulate the gaze of the character. The tool will automatically
link the gaze to either the audience if it is the only character or
another character if there is more than one. The user can change it
to point to the audience, up, down, left, right, or any combination
of directions (ie. left and down). Lastly, the user can manipulate the
body position and location of each character in the panel. Similar
to facial expressions, the user can search through a menu of body
positions using predened terms or their own. They can use actions
and emotions in their search terms and choose the direction of the
way the character is facing (either left, right, or head on). The user
can drag characters around the frame and set them in relation to
another character as well as to the background and foreground
components. Depending on the selected background, the tool often
limits the users placement of the characters to a particular plane (ie.
only allowing characters to be moved horizontally on the ground).
3.3.3 Textual Components. The nal addition to the comic panels
that users can manipulate is the textual components. These come in
the form of captions and four forms of text bubbles: speech, thought,
shout, and whisper. The thought bubble makes the bubble look like
a cloud, the shout creates jagged edges to the bubble and turns it an
orange color, and the whisper makes the bubble semi-transparent
and puts a dotted line around it. The bubbles move automatically
with the characters position on the screen, and the user cannot
change it. The captions can be placed at the top, bottom, or center
of the panel. For the top and bottom choices the rest of the panel
remains, but in the case of the center it takes up the whole thing
removing the characters and background, allowing you to set an
aesthetic background instead.
3.4 Data and Analysis
3.4.1 Data Reasoning. To answer the question, what kinds of rea-
soning do students engage in through their creation of data comics,
we pulled from three dierent data sources: (1) data comic artifacts,
(2) group interviews with four students, and (3) individual inter-
views after the implementation with each of the art and the math
teacher. We also triangulated this data through cross-referencing
and contextualizing with our co-design meeting notes and artifacts
that we draw on.
The 33 data comics created by the students were coded (see Table
3) through a social moderation approach [
25
] where a team of four
researchers made independent ratings then discussed their codes to
come to a consensus around social-emotional and data reasoning.
Two group interviews with a total of four students were con-
ducted remotely through Zoom. The students were selected to be
interviewed based on provided assent and parental consent, and rec-
ommendations from their teacher, based on whom they felt would
be good at talking about their work and reasoning. The students’
comics were representative of the comics created by the class, in
that the codes mirrored the most prevalent codes. In the group
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
interviews students were prompted to walk through their comic
and then answer questions that focused on their reasoning: (i) what
did you want the comic to communicate, (ii) why did you choose
this statistic as a focus, (iii) how did the project impact your under-
standing of friendship, and (iv) what was your process for coming
up with a narrative for your data comic?
In analyzing the student and teacher interviews, we looked
through the interview transcripts for statements that would provide
context for the data reasoning patterns found in our analysis of
the comics. In this way, we sought to better triangulate the data
to give access to provide a more complex and rich description of
data and social-emotional reasoning, including instances that were
dissonant or complimentary to coded comics [23].
3.4.2 Social-Emotional Reasoning. To answer the question, what
kinds of social-emotional reasoning do students engage through
their creation of data comics, we also analyzed the three data
sources: (1) data comic artifacts, (2) four student interviews, and (3)
two individual teacher interviews.
For coding verbal SEL components, the researchers applied a
framework from the non-prot Collaborative for Academic, Social,
and Emotional Learning (CASEL) [
24
]. Their framework identies
ve components of SEL: (1) self-awareness, (2) self-management, (3)
social awareness, (4) relationship skills, and (5) responsible decision
making. Through the coding process, the researchers rened the
denitions oered in CASEL to those found in the nal codebook
(see Table 4). Three researchers went through four iterations of
coding verbal SEL components, starting with individually coding
then shifting to collectively coding the rst subset (10 comics) and
eventually the full data set (33 comics) rening the denitions at
each iteration ensuring that the denitions supported the ways in
which they were being applied to the comics. Once a consensus
was reached on how the SEL categories were applied to the comics,
the researchers tagged each comic for the ve categories of SEL
identifying which panels they recognized in SEL reasoning. The
researchers looked for consensus across the comics themselves as
to what SEL competencies they integrated rather than consensus
across the panels since the codes typically spanned multiple panels
and depending on how one interpreted the beginning or end of the
social interaction it could be dierent. The researchers cared most
about what types of competencies were present within the comic
as a whole. The teacher and student interview data were analyzed
for statements that would provide context to the SEL reasoning
patterns found in our analysis of the comics.
3.4.3 Non-Verbal Social-Emotional Reasoning. We analyzed the
data comic artifacts using a coding scheme focused on understand-
ing how learners leveraged Pixton’s ability to manipulate the facial
expressions and postures of the characters in order to commu-
nicate aective, non-verbal components of their narratives. Two
researchers conducted three rounds of coding the comics for the
number of: (1) facial expressions used in support of communicating
emotional states of their characters, and (2) body postures used
in communicating social or emotional states of their characters
(excluding those only connected with actions).
4 RESULTS
4.1 Data Reasoning
Our analyses demonstrate that students were able to engage in
various types of data reasoning within their comics (Figure 3). All
33 of the comics integrated some form of data reasoning. The most
prevalent form of data reasoning was the descriptive integration of
data, which involved integrating a statistic within the comic narra-
tive. This was arguably reasoning that required the least amount of
inference, as it required stating a relevant statistic. However, only
three comics stayed at this level of data reasoning, while the rest
integrated more advanced forms. The next two most common types
of data reasoning tagged within the data comics were contextual-
izing the data (78.79% | 26 comics) and reecting on an individual
experience representative of a data point in relation to the relation-
ship communicated in the data (72.73% | 24 comics). The nal three
forms were less prevalent: reecting on the implications of the data
(33.33% | 11 comics), articulating inquiries about related data (30.3%
| 10 comics), and making proportional comparisons between two
data categories (9.09% | 3 comics). Almost none of the students used
probabilistic reasoning about certainty or uncertainty. On top of
the data reasoning that we identied, there were also mistakes inte-
grated into how some of the comics were referencing or reasoning
about the data. Of the 33 comics, 9 (27.27%) had some form of in-
correct interpretation. We will examine how these codes manifested
within the Findings section.
4.2 Social-Emotional Reasoning
As indicated in Figure 4, almost all comics included a character in
the narrative demonstrating self-awareness (90.91% | 30 comics). Ap-
proximately two-thirds of comics reected relationship skills (69.70%
| 23 comics) and social awareness (66.67% | 22 comics). Students were
least likely illustrate behaviors within the comic narrative—i.e. self-
management (21.21% | 7 comics) and responsible decision-making
(12.12% | 4 comics).
An analysis was also conducted to explore what percentage of
data comics coded with a certain data reasoning code, also had a
co-occurring SEL code.
In addition, students leveraged the visual components of the
comic media to integrate non-verbal indicators of social-emotional
reasoning. This is distinctive in setting comics apart from other
forms of narratives that do not have a visual component tightly
integrated with the textual component. Across the 33 comics there
were a total of 238 facial expressions and 155 body postures coded
for supporting the social-emotional components of students narra-
tives. The average number of facial expressions coded per comic
was 7, with a minimum of 2 and a maximum of 15. For the body
postures, there was an average of 5 coded per comic with a min-
imum of 0 and maximum of 14 used within the comics. The data
indicates that students took advantage of the visual manipulations
of the characters using facial expressions slightly more than body
positions as they built their narratives.
5 FINDINGS
Our ndings are organized based on the types of data reasoning
we found within students’ comics (RQ1). Through examples, we
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
Table 3: Data Reasoning Codebook
Code Description
Incorrect Interpretation
Incorrect or partially incorrect interpretation of the data in support of
an argument
Descriptive Lists a data point, percentage, or descriptor of the data
Contextualizing the Data
Interpretation of the data relationships in relation to context. General-
izing to context
Individual Experience
Reecting on an individual experience (or data point) in relation to the
relationship communicated in the data
Proportional Comparisons Uses percents to compare dierent categories in the data.
Implications of the Data
Reason about how data or other evidence support a hypothesis or claim
Inquiry about Related Data
Engages in inquiry or makes an assumption about a data relationship
by connecting it to another variable or relationship that is not present
in the data
Table 4: Social Emotional Learning Codebook Based on CASEL Framework [24]
Code Description
Self-Awareness
Any character describes or reects on their own emotions, thoughts and values
Self-Management
Any character engages in or reects on behaviors to deal with their emotions
over time or across contexts
Social Awareness Any character reects on other characters or audience members perspectives
Relationship Skills
Reection on a specic skill that is important for establishing and maintaining
relationships
Responsible Decision
Making
Reection or demonstration of characters making caring and constructive
choices about personal behavior and social interactions across diverse situations
Figure 3: Graph representing the percentage of comics coded as exhibiting one or more data reasoning competencies
highlight how particular types of data reasoning and associated
social-emotional reasoning were apparent within the comic narra-
tives (RQ2).
5.1 Describing
In comics coded as “descriptive,” the narratives described a data
point—most often, a percentage or a few related percentages from
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Figure 4: Graph representing the percentage of comics coded as exhibiting one or more social-emotional learning competencies
Figure 5: Graph representing the percentage of comics coded with a certain data reasoning code, that also had a co-occurring
SEL code
the student or national survey data. Although most comics included
descriptive reasoning about data along with other types of data
literacy reasoning, three comics focused only on describing the
data. The example in Figure 6 is demonstrative of this type of story
in which the data is present (integrated into the speech of one of
the characters in panel 2 and 3), but does not have a bearing on
the narrative—i.e. they could have had data about anything and it
would not have changed the story.
While the data reasoning stayed at the descriptive level in this
example, this comic is indicative of how the comic construction
process around the theme of friendship still presented room for
social-emotional reasoning. The comic centers on a ght between
two characters, which is caused by one character rattling o irrele-
vant data and punching the character that is annoying him. While
the response seems a bit exaggerated given the context, the charac-
ter that started the physical ght engages in self-management skills
by apologizing for their past actions, and both characters demon-
strate relationship skills through their resolution of the conict.
The author of this comic integrated 12 dierent facial expressions
and 12 dierent body positions to bring the social-emotional com-
ponents of their narrative to life through non-verbal cues. This
example demonstrates how designing comics around friendships
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
Figure 6: Data comic (C14) on counting on friends when things go wrong
supports opportunities to engage in thinking about SEL. However,
the lack of a meaningful attention to data seemed to lead to shal-
lower reasoning about both SEL and data. In the remainder of the
sections, we identify how the data reasoning codes connected to
some of the SEL codes to demonstrate the ways in which they were
mutually supportive of one another (RQ1 & R2).
5.2 Narratives to Contextualize the Data
“Contextualizing the data” emerged within the majority of the comic
narratives (78.7%) and often consisted of students illustrating the
kinds of friendship and interactions that the data might represent.
We found that contextualization was well supported by the social-
emotional reasoning code “social awareness” with 16 of the 26
“contextualizing the data” comics (61.54%) co-occurring with this
code. For instance, in the comic in Figure 7, the student uses the data
on student responses to the question: “do you nd it easy to make
friends?” and uses her narrative to consider reasons that inuence
student responses. They develop a narrative around making friends
“at a young age”, engaging in a broad inquiry of the importance of
the variable “when and how you met your friends” as it relates to
how easy or hard you nd it to make friends. In an interview, the
student author describes how they reasoned around related data:
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Figure 7: Data comic (C7) on making new friends and long-term friendships
“I kind of wasn’t surprised because a lot of the kids here
became friends at a young age because we were in the
same classes and everything. So when I saw “kind of”, I
kind of thought maybe that’s because certain kids, they
make friends at a young age age.
In further describing their reasoning process, the student also en-
gages in perspective taking as they try and examine why students
might reply a certain way, and what they may have felt, as it relates
to the survey response,
“...they make friends at a young age. They don’t know
how it was to make the friendships. I don’t know if it’s
hard or easy. They kind of just did it because they were
young. They weren’t really nervous at the time and they
didn’t really have to care that much.
In the comic, in panel ve, the student contrasts facial expres-
sions, icon usage, and body positions between two of her characters
to reinforce the verbal narrative as she questions the data in context,
demonstrating one character’s internal confused mental state and
the other’s social engagement. Additionally, the thought bubble
demonstrates the student making sense of the data as she conveys
her uncertainty of whether 2.2% is signicant.
In other instances of contextualization and social awareness,
students considered the role of race and other aspects of identity in
their relationship to making friendships, and reasoning about the
statistical interpretations within their data. When asked if there
were other things that the survey on friendship was missing, one
student responded:
“I feel like they need bullying and maybe something
close to friendships with how you make the friendship
in life when it works out. For example, if you’re maybe
a person of color it could be dicult for you to make
friends, or if you have a disability or reasons why you
might be excluded or reasons why you might not feel
like you can make friends.
Though these reections, it is clear that some students leveraged
the data to think about the contextual characteristics that might
make it harder to make friends in certain settings.
5.3 Situating the Self In Data
The second most common type of data reasoning involved students
describing an “individual experience” (data point) within the data,
that is, situating a personal experience within a broader pattern
shown in the data (72.73%). The 24 comics coded with this code
were also coded with the SEL competency of “self-awareness”, and
in their narratives it became clear how these supported one another.
The narratives often integrated a character that reected on the
data and where they fell within it. For instance, in the data comic
in Figure 8 the character reects on their own perspectives of the
diculties making friends because it’s scary and nerve wracking
which makes them part of 47.8% that nds it “kind of easy, but also
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
kind of hard” to make friends. The author of this comic uses a range
of facial expressions and body postures to convey the hesitancy of
this character (panel 1-4) that shifts into determination to change
their behavior (panel 5) and nally into success at overcoming
their fear and becoming friends (panel 6). The author contrasts
this character with the other who demonstrates laid back and open
postures with excited and happy facial expressions.
Through their comics, students demonstrated an awareness of
the reasons for their characters’ feelings and actions. For instance,
a character describes that 49.35% of students that responded to the
survey noted that shyness was the reason they felt it was hard to
make friends. They go on to state, “But when I was growing up I
didn’t really talk to people much and I think that’s why I got pretty
shy talking to people, does it aect me making best friends?”
Other students used their comics to illustrate things that were
surprising based on their own experiences. For example, in one
comic, a character reects on how they would always get in trouble
because they never had anyone to “back [them] up”. They ash
back to a 4 panel memory sequence of them getting suspended
because another kid started a ght with them. The character ends
by identifying how they were “surprised to see that according to
<school>’s survey that 54.3% of students believe that they could
count on their friends when things go wrong.
The interviews we conducted suggest that, for at least our four
interviewees, their comics reect students’ own perspectives and
emotional states. In one case, the comic helped the art teacher
identify an get help for a student that was potentially struggling
with depression:
“yeah one of the kids in her one panel came out saying
like...‘I don’t know why I’m so sad all the time’, so sad...
and I was like ‘oh Okay’, so I reached out to her and
guidance and everything so...I’m hopeful that you know,
that was an outlet for her.
The teacher further reected on their hope that this exercise also
helped students to reect on their friendships in a positive way.
Finally, although this rarely occurred, when students went beyond
the individual experience to describe data in the aggregate–such as
how others felt or related to others in contrasting ways–they also
engaged in social awareness and perspective taking.
The utility of data on self-awareness was also reected on by one
of the students, when asked what kinds of data might be relevant
in future data comics on friendship,
“Maybe possible data on how a person acts and stu or
how they see themselves, maybe they see themselves as
more of a social person, I guess, or maybe more of a type
of person that maybe gets angry that thinks a lot, and
then say how many friends they have or something, or
how well they have friends. And so, I feel like data like
that could help me, I don’t know, improving and stu.
5.4 Going Beyond Data
The process of creating the data comic prompted students to go
beyond the data by either inquiring about the data, or to consider
the implications of the data analyses. For instance in one comic, the
student considers the statistics around teens with online friends
and relates it to a hypothesis for the future, “So now that life is
coming more technology based, even with the pandemic, I wonder
if we are going to see more teens start to make online friends.
They then link this to their understanding of how this changes how
friendships manifest, “[online friendships] are a lot dierent than
normal friendships, it is usually calls, and text messages, and not
seeing their face.
In most instances, reasoning beyond the data led to the con-
sideration of behavioral changes and associated social-emotional
competencies such as self-management and responsible decision
making. For example, comics illustrated characters shifting per-
spectives in things such as how they will approach new friendships
and how they will change how they engage with others online. In
Figure 9, the student reasons about implications of the data as a
means for self-management.
In the narrative, the author creates camaraderie between three
friends, two of which are empathizing (verbally and non-verbally)
with one character’s experience getting bullied. In panel 5-6, one of
the character nds out a statistic that they use as a means helping
the other character rethink their emotional reaction to bullying.
They connect the high frequency of “drama” to the need to decrease
their emotional reactivity to being bullied (e.g., “no big deal”). The
character who was bullied ends the comic, with a changed per-
spective using the data to self-manage their feelings around being
bullied. This connection between the use of data to reect on how
one might manage their emotions was not just enacted through
comic narratives, but in the way student authors reected on their
reasoning process. For instance, one student shared:
“This made me see that I’m not the only one that’s
struggling and it’s okay because there’s still kids that
are going to nd it easy and they’re going to know how
to become your friend properly. It shows that basically
everyone has a little bit of trouble for the most part
and I’m not alone when it comes to how I view making
friends.
In addition, “going beyond the data” codes frequently co-occurred
with and supported engaging in thinking about external, relationship-
related skills. In the data comic in Figure 10, the student not only
engages in reasoning about the implications of the data, but extends
those implications to making decisions about their social media use.
They rst identify how 15.2% of the class does not use social media,
then go on to pose a set of questions related to the eects of social
media, why people might not use it, and whether there are benets
to not using it. Through the narrative the author demonstrates an
evolution through time in how the main character spends their
time.
6 DISCUSSION
This study explores the role that comics can play in supporting data
reasoning and social-emotional learning. Our results demonstrate
that students created rich narratives that reected a broad variety
of contexts, characters and emotions. Moreover, comic-making en-
gaged students in identifying what a data point means and connect
it to context through describing and illustrating the data within
their narratives.
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Figure 8: Data comic (C05) on counting on friends when things go wrong
6.1 Making Comics and Data Reasoning
The kinds of data reasoning students engage in through their cre-
ation of data comics, are the kinds that “humanize the personal
narratives behind the numbers” [
1
]. Narratives that are constructed
from processes such as situating oneself in the data, considering the
contexts in which data lives, and reecting on the implications that
data may have. The construction of data comics enabled learners to
explore their own narratives and social-emotional lives that the data
represent and hypothesize about other narratives implicit in the
data. The students brought context to the data and communicated
about it. Using informal inference, students engaged in theorizing
to explain or account for the data in relation to the context. In
line with Cobb
[14]
’s verbal/interpretive conception of statistical
reasoning, the narrative paved the way for students to engage in
reasoning about other factors or measures that might explain the
observed pattern.
In addition, the interviews with students and teachers indicated
that many of the students reected on their own situativity within
the data. Prior work has explored the advantages of rst-person “ac-
tor perspective-taking” in data reasoning such as work by Roberts
and Lyons
[55]
where learners imagine themselves represented
visuo-spatially inside datasets, and work by Kahn
[35]
where stu-
dents situate themselves in longitudinal data about family histories.
The comic medium supported the students in unpacking, where
they or their characters fell within the data and what types of
experiences may have led to a particular position.
Furthermore, we saw evidence of students reasoning around how
the data may be implicated in the future of the characters and their
behavior or actions around building and maintaining friendships.
Similar to what Curcio
[15]
called reading beyond the data, here
students made inferences from the data (e.g., pandemics might
increase online friendships) by tapping their existing schema for
information (e.g., online friendships are a lot dierent than normal
friendships) that is not explicitly stated in the data.
One of the challenges for data reasoning is nding a signal from
the noise and variability in the data [
41
]. The comic format enabled
the students to incorporate more than one narrative within the data,
such as that of a character that nds it easy to make friends and a
character that nds it dicult. In their stories, students explored
how these dierences in experiences might lead to characters’ dif-
ferent responses to their comics’ narrative circumstances. Through
the digital storytelling learners confronted these uncertainties as
they created their characters, scenes, and short stories.
In some cases, students were able to move from reasoning about
how they t with the data to describing others’ perspectives on
friendship. However, these descriptions were rare, as were propor-
tional comparisons that required identifying and then comparing
groups. When students engaged in this kind of sophisticated rea-
soning, social awareness was also present. Furthermore, some of
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
Figure 9: Data comic (C22) on online bullying and frequency of drama
the narratives integrate assumptions and biases about the data and
data points that are not supported, and in a few cases, students
made errors in their interpretations. The comic form provides the
opportunity to elaborate on what is not there. However, we see this
as a potential jumping o point to use the comic artifacts as not
the nal conclusions drawn from the data, but as a tool within the
sense-making process. In line with research by Hancock et al
. [30]
,
there are a variety of challenges in connecting students’ statistical
questions to the data needed as evidence, and then again linking
their conclusions back to the questions under investigation. Because
the students were in remote classes, the types of discussions using
the comics were limited, but in future work there is promise for
them to augment the discussions surrounding claims within their
data stories, and questions that the data has not yet answered. In
some instances, students did frame their questions as questions, and
in others we saw the comic integrating assumptions, suggesting
that students need more support connecting data to evidence.
6.2 Data Reasoning and Social-Emotional
Learning (SEL)
While the process of constructing a narrative supported dierent
kinds of data reasoning, the content focus on data about teen friend-
ships meant reasoning about the data enacted a variety of social-
emotional reasoning competencies such as self-awareness, social
awareness, and relationship skills. For instance, in reasoning about
how they might be situated in the data, many students engaged in
a form of self-awareness, considering the relevant thoughts, emo-
tions, and values and how they inuence behavior across contexts.
In reasoning about the context in which data lives, students con-
sidered related data that might not be in the dataset, yet present
interesting connections. For instance, students thought about the
historical context of friendship making—who already had friends
and in turn might nd it easier to make friends? This reasoning,
situated within the context of teen friendship data, manifested as
a form of social awareness, where they sought to understand the
perspectives of and empathize with others, including those from
diverse backgrounds, cultures, and contexts. In turn, some students
asked themselves—how might racial, gender, or disability aspects of
an individual inuence their perception of friend making or sustain-
ing friendships? In constructing contexts and considering related
data, students engaged in a form of perspective taking, commonly
described as a form of social awareness.
Further, the inherent textual and visual components of comics
as an art form enabled students’ to integrate emotions of their char-
acters through what Lipton and Nowicki
[43]
referred to as verbal
and non-verbal "emotional signals." While verbal signals include
recognizing cues that are expressed aloud, non-verbal signals in-
clude, facial expressions, postures, gestures, tones of voice, distance
in personal space, rhythm of interaction, apparel, and touch [
43
].
These non-verbal signals enabled students to express emotional
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Figure 10: Data comic (C19) on social media use
information of their characters to build the context within their
narratives.
Some of our ndings suggest that incorporating certain kinds
of data literacy skills, such as reasoning about groups or data in
the aggregate, might lead to specic kinds of SEL reasoning, such
as perspective taking and social awareness. Similarly, situating
oneself in the data seems to necessitate self-awareness. The reverse
may also be true—that asking students to reason about dierences
between groups of people might also lead to more sophisticated
data literacy skills.
Other forms of data reasoning such as considering the implica-
tions of the data many of the students drew on social-emotional
components such as self-management and responsible decision-
making, which are behavior-oriented rather than purely cognitive
or aective. In other words, as students grappled with what the data
might mean, their narratives exemplied ways in which the data
could be acted upon through actively managing one’s emotions
or thoughts, or being explicit about making caring and construc-
tive choices about personal behavior and social interactions. This
drawing of connections between data analysis, personal contexts,
and behavior echoes the work of Stornaiuolo
[59]
which explores
how engaging high school learners in the intentional construction
and analysis of data about their everyday activities helped them
develop agentive orientations toward data.
In drawing an intersection between data reasoning and social-
emotional reasoning we draw connections between a rapidly grow-
ing area of interest in STEM (data literacy) and use an art-based
approaches to connect to a social-emotional learning. Despite play-
ing a critical role in improving children’s academic performance
and lifelong learning [
63
], the integration of SEL competencies into
STEM curriculum, rather than as a stand-alone curriculum, remains
an area for further exploration. Especially within the current con-
text in which, according to data from the Centers for Disease Con-
trol and Prevention’s Youth Risk Behavior Survey (YRBS), feelings
of sadness and hopelessness and adolescent suicide-related behav-
iors have signicantly increased in the past decade Ivey-Stephenson
et al
. [34]
, creating more curricular options that integrate SEL is
important. As we seek to draw connections between data liter-
acy and socio-scientic argumentation around topics like climate
change [
33
], so to should we leverage the opportunity to draw data
literacy connections with mental health topics that may expand
social-emotional reasoning capacities, and in turn support other
mental health eorts in schools.
7 LIMITATIONS & FUTURE WORK
In our project the data was made available to the students through
graphs prepared by the research team which students could freely
explore. While such an approach may minimize technical impedi-
ments to creating data representations directly from the primary
"I happen to be one of 47.8%": Social-Emotional and Data Reasoning in Middle School Students’ Comics about Friendship CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA
data sources (often called “data wrangling” in the Information Sci-
ences; e.g., [
36
]), it also limits an important component of data
reasoning. In what Erickson et al
. [22]
refers to as “data moves”,
allowing students to lter, group, calculate, merge, summarize etc.,
reinforces the non-neutrality concept of data, in that by allowing
data to be manipulated, students can reect on how decisions about
data can enable or constrain the types of investigations.
Another limitation of this project were the limited opportunities
to iterate on narratives that embody certain kinds of data reasoning.
For instance, as students theorized about related datasets, situated
themselves in that data and reected on their own assumptions, cy-
cles of construction and analysis would have helped assess bias and
evidence and possibly support more complex data argumentation.
Lastly, there were a variety of statistical reasoning competencies
that were not evident in the comics. For instance, probabilistic lan-
guage in the comics was minimal. Few students compared percent-
ages across groups or used others statistical measures. Future work,
may consider providing scaolded structures that pairs narratives
constraints with certain kind of data reasoning to target particular
data reasoning competencies. For instance, perhaps prompts to
have characters in the comics imagine future scenarios using the
data, might prompt probabilistic language and a greater connection
with future-oriented SEL competencies like self-management and
responsible decision-making.
8 CONCLUSION
As data literacy is increasingly viewed as integral to civic engage-
ment and critical literacies in a technocentric society, the question
of how we can move towards a more humanistic and interdisci-
plinary view of data literacy will only grow in the coming years.
Echoing Tygel and Kirsch
[62]
, who draws a parallel between data
literacy and the critical pedagogy of Brazilian educator Freire
[26]
,
literacy goes beyond just reading words, but about reading the
world. A centering of data reasoning connected to argumentation
through creative expression can be part of other approaches that
seek to build one’s capacity to read, manipulate, communicate, and
produce data in ways that expose its non-neutrality. Such an un-
derstanding can empower learners to develop social understanding
and action as they become users, creators, and critical thinkers
with data rather than merely subjects of data [
18
]. As a form of
expressive construction [
56
], art-making can fuse a personally and
culturally relevant discipline to data literacy, and invite learners to
bring their own interests, experiences and skills to making meaning.
Integrating such an art-based approach that centers data on rela-
tionships, can connect data reasoning to social-emotional learning
in ways we hope will be further explored in future work.
ACKNOWLEDGMENTS
We thank the students and teachers that made this work possible
along with the funding provided by National Science Foundation
(NSF) grant 1908557 and 1908142.
REFERENCES
[1]
Aria Alamalhodaei, Alexandra P Alberda, and Anna Feigenbaum. 2020. 21.
Humanizing data through ‘data comics’: An introduction to graphic medicine
and graphic social science. In Data Visualization in Society. Amsterdam University
Press, 347–366.
[2]
Farah Nadia Azman, Syamsul Bahrin Zaibon, and Norshuhada Shiratuddin. 2015.
Digital storytelling tool for education: An analysis of comic authoring environ-
ments. In International Visual Informatics Conference. Springer, 347–355.
[3]
Benjamin Bach, Nathalie Henry Riche, Sheelagh Carpendale, and Hanspeter
Pster. 2017. The emerging genre of data comics. IEEE computer graphics and
applications 37, 3 (2017), 6–13.
[4]
Benjamin Bach, Zezhong Wang, Matteo Farinella, Dave Murray-Rust, and
Nathalie Henry Riche. 2018. Design patterns for data comics. In Proceedings
of the 2018 chi conference on human factors in computing systems. 1–12.
[5]
Natasha H Bailen, Lauren M Green, and Renee J Thompson. 2019. Understanding
emotion in adolescents: A review of emotional frequency, intensity, instability,
and clarity. Emotion Review 11, 1 (2019), 63–73.
[6]
Arthur Bakker and Jan Derry. 2011. Lessons from inferentialism for statistics
education. Mathematical thinking and learning 13, 1-2 (2011), 5–26.
[7]
Arthur Bakker and Koeno P E Gravemeijer. 2004. LEARNING TO REASON
ABOUT DISTRIBUTION.
[8]
Dani Ben-Zvi, Katie Makar, and Joan Gareld. [n.d.]. Springer International
Handbooks of Education International Handbook of Research in Statistics Education.
http://www.springer.com/series/6189
[9] Rahul Bhargava, Erica Deahl, Emmanuel Letouzé, Amanda Noonan, David San-
gokoya, and Natalie Shoup. 2015. Beyond data literacy: Reinventing community
engagement and empowerment in the age of data. (2015).
[10]
Rahul Bhargava, Ricardo Kadouaki, Emily Bhargava, Guilherme Castro, and
Catherine D’Ignazio. 2016. Data murals: Using the arts to build data literacy. The
Journal of Community Informatics 12, 3 (2016).
[11]
Danah Boyd and Kate Crawford. 2012. Critical questions for big data: Provo-
cations for a cultural, technological, and scholarly phenomenon. Information
Communication and Society 15 (6 2012), 662–679. Issue 5. https://doi.org/10.1080/
1369118X.2012.678878
[12]
Paola Cabrera, Luz Castillo, Paúl González, Ana Quiñónez, and César Ochoa.
2018. The Impact of Using" Pixton" for Teaching Grammar and Vocabulary in the
EFL Ecuadorian Context. Teaching English with Technology 18, 1 (2018), 53–76.
[13] Pew Research Center. 2015. Teens, Social Media, and Technology. (2015).
[14]
George W Cobb. 1997. Mere literacy is not enough. Why numbers count: Quanti-
tative literacy for tomorrow’s America (1997), 75–90.
[15]
Frances R Curcio. 1989. Developing Graph Comprehension. Elementary and Middle
School Activities. ERIC.
[16]
Erica Deahl and B A Art. 2007. Better the Data You Know: Developing Youth
Data Literacy in Schools and Informal Learning Environments. (2007).
[17]
Catherine D’Ignazio and Rahul Bhargava. 2016. DataBasic: Design principles,
tools and activities for data literacy learners. The Journal of Community Infor-
matics 12, 3 (2016).
[18] Catherine D’Ignazio and Laura F. Klein. 2020. Data Feminism. MIT Press.
[19]
Will Eisner. 2008. Comics and sequential art: Principles and practices from the
legendary cartoonist. WW Norton & Company.
[20]
Katie Eklund, Kayla D Kilpatrick, Stephen P Kilgus, and Aqdas Haider. 2018. A
systematic review of state-level social–emotional learning standards: Implications
for practice and research. School Psychology Review 47, 3 (2018), 316–326.
[21]
Maurice J Elias, Joseph E Zins, Roger P Weissberg,Karin S Frey, Mark T Greenberg,
Norris M Haynes, Rachael Kessler, Mary E Schwab-Stone, and Timothy P Shriver.
1997. Promoting social and emotional learning: Guidelines for educators. Ascd.
[22]
Tim Erickson, Michelle Wilkerson, William Finzer, and Frieda Reichsman. 2019.
Data moves. Technology Innovations in Statistics Education 12, 1 (2019).
[23]
Uwe Flick. 1992. Triangulation revisited: strategy of validation or alternative?
Journal for the theory of social behaviour (1992).
[24]
Collaborative for Academic Social and Emotional Learning. 2015. 2015 CASEL
guide: Eective social and emotional learning programs. Middle and high school
edition. (2015).
[25]
John R Frederiksen, Mike Sipusic, Miriam Sherin, and Edward W Wolfe. 1998.
Video portfolio assessment: Creating a framework for viewing the functions of
teaching. Educational Assessment 5, 4 (1998), 225–297.
[26] Paulo Freire. 2018. Pedagogy of the oppressed. Bloomsbury publishing USA.
[27]
Chris Gavaler and Leigh Ann Beavers. 2020. Clarifying closure. Journal of graphic
novels and Comics 11, 2 (2020), 182–211.
[28]
Mark T Greenberg, Roger P Weissberg, Mary Utne O’Brien, Joseph E Zins, Linda
Fredericks, Hank Resnik, and Maurice J Elias. 2003. Enhancing school-based
prevention and youth development through coordinated social, emotional, and
academic learning. American psychologist 58, 6-7 (2003), 466.
[29]
Nancy G Guerra and Catherine P Bradshaw. 2008. Linking the prevention of prob-
lem behaviors and positive youth development: Core competencies for positive
youth development and risk prevention. New directions for child and adolescent
development 2008, 122 (2008), 1–17.
[30]
Chris Hancock, James J Kaput, and Lynn T Goldsmith. 1992. Authentic inquiry
with data: Critical barriers to classroom implementation. Educational Psychologist
27, 3 (1992), 337–364.
[31]
John Harding. [n.d.]. Pixton Comics wins $30,000 award. BC Local news ([n. d.]).
https://www.bclocalnews.com/business/pixton-comics- wins-30000- award/
[32]
Barbara Hug and Katherine L. McNeill. 2008. Use of rst-hand and second-hand
data in science: Does data type inuence classroom conversations? International
CHI ’22, April 29-May 5, 2022, New Orleans, LA, USA Vacca and DesPortes, et al.
Journal of Science Education 30 (10 2008), 1725–1751. Issue 13. https://doi.org/10.
1080/09500690701506945
[33]
Golnaz Arastoopour Irgens, Knight Simon, Alyssa Wise, Thomas Philip, Maria C
Olivares, Sarah Van Wart, Sepehr Vakil, Jessica Marshall, Tapan S Parikh,
M Lisette Lopez, et al. 2020. Data literacies and social justice: Exploring critical
data literacies through sociocultural perspectives. (2020).
[34]
Asha Z Ivey-Stephenson, Zewditu Demissie, Alexander E Crosby, Deborah M
Stone, Elizabeth Gaylor, Natalie Wilkins, Richard Lowry, and Margaret Brown.
2020. Suicidal ideation and behaviors among high school students—Youth Risk
Behavior Survey, United States, 2019. MMWR supplements 69, 1 (2020), 47.
[35]
Jennifer Kahn. 2020. Learning at the Intersection of Self and Society: The Family
Geobiography as a Context for Data Science Education. Journal of the Learning
Sciences 29 (1 2020), 57–80. Issue 1. https://doi.org/10.1080/10508406.2019.1693377
[36]
Sean Kandel, Jerey Heer, Catherine Plaisant, Jessie Kennedy, Frank Van Ham,
Nathalie Henry Riche, Chris Weaver, Bongshin Lee, Dominique Brodbeck, and
Paolo Buono. 2011. Research directions in data wrangling: Visualizations and
transformations for usable and credible data. Information Visualization 10, 4
(2011), 271–288.
[37]
DaYe Kang, Tony Ho, Nicolai Marquardt, Bilge Mutlu, and Andrea Bianchi. 2021.
ToonNote: Improving Communication in Computational Notebooks Using Inter-
active Data Comics. In Proceedings of the 2021 CHI Conference on Human Factors
in Computing Systems. 1–14.
[38]
Nam Wook Kim, Nathalie Henry Riche, Benjamin Bach, Guanpeng Xu, Matthew
Brehmer, Ken Hinckley, Michel Pahud, Haijun Xia, Michael J McGun, and
Hanspeter Pster. 2019. Datatoon: Drawing dynamic network comics with pen+
touch interaction. In Proceedings of the 2019 CHI Conference on Human Factors in
Computing Systems. 1–12.
[39]
Rob Kitchin. 2014. The data revolution : big data, open data, data infrastructures
and their consequences.
[40]
Cliord Konold, Traci Higgins, Susan Jo Russell, and Khalimahtul Khalil. 2015.
Data seen through dierent lenses. Educational Studies in Mathematics 88 (3
2015), 305–325. Issue 3. https://doi.org/10.1007/s10649-013- 9529-8
[41]
Cliord Konold and Alexander Pollatsek. 2002. Data analysis as the search for
signals in noisy processes. Journal for research in mathematics education 33, 4
(2002), 259–289.
[42]
Victor R. Lee, Joel Drake, Ryan Cain, and Jerey Thayne. 2021. Remembering
What Produced the Data: Individual and Social Reconstruction in the Context of
a Quantied Self Elementary Data and Statistics Unit. Cognition and Instruction
(2021). https://doi.org/10.1080/07370008.2021.1936529
[43]
M Lipton and Stephen Nowicki. 2009. The social emotional learning frame-
work (SELF): A guide for understanding brain-based social emotional learning
impairments. Journal of Developmental Processes 4, 2 (2009), 99–115.
[44]
Deborah Lupton. 2017. Feeling your data: Touch and making sense of personal
digital data. New Media & Society 19, 10 (2017), 1599–1614.
[45]
Katie Makar and Andee Rubin. 2009. A framework for thinking about informal
statistical inference. Statistics Education Research Journal 8, 1 (2009).
[46]
Katie Makar and Andee Rubin. 2018. Learning about statistical inference. Inter-
national handbook of research in statistics education (2018), 261–294.
[47]
Ann S Masten and J Douglas Coatsworth. 1998. The development of competence
in favorable and unfavorable environments: Lessons from research on successful
children. American psychologist 53, 2 (1998), 205.
[48]
Scott McCloud. 1993. Understanding comics: The invisible art. Northampton,
Mass (1993).
[49]
Ann Miller. 2007. Reading bande dessinée: critical approaches to French-language
comic strip. Intellect Books.
[50]
Erik C Nook, Stephanie F Sasse, Hilary K Lambert, Katie A McLaughlin, and
Leah H Somerville. 2017. Increasing verbal knowledge mediates development of
multidimensional emotion representations. Nature human behaviour 1, 12 (2017),
881–889.
[51]
Erik C Nook, Stephanie F Sasse, Hilary K Lambert, Katie A McLaughlin, and
Leah H Somerville. 2018. The nonlinear development of emotion dierentiation:
Granular emotional experience is low in adolescence. Psychological science 29, 8
(2018), 1346–1357.
[52]
Maxine Pfannkuch, Matt Regan, Chris Wild, and Nicholas J Horton. 2010. Telling
data stories: Essential dialogues for comparative reasoning. Journal of Statistics
Education 18, 1 (2010).
[53]
Thomas M. Philip, Sarah Schuler-Brown, and Winmar Way. 2013. A framework
for learning about big data with mobile technologies for democratic participation:
Possibilities, limitations, and unanticipated obstacles. Technology, Knowledge and
Learning 18 (10 2013), 103–120. Issue 3. https://doi.org/10.1007/s10758- 013-9202-
4
[54]
Henry John Pratt. 2009. Narrative in comics. The Journal of Aesthetics and Art
Criticism 67, 1 (2009), 107–117.
[55]
Jessica Roberts and Leilah Lyons. 2020. Examining spontaneous perspective tak-
ing and uid self-to-data relationships in informal open-ended data exploration.
Journal of the Learning Sciences 29, 1 (2020), 32–56.
[56]
Jeremy Roschelle, Wendy Martin, June Ahn, and Patricia Schank. 2017. Cyber-
learning community report: The state of cyberlearning and the future of learning
with technology. Technical Report. SRI International.
[57]
Andreas Schreiber and Regina Struminski. 2017. Tracing personal data using
comics. In International Conference on Universal Access in Human-Computer
Interaction. Springer, 444–455.
[58]
Blaine E Smith, Ji Shen, and Shiyan Jiang. 2019. The science of storytelling:
Middle schoolers engaging with socioscientic issues through multimodal science
ctions. Voices from the Middle 26, 4 (2019).
[59]
Amy Stornaiuolo. 2020. Authoring Data Stories in a Media Makerspace: Adoles-
cents Developing Critical Data Literacies. Journal of the Learning Sciences 29 (1
2020), 81–103. Issue 1. https://doi.org/10.1080/10508406.2019.1689365
[60]
Katie Headrick Taylor and Rogers Hall. 2013. Counter-mapping the neighborhood
on bicycles: Mobilizing youth to reimagine the city. Technology, Knowledge and
Learning 18 (7 2013), 65–93. Issue 1-2. https://doi.org/10.1007/s10758-013-9201- 5
[61]
Tin Lam Toh, Lu Pien Cheng, Siew Yin Ho, Heng Jiang, and Kam Ming Lim.
2017. Use of comics to enhance students’ learning for the development of the
twenty-rst century competencies in the mathematics classroom. Asia Pacic
Journal of Education 37, 4 (2017), 437–452.
[62]
Alan Freihof Tygel and Rosana Kirsch. 2016. Special Issue on Data Literacy:
Articles Contributions of Paulo Freire to a Critical Data Literacy: a Popular
Education Approach. The Journal of Community Informatics 12 (2016), 108–121.
Issue 3. www.eita.org.br.
[63]
Sarah K Ura, Sara M Castro-Olivo, and Ana d’Abreu. 2020. Outcome measurement
of school-based SEL intervention follow-up studies. Assessment for Eective
Intervention 46, 1 (2020), 76–81.
[64]
Phil Vahey, Louise Yarnall, Charles Patton, Daniel Zalles, Karen Swan, and San
Francisco. 2006. Mathematizing Middle School Mathematizing Middle School:
Results From a Cross-Disciplinary Study of Data Literacy.
[65]
Zezhong Wang, Harvey Dingwall, and Benjamin Bach. 2019. Teaching data
visualization and storytelling with data comic workshops. In Extended Abstracts
of the 2019 CHI Conference on Human Factors in Computing Systems. 1–9.
[66]
Zezhong Wang, Jacob Ritchie, Jingtao Zhou, Fanny Chevalier, and Benjamin Bach.
2020. Data Comics for Reporting Controlled User Studies in Human-Computer
Interaction. IEEE Transactions on Visualization and Computer Graphics 27, 2
(2020), 967–977.
[67]
Zezhong Wang, Shunming Wang, Matteo Farinella, Dave Murray-Rust, Nathalie
Henry Riche, and Benjamin Bach. 2019. Comparing eectiveness and engagement
of data comics and infographics. In Proceedings of the 2019 CHI Conference on
Human Factors in Computing Systems. 1–12.
[68]
Roger P Weissberg, Karol L Kumpfer, and Martin EP Seligman. 2003. Prevention
that works for children and youth: An introduction. Vol. 58. American Psychological
Association.
[69]
Michelle Hoda Wilkerson and Joseph L. Polman. 2020. Situating Data Science:
Exploring How Relationships to Data Shape Learning. Journal of the Learning
Sciences 29 (1 2020), 1–10. Issue 1. https://doi.org/10.1080/10508406.2019.1705664
[70]
Annika Wol, Daniel Gooch, Jose J Cavero Montaner, Umar Rashid, and Gerd
Kortuem. 2016. Creating an understanding of data literacy for a data-driven
society. The Journal of Community Informatics 12, 3 (2016).
[71]
Joseph E Zins. 2004. Building academic success on social and emotional learning:
What does the research say? Teachers College Press.
... Classroom data varied based on school-specific IRB permission, and included student artefacts (artworks, written responses to reflective prompts), classroom observations, prepost surveys, and student interviews. Further detail, and unit-specific findings based on these data are reported elsewhere DesPortes et al., 2022;Vacca et al., 2022). ...
Article
Full-text available
Data-art inquiry is an arts-integrated approach to data literacy learning that reflects the multidisciplinary nature of data literacy not often taught in school contexts. By layering critical reflection over conventional data inquiry processes, and by supporting creative expression about data, data-art inquiry can support students' informal inference-making by revealing the role of context in shaping the meaning of data, and encouraging consideration of the personal and social relevance of data. Data-art inquiry additionally creates alternative entry points into data literacy by building on learners' non-STEM interests. Supported by technology, it can provide accessible tools for students to reflect on and communicate about data in ways that can impact broader audiences. However, data-art inquiry instruction faces many barriers to classroom implementation, particularly given the tendency for schools to structure learning with disciplinary silos, and to unequally prioritize mathematics and the arts. To explore the potential of data-art inquiry in classroom contexts, we partnered with arts and mathematics teachers to co-design and implement data-art inquiry units. We implemented the units in four school contexts that differed in terms of the student population served, their curriculum priorities, and their technology infrastructure. We reflect on participant interviews, written reflections, and classroom data, to identify synergies and tensions between data literacy, technology, and the arts. Our findings highlight how contexts of implementation shape the possibilities and limitations for data-art inquiry learning. To take full advantage of the potential for data-art inquiry, curriculum design should account for and build on the opportunities and constraints of classroom contexts.
Article
Full-text available
Society is become increasingly reliant on data, making it necessary to ensure that all citizens are equipped with the skills needed to be data literate. We argue that the foundations for a data literate society begin by acquiring key data literacy competences in school. However, as yet there is no clear definition of what these should be. This paper explores the different perspectives currently offered on both data and statistical literacy and then critically examines to what extent these address the data literacy needs of citizens in today’s society. We survey existing approaches to teaching data literacy in schools, to identify how data literacy is interpreted in practice. Based on these analyses, we propose a definition of data literacy that is focused on using data to understand real world phenomena. The contribution of this paper is the creation of a common foundation for teaching and learning data literacy skills.
Article
Given growing interest in K-12 data and data science education, new approaches are needed to help students develop robust understandings of and familiarity with data. The model of the quantified self—in which data about one’s own activities are collected and made into objects of study—provides inspiration for one such approach. By drawing on what one already knows about their self and their prior experiences, it may be possible to bootstrap students’ abilities to interpret and make sense of data. Taking that possibility seriously, this article describes some of the gains observed in students’ statistical reasoning following a quantified self, wearables-based elementary statistics unit and provides a theoretical framework drawing from cognitive psychology, embodiment, and situative perspectives to characterize how prior experience is used as a resource in data sense-making when the data are about students’ own physical experiences. This framework centralizes and interrogates the work of “remembering” prior experiences and articulates how remembering is involved in interpreting quantified self data. Specifically, the framework emphasizes that remembering in service of data interpretation is a reconstructive act that draws from both general and specific embodied resources and that the work of reconstructive remembering in the classroom is both individual and multi-participant work. To demonstrate measured learning gains and illustrate the framework, written assessment results and descriptive cases of student and teacher discussions about quantified self data from two sixth-grade classes participating in a classroom design experiment are provided. Both a discussion of and recommendations for ethical considerations related to quantified self data in education are also provided.
Conference Paper
Computational notebooks help data analysts analyze and visualize datasets, and share analysis procedures and outputs. However, notebooks typically combine code (e.g., Python scripts), notes, and outputs (e.g., tables, graphs). The combination of disparate materials is known to hinder the comprehension of notebooks, making it difficult for analysts to collaborate with other analysts unfamiliar with the dataset. To mitigate this problem, we introduce ToonNote, a JupyterLab extension that enables the conversion of notebooks into “data comics.” ToonNote provides a simplified view of a Jupyter notebook, highlighting the most important results while supporting interactive and free exploration of the dataset. This paper presents the results of a formative study that motivated the system, its implementation, and an evaluation with 12 users, demonstrating the effectiveness of the produced comics. We discuss how our findings inform the future design of interfaces for computational notebooks and features to support diverse collaborators.
Article
Paulo Freire is the patron of education in Brazil. His main work - the Popular Education pedagogy - influences many educators all over the world who believe in education as a way of liberating poor oppressed people. One of the outcomes of Freire's work is a literacy method, developed in the 1960's. In this paper, we propose the adoption of elements of Freire's Literacy Method for use in a pedagogical pathway towards data literacy. After tracing some parallels between literacy education and data literacy, we suggest some data literacy strategies inspired on Freire's method. We also derive from it a definition for critical data literacy.
Article
The growing number of tools for data novices are not designed with the goal of learning in mind. This paper proposes a set of pedagogical design principles for tool development to support data literacy learners. We document their use in the creation of three digital tools and activities that help learners build data literacy, showing design decisions driven by our pedagogy. Sketches students created during the activities reflect their adeptness with key data literacy skills. Based on early results, we suggest that tool designers and educators should orient their work from the outset around strong pedagogical principles.
Article
Current efforts to build data literacy focus on technology-centered approaches, overlooking creative non-digital opportunities. This case study is an example of how to implement a Popular Education-inspired approach to building participatory and impactful data literacy using a set of visual arts activities with students at an alternative school in Belo Horizonte, Brazil. As a result of the project data literacy among participants increased, and the project initiated a sustained interest within the school community in using data to tell stories and create social change.