Content uploaded by Rama Adithya Varanasi
Author content
All content in this area was uploaded by Rama Adithya Varanasi on Mar 15, 2021
Content may be subject to copyright.
Tag a Teacher: A alitative Analysis of WhatsApp-Based
Teacher Networks in Low-Income Indian Schools
Rama Adithya Varanasi
Department of Information Science
Cornell University
New York, NY, USA
Aditya Vashistha
Department of Information Science
Cornell University
New York, NY, USA
Nicola Dell
Department of Information Science
Jacobs Institute, Cornell Tech
New York, NY, USA
ABSTRACT
Although WhatsApp-based communication is playing an increas-
ingly large role in the professional lives of teachers in low-income
schools, the nature of the interactions that occur and how these
interactions enable cooperative work are not well understood. We
contribute a qualitative analysis of 26 existing WhatsApp groups
(35,567 messages) that examines (1) the strategies used to encour-
age interaction within teacher-focused WhatsApp groups, and (2)
how these interactions are sustained by teachers, management,
and organizations over a period of time. We use teacher networks
and activity awareness model to make sense of WhatsApp-based
interactions and show how WhatsApp narrows the gap between
management and teachers, leading to additional work and stress
for teachers. WhatsApp was also used to circulate polarizing and
malicious information, leading to a variety of interesting content
moderation strategies. Our ndings expand the scope of research
on teacher networks to low-income contexts and will inform future
interventions that enable cooperative work among teachers.
CCS CONCEPTS
•Human-centered computing →Empirical studies in HCI.
KEYWORDS
Education; teacher networks; whatsapp; teacher development; co-
operative work; ICTD; HCI4D
ACM Reference Format:
Rama Adithya Varanasi, Aditya Vashistha, and Nicola Dell. 2021. Tag a
Teacher: A Qualitative Analysis of WhatsApp-Based Teacher Networks
in Low-Income Indian Schools. In CHI Conference on Human Factors in
Computing Systems (CHI ’21), May 8–13, 2021, Yokohama, Japan. ACM, New
York, NY, USA, 16 pages. https://doi.org/0.1145/3411764.3445221
1 INTRODUCTION
The rapid proliferation of smartphone devices and falling costs of
mobile data across the world, and especially in India, has resulted
in online communication platforms (e.g., WhatsApp) playing an in-
creasingly large role in people’s personal [
17
,
49
,
56
,
84
] and profes-
sional lives [
33
,
53
,
55
,
79
,
88
,
94
,
109
]. Many new smartphone users
in the Global South, who are often interacting with digital technolo-
gies and the Internet for the rst time, possess only one (or a shared)
Permission to make digital or hard copies of part or all of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for prot or commercial advantage and that copies bear this notice and the full citation
on the rst page. Copyrights for third-party components of this work must be honored.
For all other uses, contact the owner/author(s).
CHI ’21, May 8–13, 2021, Yokohama, Japan
©2021 Copyright held by the owner/author(s).
ACM ISBN 978-1-4503-8096-6/21/05.
https://doi.org/0.1145/3411764.3445221
device and are expected to quickly learn to balance their use of this
device and associated online platforms to coordinate both personal
and work communications. In these contexts, it is important to
study the nature of the interactions that occur via online platforms
to better understand how the technologies may (or may not) enable
cooperative work among groups of people and the strategies that
have been developed to curate, share, and moderate information.
In this paper, we examine how teachers from low-income Indian
schools, school administrators, and sta from education-focused
organizations use WhatsApp groups to communicate with each
other and provide pedagogical support to teachers. We use the
concept of teacher networks [
65
] to examine WhatsApp-based in-
teractions among existing groups of teachers in low-income Indian
schools. Teacher networks have long been used to explore how
teachers with diverse backgrounds come together for shared ac-
tivities and experiences to achieve common goals in their work
[
65
]. For example, prior work has shown how teacher networks
with strong support structures and cohesive interactions can lead
to interventions and policies that improve communal relations [
65
],
collective agency [
82
], and professional development [
95
]. However,
most prior work on teacher networks has focused on educators
in Western contexts. There is a severe scarcity of research that
examines teacher networks in low-resource environments in the
Global South.
Our research lls this gap by examining how these networks
are enacted via WhatsApp group communications among teachers
in low-income Indian schools, where both smartphone adoption
by teachers and use of WhatsApp for professional purposes are
relatively recent phenomena. In particular, we sought to answer
the following research questions:
RQ1:
What strategies are used
to encourage interaction within existing WhatsApp-based teacher
networks in low-income Indian schools? and
RQ2:
How are these
interactions sustained by teachers, management, and organizations
over period of time?
To answer these questions, we conducted a qualitative study
in which we collected and analyzed conversations from 26 exist-
ing WhatsApp groups (a total of 35,567 messages) that took place
between teachers, school administrators, and sta from education-
focused organizations in India. Our dataset consists of three types
of WhatsApp groups: (1) groups that are created and managed by
321 [
1
], an organization that oers training workshops and support
to improve teachers’ capacities; (2) groups managed by Meghshala
[
74
], an organization that markets a mobile platform to improve
teachers’ pedagogical knowledge; and (3) groups that are adminis-
tered by schools’ higher management.
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
To examine the nature of cooperative work manifesting on these
dierent types of WhatsApp groups, we use Neale et al.’s [
80
]
activity awareness model as an analytical lens. While theoretical
foundations of teacher networks are useful to conceptually under-
stand these WhatsApp-based teacher networks as a whole, activity
awareness model enables us to study and categorize low-level indi-
vidual interactions among stakeholders in these networks, thereby
allowing us to compare cohesiveness and cooperative work across
dierent groups.
For RQ1, we show how teachers’ WhatsApp conversations em-
ployed creative structures (e.g., quizzes, puzzles) that often repur-
posed WhatsApp’s built-in features (e.g., emojis, image editor) in
innovative ways to engage teachers and encourage them to develop
professional skills. The groups were also used to actively recognize
and celebrate teachers’ work. For RQ2, we show how using What-
sApp to share professional resources helped sustain the interactions
but reduced the gap between school administrators and teachers,
potentially leading to extra work and stress for teachers. In ad-
dition, WhatsApp groups were used to circulate misinformation,
malicious spam, and religiously and politically polarizing infor-
mation, exposing teachers to a range of possible digital harms. In
response, participants used a variety of content moderation strate-
gies to keep conversations focused on education and reprimand
those who posted unacceptable content.
Lastly, we highlight opportunities for future research to (1) mea-
sure the impact of WhatsApp use on teachers’ wellbeing, and (2)
further analyze the spread of misinformation on teacher-focused
WhatsApp groups. In summary, we make the following contribu-
tions to the HCI(4D) community:
(1)
We expand existing knowledge on teacher networks to the
Global South by showing how formal and informal teacher
networks are enacted via WhatsApp group conversations
between teachers, administrators, and sta from education-
focused organizations in low-income Indian schools.
(2)
We highlight the strategies used to encourage interaction
within WhatsApp-based teacher networks and demonstrate
the kinds of cooperative work currently achieved within
these groups.
(3)
We reveal how interactions are sustained on teacher net-
works through content curation, sharing, and moderation on
WhatsApp-based teacher groups, including how WhatsApp
use may create additional work, stress, and risks for teachers,
and we discuss the potential eects on teachers’ wellbeing.
2 BACKGROUND AND RELATED WORK
Since its launch in 2009, WhatsApp has become a globally popular
instant messaging platform. A large body of research has examined
people’s use of WhatsApp in a wide range of settings, including
everyday activities [
85
], interactions with family and friends [
84
],
how WhatsApp extends and enacts physical relationships [
56
], and
comparisons of WhatsApp to other messaging modalities, such as
SMS [
24
]. Studies have also examined how WhatsApp impacts, for
example, stress [
102
], distraction [
2
], and privacy [
17
,
49
]. Beyond
personal use, research has also examined WhatsApp in professional
settings, including health [
53
,
57
,
58
], politics [
21
,
94
], and commu-
nity engagement [63].
In educational contexts, which are closest to our study, recent
work analyzed how WhatsApp might improve student outcomes.
Bouhnik et al. [
15
] explored WhatsApp communications between
teachers and students. Cetinkaya [
20
] suggested that WhatsApp
might improve students’ structured learning, while Barhoumi [
9
]
found that WhatsApp may help students to discover peer-generated
resources, thereby promoting context-free learning.
A recent cluster of studies has also specically examined the use
of WhatsApp in educational settings in non-Western and/or HCI4D
contexts [
50
,
114
]. For example, Willemse [
113
] examined how
WhatsApp discussions may improve the education of undergrad-
uate nursing students in South Africa. Poon et al. [
89
] compared
the utility of SMS and WhatsApp in delivering revision quizzes
to high-school students in Cameroon. Mudliar and Rangaswamy
[
76
] documented how WhatsApp-based interactions may help to
reduce gender gaps that exist in classrooms in India. Nedungadi
et al. [
81
] studied how WhatsApp communication might reduce
teacher and student absenteeism and improve student performance
in rural Indian schools. Most relevant to our work on teacher What-
sApp groups is a recent study by Varanasi et al. [
109
] that found
how teachers in India recongure their work practices around a
teacher-oriented technology intervention. Their study briey high-
lights that teachers use WhatsApp to share resources, but does not
analyze the content and nature of teachers’ WhatsApp-based com-
munications. More work is needed to better understand how teacher
networks are enacted on technology platforms like WhatsApp.
Our study expands the literature on teachers’ WhatsApp use via
a qualitative analysis of group conversations that occurred between
teachers in low-income Indian schools, school administrators, and
sta from education-focused organizations. We now situate our
research within prior work on cooperative work and teacher net-
works.
Technology-Mediated Cooperative Work. The HCI and CSCW
communities have a rich history of examining how technology
mediates cooperative work, with much of the early work in this
space focused on oce-based contexts [
8
,
12
,
39
]. Although school
environments are undoubtedly dierent from oce environments,
research in educational contexts has argued for schools to be seen
as social organizations where work is done [
30
,
36
]. This line of
thinking emphasizes the need to take into account the socio-cultural
complexities produced by such organizations, and to understand
interventions in the context of broader school environments [99].
A variety of theoretical frameworks have been proposed to char-
acterize or evaluate technology-mediated cooperative work (e.g.,
[
72
,
80
,
86
]). In our study, we nd Neale et al.’s [
80
] model for
evaluating activity awareness especially useful for characterizing
the kinds of cooperative work that manifest in teachers’ What-
sApp groups. Activity awareness model helps to systematically
analyze individual activities within complex cooperative networks.
To achieve this, the model outlines ve types of activities that ex-
plain how tightly coupled the work is: lightweight interactions (most
loosely coupled), information sharing, coordination, collaboration,
and cooperation (most tightly coupled). The more of these activities
a group is able to achieve, the stronger their cooperative work.
The rst two layers of the framework refer to loosely-coupled
activities. Light-weight interactions are only loosely tied to the
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
work itself and encompass both casual social interactions and
communication about the work, often providing information and
background that helps to understand the work context and contex-
tualize behavior and group interactions [
80
]. These interactions are
reminiscent of Nardi’s work on “outeractions”: social and informal
communications that occur in formal work settings [
78
]. The next
layer is information sharing, which may occur in one direction (e.g.,
someone shares information with no response/acknowledgment) or
in share-response pairs. Prior work has examined the complexities
associated with information sharing [
38
,
115
], such as research on
common information spaces that looked at how actors represented
and attributed meaning to the information in work spaces [14].
The other three layers refer to more tightly-coupled activities.
Coordination requires group members to coordinate the content of
the work and the process involved in carrying it out [
32
,
37
,
80
].
Collaboration involves group members working toward a common
goal, with individuals often doing separate tasks (that are interde-
pendent) but with shared goals and knowledge [
23
,
77
,
80
]. Finally,
cooperation is the most tightly coupled activity, involving “shared
goals, common plans, shared tasks, and signicant consultation with
others about how to proceed with the work” [80].
We use the activity awareness model as an analytical lens to
examine our research questions. In our ndings, we highlight which
types of activities manifest in the WhatsApp groups we study. We
then discuss in Section 7 the characteristics of the groups that may
(or may not) have facilitated cooperative work. In doing so, we build
on prior research that suggests that groups of teachers collaborate
and co-exist as part of larger professional networks that ultimately
seek to enable cooperative work [46].
Teacher Networks. In addition to exploring how the WhatsApp
groups in our study may enable cooperative work, we also exam-
ine the extent to which these WhatsApp groups constitute teacher
networks. Teacher networks are a concept that, initially, were con-
sidered to be loose and borderless social constructs (e.g., Lieberman
[
65
]). These initial networks covered both formal and informal
aspects of teacher work, similar to group work in other profes-
sional settings [
78
], with the aim of examining teacher connections
through the lens of broader social constructs (e.g. social capital)
[46, 65].
Acknowledging the importance of teacher networks, prior re-
search has explored a range of dierent models of these networks or
communities. One example is knowledge communities [
87
], which
seek to understand how teachers come together physically and
virtually to co-create knowledge. Such communities are based on
Lave and Wenger’s [64] concept of communities of practice, which
has been extended to include virtual communities [
42
,
45
]. Prior
research has also explored how dierent learning communities can
push for self- and peer-based reection within groups, via strategies
such as sharing success stories or listening and responding to oth-
ers’ experiences [
82
,
107
,
117
]. However, most community-based
studies of teacher networks have focused on the formal aspects of
learning; very few studies emphasize the informal interaction that
happen among networks without facilitation (e.g. starooms) [
71
].
Both teacher networks and the activity awareness model share
common roots in activity theory and are extensively used to study
teacher development [
65
]. Teacher networks are useful to examine
broader network characteristics and their interrelationships, and
are synergistic with the activity awareness model’s pragmatic focus
on examining low-level interactions in these networks. Drawing
on this synergy, we use both teacher networks and the activity
awareness model to deeply examine WhatsApp groups at multiple
levels. In particular, we use teachers networks as a theoretical lens to
examine the role of dierent stakeholders in WhatsApp networks
as a whole whereas we use the activity awareness model as an
analytical lens to examine low-level individual interactions between
stakeholders in these networks.
Although a cluster of studies in HCI [
6
,
13
,
19
,
34
], CSCW, and
CSCL [
60
,
116
] has examined the role of technology in teacher
networks, these studies have focused on Western communities in
developed countries. Our study contributes a new perspective to
this literature by examining how both formal and informal teacher
networks are enacted via a novel medium: WhatsApp group conver-
sations, and in a novel context: low-income Indian schools. Specif-
ically, we conducted a qualitative study to answer the following
research questions:
RQ1:
What strategies are used to encourage in-
teraction within WhatsApp-based teacher networks in low-income
Indian schools? and
RQ2:
How are these interactions sustained by
teachers, management, and organizations over a period of time?
3 DESCRIPTION OF WHATSAPP GROUPS
Before describing our study methods, we provide background on
the three types of WhatsApp groups in our study to appropriately
contextualize our ndings. All of the groups already existed at the
time of our study (i.e., we did not create them) and all were set up
to facilitate communication and coordination between groups of
teachers in low-income settings. Two types of groups were created
and managed by sta at education-focused organizations (321 and
Meghshala) while the third was set up and administered by school
management.
321 WhatsApp Groups. 321 [
1
] is an education-focused organiza-
tion in Hyderabad, Mumbai, and Bangalore that aims to improve
teachers’ capacities via a two-year support model that (1) conducts
workshops to teach classroom management and pedagogy, (2) pro-
vides one-on-one teacher coaching on specic topics (e.g., observa-
tion, creating assessments), and (3) organizes events to celebrate
participating teachers (e.g., with certicates).
The WhatsApp groups that 321 administers were set up to com-
plement their in-person workshops and coaching, and teachers who
participated in their workshops were invited to join a group consist-
ing of teachers from their school. The average size of the 321 groups
in our data was 17 members, with an average of 86 messages per
group per month (see Table 1). Groups consisted mostly of Hindi-
and English-speaking teachers (about 90% women) and with one or
two 321 sta. The content posted to the group is highly structured
and curated by 321 sta. A content design team creates customized
media and messages that are posted to the groups by training sta
who have previously interacted with the teachers face-to-face (e.g.,
in workshops). Messages are designed to prompt responses from
teachers by including activities like quizzes, puzzles, and requests
to share content.
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
Meghshala WhatsApp Groups. Meghshala is a Bangalore-based
non-prot organization whose objective is to improve teachers’
capacity by building their pedagogical knowledge. To achieve this,
Meghshala provides a mobile app-based intervention delivered to
teachers via an Android device. The app provides content modules,
developed by Meghshala’s team, that are carefully contextualized
to the state government’s curriculum and pedagogical philosophy
while also incorporating new ideas and practices to build teacher
capacity. In addition to the app, teachers receive support from
Meghshala’s sta via weekly in-person visits and organization-run
WhatsApp groups.
Similar to 321, Meghshala’s WhatsApp groups explicitly aim to
complement their in-person support by improving connectedness
and providing technical support to aid adoption of Meghshala’s
app by teachers. The groups had an average of 48 members (a mix
of men and women), mostly government school teachers from a
range of dierent schools who spoke Kannada, Hindi, Marathi,
and English. The groups averaged 173 messages per group per
month (see Table 1). WhatsApp interactions in these groups are
unstructured and open-ended. Teachers use the group to report
issues or give feedback to Meghshala sta on app usage. To aid
further motivation for teacher interaction, several groups have
Meghshala management as members.
School WhatsApp Groups. The third type of WhatsApp group con-
sists of government teachers and higher management (e.g., Cluster
Resource Ocers and Block Ocers
1
). These groups are adminis-
tered by cluster resource ocers (who rank above principals). The
groups are typically large, with over 75 members who are mostly
teachers (a mix of men and women) from several schools. These
were the most active groups in our dataset, with an average of 728
messages per group per month. These groups provide a platform
for higher management to streamline and better manage school
administration. Cluster and Block Ocers use the groups to share
information with teachers (e.g., announcements and reminders).
They also often send detailed instructions that teachers are required
to follow and resources intended to aid teachers’ work.
Overall, Meghshala and 321 groups focused more on capacity
building of teachers through development of student-centered ped-
agogical techniques and resources that emphasized on learning
through lived experiences [
29
]. School groups, on the other hand,
took a more teacher-centered approach, focusing on providing ev-
eryday support for teaching and managing classrooms [98].
4 METHODOLOGY
To answer our research questions, we conducted an IRB-approved
study in which we collected and analyzed conversations from 26
existing WhatsApp groups (35,567 messages) that took place be-
tween teachers in low-income schools (in Telengana and Karnataka),
school administrators, and support sta from education-focused
organizations.
Collecting WhatsApp Data. Prior work [
109
] suggests that What-
sApp plays an important role in teachers’ work, including school
1
In several states of India, schools are grouped into clusters, and clusters into blocks.
Each block contains several schools.
Group No. Size Duration Messages Messages
type groups (people) (months) per group per month
Min: 16 Min: 3 Min: 114 Min: 28.5
321 13 Max: 18 Max: 7 Max: 713 Max: 112
Avg: 17 Avg: 5.4 Avg: 438 Avg: 86
Min: 16 Min: 2 Min: 208 Min: 21
Meghshala 5 Max: 78 Max: 11 Max: 4435 Max: 443
Avg: 48 Avg: 8.4 Avg: 1691 Avg: 173
Min: 50 Min: 0.5 Min: 40 Min: 80
Schools 8 Max: 100 Max: 7 Max: 9340 Max: 2335
Avg: 75 Avg: 2.3 Avg: 2678 Avg: 728
Table 1: Summary of WhatsApp groups in our data set.
communications (e.g., from principals or managers) and profes-
sional development programs run by external educational organi-
zations. We wanted to examine WhatsApp use by both kinds of
groups and so reached out, via email and WhatsApp, to schools
and organizations that work with teachers in low-income Indian
schools. We engaged interested respondents in discussions where
we explained the goals of our work, methods, data privacy, etc.
Ultimately, the CEOs of two organizations—321 [
1
] and Meghshala
[
74
]—agreed to provide data from teacher-focused WhatsApp groups
run by their organizations (13 groups from 321 and ve from
Meghshala). The logs of group conversations were exported and
provided to us by organization managers. In addition to data from
organizations’ WhatsApp groups, we received permission to col-
lect WhatsApp data from school groups from a Block Ocer in
Karnataka. These groups consisted of school management (e.g., prin-
cipals) and teachers from dierent clusters of government schools.
The logs of eight WhatsApp groups were provided by two Cluster
Resource Ocers.
Table 1 summarizes the WhatsApp groups in our data set, includ-
ing the number of group members, duration of logs, and number of
messages sent. All of the groups consisted of teachers who taught
grades 1-8. Before we collected any data, we asked each group’s ad-
ministrators to publicize the study on the WhatsApp groups, share
our consent form, and post information about the study’s objective.
They also explicitly provided instructions on how group members
could opt out of the study by requesting that their messages or
posts be removed from the dataset, and emphasized that opting out
would not aect the teacher’s employment status or relationship
with the organizations. Nevertheless, no group members opted out
of the study.
Analyzing WhatsApp Data. Exporting logs of WhatsApp group
conversations for analysis presented several challenges. Since con-
versations consist of text and media (i.e., images, videos), the export
function provides an option to include media les along with text-
based messages. However, this process results in the media les
becoming decontextualized from the conversations, since What-
sApp exports all the media les separately in a .zip folder and
replaces them in the conversations with a <media omitted> tag (see
Fig. 1.A). To circumvent this issue, we asked participants to use a
screen recording app to record their screen as they scrolled through
the group conversations. We then analyzed the video recording
of the media associated with a conversation in parallel to the
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
Figure 1: (A, B) Excerpt of WhatsApp log and video content analysis showing how we analyzed WhatsApp data.
text-based conversation log, which enabled us to view the media in
the context of the conversations (see Figure 1.B).
We began our analysis of text conversation and video record-
ings of media by cleaning the text les and converting them to
standardized UTF-8 encoding to accommodate text written in lo-
cal languages (Kannada, Hindi, and Marathi) and emojis. We then
used inductive thematic analysis to analyze our data [
16
] with
the coding conducted by the rst author. We began by reading
through the WhatsApp logs (and scrolling through the correspond-
ing video recordings). We then conducted multiple rounds of open
coding. We avoided using any preconceived codes and instead al-
lowed the codes and categories to emerge from the data. Our unit
of analysis was a single message sent by a participant. If partici-
pants broke messages into multiple lines, they were treated as one
message. Credibility of our analysis was established by prolonged
engagement with the data and multiple coding iterations, with
peer-debrieng sessions with the research team after each coding
pass [
27
]. Our analysis resulted in 53 codes (e.g., sharing highlights,
policing norms, peer interactions), which were organized into dif-
ferent themes. Finally, following Pierce’s pragmatic philosophy, we
used an abductive approach [
108
] to further map, categorize, and
structure themes using the activity awareness model. Our anal-
ysis yielded nal nine themes, namely professional interactions
(23%), online-oine bridge (14%), contextualization, (11%), bottom-
up support (7%), community care (2%), top-down support (18%),
professional wellbeing (13%), capacity improvement (9%), and secu-
rity (2%). Appendix A provides our complete codebook, along with
the prevalence of each theme and code.
Qualitative Interviews. To corroborate our analysis of WhatsApp
logs, we conducted 12 post-analysis semi-structured interviews
with school management and sta who participated in the three
groups (4 each from Meghshala, 321, and school groups). Our moti-
vation for conducting these interviews was to triangulate ndings
from our analysis of WhatsApp logs, obtain additional context for
the group conversations, and understand oine activities that may
have motivated actions on the digital platform. To recruit inter-
viewees, we reached out to group administrators via WhatsApp.
Administrators sent out a message inviting the group members
to participate in the interview. Those who responded were inter-
viewed for roughly 30 minutes in their local language.
Our 12 participants (eight women) included nine teachers and
three principals. The average age of the participants was 35 years
(min=23, max=47, SD=6.3). Participants had an average of 9.8 years
of experience (min=2, max=30, SD=7.3). Interviews were conducted
in-person or over WhatsApp audio calls. Our interview protocol
sought an understanding of (1) how participants currently use
WhatsApp in their work, (2) reasons for interacting (or not) in the
WhatsApp groups, (3) challenges and issues they experienced when
using WhatsApp, and (4) how WhatsApp enabled (or not) com-
munication and collaboration with peer teachers, sta, or higher
management. After each interview, we revised our questions to add
new probes, stopping when we reached saturation in our interview
data.
Interviews were audio-recorded (with consent), translated into
English, transcribed, and analyzed using MAXQDA. We used the-
matic analysis to analyze our interview data [
16
]. We performed
multiple passes over the data, allowing codes to emerge freely. After
each round of coding, we used peer-debrieng [
27
] with two co-
authors to iterate on the codes and improve consistency. Our nal
codebook consisted of 38 codes (e.g., local contextualization, sharing
resources, celebrating promotions). These codes were then clus-
tered into nine themes (e,g., peer collaboration, bottom-up support,
and misinformation) and collated with themes from our analysis
of WhatsApp logs. Throughout our ndings, we deliberately inter-
weave analyses of WhatsApp logs with analyses from interviews,
using the interview data to provide additional context and insights
to the log data, rather than presenting the interview data separately.
Ethical Considerations. We received IRB approval for our study
as well as approval from 321’s and Meghshala’s management and
appropriate school Block Ocers. We also took several steps to
safeguard the privacy and interests of the teacher members of
the WhatsApp groups. We asked group administrators (Cluster
Resource Ocers and organization managers) to publicize our study
on their groups and inform all group members about our research.
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
As part of this process, we provided teachers with the ability to
opt out of the study (i.e., remove their messages from our data set).
However, we did not receive any such requests from teachers, which
may be because all of the messages sent were already accessible
to many group members and visible to their school’s management
and organization personnel (i.e., they were comfortable sharing this
information with a large group and our study did not change that).
In reporting our ndings, we use pseudonyms for participants
and anonymize quotes and messages. We have replaced all poten-
tially identifying information, including names, phone numbers,
and addresses. However, we chose to keep the day and time at
which messages were sent unaltered, since this provides context
for understanding the interactions.
5 STRATEGIES TO ENCOURAGE
INTERACTION WITHIN WHATSAPP
GROUPS
We organize our ndings around our two research questions. To
answer the rst question, we (1) discuss a set of creative structures
that groups used to engage teachers and promote the development
of professional skills, and (2) show how the groups encouraged
teachers by actively recognizing and celebrating teachers for their
eorts. As discussed in Section 2, we use Neale et al.’s [
80
]activity
awareness model as an analytical lens to link our ndings with
theory on technology-mediated cooperative work.
5.1 Creative Structures to Encourage
Interaction
Our analysis reveals a range of creative strategies that organization
sta and school management employed to promote interaction in
the WhatsApp groups. Organization-run groups used WhatsApp’s
built-in features in innovative ways to encourage teachers to post
and respond to messages. One strategy that we saw in 321 groups
was for sta to instruct teachers to interact with activities using
only emojis. For example, Rubina, a 321 sta member, posted:
Feb 18, 10:50 PM. 321 sta:
Give us a if you’re
excited to receive your certicates. Are you wondering
- how is a PARTICIPATION dierent from a COMPLE-
TION? Or how is a RECOGNITION dierent from an
EXCELLENCE? Stay tuned, we’ll share more info about
certicates in the coming weeks. DM us any questions.
Waiting for your
In response to this message, seven teachers posted a emoji. As
Rubina described in a followup interview, emoji-only interaction
enabled shy teachers, who may not otherwise feel comfortable,
to take a risk and post in the group. It also enabled participation
from those teachers who were hesitant to communicate in English
because they feared that they might make mistakes and embarrass
themselves in front of others.
Emoji-only interactions extended beyond single questions to
multiple-choice quizzes that asked teachers to post answers to
several questions (e.g., matching classroom management techniques
with appropriate classroom resources) using dierent combinations
of emojis (see Fig. 2.A). Some teachers struggled to locate emojis
due to limited WhatsApp know-how. They shared their answers
by drawing the relevant emojis on a piece of paper and taking a
photo of it that was shared with the group. To ensure that correct
answers were not posted before most teachers had a chance to think
about the questions, organization sta members asked teachers to
refrain from posting answers until they sent a second message
requesting responses. WhatsApp features such as the delete for all
functionality also allowed teachers to retract their answer from
everyone’s phones when they wanted another chance to answer
the questions.
We recognize these emoji-only activities as lightweight inter-
actions [
80
] involving casual and fun communications within the
group. These ndings also connect to prior work on teachers’ coop-
erative work (e.g., Dunlop et al. [
34
]) which showed that technology-
mediated communication with complementary and contextual inter-
actions (especially risk taking) can encourage fruitful participation
and communication at a teacher’s own pace.
Along these lines, another common strategy sta used to encour-
age teacher interaction was to post a message or announcement to
the group in English followed by an audio recording of the same
message in spoken Hindi, with the goal of reducing teachers’ hes-
itations to participate if they lacked condence communicating
in English. This allowed teachers to freely express themselves by
typing or recording replies in their local language.
Another creative use of WhatsApp features that we observed
involved using the built-in image editor to interact with content in
innovative ways. For example, a 321 sta member shared a picture
of a wordnder puzzle in their group and asked teachers to use the
built-in image editor to draw their answers over the picture and
reshare it with the group (see Fig. 2.B). WhatsApp’s image editing
tools were also useful for allowing teachers to provide feedback on
Meghshala’s app content. For instance, we saw occasions where
teachers in the Meghshala group used the image-editing feature to
notify Meghshala’s sta about incorrect mathematical notation in
Kannada (the local language). Teachers drew boxes around inaccu-
rate content and shared it with Meghshala sta in the group.
Peer-based Activities. Another set of structures that organizations
created aimed to promote interaction via peer-based activities. One
type of activity encouraged teachers to tag peers (using WhatsApp’s
function) and engage in an activity with them oine that was then
shared with the group. As one 321 sta member posted:
Sep 29, 1:10 PM. 321 sta:
Hi teachers, Welcome
back! We had a refreshing holiday . We are all
set to share more winning solutions and come visit your
classrooms once again. If you are as excited as we
are,
∗
tag a teacher who you saw doing something
interesting in her/his classroom
∗
We look forward to
hear from you!
Among the six responses that teachers posted, one read:
Sep 09, 2:02 PM. Teacher:
These days are very im-
portant because its revision time. My friend Yasmeen is
making her revision time very interesting for students
using the practice class chant.
However, these planned initiatives to include teachers did not
always work. There were a few instances where teachers felt de-
jected that they had not been tagged by peers. For instance, when
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
Figure 2: (A) A teacher uses emojis to respond to a quiz; (B)Teachers in Meghshala & 321 groups using WhatsApp’s built-in
drawing tool; (C) A teacher signs her name when using a peer’s phone
no member of the group tagged her, a teacher posted, “There is no
one to take my name ”. In such situations, 321 sta tried to tag
these teachers and provide positive encouragement. In addition, 321
sta often used tagging to increase engagement with questions that
received few responses, such as by nudging teachers who did re-
spond to tag ve more teachers. Content designers at 321 described
that the explicit action of digitally tagging other teachers encour-
aged action-oriented peer reection, similar to in-class peer-based
activities that they conducted during their training workshops.
Finally, not all teachers owned or had full-time access to smart-
phones. This is in line with prior HCI4D literature that extensively
documents shared smartphone use in the Global South [
3
,
18
]. To
overcome this barrier, 321 sta encouraged teachers who did not
possess a smartphone (and therefore WhatsApp) to participate by
borrowing a peer’s smartphone and using it to participate in the
group. To tell the dierence between multiple teachers using the
same device, teachers adopted the practice of signing their name
at the end of the message to indicate the sender (see Fig. 2.C). We
found four teachers that utilized this practice. Using the activity
awareness model, we see how these peer-based activities constitute
coordination, particularly activities such as sharing a smartphone
with others to complete activities [
80
]. This suggests that the 321
WhatsApp groups engaged in relatively tightly-coupled work [
59
].
5.2 Celebrating and Recognizing Teachers
Recognizing Teachers’ Professionalism. We found that these What-
sApp groups actively celebrated teachers’ achievements and rec-
ognized them for incorporating better pedagogical techniques to
encourage interaction. The WhatsApp groups provided a space for
promoting and acknowledging teaching as an important profes-
sion and emphasizing teachers as professionals. In 321 groups, sta
incorporated specic keywords in their messages to achieve this,
such as modern professionals and nation builders. When asked to
reect on their own role, teachers often used the same terminology
in their responses:
Nov 10, 4:24 PM. Teacher:
Hello! I think being a Mod-
ern Professional and A Nation builder I use all the skills
in one or other way. I think all students are unique
because when we used all these skills in our teaching
they collaborate with each other they openly accept the
challenges, practice their knowledge with kindness and
build a new creativity.
Other messages sought to align teaching with other important
professions, such as doctors. For example, a 321 sta member posted,
“Can you share with us what tools you use as modern professionals? For
example, a doctor uses thermometer, teachers use textbooks.” Another
strategy we saw related teachers to famous leaders, such as Abdul
Kalam (a scientist and former Indian president), to highlight positive
leadership qualities that teachers should strive for.
Teachers were also recognized for the important role they play
in teaching values to students. Organization sta and higher man-
agement encouraged teachers to adopt activities that made students
happy and share their experiences of teaching values with the group.
Teachers responses to these requests suggest that they too saw the
importance of their role in shaping the happiness and success of
their students. For example, one teacher likened herself to a super-
hero, posting, “I wear an invisible superhero crown everyday that
spreads joy and happiness while teaching.” Another teacher
reected on her responsibility to teach students to be good citizens:
Nov 11, 10:47 PM. Teacher:
I teach them how to
share their feelings with others which will help children
not only discover personal success, but also contribute
to the betterment of society by improving them. My
lessons also let students practice kindness which is re-
ally essential because i had learned a nation is known
by the character of its citizens.
Celebrating and Praising Teachers’ Eorts. Beyond recognizing
their professionalism, the WhatsApp groups also provided im-
portant spaces where teachers could be explicitly celebrated and
praised for their hard work. 321 sta, in particular, took care to
emphasize and praise teachers’ eorts to integrate new techniques
in their teaching, rather than focusing on success. For example, a
sta member posted, “Thank you all the teachers who have already
made an attempt to answer how they are integrating these new tech-
niques in their classroom .” Sta frequently posted encouraging
messages that made heavy use of emojis and WhatsApp’s built-in
GIF feature to explicitly acknowledge teachers for participating in
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
group activities, trying to answer questions, or sharing evidence
of progress integrating new pedagogical practices in their class-
rooms (e.g., sharing pictures that demonstrate student-centered
classroom management strategies using mnemonics learned from
in-person training workshops). In addition, 321 sta created custom
celebratory structures, such as special claps ( ), that they
‘awarded’ to teachers who engaged with their puzzles, quizzes, or
shared evidence of their progress, for example, by posting a photo
of an in-class activity.
Our interview ndings suggest that these WhatsApp-based struc-
tures complemented celebrations in the physical world, such as
teachers receiving printed certicates, or events that celebrated
teachers in their community:
Nov 29, 4:33 PM. 321 sta
:Dear Teachers! .. .all of
you have *worked extremely hard and showed exemplar
growth* in your classrooms. It is *time to celebrate!
Double the eorts, double the celebrations!* Let us
all *meet tomorrow from 12:45 - 1:45 pm* to celebrate
yourself and each other for all the amazing work you
have put into your classrooms, schools and students!
*Looking forward to seeing you all*
In Meghshala groups, organization sta frequently praised teach-
ers who shared photos of themselves using Meghshala’s app in
class. This applause typically highlighted how teachers’ actions
positively impacted students. For example, in response to a teacher
who shared a photo of her students in class, a school administrator
posted, “Also please all observe the smile and condence on the face
of the children ”. The sta also added their founder and CEO to
a few WhatsApp groups to further motivate teachers and improve
morale. The addition of the CEO led to many enthusiastic responses
from teachers, such as “Welcome respected Jaya maam”. The CEO
also posted replies to teachers who shared their experiences adopt-
ing the app. For example, when Gita, a fth grade teacher, shared
a photo of herself teaching Math with the app, the CEO replied,
“Fabulous to see the kids so interested! I want to be like you in class,
Madam!”
We also discovered instances where, without prompting, teachers
shared achievements with the group that were unrelated to the
group’s ocial purpose. Teachers posted messages highlighting
their own and their students achievements, inside and outside the
school. For example, one teacher in a school group posted a message
celebrating a student who was selected for a prestigious interstate
music competition. Another teacher shared a picture of herself
receiving a teaching award. Such milestone events usually received
recognition in the group from teachers and higher management.
Prior work has shown that these kinds of intrinsic rewards,
such as positive feedback, may contribute to teacher motivation
and the development of self-ecacy [
96
,
97
]. Since these personal
and celebratory messages do not necessarily constitute work, but
rather motivate teachers to do the work better, we see them as
lightweight interactions [
80
]. Moreover, we see the smileys awarded
by 321 and Meghshala as mechanisms that reward the interactions
and types of cooperative work they valued in the group, which
is synonymous with prior work on the types of rewards given in
online communities like Wikipedia [
61
]. For example, Kriplean
et al. studied barnstars, tokens given to contributors for valued
cooperative work [
61
]. Their study shows how people who received
these rewards cherished them. Similarly, teachers we interviewed
described how they cherished the smileys awarded to them and
saved screenshots of them.
6 CONTENT CURATION, SHARING, AND
MODERATION VIA THE GROUPS
We turn now to RQ2: How are interactions sustained by teachers,
management, and organizations over time? We (1) discuss how
WhatsApp groups were used to share professional resources. We
also (2) describe how the groups were used to circulate misinfor-
mation, malicious spam, and polarizing information, before (3) cov-
ering content moderation strategies used by group admins to keep
conversations focused on education.
6.1 Sharing Professional Resources
Our data shows that all the WhatsApp groups were heavily utilized
to share professional resources, often in interesting or innovative
ways. In school WhatsApp groups, higher management (e.g., prin-
cipals, cluster resource ocers, block ocers) used WhatsApp to
make bureaucratic procedures more ecient, such as distributing
school circulars or requesting information from teachers via What-
sApp instead of paper. In most cases, the higher management would
take a picture of a paper-based circular and share the photograph.
Interestingly, when teachers responded to these requests, they too
would write the information on paper and share a photograph of it
with the WhatsApp group. This is because many of these circulars
required teachers to capture lengthy and complex details, such as
students’ progress in dierent subjects, from multiple sources, mak-
ing it easier for them to write the required information on a paper
than typing it in WhatsApp. For example, we saw how teachers
who were asked to share their lesson plans for a periodic review
wrote the requested information in the required format on A4 pa-
per, photographed it, and sent the photo to the higher management
via the group. This nding builds on prior HCI literature exam-
ining the ow of information across digital and physical spaces
[22, 28, 40, 47].
While WhatsApp made it easy for administrators and teach-
ers to share information, it added more work to teachers. Unlike
paper-based circulars, which needed to be physically sent to each
teacher for a signature during school hours, using WhatsApp for
administrative work enabled higher management to reach out to
teachers outside of school hours and request that they do work.
Moreover, when sending their requests on these groups, manage-
ment frequently included the words URGENT or VERY URGENT
to indicate the urgency of the messages, often with accompanying
text that explicitly called out their intended audience. For example,
a cluster resource ocer sent:
Jan 09, 5:04 PM. Cluster resource ocer:
All, send
urgently before 12 pm by WhatsApp your [students’]
information . . . Raman sir, Prashant sir please do it.
In addition to responding to bureaucratic requests from higher
management, several teachers created and shared online resources
like YouTube videos with step-by-step instructions to help their less
tech-savvy peers complete bureaucratic requirements. For example,
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
one teacher made a video showing how to upload students’ scores
to a government website and promote them to next grade.
Sharing Event Highlights to Broaden Inclusivity. Another preva-
lent practice in our data was for participants who attended in-
person teacher-development events and workshops to share high-
lights and summaries of these events, usually with photos or videos,
so that teachers who were unable to attend could benet from the
information. For example, a teacher in 321 group, Sahana, shared
pictures of workshop activities that involved creating classroom ma-
terials, preceded by a textual summary of the activities in question.
We saw similar practices in Meghshala groups, where organiza-
tion sta often shared important snippets of Meghshala training
workshops via videos or photos. Sta members told us how this
allowed them to reach a broad audience of teachers who were not
necessarily part of the in-person training workshop.
According to a cluster resource ocer, Suresh, sharing meeting
summaries via WhatsApp promoted transparency regarding the
activities and provided a written record of decisions made. For
example, another cluster resource ocer in a school group shared:
March 05, 5:21 PM. Cluster resource ocer:
High-
lights from today’s meeting:
∗
Examination date has
been xed and can not be changed . . .
∗
Now linking
of state scholarship with Adhaar is required
∗
We will
purchase books from the Department with the cumula-
tive funds.
∗
Meeting of selected school teachers with
Meghshala on 8th at [location].
Sharing Teaching Resources. The WhatsApp groups were also
used to share classroom management and pedagogical techniques
with teachers. Activities and content shared by organization sta
on Meghshala and 321 groups was often tailored to the teachers’
specic contexts. In a Meghshala group we saw sta post ideas for
summer projects that teachers could assign to students during the
holidays. In a 321 group, a sta member posted YouTube links of
popular children’s songs sorted into a “Bookmark of Calmers and
Energizers” to help teachers better manage their classroom.
Teachers also shared their own content and resources with peers.
For instance, teachers in Meghshala groups shared strategies for
troubleshooting technological challenges encountered when us-
ing Meghshala’s app. As an example, one teacher reached out to
a Meghshala WhatsApp group sharing the struggles she encoun-
tered trying to show Meghshala content via a tablet with a small
screen to her large number of students, and asking for alternative
suggestions. In response, another teacher shared detailed instruc-
tions and resources for how to repurpose an old LCD TV monitor
as a display device by using Meracast (similar to Chromecast) to
cast Meghshala’s app content onto the LCD TV screen. As another
example, a few teachers in Meghshala shared screenshots of their
experiments using a new augmented reality app related to their
syllabus on planets (see Fig. 3.A). They also shared a web-based
resource on how to install and use the app in the classroom. Other
teachers followed their guidelines and in turn shared their results
with the group.
All of the activities described here (sharing event highlights,
sharing teaching resources, etc.) constitute information sharing
within the activity awareness model [
80
], and we see cases in which
this sharing is both unidirectional (e.g., sending a circular without
a response) and in the form of share/response pairs (e.g., requesting
a response from teachers).
6.2 Sharing of Malicious Forwards,
Misinformation, and Spam
In addition to professional content and resources, teachers also
posted non-work-related information. A large proportion of these
posts (34%) were spam or malicious messages forwarded from other
conversations. Such messages were relatively easy to identify from
their long length, formal language, repetition across groups, and
availability via online searches.
Malicious Forwards. One type of forward common in our data
consisted of malicious messages that contained suspicious links.
These messages drew readers’ attention by providing unrealistic
oers that enticed people to click on the link, which was often
designed to closely resemble a legitimate link (e.g., amazon.com-
deal.in). Several such messages in our data were explicitly worded
to appear as government education schemes, thereby specically
targeting teachers. For example, one such forward that advertised
a non-existent educational scheme read:
July 06, 05:30 PM. Teacher:
PM Modi is providing
free mini laptop to every student .. . Regiser now to get
free laptop. Visit Here
∗
http://pmyojna.com
∗
Share
with your friends and groups so they can also apply for
∗Free Laptop∗
Clicking on the link directs users to malicious websites that
utilize a variety of harmful practices, such as prompting installation
of malware (see Fig. 3.B) or tricking users into installing spam
applications that earn the spammer referral income. Other links
lead to forms that prompt users to enter personal information, such
as date of birth, phone number, Aadhaar (ID) number, email address,
and physical address. These websites often then sell such data to
third parties.
Misinformation. Another common type of forward we saw was
fake information on a variety of popular topics, including science,
technology, and current aairs. These messages often promoted a
specic message or agenda. For example, one forward circulated
on Meghshala’s groups was a link to a YouTube video that claimed
“mobile phone radiation lead to brain cancer”. The authenticity of
such messages was often disputed by other teachers in the group.
For example, a teacher in a Meghshala group shared a forward with
fake photos that claimed to show pictures of the moon taken by
Chandrayaan2, India’s recent lunar mission. Other teachers quickly
responded that the images were fake and Chandrayaan2 had not
yet released any pictures. Some teachers also proactively shared
warnings about other forwards that had not yet been posted in the
group. For example, one teacher warned the group about a forward
that they had seen elsewhere that they believed to be promoting
terrorism (although was in fact also misinformation):
April 11, 03:26 PM. Teacher:
There is a WhatsApp
group called *Interschools*. If invited, don’t join. It be-
longs to Daesh (ISIS). If you join you will not be able to
exit from it. Be vigilant. My dear colleague send it to
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
Figure 3: (A) A teacher shares use of an augmented-reality app; (B) Content from a malicious link in a Meghshala group; (C) A
block ocer moderates content on a WhatsApp group.
your relatives and children on WhatsApp so they will
also be careful.
Similar misinformation sharing practices are also generalizable
across other forms of cooperative work, such as nance [
112
], law,
governance, and similar social institutions [
35
]. Moreover, Resende
et al.’s [
93
] work shows that misinformation has the tendency to
be shared more frequently on informal tools like WhatsApp than
other platforms.
Religious and Politically Motivated Forwards. Another category
of controversial forwards we saw were religiously or politically
motivated. For example, some teachers in school WhatsApp groups
forwarded political posts that promoted propaganda regarding spe-
cic political activities. On a few occasions, such forwards combined
themes of politics and religion to make their point. For instance, a
teacher in a Meghshala group shared a forward that gave an unveri-
ed account of Narendra Modi, India’s prime minister, in a religious
context, thereby indirectly promoting both the religion and Modi.
Administrators were usually quick to respond and reprimand peo-
ple who forwarded such messages. We discuss strategies for group
management and content moderation in more detail next.
6.3 Strategies for Group Moderation
We observed a range of group moderation and policing strategies
in our data. Higher management (e.g., block ocers) were usually
quick to respond and point out when a message violated the group’s
norms or purpose (e.g., was not related to education), often calling
out the oending teacher by name and warning them not to post
such messages. For example, when a teacher, Prerna, forwarded an
irrelevant message in the group, a block ocer immediately posted,
“Prerna, not needed such messages in this group...Pl[ease] take care.”
(see Fig. 3.C). For oensive religious or political content, higher
management and organization sta adopted a more aggressive tone,
making it clear that teachers who posted such messages will be
removed from the group. In some cases, oensive messages caused
teachers to leave the group before administrators were able to
moderate the conversation. For example, the following conversation
took place after a teacher posted a polarizing religious message
that hurt other teachers, some of whom left the group:
6/24/18, 4:41 PM. Teacher-1
:Manika madam, you
may call your religion a treasure, we do not have any
issue. But, please do not refer other religion as trash. In
this group there are teachers from all religions. Please
think before sending the messages. A lot of people get
hurt.
6/24/18, 5:30 PM. Teacher-2]
:I no longer feel good
about being part of such group. Sorry teachers, I am
leaving from this group. [Left the group; two other
teachers also left.]
6/24/18, 6:29 PM. Cluster resource ocer
:Yes such
msgs will not be sheared [sic] to any group. It is wrong
We have given many warnings for sending such
messages in the group. Manika maam unknowingly sent
this msg and maximum teachers in our cluster know
that Manika maam always respect others feelings..so
plz excuse her this time.
Unlike other instant messaging and social media services (e.g.,
Facebook), WhatsApp’s underlying encryption makes it dicult
to automate content moderation and ltering [
5
], making social
moderation strategies more eective. Beyond moderating interac-
tions around oensive content, administrators also discouraged
teachers from using the group for personal chatting. A 321 sta
member posted the following after several teachers used the group
for personal messages:
2/14/19, 6:00 PM. 321 sta
:Teachers, Hope you had
a good day. We will not use this WhatsApp group for
personal chat. If you want to use it for personal chat
you can make a separate group. Why? Because here we
are going to post messages related to classroom, learn-
ing, and education and conduct discussions around it.
Otherwise, we will miss them.
We also noted an interesting dierence in the moderation mes-
sages posted by school block ocers, which typically called out
the oending teacher by name, and those posted by organization
sta, which usually targeted all teachers in the group. Interviews
with sta revealed that they adopted this strategy to avoid causing
teachers personal embarrassment.
Apart from content moderation in the groups, teachers in school
groups also received instructions from higher management on ap-
propriate teacher behavior in the physical world. Cluster resource
ocers typically sent several messages per month reinforcing rules
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
that teachers had to follow in their classroom. For example, the
following message was sent to a school group in anticipation of an
upcoming surprise inspection by government ocers:
March 13, 1:39 PM. Block ocer:
*Greetings teach-
ers* There will be a study of classes across the state
again from March 18, 2019 .. . observers visited a few
schools and made the classroom observation manda-
tory...So it is very necessary for your classes to be in this
way. 1) The wall slate should be written by the child’s
name and assigned . . .2) The activities of the children
should be recorded and signed and dated. 3) Appliances,
Puppet Screen, and Puppet must be compulsory. 4) The
children should take the card and sit in the prescribed
learning boards. . .
We see the various moderation strategies used in the groups as
information sharing activities [
80
]. In general, moderation messages
made clear the possibility of negative consequences and/or teachers’
expected behavior in the groups. When teachers responded, they
usually simply stated that they would be more careful in the future.
We did not see any instances where teachers tried to refute or justify
their messages. These ndings are similar to Hun’s [
51
] work on
moderating online health forums, which also showed how such
‘template responses’ sent by higher management may negatively
impact morale, with participants who are being moderated feeling
discouraged from participating.
7 DISCUSSION
Having presented ndings around our research questions, we now
synthesize key takeaways for the HCI community by discussing
how teacher networks are enacted via WhatsApp and how dierent
WhatsApp groups enable cooperative work on these online teacher
networks in low-resource settings. We also highlight fruitful areas
for future research, including analyzing the impact of WhatsApp
on teachers’ wellbeing, and exploring the role of professional What-
sApp groups in spreading misinformation.
As discussed in Section 2, research in Western contexts has ex-
amined the role of technology in teacher networks [
6
,
13
,
19
,
34
,
60
,
116
], suggesting that these networks provide fruitful spaces for un-
derstanding teachers’ work practices and professional interactions.
Our study expands this research to the Global South by examining
how teacher networks are enacted via dierent types of teacher-
centered WhatsApp groups in India, contributing a high level per-
spective that is particularly important in light of the rapid adoption
of WhatsApp-based communication in schools [
76
,
109
,
110
], higher
education [
83
,
92
,
113
], and other professional domains (e.g., health
workers [
44
,
52
,
57
]) across HCI4D contexts. At the same time, the
activity awareness model [
80
] provides complementary low-level
insights on dierent types of cooperative work that these teacher
networks exhibit, helping us to understand their value addition in
teacher professionalization. Drawing on these frameworks, we now
discuss the dierent types of cooperative work we found on these
teacher-focused WhatsApp networks.
321 groups: formal teacher networks with tightly coupled cooper-
ative work. 321 groups exhibited characteristics of formal teacher
networks, since they were highly structured and curated by or-
ganization sta. Several of the structures used, such as tagging,
encouraged peer-based activities and action-oriented peer reec-
tion. In doing so, these groups became coaching networks in digital
spaces [
95
]. Coaching networks aim to support teachers by enhanc-
ing their teaching and managerial skills via systematic reection
[
111
]. Such coaching networks are structured and nourished by
a facilitator or a coach (321 sta). We saw how 321’s content de-
signers maintained a uniform coaching structure across all their
groups, and how facilitators encouraged peer-based interactions via
activities. Prior literature has suggested that such collaborative re-
ections between peers in groups can augment professional growth
[
10
]. Facilitators led engagement in these coaching networks, re-
sulting in groups engaging in tightly coupled activities, namely
lightweight interactions, information sharing, and coordination
within activity awareness framework’s ve layers. For instance,
peer-based activities that teachers participated in contained inter-
actions that reected a specic plan to answer questions on teacher
professionalization.
School Groups as knowledge communities with loosely coupled co-
operative work. By contrast, the conversations in the school groups
originated from teachers. These conversations represent a more
informal network in which teachers were comfortable sharing their
own and their students’ achievements via messages, photos, and
videos. These acts of sharing everyday experiences suggest these
networks are operating as knowledge communities [
87
]. Knowledge
communities are spaces where educators can share ‘amateurish-
ness’ experiences (or ‘legitimate tellings’) [
75
] and react to each
other’s experiences openly and honestly. This was also reective
in the cooperative work in the group that was primarily limited
to loosely coupled activities, such as information sharing activi-
ties around teachers’ work. In addition to the informal nature of
these knowledge communities, school groups also exhibited for-
mal structures. The hierarchical and bureaucratic ways in which
higher management used these groups to facilitate certain types
of exchanges (e.g. sharing teaching instructions, sending circulars)
constitute a formal network in which management prescribed au-
thorized versions of teacher development that made explicit what
behaviors are "right" and "wrong" [25].
Meghshala groups as informal networks with loosely coupled co-
operative work. Lastly, Meghshala groups’ absence of an imposed
structure enabled these groups to operate as informal learning net-
works [
70
]. We adopt Livingstone’s [
66
] perspective of informal
learning as exchanges that result in understanding, knowledge, or
skill without externally-imposed structures. The absence of such
structures promotes open discussion and easy sharing of knowledge.
In our data, teachers were comfortable providing feedback on incor-
rect Meghshala content. They also freely shared troubleshooting
tips to help their peers develop technical skills (e.g. how to cast con-
tent on a TV screen). We saw how peers in these informal networks
customize their interactions to the context of their community,
thereby increasing active participation [
48
]. Of course, the unstruc-
tured nature of the networks also means that not all exchanges are
relevant to teaching [70], with teachers forwarding messages that
were often unrelated to their work. In these cases, group moder-
ation helps keep the network focused on teachers’ instructional
practices [
71
]. In contrast to 321 groups, Meghshala and school
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
groups only exhibited two layers of interactions: lightweight inter-
actions and information sharing, suggesting these groups are more
loosely coupled. For instance, school networks engaged in infor-
mation sharing activities around teachers’ work, while Meghshala
networks were prominent for their lightweight interactions around
encouragement.
We hypothesize that dierences in the groups’ network struc-
tures may have contributed to the dierent types of activities we
saw. 321 groups were typically small (avg 17 users) with teachers
from a single school. This ensured that teachers in the groups pos-
sessed shared background and contextual understanding, which
may have enabled them to carry out activities that involved coor-
dination [
26
]. 321 groups were also formal networks, with highly
structured and curated interactions (e.g., quizzes) that provided
professionalization opportunities to teachers. However, this also
meant that teachers’ participation was largely dependent on active
guidance from the organization sta.
By contrast, both Meghshala and school groups were larger (avg
48 and 75 users, respectively) with teachers from several schools.
The groups were also more informal networks, without the struc-
tured interactions prevalent in 321 groups. Although these charac-
teristics may have made it more dicult to achieve tightly-coupled
work, they do enable professional interactions that are currently
not possible in the physical world. For example, these groups pro-
vide opportunities for teachers to communicate with peers from
dierent schools, something feasible only via WhatsApp. Moreover,
we saw how the informal structure of these groups (in contrast to
321’s highly structured groups) allowed teachers to freely share
information and teaching resources with each other, contributing to
loosely-coupled interactions that provided short-term, in-context
solutions to professional issues experienced by teachers.
These insights are relevant for HCI researchers interested in
creating future WhatsApp-based interventions that promote co-
operative work, for both teachers and workers in other domains
(e.g., health [
52
]), by suggesting specic group characteristics (size,
composition) and types of activities (structured) that may lead to
more tightly-coupled vs. loosely-coupled work. At the same time,
none of the groups in our study showed evidence of tightly-knit
collaboration or cooperation [
80
]. One possible reason could be
that WhatsApp group messages are displayed as a single long list,
rather than, say, threaded forums with searchable topics, categories,
etc. This makes it challenging for users to go back to past messages
or separate dierent threads of a conversation (without needing a
separate WhatsApp group). This communication style may hinder
more complex activities like collaboration and coordination. An
interesting area of future work is to study how tightly-coupled
layers of cooperative work may (or may not) be achieved via What-
sApp.
Analyzing how WhatsApp use impacts teachers’ professional well-
being. Our analysis provides preliminary evidence for how What-
sApp groups could have both positive and negative eects on teach-
ers’ wellbeing. Dodge et al. [
31
] dene wellbeing as an equilibrium
between challenges that aect a person and resources that help
the person to cope with those challenges. For teachers, lack of
resources and constant challenges have been shown to result in ab-
senteeism, burnout, and stress [
11
,
41
,
62
,
100
,
103
,
105
]. However,
how technology contributes to these issues has received limited
attention. A few studies have described technology as a challenge
that creates technostress [
4
,
54
]. Relevant to our ndings, Sherno
et al. [
101
] and Skalvik & Skalvik [
104
] showed that excessive work-
load created by higher management results in increased emotional
stress [
43
]. In our data, higher management’s use of WhatsApp
for numerous bureaucratic activities and administrative policing,
as well as their control over setting priorities and deadlines for
teachers, increased teachers’ workload and stress and negatively
impacted their wellbeing.
However, we also see ways in which WhatsApp groups might
improve teacher wellbeing. For example, Meghshala praised teach-
ers who shared their attempts to integrate technology into their
classrooms. Similarly, 321 created structures (e.g., special claps)
to motivate and celebrate teachers, as well as praise teachers for
sharing attempts to implement new pedagogical strategies in their
classrooms. Such positive feedback structures could promote psy-
chological wellbeing among teachers [
91
]. 321 also encouraged
teachers to share even unsuccessful attempts to implement what
they have learnt in their classrooms and praised them for doing
so. Prior work has shown that learning by failure is important
and can provide positive learning benets and advancement for
individuals [
73
], which suggests that this too has the potential to
improve teacher wellbeing. Lastly, teachers getting encouraging
responses when they share messages describing their achievements
may also promote wellbeing, especially since these messages are
teacher-driven.
These insights suggest a need for future work that measures
the impact of WhatsApp use on teacher wellbeing, including by
adapting validated scales for assessing wellbeing (e.g., [
67
,
90
]) to
low-income Indian contexts. We also see potential for WhatsApp-
based interventions that explicitly promote teacher wellbeing in
low-income Indian schools.
Exploring how teachers discover, propagate, and mitigate the spread
of misinformation on WhatsApp. Our ndings showed how teacher
WhatsApp groups were used to circulate misinformation, malicious
spam, and religiously and politically polarizing information, expos-
ing teachers to a range of possible digital harms. These ndings
support and extend recent studies in HCI and HCI4D that show
the prevalence of disinformation, misinformation, and polarizing
content across social media platforms, including Twitter [
106
], Face-
book [68], and WhatsApp [93].
Particularly relevant to our study, Machado [
69
] and Banaji [
7
]
demonstrated a growing concern around the spread of misinforma-
tion on WhatsApp and discussed the need for group members to
increase accountability and reduce the spread of misinformation
via gate-keeping and moderation. Our study shows dierent mod-
eration behaviors, with group administrators (not group members)
primarily responsible for moderating content and reprimanding
oenders. It is possible that, since the groups were created and
administered by organization sta or school higher management,
individual teachers did not feel that it was their responsibility (or
right) to engage in content moderation. Regardless, our data uncov-
ers a need for future research to examine teachers’ mental models of
misinformation. Specically, future studies should aim to examine
ways in which teachers discover, engage, propagate, or mitigate
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
the spread of misinformation. Developing this understanding is es-
pecially important on encrypted platforms, like WhatsApp, where
automated tracing of misinformation is challenging.
8 CONCLUSION
Our study examined WhatsApp group conversations that occurred
between teachers in low-income Indian schools, school adminis-
trators, and sta from education-focused organizations. We an-
alyzed the strategies employed to encourage interaction within
these WhatsApp-based teacher networks and revealed how content
is curated, shared, and moderated. Based on these ndings, we
discussed how teacher networks manifest via WhatsApp groups
and explored how these groups achieve cooperative work. We also
uncovered interesting directions for future work to measure the
impact of WhatsApp use on teachers’ wellbeing and explore the
role of WhatsApp groups in spreading misinformation. Taken to-
gether, our ndings will help HCI researchers and practitioners to
design future interventions that better support cooperative work
and wellbeing for teachers, and workers more broadly, in other
domains across HCI and HCI4D.
ACKNOWLEDGMENTS
This work was funded in part by the Einaudi center for International
Studies. We thank the non-prot organizations and schools for
providing us the data and Thejaswi PC for help in recruitment and
logistics. We are grateful to teachers for participating in our study.
REFERENCES
[1] 321. 2012. http://www.321-foundation.org/
[2]
Annie Dayani Ahad and Syamimi Md Ari Lim. 2014. Convenience or nuisance?:
The ‘WhatsApp’dilemma. Procedia-Social and Behavioral Sciences 155 (2014),
189–196.
[3]
Syed Ishtiaque Ahmed, Md Romael Haque, Jay Chen, and Nicola Dell. 2017. Dig-
ital privacy challenges with shared mobile phone use in bangladesh. Proceedings
of the ACM on Human-Computer Interaction 1, CSCW (2017), 1–20.
[4]
Mohammed Al-Fudail and Harvey Mellar. 2008. Investigating teacher stress
when using technology. Computers & Education 51, 3 (2008), 1103–1110.
[5]
Hunt Allcott and Matthew Gentzkow. 2017. Social media and fake news in the
2016 election. Journal of economic perspectives 31, 2 (2017), 211–36.
[6]
Pengcheng An, Saskia Bakker, Sara Ordanovski, Ruurd Taconis, Chris LE Paf-
fen, and Berry Eggen. 2019. Unobtrusively Enhancing Reection-in-Action of
Teachers through Spatially Distributed Ambient Information. In Proceedings of
the 2019 CHI Conference on Human Factors in Computing Systems. ACM, 91.
[7]
Shakuntala Banaji, Ramnath Bhat, Anushi Agarwal, Nihal Passanha, and Mukti
Sadhana Pravin. 2019. WhatsApp vigilantes: an exploration of citizen reception
and circulation of WhatsApp misinformation linked to mob violence in India.
(2019).
[8]
Liam J Bannon and Kjeld Schmidt. 1989. CSCW: Four characters in search
of a context. In ECSCW 1989: Proceedings of the First European Conference on
Computer Supported Cooperative Work. Computer Sciences Company, London.
[9]
Chokri Barhoumi. 2015. The Eectiveness of WhatsApp Mobile Learning Activ-
ities Guided by Activity Theory on Students’ Knowledge Management. Con-
temporary Educational Technology 6, 3 (2015), 221–238.
[10]
Brenda R Beatty. 2000. Teachers leading their own professional growth: Self-
directed reection and collaboration and changes in perception of self and work
in secondary school teachers. Journal of In-Service Education 26, 1 (2000), 73–97.
[11]
Fernando Doménech Betoret. 2006. Stressors, self-ecacy, coping resources,
and burnout among secondary school teachers in Spain. Educational psychology
26, 4 (2006), 519–539.
[12]
Uwe M Borgho and Johann H Schlichter. 2000. Computer-supported coopera-
tive work. In Computer-supported cooperative work. Springer, 87–141.
[13]
Peter Börjesson, Wolmet Barendregt, Eva Eriksson, Olof Torgersson, and Tilde
Bekker. 2019. Teachers’ Expected and Perceived Gains of Participation in
Classroom Based Design Activities. In Proceedings of the 2019 CHI Conference
on Human Factors in Computing Systems. ACM, 157.
[14]
Claus Bossen. 2002. The parameters of common information spaces: The het-
erogeneity of cooperative work at a hospital ward. In Proceedings of the 2002
ACM conference on Computer supported cooperative work. 176–185.
[15]
Dan Bouhnik, Mor Deshen, and R Gan. 2014. WhatsApp goes to school: Mo-
bile instant messaging between teachers and students. Journal of Information
Technology Education: Research 13, 1 (2014), 217–231.
[16]
Virginia Braun and Victoria Clarke. 2006. Using thematic analysis in psychology.
Qualitative Research in Psychology 3, 2 (2006), 77–101. https://doi.org/10.1191/
1478088706qp063oa
[17]
Andreas Buchenscheit, Bastian Könings, Andreas Neubert, Florian Schaub,
Matthias Schneider, and Frank Kargl. 2014. Privacy Implications of Pres-
ence Sharing in Mobile Messaging Applications. In Proceedings of the 13th
International Conference on Mobile and Ubiquitous Multimedia (Melbourne,
Victoria, Australia) (MUM ’14). ACM, New York, NY, USA, 20–29. https:
//doi.org/10.1145/2677972.2677980
[18]
Jenna Burrell. 2010. Evaluating Shared Access: social equality and the circulation
of mobile phones in rural Uganda. Journal of computer-mediated communication
15, 2 (2010), 230–250.
[19]
John M Carroll, Chun Wei Choo, Daniel R Dunlap, Philip L Isenhour, Stephen T
Kerr, Allan MacLean, and Mary Beth Rosson. 2003. Knowledge management
support for teachers. Educational Technology Research and Development 51, 4
(2003), 42–64.
[20]
Levent Cetinkaya. 2017. The impact of WhatsApp use on success in education
process. International Review of Research in Open and Distributed Learning 18, 7
(2017).
[21]
Drupa Dinnie Charles, Azhagu Meena, Simiran Lalvani, Syeda Zainab Akbar,
Divya Siddharth, and Joyojeet Pal. 2020. Performing Gender, Doing Politics:
Social Media and Women Election Workers in Kerala and Tamil Nadu. In Pro-
ceedings of the 2020 International Conference on Information and Communica-
tion Technologies and Development (Guayaquil, Ecuador) (ICTD2020). Associ-
ation for Computing Machinery, New York, NY, USA, Article 20, 11 pages.
https://doi.org/10.1145/3392561.3394648
[22]
Kuang Chen, Akshay Kannan, YoriyasuYano, Joseph M. Hellerstein, and Tapan S.
Parikh. 2012. Shreddr: pipelined paper digitization for low-resource organiza-
tions. In ACM Annual Symposium on Computing for Development, ACM DEV ’12,
Atlanta, GA, USA - March 10 - 11, 2012. 3:1–3:10.
[23] Prerna Chikersal, Maria Tomprou, Young Ji Kim, Anita Williams Woolley, and
Laura Dabbish. 2017. Deep structures of collaboration: physiological correlates
of collective intelligence and group satisfaction. In Proceedings of the 2017 ACM
Conference on Computer Supported Cooperative Work and Social Computing.
873–888.
[24]
Karen Church and Rodrigo De Oliveira. 2013. What’s up with whatsapp?: com-
paring mobile instant messaging behaviors with traditional SMS. In Proceedings
of the 15th international conference on Human-computer interaction with mobile
devices and services. ACM, 352–361.
[25]
D Jean Clandinin and F Michael Connelly. 1996. Teachers’ professional knowl-
edge landscapes: Teacher stories—-stories of teachers—-school stories—-stories
of schools. Educational researcher 25, 3 (1996), 24–30.
[26] Herbert H Clark. 1996. Using language. Cambridge university press.
[27]
John W Creswell and Dana L Miller. 2000. Determining validity in qualitative
inquiry. Theory into practice 39, 3 (2000), 124–130.
[28]
Nicola Dell, Trevor Perrier, Neha Kumar, Mitchell Lee, Rachel Powers, and
Gaetano Borriello. 2015. Paper-Digital Workows in Global Development Or-
ganizations. In Proceedings of the 18th ACM Conference on Computer Supported
Cooperative Work & Social Computing, CSCW 2015, Vancouver, BC, Canada,
March 14 - 18, 2015. 1659–1669.
[29] John Dewey. 1897. My pedagogic creed. Number 25. EL Kellogg & Company.
[30]
John Dewey. 1923. Democracy and education: An introduction to the philosophy
of education. Macmillan.
[31]
Rachel Dodge, Annette P Daly, Jan Huyton, and Lalage D Sanders. 2012. The
challenge of dening wellbeing. International journal of wellbeing 2, 3 (2012).
[32]
Paul Dourish and Victoria Bellotti. 1992. Awareness and coordination in shared
workspaces. In Proceedings of the 1992 ACM conference on Computer-supported
cooperative work. 107–114.
[33]
Kim Doyle et al
.
2015. Facebook, Whatsapp and the Commodication of Aec-
tive labour. Communication, politics & culture 48, 1 (2015), 51.
[34]
Daniel R Dunlap, Dennis C Neale, and John M Carroll. 2000. Teacher collabora-
tion in a networked community. Journal of Educational Technology & Society 3,
3 (2000), 442–454.
[35]
Murray Edelman and Murray Jacob Edelman Edelman. 2001. The politics of
misinformation. Cambridge University Press.
[36] Michael Fullan. 2001. The new meaning of educational change. Routledge.
[37]
Susan R Fussell, Robert E Kraut, F Javier Lerch, William L Scherlis, Matthew M
McNally, and Jonathan J Cadiz. 1998. Coordination, overload and team per-
formance: eects of team communication strategies. In Proceedings of the 1998
ACM conference on Computer supported cooperative work. 275–284.
[38]
Mary L Gray, Siddharth Suri, Syed Shoaib Ali, and Deepti Kulkarni. 2016. The
crowd is a collaborative network. In Proceedings of the 19th ACM conference on
computer-supported cooperative work & social computing. 134–147.
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
[39]
Jonathan Grudin. 1988. Why CSCW applications fail: problems in the design and
evaluationof organizational interfaces. In Proceedings of the 1988 ACM conference
on Computer-supported cooperative work. 85–93.
[40]
François Guimbretière. 2003. Paper augmented digital documents. In Proceedings
of the 16th Annual ACM Symposium on User Interface Software and Technology,
Vancouver, Canada, November 2-5, 2003. 51–60. https://doi.org/10.1145/964696.
964702
[41]
Jari J Hakanen, Arnold B Bakker, and Wilmar B Schaufeli. 2006. Burnout and
work engagement among teachers. Journal of school psychology 43, 6 (2006),
495–513.
[42]
Noriko Hara and Khe Foon Hew. 2007. Knowledge-sharing in an online commu-
nity of health-care professionals. Information Technology & People 20, 3 (2007),
235–261.
[43]
Andy Hargreaves. 1998. The emotional practice of teaching. Teaching and
teacher education (1998).
[44]
Jade Vu Henry, Niall Winters, Alice Lakati, Martin Oliver, Anne Geniets, Si-
mon M Mbae, and Hannah Wanjiru. 2016. Enhancing the supervision of com-
munity health workers with WhatsApp mobile messaging: qualitative ndings
from 2 low-resource settings in Kenya. Global Health: Science and Practice 4, 2
(2016), 311–325.
[45]
Khe Foon Hew and Noriko Hara. 2007. Empirical study of motivators and
barriers of teacher online knowledge sharing. Educational Technology Research
and Development 55, 6 (2007), 573.
[46]
Roelande H Hofman and Bernadette J Dijkstra. 2010. Eective teacher pro-
fessionalization in networks? Teaching and Teacher education 26, 4 (2010),
1031–1040.
[47]
David Holman, Roel Vertegaal, Mark Altosaar, Nikolaus F. Troje, and Derek
Johns. 2005. Paper windows: interaction techniques for digital paper. In Pro-
ceedings of the 2005 Conference on Human Factors in Computing Systems, CHI
2005, Portland, Oregon, USA, April 2-7, 2005. 591–599.
[48]
Brian Holmes. 2013. School Teachers’ Continuous Professional Development in
an Online Learning Community: lessons from a case study of an e T winning
Learning Event. European Journal of Education 48, 1 (2013), 97–112.
[49]
Roberto Hoyle, Srijita Das, Apu Kapadia, Adam J. Lee, and Kami Vaniea. 2017.
Was My Message Read?: Privacy and Signaling on Facebook Messenger. In
Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems
(Denver, Colorado, USA) (CHI ’17). ACM, New York, NY, USA, 3838–3842. https:
//doi.org/10.1145/3025453.3025925
[50]
Lorraine Hudson, Clement Amponsah, Josephine Ohenewa Bampoe, Julie
Marshall, Nana Akua Victoria Owusu, Khalid Hussein, Jess Linington, Zoe
Banks Gross, Jane Stokes, and Róisín McNaney. 2020. Co-designing Digital
Tools to Enhance Speech and Language Therapy Training in Ghana. In Pro-
ceedings of the 2020 CHI Conference on Human Factors in Computing Systems.
1–13.
[51]
Jina Huh. 2015. Clinical Questions in Online Health Communities: The Case
of “See Your Doctor” Threads. In Proceedings of the 18th ACM Conference on
Computer Supported Cooperative Work & Social Computing (Vancouver, BC,
Canada) (CSCW ’15). Association for Computing Machinery, New York, NY,
USA, 1488–1499. https://doi.org/10.1145/2675133.2675259
[52]
Azra Ismail and Neha Kumar. 2019. Empowerment on the Margins: The Online
Experiences of Community Health Workers. In Proceedings of the 2019 CHI
Conference on Human Factors in Computing Systems. 1–15.
[53]
Maximilian J Johnston, Dominic King, Sonal Arora, Nebil Behar, Thanos Athana-
siou, Nick Sevdalis, and Ara Darzi. 2015. Smartphones let surgeons know What-
sApp: an analysis of communication in emergency surgical teams. The American
Journal of Surgery 209, 1 (2015), 45–51.
[54]
Young Ju Joo, Kyu Yon Lim, and Nam Hee Kim. 2016. The eects of secondary
teachers’ technostress on the intention to use technology in South Korea. Com-
puters & Education 95 (2016), 114–122.
[55]
KC Kamani et al
.
2016. Empowering Indian agriculture with WhatsApp–a
positive step towards digital India. International Journal of Agriculture Sciences,
ISSN (2016), 0975–3710.
[56]
Evangelos Karapanos, Pedro Teixeira,and Ruben Gouveia. 2016. Need fulllment
and experiences on social media: A case on Facebook and WhatsApp. Computers
in Human Behavior 55 (2016), 888–897.
[57]
Naveena Karusala, Ding Wang, and Jacki O’Neill. 2020. Making Chat at Home
in the Hospital: Exploring Chat Use by Nurses. In Proceedings of the 2020 CHI
Conference on Human Factors in Computing Systems (Honolulu, HI, USA) (CHI
’20). Association for Computing Machinery, New York, NY, USA, 1–15. https:
//doi.org/10.1145/3313831.3376166
[58]
Jasmeet Kaur, Asra Sakeen Wani, and Pushpendra Singh. 2019. Engagement of
Pregnant Women and Mothers over WhatsApp: Challenges and Opportunities
Involved. In Conference Companion Publication of the 2019 on Computer Supported
Cooperative Work and Social Computing. 236–240.
[59]
G Klein. 2001. Features of team coordination. New trends in cooperative activities:
Understanding system dynamics in complex environments (2001), 68–95.
[60]
Elizabeth Koh and Helen Hong. 2017. Developing professional competency in
a CSCL environment for teamwork: Two TPACK case studies of teachers as
co-designers. (2017).
[61]
Travis Kriplean, Ivan Beschastnikh, and David W McDonald. 2008. Articula-
tions of wikiwork: uncovering valued work in wikipedia through barnstars. In
Proceedings of the 2008 ACM conference on Computer supported cooperative work.
47–56.
[62]
Chris Kyriacou. 2001. Teacher stress: Directions for future research. Educational
review 53, 1 (2001), 27–35.
[63]
Daniel Lambton-Howard, Robert Anderson, Kyle Montague, Andrew Garbett,
Shaun Hazeldine, Carlos Alvarez, John A Sweeney, Patrick Olivier, and Ahmed
Kharrufa. 2019. WhatFutures: Designing Large-Scale Engagements on What-
sApp. In Proceedings of the 2019 CHI Conference on Human Factors in Computing
Systems. ACM, 159.
[64]
Jean Lave, Etienne Wenger, et al
.
1991. Situated learning: Legitimate peripheral
participation. Cambridge university press.
[65]
Ann Lieberman and Milbrey W McLaughlin. 1992. Networks for educational
change: Powerful and problematic. Phi delta kappan 73, 9 (1992), 673.
[66]
David W Livingstone. 2001. Adults’ informal learning: Denitions, ndings,
gaps and future research. (2001).
[67]
Brenda H Loyd and Douglas E Loyd. 1985. The reliability and validity of an
instrument for the assessment of computer attitudes. Educational and psycho-
logical measurement 45, 4 (1985), 903–908.
[68]
Zhicong Lu, Yue Jiang, Cheng Lu, Mor Naaman, and Daniel Wigdor. 2020. The
Government’s Dividend: Complex Perceptions of Social Media Misinforma-
tion in China. In Proceedings of the 2020 CHI Conference on Human Factors in
Computing Systems. 1–12.
[69]
Caio Machado, Beatriz Kira, Vidya Narayanan, Bence Kollanyi, and Philip
Howard. 2019. A Study of Misinformation in WhatsApp groups with a fo-
cus on the Brazilian Presidential Elections.. In Companion proceedings of the
2019 World Wide Web conference. 1013–1019.
[70]
Maria Macià and Iolanda García. 2016. Informal online communities and net-
works as a source of teacher professional development: A review. Teaching and
teacher education 55 (2016), 291–307.
[71]
Lynnette Mawhinney. 2010. Let’s lunch and learn: Professional knowledge
sharing in teachers’ lounges and other congregational spaces. Teaching and
Teacher Education 26, 4 (2010), 972–978.
[72]
Joseph Edward McGrath. 1984. Groups: Interaction and performance. Vol. 14.
Prentice-Hall Englewood Clis, NJ.
[73]
Rita Gunther McGrath. 1999. Falling forward: Real options reasoning and
entrepreneurial failure. Academy of Management review 24, 1 (1999), 13–30.
[74] Meghshala. 2019. http://meghshala.online/
[75] E.G. Mishler. 1986. Research interviewing: Context and narrative. (1986).
[76]
Preeti Mudliar and Nimmi Rangaswamy. 2015. Oine strangers, online friends:
Bridging classroom gender segregation with whatsapp. In Proceedings of the
33rd Annual ACM Conference on Human Factors in Computing Systems. ACM,
3799–3808.
[77]
Bonnie Nardi and Justin Harris. 2006. Strangers and friends: Collaborative play
in World of Warcraft. In Proceedings of the 2006 20th anniversary conference on
Computer supported cooperative work. 149–158.
[78] Bonnie A Nardi, Steve Whittaker, and Erin Bradner. 2000. Interaction and out-
eraction: instant messaging in action. In Proceedings of the 2000 ACM conference
on Computer supported cooperative work. 79–88.
[79]
PS Naruka, Shilpi Verma, SS Sarangdevot, CP Pachauri, Shilpi Kerketta, and JP
Singh. 2017. A study on role of WhatsApp in agriculture value chains. Asian
Journal of Agricultural Extension, Economics & Sociology 20, 1 (2017), 1–11.
[80]
Dennis C Neale, John M Carroll, and Mary Beth Rosson. 2004. Evaluating
computer-supported cooperative work: models and frameworks. In Proceedings
of the 2004 ACM conference on Computer supported cooperative work. 112–121.
[81]
Prema Nedungadi, Karunya Mulki, and Raghu Raman. 2018. Improving ed-
ucational outcomes & reducing absenteeism at remote villages with mobile
technology and WhatsAPP: Findings from rural India. Education and Informa-
tion Technologies 23, 1 (2018), 113–127.
[82]
Tricia Niesz. 2010. Chasms and bridges: Generativity in the space between
educators’ communities of practice. Teaching and Teacher Education 26, 1 (2010),
37–44.
[83]
Davidivitch Nitza and Yavich Roman. 2016. WhatsApp Messaging: Achieve-
ments and Success in Academia. International Journal of Higher Education 5, 4
(2016), 255–261.
[84]
Midas Nouwens, Carla F Griggio, and Wendy E Mackay. 2017. WhatsApp is for
family; Messenger is for friends: Communication Places in App Ecosystems. In
Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems.
ACM, 727–735.
[85]
Kenton P O’Hara, Michael Massimi, Richard Harper, Simon Rubens, and Jessica
Morris. 2014. Everyday dwelling with WhatsApp. In Proceedings of the 17th
ACM conference on Computer supported cooperative work & social computing.
ACM, 1131–1143.
[86]
Gary M Olson and Judith S Olson. 2001. Technology support for collaborative
workgroups. Coordination theory and collaboration technology (2001), 559–584.
Tag a Teacher: WhatsApp-Based Teacher Networks in Low-Income Indian Schools CHI ’21, May 8–13, 2021, Yokohama, Japan
[87]
Margaret R Olson and Cheryl J Craig. 2001. Opportunities and challenges in the
development of teachers’ knowledge: The development of narrative authority
through knowledge communities. Teaching and teacher education 17, 6 (2001),
667–684.
[88]
Massimo Petruzzi and Michele De Benedittis. 2016. WhatsApp: a telemedicine
platform for facilitating remote oral medicine consultation and improving clini-
cal examinations. Oral surger y, oral medicine, oral pathology and oral radiology
121, 3 (2016), 248–254.
[89]
Anthony Poon, Sarah Giroux, Parfait Eloundou-Enyegue, François Guimbretiere,
and Nicola Dell. 2019. Engaging High School Students in Cameroon with
Exam Practice Quizzes via SMS and WhatsApp. In Proceedings of the 2019 CHI
Conference on Human Factors in Computing Systems (Glasgow, Scotland Uk)
(CHI ’19). ACM, New York, NY, USA, Article 482, 13 pages. https://doi.org/10.
1145/3290605.3300712
[90]
TS Ragu-Nathan, Monideepa Tarafdar, Bhanu S Ragu-Nathan, and Qiang Tu.
2008. The consequences of technostress for end users in organizations: Concep-
tual development and empirical validation. Information systems research 19, 4
(2008), 417–433.
[91]
Vimala Ramachandran, Madhumita Pal, Sharada Jain, Sunil Shekar, Jitendra
Sharma, et al
.
2005. Teacher motivation in India. Technical Report. Discussion
Paper,(Azim Premji Foundation, Bangalore, 2005).
[92]
Patient Rambe and Crispen Chipunza. 2013. Using mobile devices to leverage
student access to collaboratively-generated resources: A case of WhatsApp
instant messaging at a South African University. In 2013 International Conference
on Advanced ICT and Education (ICAICTE-13). Atlantis Press.
[93]
Gustavo Resende, Philipe Melo, Hugo Sousa, Johnnatan Messias, Marisa Vas-
concelos, Jussara Almeida, and Fabrício Benevenuto. 2019. (Mis) Information
Dissemination in WhatsApp: Gathering, Analyzing and Countermeasures. In
The World Wide Web Conference. 818–828.
[94]
Gustavo Resende, Johnnatan Messias, Márcio Silva, Jussara Almeida, Marisa
Vasconcelos, and Fabrício Benevenuto. 2018. A System for Monitoring Public
Political Groups in WhatsApp. In Proceedings of the 24th Brazilian Symposium
on Multimedia and the Web. ACM, 387–390.
[95]
Christopher Rhodes and Sandra Beneicke. 2002. Coaching, mentoring and
peer-networking: Challenges for the management of teacher professional devel-
opment in schools. Journal of in-service education 28, 2 (2002), 297–310.
[96]
Piety Runhaar, Karin Sanders, and Huadong Yang. 2010. Stimulating teachers’
reection and feedback asking: An interplay of self-ecacy, learning goal ori-
entation, and transformational leadership. Teaching and teacher education 26, 5
(2010), 1154–1161.
[97]
Richard M Ryan and Edward L Deci. 2000. Self-determination theory and the
facilitation of intrinsic motivation, social development, and well-being. American
psychologist 55, 1 (2000), 68.
[98]
DM Sadker and KR Zittleman. 2006. Teacher-centered philosophies. Retrieved
June 7 (2006), 2013.
[99]
Mark S Schlager and Judith Fusco. 2003. Teacher professional development,
technology, and communities of practice: Are we putting the cart before the
horse? The information society 19, 3 (2003), 203–220.
[100]
Bo Shen, Nate McCaughtry, Jerey Martin, Alex Garn, Noel Kulik, and Mar-
iane Fahlman. 2015. The relationship between teacher burnout and student
motivation. British Journal of Educational Psychology 85, 4 (2015), 519–532.
[101]
Elisa S Sherno, Tara G Mehta, Marc S Atkins, Raechel Torf, and Jordan Spencer.
2011. A qualitative study of the sources and impact of stress among urban
teachers. School mental health 3, 2 (2011), 59–69.
[102]
In-geon Shin, Jin-min Seok, and Youn-kyung Lim. 2018. Too Close and Crowded:
Understanding Stress on Mobile Instant Messengers Based on Proxemics. In
Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems.
ACM, 615.
[103]
Anil Shukla and Tripta Trivedi. 2008. Burnout in Indian teachers. Asia Pacic
Education Review 9, 3 (2008), 320–334.
[104]
Einar M Skaalvik and Sidsel Skaalvik. 2015. Job Satisfaction, Stress and Coping
Strategies in the Teaching Profession-What Do Teachers Say?. International
education studies 8, 3 (2015), 181–192.
[105]
Einar M Skaalvik and Sidsel Skaalvik. 2016. Teacher stress and teacher self-
ecacy as predictors of engagement, emotional exhaustion, and motivation to
leave the teaching profession. Creative Education 7, 13 (2016), 1785–1799.
[106]
Kate Starbird, Jim Maddock, Mania Orand, Peg Achterman, and Robert M Mason.
2014. Rumors, false ags, and digital vigilantes: Misinformation on twitter after
the 2013 boston marathon bombing. IConference 2014 Proceedings (2014).
[107]
Jeanne Swaord. 2000. Teachers supporting teachers through peer coaching.
Leading professional development in education (2000), 105–115.
[108]
Stefan Timmermans and Iddo Tavory. 2012. Theory construction in qualitative
research: From grounded theory to abductive analysis. Sociological theory 30, 3
(2012), 167–186.
[109]
Rama Adithya Varanasi, Rene F.Kizilce c, and Nicola Dell. 2019. How Teachers in
India Recongure their Work Practices around a Teacher-Oriented Technology
Intervention. Proceedings of the ACM on Human-Computer Interaction 3 (2019),
Article 220. Issue CSCW.
[110]
Rama Adithya Varanasi, Aditya Vashistha, Tapan Parikh, and Nicola Dell. 2020.
Challenges and Issues Integrating Smartphones into Teacher Support Programs
in India. In Proceedings of the 2020 International Conference on Information and
Communication Technologies and Development (Guayaquil, Ecuador) (ICTD2020).
Association for Computing Machinery, New York, NY, USA, Article 10, 11 pages.
https://doi.org/10.1145/3392561.3394638
[111]
Simon Veenman, Hanneke De Laat, and Corine Staring. 1998. Evaluation of a
coaching programme for mentors of beginning teachers. Journal of In-service
education 24, 3 (1998), 411–431.
[112]
Tianyi Wang, Gang Wang, Bolun Wang, Divya Sambasivan, Zengbin Zhang,
Xing Li, Haitao Zheng, and Ben Y Zhao. 2017. Value and misinformation in
collaborative investing platforms. ACM Transactions on the Web (TWEB) 11, 2
(2017), 1–32.
[113]
Juliana J Willemse. 2015. Undergraduate nurses reections on Whatsapp use in
improving primary health care education. curationis 38, 2 (2015), 1–7.
[114]
Marisol Wong-Villacres, Hayley Evans, Danielle Schechter, Betsy DiSalvo, and
Neha Kumar. 2019. Consejero Automatico: Chatbots for Supporting Latino Par-
ents’ Educational Engagement (ICTD ’19). Association for Computing Machinery,
New York,N Y,USA, Article 53, 5 pages. https://doi.org/10.1145/3287098.3287149
[115]
Svetlana Yarosh, Tara Matthews, Michelle Zhou, and Kate Ehrlich. 2013. I Need
Someone to Help! A Taxonomy of Helper-Finding Activities in the Enterprise.
In Proceedings of the 2013 Conference on Computer Supported Cooperative Work
(San Antonio, Texas, USA) (CSCW ’13). Association for Computing Machinery,
New York, NY, USA, 1375–1386. https://doi.org/10.1145/2441776.2441931
[116]
Maxwell Yurkofsky, Sarah Blum-Smith, and Karen Brennan. 2016. Expand-
ing Outcomes: Exploring Varied Forms of Teacher Learning in an Online Pro-
fessional Development Experience. Singapore: International Society of the
Learning Sciences.
[117]
Rosanne C Zwart, Th Wubbels, Sanneke Bolhuis, and Th CM Bergen. 2008.
Teacher learning through reciprocal peer coaching: An analysis of activity
sequences. Teaching and teacher education 24, 4 (2008), 982–1002.
CHI ’21, May 8–13, 2021, Yokohama, Japan R. Varanasi, Aditya V., and N. Dell
A CODEBOOK FROM ANALYSIS OF WHATSAPP LOGS
Theme / Code Count Theme / Code Count
Professional interactions (23.40%) 3348 Top-down support (18.34%) 2624
Irrevelant forward 601 Assistance 670
Sharing highlights 494 School-related forward 446
Sharing memories 438 Top-down structures 436
Sharing learning resource 352 Micromanagement 381
Query 341 Encouragement / nudge 280
Sharing information 326 Policing norms 221
Motivational forward 256 Challenges / issues 178
Forwarding event/occasions 151 Technology support 12
Sharing Progress 146 Professional well-being (13.08%) 1872
Religious forward 126 Support sta appreciation 702
Forwarding general information 70 Sharing achievements 447
Sharing a promotion 47 Values 175
Online-oline bridge (13.78%) 1971 Celebration 149
Tech challenges 797 Engaging 148
Digitization 567 Enquiring wellbeing 101
Oine activity 453 Enjoying 83
Tech requirement 145 Happiness 40
Struggle to write 9 Stress 27
Contextualization (11.28%) 1614 Capacity improvement (9.28%) 1328
Language 682 Classroom management 326
Local 535 Pedagogy observation 225
India 221 Coaching 217
Western 176 Pedagogy strategies 202
Bottom-up support (7.42%) 1061 Pedagogy activities 122
Peer interactions 552 Pedagogy challenges 122
Peer appreciation 278 Pedagogy rationale 114
Teaching relevant forward 231 Security (1.87%) 268
Community care (1.55%) 222 Political forward 185
Parent management 114 Fake forward 51
Community challenges/issues 108 Malicious forward 32
Table 2: The complete codebook that resulted from our analysis of WhatsApp logs, showing our nine themes (bold) and 53
codes, including the prevalence (%) for each theme, and the total count for each theme/code. (The count for each theme is the
sum of the counts of all codes within that theme.)