Strengthening Learning Outcomes for Migrant Children: An Evaluation of the Learning and Migration Program (LAMP)

AMERICAN INDIA FOUNDATION
10th Floor, DLF City Court, MG Road, Near Sikanderpur Metro Station, Gurgaon 122002
216 E. 45th Street, 7th Floor New York, NY 10017
530 Lytton Avenue, Palo Alto, CA 94301
First published in 2020
This publication is not for sale. It may not be circulated in any other binding or cover. All
rights reserved.
Permission is required to reproduce material from this publication. Permission will be
freely granted to educational or non-profit organizations.
About the American India Foundation
The American India Foundation is committed to catalyzing social and economic change
in India, and building a lasting bridge between the United States and India through
high-impact interventions in education, livelihoods, public health, and leadership
development. Working closely with local communities, AIF partners with NGOs to develop
and test innovative solutions and with governments to create and scale sustainable
impact. Founded in 2001 at the initiative of President Bill Clinton following a suggestion
from Indian Prime Minister Vajpayee, AIF has impacted the lives of 6.7 million of India’s
poor. Learn more at www.AIF.org
About the Learning and Migration Program
AIF’s Learning and Migration Program (LAMP) provides access to quality education
opportunities to children in areas of seasonal migration, while also advocating to
communities and governments the universal right to education. Since 2004, LAMP has
reached 583,877 children with quality education opportunities in over 2,279 villages
across 13 states of India.
Publication Oversight and Management
Rowena Kay Mascarenhas, Director, Communications and Advocacy, American India
Foundation
LAMP Team
Swati Jha, Project Director, Learning and Migration Program, American India Foundation
Anindya Dutta Gupta, Program Oﬃcer, Learning and Migration Program, American India
Foundation
Tapas Satpathy, State Program Manager, Learning and Migration Program, American
India Foundation
Evaluation and Research Team
Anika Badyal Basu, Director, Learning Evaluation and Impact, American India Foundation
Ishika Kumar, Manager, Learning Evaluation and Impact, American India Foundation
Srikant Patel and Richa Shivhare, MRSS India
Editorial, Visualization, and Publishing Support
Lucid Solutions, New Delhi. www.lucidsolutionsonline.com
Photo Credits: All photos in this report are © Prashant Panjiar for AIF, unless otherwise
mentioned.

EXECUTIVE SUMMARY iii
Contents
Foreword vii
Acknowledgments ix
Abbreviations and Acronyms x
Glossary xi
Executive Summary xiii
1 INTRODUCTION 1
1.1 The Context of Seasonal Labor Migration 2
1.2 Migration from Nuapada and its Impact on the Children 3
1.3 The Learning Resource Center within the Learning and
Migration Program 4
1.3.1 Seasonal Hostels 5
1.3.2 Learning Enrichment Program 5
1.3.3 Learning Resource Centers 6
1.4 Purpose of the Study 8
2 METHODOLOGY 11
2.1 Evaluation Questions 12
2.2 Evaluation Design 12
2.3 Sample Size and Sampling Plan 13
2.3.1 School Sample Size 13
2.3.2 Student Sample Size 13
2.3.3 Sample Size for SMC Members 14
2.3.4 Sample for Qualitative Research 14
2.4 Research Instruments 14
2.5 Field Work Execution 18
2.5.1 Quantitative Research 18
2.5.2 Qualitative Research 18
2.6 Data Entry and Analysis 18
3 STATISTICAL METHODS 23
3.1 Assigning Weights to Questions 24
3.2 Assigning Weights to LOs 24
3.3 Comparing Standardized Scores of Students 26
3.3.1 The t-Test 26
3.3.2 Analysis of Variance 26

iv STRENGTHENING LEARNING OUTCOMES
4 RESULTS OF THE LEARNER’S ACHIEVEMENT TESTS 29
4.1 Learning Outcomes: Mathematics 30
4.1.1 Teaching–Learning Material for Mathematics 30
4.1.2 Learning Outcomes Identified for Mathematics 30
4.1.3 Mathematics Learning Achievements:
Endline versus Baseline 30
Treatment versus Control Group 31
Hub versus Spoke 32
4.2 Learning Outcomes: Science 33
4.2.1 Teaching–Learning Material for Science 33
4.2.2 Learning Outcomes Identified for Science 33
4.2.3 Science Learning Achievements: Endline
versus Baseline 34
4.2.4 Science Learning Achievements: Treatment
versus Control Group 35
4.2.5 Science Learning Achievements: Hub versus Spoke 36
4.3 Learning Outcomes: Language (Odia) 37
4.3.1 Teaching–Learning Material for Language (Odia) 37
4.3.2 Learning Outcomes Identified for Language (Odia) 37
4.3.3 Language (Odia) Learning Achievements:
Hub versus Spoke 38
4.4 Learning Outcomes: School Readiness 39
4.4.1 Learning Outcomes Identified for School Readiness 40
4.4.2 School Readiness Learning Achievements:
4.4.3 School Readiness Learning Achievements:
4.5 Understanding the Results 41
5 IMPACT AREAS OF LRCS: RESULTS FROM THE QUANTITATIVE
AND QUALITATIVE SURVEYS 44
5.1 Perspective of School Management Committees 45
5.2 Perspective of Head Teachers of Hub Schools 48
5.3 Perspective of Class I and II Teachers from Hub and
Spoke Schools 49
5.4 Perspective of Parents 50
5.5 Visits to the LRCs: Perspective of LRC Facilitators 50
5.1.1 Way of Teaching 50
5.1.2 Self-learning Environment in LRCs 51
5.1.3 Feedback from Facilitators 51

EXECUTIVE SUMMARY v
5.6 Sustainability of LRCs 51
5.7 Classroom Observations 52
6. CONCLUSION 57
ANNEXES 62
STORIES, TABLES, ANNEX TABLES, AND FIGURES
STORIES
I. Better Learning Outcomes 9
II. Higher Transition Rates 21
III. Unlocking Government Resources 27
IV. Improved Enrolment and Attendance 43
V. Child-Migration-Free Villages 55
VI. Technical Support to Seasonal Hostels 61
TABLES
1.1 Migrant Children under LAMP 5
2.1 Sample School Coverage and Geographical Spread 14
2.2 Number of Sample Students by Type of School and Class 15
2.3 Number of Sample Students by Class, Subject, and Type of School 15
2.4 SMC Sample Coverage 15
2.5 Participants in IDIs and FGDs 15
4.1 The Mathematics LO in Which Students Showed Greatest
Improvement (by Class): Endline versus Baseline 31
Improvement (by Class): Treatment versus Control Group 32
Improvement (by Class): Hub versus Spoke 33
4.4 The Science LO in Which Students Showed Greatest Improvement
(by Class): Endline versus Baseline 35
(by Class): Treatment versus Control Group 36
(by Class): Hub versus Spoke 37
4.7 The Language (Odia) LO in Which Students Showed Greatest
Improvement (by Class): Endline versus Baseline 38
Improvement (by Class): Treatment versus Control Group 39
Improvement (by Class): Hub versus Spoke 40
5.1 Classroom Observation: Performance Parameters Studied
Across 27 Classrooms 53

vi STRENGTHENING LEARNING OUTCOMES
ANNEX TABLES
A 4.1: Learning Outcomes: Mathematics 62
A4.2: Mean Difference in Mathematics LO Scores: Endline versus Baseline 62
A4.3: Mean Difference in Mathematics LO Scores: Treatment versus
Control Group 63
A4.4: Mean Difference in Mathematics LO Scores: Hub versus Spoke 63
A4.5: Learning Outcomes: Science 63
A4.6: Mean Difference in Science LO Scores: Endline versus Baseline 64
A4.7: Mean Difference in Science LO Scores: Treatment versus
Control Group 64
A4.8: Mean Difference in Science LO Scores: Hub versus Spoke 64
A4.9: Learning Outcomes: Language (Odia) 65
A4.10: Mean Difference in Language (Odia) LO Scores: Endline
versus Baseline 65
A4.11: Mean Difference in Language (Odia) LO Scores: Treatment versus
Control Group 66
A4.12: Mean Difference in Language (Odia) LO Scores: Hub versus Spoke 66
A4.13: Learning Outcomes: School Readiness 66
A4.14: Mean Difference in School Readiness LO Scores: Endline versus
Baseline 66
A4.15: Percentage of Students (Classes I and II) Achieving School
Readiness Learning Outcomes (Endline versus Baseline) 67
A4.16: Weighted School Readiness Mean Scores: Treatment versus
Control Group 67
FIGURES
4.1 Weighted Mean Mathematics LO Scores: Endline versus Baseline 31
4.2 Weighted Mean Mathematics LO Scores: Treatment versus
Control Group 32
4.3 Weighted Mean Mathematics LO Scores: Hub versus Spoke 33
4.4 Weighted Mean Science LO Scores: Endline versus Baseline 34
4.5 Weighted Mean Science LO Scores: Treatment versus
Control Group 35
4.6 Weighted Mean Science LO Scores: Hub versus Spoke 36
4.7: Weighted Mean Language (Odia) LO Scores: Endline versus Baseline 38
4.8: Weighted Mean Language (Odia) LO Scores: Treatment versus
Control Group 39
4.9: Weighted Mean Language (Odia) LO Scores: Hub versus Spoke 39
4.10: Weighted Mean School Readiness LO Scores: Endline
versus Baseline 41
5.1 Summary Responses of SMC Members during In-depth Interviews 48

Foreword
This report brings back old and
fond memories of my numerous
visits to Nuapada to spend time with
the NGO Lokadrusti led by Abani
Panigrahi, an early partner of the
Learning and Migration Program
(LAMP) of the American India
Foundation (AIF). I am delighted to be
asked to write its foreword.
Seasonal migration is not new
to India and as long as parts of the
country constitute a rain-fed single-
crop economy, it will continue. And,
as AIF discovered during background
research for LAMP, seasonal migration
is a pan-India phenomenon. It links
workers to myriad occupations that
start immediately after the monsoons
and continue till it is time for the
kharif sowing. Sugarcane cutting, salt
pan work, and brick kilns
(AIF’s focus sectors) are but three
such occupations designed around
seasonal migration – they just shut
down during the monsoons and
restart immediately after.
Odisha, particularly western
Odisha is without a doubt, most prone
to seasonal migration. It presents
the deadly combination of being a
rain-fed and drought-prone land,
peopled mainly by impoverished
Adivasi communities who are either
landless or own marginal lands
in the slopes and so are heavily
dependent on agriculture labour for
their livelihood. So, once the kharif
crop is harvested and work dries up,
migration is almost the only option
for survival. Migration to brick kilns
is very common where work units
called pathariya (each comprising
a man, a woman, and a child) are
de facto bonded by middlemen
(through the lure of opportune
money advances) to work at brick
kilns in Andhra Pradesh.
As the child migrates during the
critical part of the school year, and
does that for a few years, getting
her back to school, especially in
an age-appropriate class becomes
more and more distant a possibility.
In this context, LAMP is not just a
relevant but a crucial intervention.
It’s first response had two key
components—one, to ensure that
the migrant child does not drop
out; and two, to persuade migrant
parents to opt for occupations that
(unlike the pathariya system) do not
require the child to move. LAMP, in
its initial phase, focused on setting
up site schools in the locations
where children migrated to, and
soon began bridge classes as well to

mainstream children on their return
from migration. Recognizing that
these were interim solutions and very
diﬃcult to ensure quality, LAMP set up
seasonal hostels in the next phase to
provide safety and support to children
who stayed back while their parents
migrated and strengthen the quality
of education in these areas. This
entailed trust building with parents to
persuade more and more migrating
families to leave their children behind
and let them attend school.
Learning Resources Centers
(LRCs) are the latest in the continuous
evolution that the LAMP has seen
in Odisha. Working on a hub-spoke
model covering 200 schools in
Nuapada, its principal objective is to
improve learning levels, not just in
the schools it covers but in the larger
education ecosystem by working
with government school teachers.
This report presents the results of an
assessment study conducted in 2018
to evaluate the impact of LRCs on the
learning outcomes of students and the
related impact on communities, LRC
facilitators, and teachers.
As the report shows, the results
have been encouraging to say the
least. Compared to the baseline, the
learning levels across mathematics,
science and Odiya has improved by a
factor of 2.6 to 2.9. Even compared to
the control schools, the schools that
are part of LAMP show significantly
higher achievement levels. Further,
the study found that students were
more “school ready” which meant they
were better prepared both technically
and socially to deal with a new
school and classroom environment.
An interesting new component is
the introduction of a hub-spoke
model, which is universally used to
cascade and scale a program. Hub
schools seem to perform better than
spoke schools, and the next level
of research would be to evaluate
how spoke schools compare with
control schools. Stakeholders –
parents, principals, teachers, SMCs
– expressed great faith and spoke
positively about the program.
This report will provide
valuable insights for educationists,
policy-makers, administrators,
multilateral development agencies,
implementation agencies, civil society
organizations, and practitioners
in the area of child development
and education. All in all, there is
every reason to believe that the
intervention is worth replicating at
scale.
I remember spending many
a fun evening with children at
the seasonal hostels. I recall their
enthusiastic and smiling faces and
was often overwhelmed by the huge
responsibility the caretakers of these
seasonal hostels carried keeping
these children, left in trust by their
parents, safe, fed, groomed and
learning. When I see the results of
this study, I feel this trust has been,
to an extent, repaid.
Shankar Venkateswaran
Founding Executive Director, AIF
Chair, Oxfam India

Acknowledgments
This report is the result of deep
collaboration across many individuals
and organizations. The evaluation
and research team wishes to thank
all those who have contributed
their time, energy, and expertise to
successfully compiling this report.
We value our longstanding
relationships with all the supporters
and donors for their unshakeable faith
in us through the 15 years journey of
LAMP and wish to thank them for their
generous contribution. In particular,
we extend our sincere gratitude to
the donor, Western Digital Foundation
(formerly Sandisk Foundation), who
made this study and report possible.
AIF credits the success of LAMP to
everyone who has been involved with
the program right from inception to
date, due to whom LAMP is a scalable
and replicable model today.
We are deeply thankful to
Lokadrusti for making themselves
available at all stages of this study and
for providing meaningful insights on
the geography of the LRCs that came
under its purview. Their steadfast
commitment to LAMP spanning
over a decade has forged with
them one of AIF’s most valued
relationships.
We are grateful to the AIF’s
Board and management for their
unflinching belief in LAMP. We
appreciate the commitment of the
AIF LAMP team, implementation
partners, LRC facilitators, resource
persons, government school
teachers and headmasters, members
of Panchayati Raj Institutions (PRIs),
Block coordinators, staﬀ of the
seasonal hostels, parents, members
of school management committees,
and the communities that LAMP
works with. We also acknowledge
the contribution of Arjun Sanyal and
Kundan Mishra, formerly with the
LAMP team, who gave their inputs at
the conceptualization of this study.
The real heroes of this success story,
are, of course, the over half a million
children who have benefitted from
LAMP till date.
It is our honest attempt to thank
everyone, and any omission is
inadvertent.

Abbreviations and
Acronyms
ADEO Additional District Education Oﬃcer
AIF American India Foundation
AWW Anganwadi worker
FGD Focused Group Discussion
IDI In-depth Interview
LAMP Learning and Migration Program
LAT Learner’s Achievement Test
LEP Learning Enrichment Program
LO Learning Outcome
LRC Learning Resource Center
NGO Nongovernmental Organization
RCC Residential Care Center
SMC School Management Committee
SRI School Readiness Instrument
TLM Teaching–Learning Material

Glossary
Hub schools The Learning Resource Center works on a hub and
spoke model wherein each hub serves as a nodal
point and is attached with 5–6 spoke schools in the
respective blocks with the LRC facilitator managing
all the activities within the cluster.
Spoke schools These are the 5–6 schools covered by the LRC
facilitator on a weekly basis, providing handholding
support to government school teachers on LAMP
pedagogy, best classroom practices and proper use
of teaching learning materials (TLMs).
Treatment schools Schools where the LRC intervention is being
undertaken either as hub or spoke
Treatment students Students covered by the LRC intervention in either
the hub or the spoke schools
Control schools/
students
Schools/students not exposed to the LRC
intervention either as hubs or spokes
Baseline data Data on learning outcomes of students collected
before the LRC intervention, that is, before the
exposure of treatment students to LRCs
Endline data Data on learning outcomes collected during the
study after the exposure of treatment students to
LRCs

xii STRENGTHENING LEARNING OUTCOMES

Executive Summary
Sarva Shiksha Abhiyan in India
has tried to universalize elementary
education. In order to achieve this
goal, the country enacted the Right to
Free and Compulsory Education Act
in the year 2009 (RTE Act 2009) which
mandates that every child in the age
of 6 to 14 years has the constitutional
right to quality, inclusive, and child-
friendly education. The RTE Act (2009)
implies that, “every child has a right
to full-time elementary education of
satisfactory and equitable quality in a
formal school which satisfies certain
essential norms and standards”.1
Nevertheless, there are many
communities in India whose children
continue to be deprived of quality
education.
Nuapada in West Odisha, one of
the most backward districts in the
country, witnesses seasonal exodus
of its villagers to other states in
search of livelihood options driven
by repeated droughts, irregular
terrain, and low levels of literacy.
Children usually migrate with their
parents, which results in increase in
school drop outs. Even those who
1 https://mhrd.gov.in/rte
stay back with siblings, relatives,
or caregivers are often irregular
in attending school. In order to
arrest the dropout of children from
schools and improve their access
to quality education, American
India Foundation (AIF) launched the
(LAMP) in five migration-prone
blocks in Nuapada district of Odisha.
Since 2004, AIF has implemented a
series of interventions under LAMP,
such as Seasonal Hostels (SHs), the
Learning Enrichment Program (LEP),
and Learning Resource Centers
(LRCs). LAMP operates in highly
underserved, remote regions, where
school dropout often occurs due
to seasonal migration by the entire
family.
With implementing partner
Lokadrusti, LRCs have reached out
to children across 200 schools in
Nuapada since 2013. Based on a
hub-and-spoke model, each hub LRC
serves as a nodal point attached with
5–6 spoke schools in the vicinity.
One LRC facilitator manages all
the activities within a (hub–spokes)
cluster, engaging with children from
grades III to VIII before and after
The LRCs began in the year
2013 and reached out to
the Nuapada district of
Odisha. The interven on is
implemented by the NGO
Lokadrus and spans across
200 schools.
“

xiv STRENGTHENING LEARNING OUTCOMES
school hours. By design, students of
hub schools receive more intensive
guidance than those in spoke schools.
The current study attempts to
evaluate the impact of LRCs on the
learning outcomes of students and
the related impact on communities,
LRC facilitators, and teachers.
The study adopted a before–after
(comparison between baseline and
endline) and after–after (comparison
between treatment [hub–spokes] and
control [non-intervention] group).
Data collection was done using a
mixed method approach. A Learner’s
Achievement Test (LAT) was conducted
for classes III–VIII.
Impact on Learning Outcomes
MathemaƟcs
1. The mean scores for Math
during the endline in hub schools
was 2.89 times the score of the
treatment students in the baseline.
The difference in the mean scores was
also found to be statistically significant
for all classes except Class III. One of
the probable reasons could be that
the Class III students were exposed to
the LRC intervention for only a year.
2. The mean Math score of
treatment students was 1.22 times
the score of control students. The
difference in the mean scores was
statistically significant for all classes.
3. Students studying in hub
schools have lower learning deficit
in comparison to students of spoke
schools in Math. On an average the
hub Math scores were more than
1.24 times the spoke Math scores.
The difference in the mean scores
was statistically significant for all
classes.
Science
4. On an average (for hub and
spoke schools together) the scores
for endline were more than double
(2.63 times) the baseline scores for
Science. The difference in the mean
scores was statistically significant for
all classes.
5. Treatment students had
lower learning deficit in comparison
to control (non-LRC) students in
Science. On an average the LRC
Science scores were more than
1.35 times the non-LRC scores.
The difference in the mean scores
classes.
schools had lower learning deficit
in Science as compared to students
in spoke schools. On an average
the Science scores of hub students
were more than 1.29 times that of
spoke Science scores. The difference
in the mean scores was statistically
significant for all classes.
Language (Odia)
7. On an average (for hub
and spoke schools together) the
Language scores for endline were
more than double (2.64 times) the
baseline scores. The difference in
the mean scores was statistically
On an average (for hub
and spoke schools
together) the scores for
endline were more than
double (2.63 mes) the
baseline scores for Science.
The difference in the mean
scores was sta s cally
“

EXECUTIVE SUMMARY xv
8. Treatment students had
lower learning deficit in Odia as
compared to control students. On an
average the Odia scores of hub and
spoke students were more than 1.3
times that of non-LRC students. The
difference in the mean scores was
statistically significant for classes III
and VII.
schools had lower learning deficit in
Language as compared to students
of spoke schools. On an average
the hub scores in Odia were more
than 1.19 times the Odia spoke
scores. The difference in the mean
all classes.
School Readiness
10. Learning outcomes of
students, post-LRC intervention,
improved as compared to the
baseline in school readiness
in classes I and II. Students of
LRCs were more school ready as
compared to the baseline. ‘Ready
children’ could successfully transition
to higher grades in school. On an
average, LRC students had more
than double the scores (2.12 times)
of baseline students. The difference
in the mean scores was statistically
significant for classes I and II.
11. Students who underwent
LRC intervention were better enabled
Students who underwent
LRC interven on were
be er enabled to transi on
to higher grades in school
as compared to non-
interven on students.
“

xvi STRENGTHENING LEARNING OUTCOMES
to transition to higher grades in school
as compared to non-intervention
students. On an average, school
readiness scores of LRC students
were 1.03 times that of the control
students. The diﬀerence in the mean
classes I and II.
Impact on Stakeholders
Of the respondent members of
School Management Committees
(SMCs), 91% reported awareness
about their roles and responsibilities
under the Right to Education Act 2009.
When probed on the deliberation
themes in SMC meetings, 85%
acknowledged discussions regarding
performance of the school children,
progress of the school, and extra- and
co-curricular activities.
Of the respondent SMC members,
75% discussed LRC interventions
in the meetings and 67.5% strongly
believed that LRC helps in retaining
children in school.
Head teachers noticed
improvement in class performance,
reduction in absenteeism, and
decrease in drop-out rate because
of LRC intervention in government
schools. Four of five teachers
believed that parents had an
increased level of awareness
regarding education of students,
they regularly sent their children to
school, and that their participation
had increased in
monthly meetings.
It was found that parents
were more intensely engaged in
the performance of their wards in
schools. They were more involved
with the teachers and LRC facilitators
in understanding the gap areas and
working towards improving learning
achievements of students. This is
mostly due to the eﬀorts of the LRC
facilitators in engaging with parents,
informing them about the progress
of their children, and involving them
in LRC management.
Field visits to LRCs revealed
teaching methods that included
creative demonstrations,
experiments, and art work in class.
Each LRC was found to adopt a
unique approach to delivering
learning through teaching–learning
materials (TLMs), group exercises,
demonstrations etc. Students were
found to be self-motivated, curious,
and engrossed in the activities
conducted.
Each LRC was found to
adopt a unique approach to
delivering learning through
teaching–learning
materials (TLMs),
group exercises, and
demonstra ons.
“

1
Introduction
In developing countries, seasonal labor
migration is a household livelihood
strategy to cope with poverty.
Mainstreaming children of migrant
families into the development process
is a big challenge in attaining the goal
of universal primary education and
inclusive growth in a country like India.
Learning and Migration Program (LAMP), a flagship program of AIF
(since 2004) operates in highly underserved and remote districts of
India, where children drop out of school owed to distress seasonal
migration and frequently end up laboring in hazardous industries such
as brick kilns.

2 STRENGTHENING LEARNING OUTCOMES
With the children and their
migrant families constantly
on the move, it is diﬃcult
to get their precise count.
However, empirical
evidence from across the
globe indicates that the
numbers are significant.
“ 1.1 The Context of Seasonal
Labor MigraƟon
In developing countries, seasonal
labor migration is a household
livelihood strategy to cope with
poverty. The children of such migrants
are either left behind in the villages
(under the care of older relatives) or
forced to move with their parents to
locations where work may be found.
Those that migrate seasonally often
remain out of school or are forced
to drop out; many are put to work as
child labor. Thus, mainstreaming
these children in the development
process is a big challenge in attaining
the goal of universal primary
education and inclusive growth in a
country like India.
With the children and their
families constantly on the move, it
is diﬃcult to get their precise count.
However, empirical evidence from
across the globe indicates that the
numbers are significant. For example,
18%–40% children in Bangladesh,
50%–60% in Tanzania, and 80% in
Mali are reported living in migrant
households in rural areas. Roughly a
million children in Indonesia and half-
a-million in Thailand are left behind by
parents working overseas. Similarly,
the wives and children are left behind
in most of the rural out-migrant
households in eastern India.1
Labor migrants based on their family
1 Archana K. Roy, Pappu Singh, and UN
Roy (2015), ‘Impact of Rural–Urban Labour
Migration on Education of Children: A Case
Study of Left Behind and Accompanied
Migrant Children in India’, Space and Culture,
2(4), p. 18.
association and nature of movement
may be grouped as:
• seasonal/temporary without-
family migrants
• semi/permanent with-family
migrants
• seasonal/temporary with-
family migrants
In the poverty-prone regions
of eastern India, male migration
is a ‘culturally accepted’ livelihood
strategy or a coping mechanism
during duress. This type of migration
potentially plays both a positive and
a negative role in the education of
children who are left behind. On
one hand, remittances are seen
to improve school enrolment/
attendance of left-behind children;
while on the other, the absence
of the father puts the children
at the risk of lax discipline, poor
performance, dropping out of
school, or being put to child labor.
In cases where the family of the
male migrant also shifts gradually to
the place of work and settles in the
mid-long-term, the children do have
a chance of being part of a formal
schooling and healthcare system.
In an alternate model, often
followed by extremely deprived or
marginalized communities such as
tribal folk, the entire family migrates
in a bid to survive. They generally
find low-paid menial work at brick
kilns or salt pans, in sugarcane
harvesting, stone quarrying,
construction, plantations, or
fishing. Since work is seasonal and
temporary, these families keep
oscillating between their village
homestead and multiple work

INTRODUCTION 3
destinations with the children bearing
the brunt of the uncertainties.
Like the rest of eastern India,
migration is a key survival tool for the
rural poor in Odisha too. About 66%
of workers from coastal Odisha and
88% from west Odisha travel beyond
the state boundaries in search of work
each year. Of the migrants from west
Odisha, about 26% find employment
in the construction, brick making,
agriculture, and transportation in
Chhattisgarh (primarily in the districts
of Raipur and Durg). Another 8% head
for Hyderabad, Ranga Reddy, and
Vishakhapatnam districts of Andhra
Pradesh, again to find work in brick
making. A tenth of the west Odisha
migrants reach the construction
sector of Gujarat, primarily Rajkot.2
1.2 MigraƟon from Nuapada
and its Impact on the Children
Noteworthy for its adverse
impact on children is the pathariya
system practiced in the districts of
west Odisha (including Nuapada).
The pathariya represents a migrating
work unit of a man, a woman, and
one or two children (6–14 years old).
2 See : http://www.aajeevika.
org/assets/pdfs/Odisha%20State%20
Migration%20Profile%20Report.pdf

Labor contractors (locally known as
dalals) lure the pathariya (basically
the family unit) with a fixed monetary
advance to the brick kilns of Andhra
Pradesh and force the children into
labor. Poor families of districts such as
Nuapada are driven by recurrent and
severe droughts in the region into this
form of modern-day slavery.
To alleviate the circumstances,
the Odisha government has initiated
a slew of measures that include the
provision of healthcare and education
to the families that the migrants leave
behind, the rescue of workers caught
in diﬃcult situations in other states,
and the setting up of help desks for
migrant Odia workers in Telangana,
Andhra Pradesh, Tamil Nadu,
Karnataka, and Delhi.
In Odisha, AIF’s Residential Care
Centers (RCCs) have been set up
to enable school-going children to
stay back safely with the community
and continue studying in the village
school when their parents migrated
out. The RCCs were first started in
2004–05 in Balangir district and then
extended to the Khariar block of
Nuapada district. Under the initiative,
a seasonal hostel was set up in a
school located such that children of
migrants from three to four nearby
villages or schools could live and
study in it. During the RCC stay, the
midday meals of the hostel inmates
were transferred from the parent
school to the school that housed
the RCC. Initially the RCCs were run
by nongovernmental organizations
(NGOs) with government funding;
eventually the state government took
over direct operations.
1.3 The Learning Resource
Center within the Learning and
MigraƟon Program
(LAMP), a flagship program of the
American India Foundation (AIF)
operates in highly underserved
and remote districts of India,
where children drop out of school
owed to distress seasonal migration
by the entire family (as opposed to
just male migration) and frequently
end up laboring in hazardous
industries such as brick kilns. LAMP
works to ensure that these children
remain in the school net, develop
strong foundational skills, and
improve retention onto secondary
school.
Since 2004, AIF has implemented
a series of interventions such
as Seasonal Hostels, Learning
Enrichment Program (LEP), and
Learning Resource Centers (LRCs)
under LAMP (Table 1.1).

INTRODUCTION 5
1.3.1 Seasonal Hostels
With the advent of LAMP, seasonal
hostels (like the RCCs) provided
space and opportunity for children
of migrant families to stay back and
continue with schooling while the
parents were away. Located within
the village, the hostel was viewed as
being part of the trusted environment
of community, relatives, and friends
where the children could be left safely.
The children were free to visit their
homes and homes of relatives. At the
same time, seasonal hostels were
administratively complex, expensive,
and required a greater degree of
community support as well as closer
and more regular interaction with
formal schools.
1.3.2 Learning Enrichment
Program
LAMP’s LEP was designed to
address the learning needs of first
generation learners. In the process
of demonstrating the enormous
learning potential of children
through the pedagogic practices
followed in LEP, there was an
attempt to influence government
school teachers, especially in early
primary grades (I–V), to adopt these
practices. The aim was to address
the endemic problem of learning
Table 1.1: Migrant Children under LAMP
Number of
Seasonal
Hostels
Children Enrolled in Seasonal
Hostels
Children Enrolled in Bridge Camps
after Returning to Home Village
Children of Migrant Parents
Living with Caregivers
Boys Girls Total Boys Girls Total Boys Girls Total
2004–05 7 112 81 193 0 0 0 0 0 0
2005–06 16 325 221 546 110 116 226 0 0 0
2006–07 30 677 426 1103 228 190 418 0 0 0
2007–08 24 417 208 625 420 325 745 0 0 0
2008–09 30 508 226 734 400 364 764 0 0 0
2009–10 35 579 260 839 0 0 0 0 0 0
2010–11 16 248 129 377 0 0 0 230 195 425
2011–12 10 166 88 254 0 0 0 396 336 732
2012–13 59 1316 943 2259 0 0 0 121 78 199
2013–14 124 2651 1786 4437 0 0 0 263 199 462
2014–15 188 3869 2732 6601 0 0 0 635 362 997
2015–16 235 3826 5054 8880 0 0 0 509 442 951
2016–17 179 3937 3141 7078 0 0 0 495 405 900
2017–18 188 4290 3392 7682 0 0 0 1059 851 1910
2018–19 155 2803 3907 6710 0 0 0 469 528 997
Total 1,296 25,724 22,594 48,318 1,158 995 2,153 4,177 3,396 7,573
Note: Seasonal hostels are run by school management committees with support of Sarva Shiksha Abhiyan and technical assistance
from AIF/Lokadrusti.

deficits among children (especially
those from Schedule Tribes) and their
exclusion from schooling.
The LEP was implemented
first inside seasonal hostels
and subsequently in or outside
government schools. Initially rolled out
as a short-term remediation measure
to bridge the competency deficit of the
migrant children, it was subsequently
extended to other children. It
specifically targeted children of classes
III–V who were not able to read, write,
or count (competencies expected in
classes I–II). Facilitators from NGOs
would conduct these special classes
before or after school hours, and
it was envisioned that some of the
good practices demonstrated in LEP
would be adopted by the government-
schoolteachers too. The LEP had a
structured curriculum, a specific set
of workbooks and materials, weekly
instructional plans, and training plans.
It was codified to an extent that made
it easy to replicate and scale, and LEP
was implemented across four states in
four different languages.
Despite the apparent successes
where the LEP did improve learning
outcomes significantly, the model had
its limitations. Having facilitators take
2-hour sessions, both before and after
school was extremely difficult. There
were only a few instances wherein
teachers witnessing the progress
made by LEP children, proactively
asked for training. Since the LEP
sessions were held either before or
after school hours, most teachers
never even got to see what these
sessions were like. This made the
institutionalization of LEP practices
very difficult.
1.3.3 Learning Resource Centers
Challenges faced during the LEP
were sought to be met through LRCs
2013 onward. Implemented by the
NGO Lokadrusti, the intervention
spans 200 schools in Nuapada,
Odisha. The LRCs run in government
school classrooms and community-
owned spaces located in villages,
which cater to children from 7 to 10
nearby schools. The centers serve
as nodal points to support learning
and encourage local children to
continue education. With a focus
on improving language (both local
and English), Mathematics, Science,
and knowledge of social issues,
LRCs are equipped with additional
learning materials and technology
(like computers, science kits, and
projectors) that address the needs of
children across grades. The centers
are based on a hub-and-spoke model
wherein the LRCs located in the hub
villages cater to children in nearby
spoke villages as well.
These centers also serve as
information hubs which keep
children updated about scholarships,
education schemes, entrance
examinations, and opportunities
available for higher/ vocational
education in the area. They are
typically staffed by local youth,
also known as LRC facilitators
who can speak the local dialect
and connect easily with the
community. Supported by the Block
coordinator, the LRC facilitator
mobilizes the community, conducts
classroom sessions beyond school
hours, reaches out to the spoke
schools, and assists or supports
teachers of classes I and II in use
LRCs serve as informa on
hubs which keep
children updated about
scholarships, educa on
schemes, entrance
examina ons, and
opportuni es available
for higher/ voca onal
educa on.
“

of teaching–learning materials
(TLMs), thereby embedding best-
practices in mainstream schooling.
The facilitators are responsible for
conducting different activities with
children, creating contextual learning
materials, focusing on innovations
and use of technology, and assisting
the nearby government schools in
adopting quality teaching–learning
practices to address the root causes
of learning deficit. Most importantly,
LRCs aim to demonstrate and
effectively communicate the idea of
‘quality’ education to the community
and drive a shift in emphasis from
mere improvement of tangible
inputs, like infrastructure, to
ensuring improvement in educational
outcomes.
The LRC has five major functions:
1. Demonstrating effective
learning remediation techniques
with model LEP classes: To
demonstrate quality practices and
effective remediation techniques
in school education, children are
enrolled in LEPs which help bridge
their learning gaps and reach
learning levels corresponding to their
respective grades/classes.
2. Conducting grade-specific
enrichment activities for children:
LRC facilitators conduct activities like
story-telling sessions, laboratory-
based science education and social
projects which support grade-specific
curricular concepts.
3. Using research and
innovation: LRCs test and develop
curricular content and pedagogical
techniques through field trials and
experiments, use technology through
videos, presentations, and interactive
software and create a video bank of
classroom practices, student projects
and their feedback. The LEP classes
are documented in detail with an
emphasis on diagnostic assessment
and evidence-based case studies.
4. Providing information
support services: LRCs serve as
information hubs with the objective
of keeping children updated about
scholarships and other education
schemes, entrance examinations
for other schools and opportunities
available for higher and vocational
education.
5. Assisting government
schools through academic
support: LRC facilitators liaise
with government schools in
nearby villages to train and assist
government school teachers in
implementing good practices in their
classrooms.
AIF’s objective is to establish a
demonstration-based ‘model’ LRC,
driven by the community, for other
states and districts to replicate.
About 25 LRCs (hub schools)
supported 175 spoke schools as
part of the intervention.
1.4 Purpose of the Study
The purpose of this study is to
evaluate the impact of the program
by examining the difference in
learning outcomes (LOs) of students
before and after its implementation.
The study also tries to understand
efficacy of the hub and spoke
model.
LRC facilitators conduct
ac vi es like story-telling
sessions, laboratory-based
science educa on, and
social projects which
support grade-specific
curricular concepts.
“

INTRODUCTION 9
Kuber Jagat, at 15, studies in Grade 9 at the Ghuchaguda High School.
He lives with his parents and sister at Fursund village in Nuapada
district. Fursund has around 1,300 residents, mostly from the Other
Backward Classes (OBC).
His father Jagatjiban and mother Nandini are migrant laborers.
For over a decade now, they have been migrating to brick kilns in
Andhra Pradesh in search of work. Earlier, Kuber and his sister
Basanti would also migrate with their parents.
In June 2013, when they returned to their village, Kuber’s friends in
the neighborhood invited him to join them at school. Kuber, however,
refused as he had dropped out of school and was not sure he would
be able to cope with the learning gap. To his surprise, the teacher
visited Kuber at home later that day, and requested his parents to
send him to the LEP class of LAMP.
That was the turning point in his life. He was in Grade III at the time,
and gradually overcame his fear of school. His parents noticed his
interest in learning, and decided not take him along when they
migrated next. The following year, Kuber’s parents enrolled their
daughter Basanti under LAMP’s LEP class as well. This year Basanti
has cleared the entrance test for Grade VI admission into Odisha
Adarsha Vidyalaya (OAV), which is a model English medium state
government school.
STAKEHOLDER: STUDENT
Better Learning
Outcomes
I
From being a
school dropout, I have gone
to scoring 74% in Grade VIII
final exams. My sister has
scored 69% in the Grade V
annual assessments. LAMP’s LEP
classes are the turning point in
our lives.
—Kuber Jagat
Kuber Jagat on his way to school
Photo Courtesy: Lokadrusti
“
”

2
Methodology
Both quantitative and qualitative data
was collected through structured
interviews, in-depth interviews, focus
group discussions, and observation
checklists.
A cross sectional, before–after research design was adopted for
the evaluation of LRCs with the core purpose of answering the key
research questions.

The research tools were
designed to arrive at
indicator values as well as
gather more informa on
related to the outcome
indicators.
“ 2.1 EvaluaƟon QuesƟons
Primarily the study answered the
following questions:
1. Does the LRC program improve
the learning outcomes (LOs) of
students (endline versus baseline and
treatment versus control)?
2. How do the LOs of students in
hub schools compare with those in
spoke schools?
3. What is the impact of the LRC
program on the engagement of
stakeholders (such as SMC members,
teachers, and parents) with education?
4. What is the impact of the
community engagement strategies
and training programs on the capacity
of SMCs?
5. What is the eﬀectiveness
of pedagogical interventions and
classroom practices at the LRCs?
6. What is the impact of training
and supporting government- school
teachers of classes I and II?
7. How can LAMP improve the
quality and outreach of LRCs?
2.2 EvaluaƟon Design
A cross sectional, before–after
research design was adopted for
the evaluation of LRCs with the
core purpose of answering the key
research questions outlined in the
terms of reference (Section 2.1). Based
on the research questions a list of
indicators was prepared along with
the sources of information for each
indicator. The research tools were
designed to arrive at indicator values
as well as gather more information
related to the outcome indicators.
Both quantitative and
qualitative data was collected
through structured interviews,
in-depth interviews (IDIs), focus
group discussions (FGDs), and
observation checklists. These
research instruments were prepared
by MRSS in consultation with AIF. In
addition, the Learner’s Achievement
Tests (LATs) conducted during the
baseline study were repeated during
the endline study in both treatment
and control schools to measure the
change in LOs with LRCs.
Scope of work included:
• Desk research: Secondary
literature and data review
• Exploratory field visit: A
joint AIF–MRSS India visit to
Nuapada in January 2017 for
formative research
• LATs for Mathematics,
Science, and Language (Odia)
a. Baseline LAT (before
the commencement of
the new LRC session)
conducted by AIF
b. Endline LAT (after the
completion of the LRC
session) conducted by
MRSS India
• Quantitative data collection
through structured
questionnaires: Structured
interviews with SMC

METHODOLOGY 13
members—sarpanch/ward
member, SMC president,
students, and parents
• Qualitative research:
a. through IDIs with:
i. Head teachers
ii. Spoke teachers
(classes I and II)
iii. LRC facilitators
iv. Block coordinators
v. Assistant District
Education Oﬃcer
(ADEO)
vi. Parents
vii. Students
b. observation of classroom
transaction
2.3 Sample Size and
Sampling Plan
2.3.1 School Sample Size
The endline study was conducted
in the 25 hub schools (across five
blocks of Nuapada) with one spoke
school each (randomly selected)
(Table 2.1). For each hub–spoke pair,
a non-intervention (control) school
from the same block (to ensure
similar conditions and context) was
randomly selected for comparison.
2.3.2 Student Sample Size
Within the sample schools a
second level of sampling was done

to select respondent students. All
the students of classes III–VIII were
covered in the hub schools to test
LOs in Mathematics, Science, and
Language (Odia). All hub-school
students of classes I and II were also
covered using the School Readiness
Instrument (SRI).1 In each of the spoke
and control schools, however, only
six students were randomly selected
from each class (III–VIII) to test LOs in
Mathematics, Science, and Language
(Odia). Similarly, six students were
randomly selected from each class
(I and II) for SRI. Number of sample
students by type of school and class
is provided in Table 2.2. Number of
sample students by class, subject, and
type of school is listed in
Table 2.3.
2.3.3 Sample Size for SMC Members
Across the five intervention
blocks in Nuapada, in the interest
of complete representation from
1 The School Readiness Instrument,
developed and standardized by World
Bank in 2009, is an activity-based tool
conceptualized as a dip stick measure for
the assessment of the readiness of children
for school at a systemic level. It specifically
measures certain cognitive and language
skills and concepts that have proven to be
good indicators of school readiness.
all stakeholders, interviews were
conducted with all the SMC
members—parents, students,
anganwadi workers (AWWs), the
SMC president, and sarpanch/ward
member (Table 2.4).
2.3.4 Sample Size for QualitaƟve
Research
As mentioned earlier, FGDs and
IDIs with key stakeholders such as
head teachers, teachers, parents,
LRC facilitators, block coordinators,
ADEO, and students were conducted
to evaluate the impact of the LRC
intervention, specifically addressing
the evaluation questions 3–6 listed in
Section 2.1 (Table 2.5).
The IDIs and FGDs were
conducted in the same five
intervention blocks as the
quantitative survey.
Apart from IDIs and FGDs, 28
classroom observations across the
five intervention blocks captured
the quality of classroom transaction,
lesson delivery, and classroom
environment.
2.4 Research Instruments
In accordance with global best
practices as well as the general
Table 2.1: Sample School Coverage and Geographical Spread
Block Hub schools Spoke schools Control schools
Boden 5 5 5
Khariar 6 6 6
Komna 4 4 4
Nuapada 5 5 5
Sinapali 5 5 5
Total 25 25 25

METHODOLOGY 15
Table 2.2: Number of Sample Students by Type of School and Class
Classes I and II Class III Class IV Class V Class VI Class VII Class VIII Total
Baseline 1,377 517 817 1,452 1,568 1,624 1,486 8,841
Endline: Hub 894 287 489 541 674 717 636 4,237
Endline: Spoke 282 118 142 139 143 142 139 1,105
Control 306 140 151 156 157 151 149 1,209
Table 2.3: Number of Sample Students by Class, Subject, and Type of School
Mathematics
Class III Class IV Class V Class VI Class VII Class VIII
Baseline 539 821 1,437 1,566 1,621 1,475
Endline: Hub 284 487 541 671 712 642
Endline: Spoke 124 152 142 146 151 141
Control 145 151 156 156 155 148
Science
Baseline 488 808 1,460 1,553 1,614 1,502
Endline: Hub 292 496 540 677 717 637
Endline: Spoke 112 134 135 135 138 135
Control 138 151 155 161 144 148
Language (Odia)
Baseline 523 823 1,459 1,584 1,637 1,481
Endline: Hub 284 483 542 673 722 629
Endline: Spoke 118 140 140 148 138 140
Control 137 150 156 155 154 150
Student Sample: SRI
Classes I and II
Baseline 1,377
Endline: Hub and Spoke 1,176
Control 306
Table 2.4: SMC Sample Coverage
SMC Members Sample Size
Anganwadi workers 12
Sarpanch/ Ward Members 11
SMC Presidents 9
Students 9
Parents 8
Total 49
Table 2.5: ParƟcipants in IDIs and FGDs
Stakeholder Sample Size Research Instrument
Head teachers 5 IDI
Teachers 5 IDI
Parents 10 IDI
LRC facilitators 5 IDI
Block coordinators 3 IDI
ADEO 1 IDI
Students 2 FGD

testing and performance standards
followed for developing the Annual
Survey of Education Report (ASER)
in India, all students were tested
for mastery over minimum learning
competencies from two classes
lower than the class they currently
occupied.2
Separate instruments were
designed for qualitative and
quantitative research. Understandably,
the quantitative tools had coded
responses with no open-ended
questions. The research instruments
included the following:
• Quantitative research
instruments:
LAT questionnaires:
The LATs for classes III
2 Annual Survey of Education Report at
www.asercentre.org.
to VIII were based on
standardized questions
adopted, prepared and
translated by practicing
teachers in Odisha. Six
questionnaires, one
for each class (III to
VIII), were developed
for each testing
area, Mathematics,
Science, and Language
bringing the total to 18
questionnaires.
SRI: World Bank’s SRI
was adapted to the local
context for assessing
the school readiness of
students currently in
classes I and II.
SMC questionnaires:
Aimed at SMC members,
the enquiry areas
included awareness

METHODOLOGY 17
about: (i) SMC conduct;
(ii) meeting protocols; (iii)
SMC functions; and (iv)
LRC and perspective on its
operations.
• Qualitative research
instruments:
Discussion guide for IDI
of the head teachers
of the hub schools:
The questionnaire was
designed to collect
information on the
head teacher’s roles
and responsibilities,
awareness regarding LRC,
feedback on the support
provided, and the head
teacher’s point of view
on the effectiveness and
sustainability LRCs.
of teachers of classes I
and II of spoke schools:
designed to collect
information on the
head teacher’s roles and
responsibilities, awareness
regarding LRC, feedback
on the support provided,
and the teacher’s point of
view on the effectiveness
and sustainability LRCs.
Discussion guide
for IDI of parents of
LRC students: The
questionnaire was
designed to collect
information on parental
understanding of LRCs,
their engagement with
the child’s education,
their views on the
school infrastructure,
engagement with the
teachers, and feedback
on the LRC.
Discussion guide for
IDI of LRC facilitator:
designed to collect
information on the roles
and responsibilities
of the LRC facilitator,
activities in the LRC,
challenges faced during
the implementation
on themes of learning
spaces, encouragement
provided to children,
word load and schedule
management, training
attended, outcomes,
and their opinion on
the sustainability of
the LRC.
of Block coordinator:
designed to collect
information about the
roles and responsibilities
of the Block coordinator,
day-to-day activities,
monitoring of LRCs,
engaging with the
communities and
government officials, etc.
Discussion guide for
IDI of the ADEO: The
questionnaire was
designed to assess the
ADEO’s awareness about
and support to LRC,
challenges faced in the
LRC’s implementation
and feedback for the
program.
The ques onnaire was
designed to collect
informa on on the head
teacher’s roles and
responsibili es, awareness
regarding LRC, feedback on
the support provided, and
the teacher’s point of view
on the effec veness and
sustainability LRCs.
“

Discussion guide for FGDs
among students: FGDs
were held among 7–10
students in each sample
school exploring their
interest in Mathematics,
Science, and Odia, most
and least favorite aspects
of LRC, and their feedback
on the program.
• Classroom observation:
Observations of hub and
spoke classrooms aimed
to capture the interaction
between students and the
teacher, lesson delivery, and
classroom environment.
2.5 Field Work ExecuƟon
2.5.1 QuanƟtaƟve Research
Field work for the LRC endline
study was conducted with checks and
verifications at multiple stages and
points to ensure data quality. The
survey tools were pre-tested during
the baseline study and the endline
questionnaire was designed on similar
lines. Pilots of both baseline and
endline assessment papers were run
to check the reliability of these sets in
assessing the learning outcomes per
their grade level. The purpose of such
testing at the baseline was to fine tune
the overall framework of questions,
assess the time required, and the
quality and quantum of information
available.
Endline research instruments
were translated to Odia and the
quantitative survey investigators and
supervisors, recruited from across
the five intervention blocks, were
trained in the last week of March
2018 in Khariar. The survey team
consisted of five supervisors and 24
investigators, managed by research
coordinators. The field work was
undertaken during April–May 2018.
The field supervisors along with
the research coordinators were
responsible for data collection and
quality assurance. Every eﬀort was
made to ensure quality data with
consistency checks and correction
incorporation at various levels.
Research coordinators visited at
least 20% of the schools with the
investigators to ensure quality. Data
from another 20% was cross checked
during repeat visits by supervisors. At
the close of each day, the supervisors
would scrutinize the questionnaires
to identify gaps and errors which
would be rectified the following day.
2.5.2 QualitaƟve Research
The qualitative research was
undertaken by a separate team
of researchers from MRSS. The
three researchers knew Odia
and were aware of the ethical
issues, operational guidelines,
and the purpose of the qualitative
research in LRC evaluation. The
qualitative research field work was
accomplished during April 2018 along
with the quantitative field work.
2.6 Data Entry and Analysis
A data entry format was
developed in MS-Excel for the
student questionnaires and test
The field supervisors
along with the research
coordinators were
responsible for data
collec on and quality
assurance. Every eﬀort was
made to ensure quality
data with consistency
checks and correc on
incorpora on at various
levels.
“

papers. Initially the format was piloted
with a small sample of completed
questionnaires to flag and weed out
errors and inconsistencies. Data entry
and cleaning was completed during
May–June 2018 and analysis of the
clean primary data accomplished
during the month of June 2018.
The data analysis consisted of the
following items:
1. Analyzing secondary data for
comparison: Average performance
of students belonging to Nuapada
district, Odisha across Mathematics,
Science, and Language was compared
with the district average drawn from
the ASER and national average drawn
from the National Achievement
Survey.3
2. Assigning weights to each
question using the baseline student
data: To normalize the responses for
comparisons, weight was assigned
to each question based on the
percentage of correct responses in
the baseline dataset. These weights
were applied to both the baseline and
endline data to arrive at a weighted
score for each student.
3. Assigning weights to each LO
using the baseline student data:
Each LO is tested by a set of
questions; an LO is marked correct
only if a student answered at least
a third of these questions correctly.
Weight was assigned to each LO
3 Annual Survey of Education Report at
www.asercentre.org. National Achievement
Survey conducted by the Ministry of Human
Resource Development, Government of India
at https://mhrd.gov.in/nas.
based on the percentage of correct
responses in the baseline dataset.
These weights were applied to both
the baseline and endline data to
arrive at an LO weighted score for
each student.
4. Statistical analysis: Based on
the LO weighted scores, statistical
tests were applied to establish if the
diﬀerences between score means,
variance, correlation, and regression
etc. of the treatment and ‘control’
groups were significant. Chapter
3 of this report is dedicated to the
statistical methods applied to arrive
at the results.
5. Comparison of LOs: The LO
weighted scores were compared
across:
a. baseline and endline data,
b. treatment and control data,
and
c. hub and spoke datasets
6. Qualitative assessment
based on FGDs and IDIs: FGD and
IDI transcripts were analyzed and
summary sheets were prepared
based on various themes such as,
benefits accrued from LRC, LOs
of students, incidence of dropout,
parent engagement with children’s
education, feedback on training
imparted to LRC facilitators and
teachers, and implementation
challenges and sustainability of
the LRC.
7. Integration of information:
All the data sources—secondary,
primary and qualitative—were
integrated to present a coherent
picture for each outcome indicator.

METHODOLOGY 21
Lambodar Bag, a resident of Fursund village of Nuapada district,
is a landless laborer. Prior to 2013, his three young children would
accompany him and his wife to the brick kilns of Andhra Pradesh
annually, dropping out of school during the migration period. In
2013, when they learned about LAMP they approached the teacher
for enrolling their children in the LRC.
The following season, Lambodar and his wife migrated to the brick
kilns, leaving their children behind with a relative. They hoped that
the LRC support would help the children to cope with studies in
school. Their son Hemant was in Grade V at the time. He has now
completed schooling and is studying at Sinapali College.
The daughters are younger and continuing their schooling.
Jyoti has passed Grade VIII scoring 70% and Sandhya has cleared
Grade VII. LAMP’s LRC classes have helped in bridging their
learning gaps and reaching learning levels appropriate for their
respective grades.
STAKEHOLDER: PARENT
Higher Transition
Rates
II
If my children
had not enrolled in the
LAMP in 2013, they may
have been migrating with
us even today. I will always
remain thankful to LAMP
for its support to my
children as well as many
others.
—Lambodar Bag
Girls from Brahmaniguda village in Nuapada, Odisha, who have transitioned to
High School. Their higher education was covered through scholarships under LAMP
Photo Courtesy: Prashant Panjiar
“
”

3
Statistical Methods
Experimental or quasiexperimental
methods are used to observe and
quantify changes in outcomes of
interest, before and after a program
is introduced, through baseline and
endline surveys and applying statistical
tests to determine whether any
changes observed between a ‘control’
and ‘project’ group are significant.
In the context of LRC, the diﬀerence found between learning achievements of
baseline and endline respondents was significant enough to indicate that it was
unlikely that the results occurred by chance and the diﬀerence found in the sample
probably also existed in the larger student population.

For evaluaƟon of the LRC
intervenƟon 2017–18, the
baseline was compared
with the endline to gauge
the change pre- and post-
intervenƟon. Comparison
was also made across hub,
spoke, and control schools
to understand the impact
under LRC.
“ Experimental or quasi-
experimental methods are used to
observe and quantify changes in
outcomes of interest, before and after
a program is introduced, through
baseline and endline surveys and
applying statistical tests to determine
whether any changes observed
between a ‘control’ and ‘project/
treatment’ group are significant.
In order to attribute impact by
comparing changes against the
‘counterfactual’, an effort is made to
measure the impact in the absence
of the intervention. By surveying a
statistically similar control group along
with the project group, comparisons
can be made not only ‘before and
after’ but also ‘with and without’ the
intervention. This is the essential logic
of the before–after and after–after:
the difference in the indicator of
interest between the project and
control group, before and after the
intervention is introduced is termed as
‘attributable’ impact.
For evaluation of the LRC
intervention 2017–18, the baseline was
compared with the endline to gauge
the change pre- and post-intervention.
Comparison was also made across
hub, spoke, and control schools to
understand the impact under LRC. To
establish the significance of results,
statistical tests were conducted.
3.1 Assigning Weights to
QuesƟons
Most examinations include a mix
of questions of low, medium, and high
difficulty and this partly influences the
student’s performance. Questionnaire
designing thus involves identification
of questions and their respective
difficulty level. Based on the selection
of questions appropriate marks are
assigned to each. Because the LRC
baseline questionnaire for each
subject had equal weightage to each
question, weights were assigned
to each question post-endline data
collection.
Weights were calculated based
on the percentage of students
correctly answering the question
and a Z score was calculated to
standardize the raw scores. These
weights were applied to all the
datasets for comparison.
3.2 Assigning Weights to LOs
Learning outcomes are
assessment standards indicating the
levels of learning that children are
expected to achieve for that class.
These outcomes are used as check
points to assess learning at different
points of time.1 For LRC evaluation,
LOs for each class and subject were
drawn from a document prepared by
the National Council of Educational
Research and Training.2
Each LO, therefore, was tested by
a set of questions. An LO was marked
correct only if a student answered
at least a third of the questions
correctly. Weight was assigned to
each LO based on the percentage
of correct responses in the baseline
dataset. These weights were applied
1 See at: http://mhrd.gov.in/draft-
learning-outcomes
2 See at: http://mhrd.gov.in/sites/
upload_ files/mhrd/files/Learning_
outcomes.pdf

to both the baseline and endline data
to arrive at an LO weighted score for
each student.
3.3 Comparing Standardized
Scores of Students
3.3.1 The t-Test
The t-test is a type of inferential
statistic which is used to determine
whether there is a significant
difference between the means of two
groups such as learning achievements
of baseline and endline respondents
in the case of LRC. The t-test indicates
with a degree of confidence if the
difference obtained between the
means of the sample groups is too
great to be a chance event. In the
context of LRC, the difference found
between learning achievements of
baseline and endline respondents was
significant enough (p-value<=0.05)
to indicate that it was unlikely that
the results occurred by chance and
the difference found in the sample
probably also existed in the larger
student population.
3.3.2 Analysis of Variance
Analysis of Variance (ANOVA) is
used to determine whether there
is a significant difference between
the means of more than two groups
at the same time. The test was
applied to determine if there is a
significant difference in learning
achievements of students belonging
to hub, spoke, and control schools.
In the context of LRC, the difference
found across learning achievements
of students belonging to hub, spoke,
and control schools in the sample
were significant (p-value<=0.05).
This signified that it was unlikely that
the results occurred by chance and
the difference found in the sample
probably also existed in the larger
student population.

STATISTICAL METHODS 27
Rohit Rout was the Sarpanch of the Boirbadhi Gram Panchayat in 2017.
Around 51% of the population of Boirbadhi belonged to ST, 45% to
OBCs and the rest were SCs. At the time, the Panchayat authorities
were more concerned with issues like roads, sanitation, and drinking
water, rather than education. However, children studying in LAMP’s
LRC approached the Sarpanch for a discussion on his plans to improve
education in the village.
This encouraged Rout to call for an SMC meeting and form an Education
Standing Committee slated to meet every quarter to plan and
implement initiatives for school infrastructure improvement, playground
development, sports activities, and management of a seasonal hostel
for children from migrating families. The chairpersons of SMCs, head
masters of schools, ward members and representatives from ICDS,
Health Department and Panchayati Raj Department were members of
this committee.
Under Rout’s leadership, the Panchayat invested more than INR 3 million
on school infrastructure and electricity. In the current year, the
Panchayat has allocated more than INR 2 million for continuing the work
in government schools.
STAKEHOLDER: SARPANCH
Unlocking Government
Resources
III
LAMP trained
Panchayat members to know
their rights and ideate on ways
to improve education in the
villages. I was also fortunate
to be part of exposure visits to
LRCs and other model schools.
The SMCs presented the needs
of the schools and I strategized
to invest more panchayat funds
on education.
—Rohit Rout
“
”
Children of an LRC conduct an interview with the Sarpanch in preparation for a Children’s
Seminar in Magudpani village near Nuapada, Odisha.
Photo Courtesy: Prashant Panjiar.

4
Results of the Learner’s
Achievement Tests
Students of Class III–VIII were evaluated
on the learning outcomes appropriate
for their level, ranging from simple
concepts such as identification
of shapes and counting to more
complex concepts related to financial
calculations, decimals, fractions, factors
and multiples.
Learning objectives identify what the learner should know and be able
to do (reliably demonstrate competency in) by the end of a course or
program. Learning outcomes mirror learning objectives and refer to
observable and measurable achievements (learner knows and can
demonstrate) in terms of knowledge, skills, and attitudes (KSA).

Research instruments
designed for LRC consisted
of quesƟons that tested
the KSA learning domains
in MathemaƟcs, Science,
Language (Odia),
and School Readiness
Instrument.
“ Each LRC is supported by trained
facilitators in the hub schools and
trained school teachers in the spoke
schools. It was observed during field
visits that each LRC was equipped with
a complete range of study materials
including pattern blocks, fraction
kits (for Mathematics), Science and
electronics kits, dictionaries, globes,
atlases, maps, charts, sports, art and
craft material. Besides the study
material, each LRC had a technology-
aided classroom (a computer and a
camera) and provided an enabling
environment for reducing learning
deficit and improving LOs.
Learning objectives identify
what the learner should know and
be able to do (reliably demonstrate
competency in) by the end of a course
or program.1 Learning outcomes
mirror learning objectives and refer
to observable and measurable
achievements (learner knows and can
demonstrate) in terms of knowledge,
skills, and attitudes (KSA). Research
instruments designed for LRC
consisted of questions that tested the
KSA learning domains in Mathematics,
Science, Language (Odia), and School
Readiness Instrument.
4.1 Learning Outcomes:
Mathema cs
4.1.1 Teaching–Learning Material
for Mathema cs
Each LRC was provided with
TLM such as Ganitmala, Rangometry,
Zodostra, fraction kit, number ball,
maan card, balance (brass), volume
measuring set, number catcher,
number tilli (matchsticks),fraction
dice, number dice, yellow and blue
clips, and story props such as tota
(parrot), along with geometry box,
textbooks, and workbooks. Besides
the classes conducted by the LRC
facilitator, students are supported by
these TLMs which are available for
use and practice, thus incorporating
a culture of self-learning.
4.1.2 Learning Outcomes
Iden fied for Mathema cs
As mentioned in the earlier
chapters, students of Class III–VIII
were evaluated on the LOs
appropriate for their level, ranging
from simple concepts such as
identification of shapes and counting
to more complex concepts related
to financial calculations, decimals,
fractions, factors, and multiples
(Annex Table A4.1).
4.1.3 Mathema cs Learning
Achievements: Endline
versus Baseline
Learning achievements of post-
intervention students improved
across all Mathematics LOs as
compared to the baseline (Annex
Table A4.2, Figure 4.1). On an average
the endline scores were almost
thrice (2.89 times) the baseline and
the improvement in the scores
classes barring Class III (which did
improve but not significantly). This
was probably because Class III
1 See at: http://liad.gbrownc.on.ca/
programs/InsAdult/currlo.htm

RESULTS OF THE LEARNER’S ACHIEVEMENT TESTS 31
Table 4.1: The Mathema cs LO in Which Students Showed Greatest Improvement (by Class): Endline versus Baseline
Class III Financial calculations: multiplication
Class IV Counting
Class V Financial calculations: addition, subtraction, multiplication, and division
Class VI Exploring patterns in numbers while doing various operations
Class VII Working with fractions
Class VIII Financial calculations: group distribution, subtraction, multiplication, and division
students had been exposed to the LRC
intervention for only one year.
According to ASER 2016 for
Nuapada, Odisha, only 19.7% of
students (Class III to V) could
do at least subtraction.2 In
comparison, 97.5% of Class III
students studying in LRCs could
do both addition and subtraction.
Across all LOs, a higher percentage
of students achieved each LO in the
endline study than the baseline. The
specific LO in which endline students
of each class performed the best (as
compared to baseline) is listed in
Table 4.1.
Achievements: Treatment versus
Control Group
Students studying in LRCs had
smaller learning deficit as compared
to students of control schools.
Average scores of LRC students were
1.22 times that of non-LRC students.
The diﬀerence in mean scores was
statistically significant for all classes
(Annex Table A4.3, Figure 4.2).
Across all LOs, a higher
percentage of students achieved
each LO in the endline study than
the control group. The specific LO in
which endline students of each class
performed the best (as compared to
control group) is listed in Table 4.2.
Figure 4.1: Weighted Mean Mathema cs LO Scores: Endline versus Baseline
Source: Study team analysis
Baseline Endline
1.4
3.17
2.36
6.61
2.79
6.06
1.83
5.87
1.93
6.71
1.38
4.73
2 http://img.asercentre.org/docs/Publications/ASER%20Reports/ASER%202016/District%20Estimates/odisha.pdf

‘AIF supports us with handholding.
It has provided a lot of facilities
to students. The LRC offers a
pleasant environment for students
to learn with the help of TLMs and
activity-based sessions.’
—Government teacher at
a spoke school
Achievements: Hub versus Spoke
By the design of the LRC
intervention, students in the hub
school received more intensive
guidance than spoke school students.
Each hub school had a dedicated
LRC facilitator who taught students
of classes III to VIII before and after
school hours. Since each LRC facilitator
was associated with 5–6 spoke
schools, he/she could visit each spoke
school only once a week. Teachers
of classes I and II of spoke schools
were trained under the LRC facilitator
to bridge the learning deficit of
students. These students also
approached the LRC facilitator during
his/her visit for clearing doubts.
While the LRC facilitators tried their
best to ensure that students of
the spoke schools attended school
regularly and performed better in all
the subjects, during the study, it was
anticipated that hub schools would
perform better than spoke schools, a
conclusion that was substantiated by
the statistical results as well.
Students studying in hub schools
had lower learning deficit than spoke
school students. Average scores of
hub students were 1.24 times that
of spoke students. The diﬀerence
in mean scores was statistically
significant for all classes (Annex
Table A4.4, Figure 4.3).
Figure 4.2 Weighted Mean Mathematics LO Scores: Treatment versus Control Group
3.17
6.61
6.06 5.87
6.71
4.73
2.55
5.24
4.74 4.96 5.05
4.48
Treatment Control
Table 4.2: The Mathema cs LO in Which Students Showed Greatest Improvement (by Class): Treatment versus
Control Group
Class III Daily tasks: days, time, and recognizing coins
Class IV Identifying shapes
Class V Daily tasks: studying the calendar, days, and time
Class VI Financial calculations: addition, subtraction, multiplication, division
Class VII Daily life situations related to unit, distance, weight, currency, time, calendar involving four basic
arithmetic operations, and understanding directions
Class VIII Applying HCF, LCM in a given situation

Overall, all hub schools had higher
LO scores as compared to spokes
with the exception of the LO ‘counting
numbers’ in Class III in which the
spoke students performed better than
the hub students.
The specific LO in which hub
students of each class performed the
best (as compared to spoke school
students) is listed in Table 4.3.
4.2 Learning Outcomes: Science
for Science
To make Science more hands-on,
each LRC is provided with TLMs such
as science kits, experiment glass,
maps (district, state, and world), atlas,
science dictionary, internet, laptop
and projector, locally available
resources, books for all classes,
magnet, electric bulb, wire, ball,
bottles, dust bin, and first aid box.
Besides the classes conducted by
the LRC facilitator, students are
supported by these TLMs which are
available for use and practice, thus
incorporating a culture of self-
learning.
4.2.2 Learning Outcomes Iden fied
for Science
Students of classes III–VIII were
evaluated on the LOs appropriate
for their level ranging from simple
concepts such as parts of the human
body to more complex concepts
related to shapes, volumes, weights,
and mixtures (Annex Table A4.5).
Table 4.3: The Mathema cs LO in Which Students Showed Greatest Improvement (by Class): Hub versus Spoke
Class III Financial calculations: multiplication
Class IV Working with 2-digit numbers: addition, subtraction
Class V Financial calculations: addition, subtraction, division, multiplication
Class VI Financial calculations: addition, subtraction, division, multiplication
Class VII Identify and count 2D and 3D shapes
Class VIII Financial calculations: group distribution, subtraction, division, and multiplication
Figure 4.3 Weighted Mean Mathematics LO Scores: Hub versus Spoke
3.32
7.04
6.32
6.95 6.95
5.02
2.83
5.24 5.06
5.6 5.6
3.43
Hub Spoke

4.2.3 Science Learning
Achievements: Endline versus
Baseline
Performance of students, post–
intervention, improved across all LOs
in Science. LRC students reported
that they enjoyed coming to class and
conducting the experiments with
the science kits independently. It
was an engaging process wherein
students assembled into groups,
conducted the experiments, and
discussed the outcomes. Instant
results such as change of color,
lighting up of a bulb, or emission
of a sound excited them. In the
words of a student, ‘I repeat this
activity as many times as I need
to. Practice helps me learn at
each step’.
On an average the endline
scores were more than double
(2.63 times) the baseline and the
improvement in the scores was
statistically significant for all
classes.
On an average the endline scores
were more than double (2.63 times)
the baseline and the improvement in
the scores was statistically significant
for all classes (Annex Table A4.6,
Figure 4.4).
Figure 4.4 Weighted Mean Science LO Scores: Endline versus Baseline
2.66
1.41
0.53
1.05 1.33
4.21
5.92
4.72
1.51 1.68
4.26
10.97
Baseline Endline
We play together and learn
new things at the LRC. It
enables us with pracƟcal
learning. Subjects that are
tough, like Science and
Math, have now become
easy to understand.
— Students at a hub school
during the FGD
“

According to National
Achievement Survey (NAS)
2017 for Nuapada about 63.18%
students of Class III in the district
could identify the relationships
with and among family members.3
In comparison, 81.9% of LRC
students in Class III answered
questions related to the same
LO correctly.
Similarly, according to NAS the
district average for Class VIII
students who could explain
process and phenomena was
41.07%.4 In comparison, 84.5% of
Class VIII LRC students answered
questions related to the same
LO correctly.
The specific LO in which endline
students of each class performed
the best (as compared to baseline) is
listed in Table 4.4.
Control Group
to students of control schools.
Average scores of LRC students were
1.35 times the average scores of
non-intervention students. The mean
diﬀerence was statistically significant
Figure 4.5).
Table 4.4: The Science LO in Which Students Showed Greatest Improvement (by Class): Endline versus Baseline
Class III Identifying simple observable features in human body
Class IV Identifying vegetables, fruits, shrubs, trees
Class V Explaining unusual features of animals and their responses
Class VI Applying scientific concepts to day-to-day life
Class VII Measuring and calculating
Class VIII Diﬀerentiating materials and organisms on the basis of their properties
Figure 4.5 Weighted Mean Science LO Scores: Treatment versus Control Group
5.92
4.72
1.51 1.68
4.26
10.97
4
3.5
0.99 1.41
3.08
8.54
Treatment Control
3 http://www.ncert.nic.in/programmes/NAS/DRC.html

Table 4.5: The Science LO in Which Students Showed Greatest Improvement (by Class): Treatment versus Control Group
Class III Identifying occupations
Class IV Voicing opinions on issues observed/experienced in family/neighborhood
Class V Recording observations in an organized manner
Class VI Applying scientific concepts to day-to-day life
Class VII Measuring and calculating
Class VIII Conducting simple investigations to seek answers to queries
Figure 4.6 Weighted Mean Science LO Scores: Hub versus Spoke
6.58
4.87
1.59 1.71
4.45
11.34
4.19 4.17
1.2 1.56
3.27
9.24
Hub Spoke
Across all LOs, a higher percentage
of students achieved each LO in the
treatment group than the control
group. The specific LO in which
treatment students of each class
performed the best (as compared to
control group) is listed in Table 4.5.
Achievements: Hub versus Spoke
had lower learning deficit than
spoke school students. Average scores
of hub students were 1.29 times
that of spoke students. The mean
Figure 4.6).
Overall, all hub schools reported
higher LO scores as compared to
spokes with the exception of the
LO ‘Recording observations about
the neighborhood and identifying
India’s festivals/ national bird/flower/
animal’ in Class IV in which the spoke
students performed better than the
hub students.
The specific LO in which hub
‘The LRC teacher uses TLMs for
Science and Math. He has taught
us topics like matter, water, light,
and human body in Science. We
understand these concepts better
when the LRC teacher visits our
school.’
- Students at a spoke school
during the FGD

4.3 Learning Outcomes:
Language (Odia)
for Language (Odia)
Teaching–learning material for
language learning (Odia) includes
drawings, library books, textbooks,
guidelines for writing essays and
letters, and dictionaries. These TLMs
are available on the hub school
premises for students to achieve
fluency in Odia and expand their
vocabulary from the resource
material.
for Language (Odia)
Students of classes III–VIII were
evaluated on the LOs appropriate
for their level ranging from basic
oral competency to more complex
comprehension of stories and poems,
sequence of ideas and events (Annex
Table A4.9).
4.3.3 Language (Odia) Learning
Baseline
Every language evolves through
history based on the influence of
culture, geography, neighbors, foreign
incursions, migration, trade relations
etc. Because Nuapada shares
borders with the Hindi-speaking (and
mainly tribal) state of Chhattisgarh,
students pick up local dialects of
both Odia and Hindi. The village-level
dialects have certain peculiarities
and are frequently difficult to
comprehend because oftentimes,
the same words convey entirely
different things or the same words
may be spoken in a completely
different way or context. Intervention
in Language (formal Odia) is thus
very relevant to students of Nuapada
district.
Overall, the students in the
endline survey performed better
than at the baseline. On an average
the endline scores were more than
double (2.64 times) the baseline and
the improvement in the scores was
statistically significant for all classes
(Annex Table A4.10, Figure 4.7).
students of each class performed
the best (as compared to baseline) is
listed in Table 4.7.
4.3.4 Language (Odia) Learning
Control Group
Table 4.6: The Science LO in Which Students Showed Greatest Improvement (by Class): Hub versus Spoke
Class III Identifying occupations
Class IV Voicing opinions on issues observed/experienced in family/neighborhood
Class V Explaining unusual features of animals & their responses
Class VI Classifying materials, organisms, and processes based on observable properties
Class VII Identifying materials and organisms on the basis of observable features
Class VIII Applying scientific concepts in day-to-day life
‘Earlier, students were
unable to understand
Odia as in this village,
the language spoken
is Chhattisgarhi/
Laria. And now, they
read well, write well,
speak well in Odia
and answer any
question in Odia
freely.’
—LRC facilitator,
Pipalchhendi village,
Nuapada block

to students of control schools. Average
scores of LRC students were 1.3 times
that of non-LRC students. The mean
for classes III and VII but not so for
the others (Annex Table A4.11, Figure
4.8). This does not mean that the LRC
intervention is not making an impact
on the Odia competence of classes
IV, V, VI and VIII. The mean diﬀerence
may be statistically insignificant for
these classes because the treatment
group includes both hub and spoke
children. It is possible that variance
between only hub and control children
would be much higher, as intensity
of engagement in LRCs is much more
than that of the spokes.
best (as compared to control group)
is listed in Table 4.8.
4.3.5 Language (Odia)
Learning Achievements: Hub
versus Spoke
had lower learning deficit than spoke
school students. Average scores of
hub students were 1.19 times that
of spoke students. The mean
Figure 4.9).
Table 4.7: The Language (Odia) LO in Which Students Showed
Greatest Improvement (by Class): Endline versus Baseline
Class III Correct the spelling/sentence
Class IV Infer meaning of word
Class V Correct the spelling/sentence
Class VI Synonyms and antonyms
Class VII Infer meaning of word
Class VIII Correct the spelling/sentence
Figure 4.7: Weighted Mean Language (Odia) LO Scores: Endline versus Baseline
0.44 0.81
2.7
1.91
2.61
4.69
1.03
2.66
7.65
5.91
4.48
11.89
Baseline Endline
According to National Achievement
Survey (NAS) 2017 for Nuapada about
50.42% students of Class VIII in the
district could read and understand
textual/non-textual material in Odia and
identify the details, characters, main
idea, and sequence of ideas and events.5
In comparison, 93.4% of LRC students in
Class VIII answered questions related to
the same LO correctly.

Table 4.8: The Language (Odia) LO in Which Students Showed Greatest Improvement (by Class): Treatment versus
Control Group
Class IV Infer meaning of word
Class V Correct the spelling/sentence
Class VI Infer meaning of word
Class VII Identify details, characters, main ideas, and sequence of ideas and events
Class VIII Infer meaning of word
Figure 4.8: Weighted Mean Language (Odia) LO Scores: Treatment versus Control Group
1.03
2.66
7.65
5.91
4.48
11.89
0.83
2.12
5.63
3.93 3.73
9.28
Treatment Control
Figure 4.9: Weighted Mean Language (Odia) LO Scores: Hub versus Spoke
1.07
2.74
7.95
6.13
4.58
12.33
0.94
2.39
6.49
4.93
3.93
9.89
Hub Spoke
Overall, all hub schools reported
higher LO scores as compared to
spokes. The specific LO in which hub
4.4 Learning Outcomes: School
Readiness
School readiness refers to
the skills and concepts which if

developed well in children in the
early pre-school years help them to
enhance their social competence,
adjust better in school, and learn the
skills of literacy and numeracy more
eﬀectively and in a more sustained
manner.6 To prepare students for
higher grades in school, art plays
an important role, and is thus
incorporated into LRCs. Material
available for students includes colors,
glue, A4 blank sheets, drawing sheets,
markers, sketch pens, newspapers,
butter paper, colored paper, chalk,
bucket, jug, mirror, brush, camera,
match box, thermocol, knife, scissors,
plaster of Paris, foam, and locally
available resources.
for School Readiness
Students of classes I and II were
evaluated on the LOs identified for
school readiness ranging from
pre-number and space concepts to
pattern making, object matching,
comparison of greater or lesser
values etc. (Annex Table A4.13).
4.4.2 School Readiness Learning
Baseline
Learning achievements of
students, post–intervention,
improved across all learning
outcomes in school readiness. LRC
students were more school-ready
than at the baseline. ‘Ready children’
transition to higher classes more
easily. On an average, LAT scores
of LRC students were more than
double (2.12 times) the scores at
the baseline. The mean diﬀerence
of these scores was also statistically
significant for classes I and II (Annex
Table A4.14, Figure 4.10).
Learning achievements of
class I and II students improved
considerably post-LRC with a much
higher percentage of students
answering LAT questions correctly in
the endline than the baseline survey.
For instance, for the LO ‘Pre-number
concept’, percentage of students
who gave correct responses at the
baseline trailed behind the endline
Table 4.9: The Language (Odia) LO in Which Students Showed Greatest Improvement (by Class): Hub versus Spoke
Class IV Synonyms and antonyms
Class V Synonyms and antonyms
Class VI Infer meaning of word
Class VII Identify details, characters, main ideas, and sequence of ideas and events
Class VIII Correct the spelling/sentence
6 See at: http://siteresources.worldbank.
org/INTSOUTHASIA/Resources/ECE_Toolkit.
pdf

respondents by 75 percentage points
(Annex Table A4.15).
4.4.3 School Readiness Learning
Control Group
Students who had undergone
LRC intervention were somewhat
better enabled to transition to higher
grades in school as compared to non-
intervention students. Average school
readiness score of LRC students was
1.03 times that of control students
implying that the school readiness of
control group was quite comparable
to the treatment group.
The mean diﬀerence of the scores
was statistically significant (Annex
Table A4.16).
4.5 Understanding the Results
The findings of this study are
relevant to understand how LAMP
is helping to improve the quality of
learning through LEP in the Nuapada
district of Odisha. The study indicated
that the focus on teaching Language,
Science, and Mathematics along with
the government school teacher
training, school governance, and
community engagement provide
an eﬀective solution to addressing
learning needs of children in
resource deficient schools/
geographies. These findings are:
• School readiness: Students
of classes I and II are today more
school ready than they were before
the intervention was initiated. The
SRI helped in preparing the children
gain social competence and develop
basic numeracy and literacy skills.
Such skills and competencies will
certainly help them cope with the
academic demands of higher
grades.
• Learning outcomes in
Language: Statistically significant
score improvement was seen
post-intervention (endline versus
baseline) at all levels of school
children with respect to language
competency LOs (mother tongue,
Odia, in this specific context). Also,
when compared with the control
group, the treatment group showed
statistically significant improvement
in mean scores across all levels.
Figure 4.10: Weighted Mean School Readiness LO Scores: Endline versus Baseline
8.1
17.2
Classes I & II
Baseline Endline

Mathematics: Improved outcomes
were witnessed in comparison to
the baseline. These were evident at
all levels and were also statistically
significant.
Science: Students receiving inputs
from the LRCs under LEP gained
significantly in comparison to baseline
indicators. Endline scores were seen
to be more than double the baseline
in Science.
Findings also revealed that
LRC students performed
significantly better than non-
intervention students across all
classes and subjects. This shows
that employing strategies such as
child-centric pedagogy, contextual
and activity-based experiential
learning, aided with interactive TLMs,
and innovative technology can lead
to improvement in learning levels
even among children who have
learning gaps.

Sudarsan Panda is Headmaster of Dudkibahal School in Boden
block since 2012 where LAMP started an LRC in 2015. Apart
from providing after-school support to LRC children, LAMP has
trained SMC members to monitor seasonal hostels and created
awareness in the community on the Right to Education Act, 2009.
Over 90% households in Dudkibahal survive below the poverty
line and about 30% migrate annually in search of employment
opportunities. Children of migrating families stay either in a
seasonal hostel or with caregivers and they are either in high
school or college. The LRCs and seasonal hostels have truly helped
in improving education in these villages and providing a bright
future for the children.
LAMP trainings and exposure visits of SMCs to LRCs have helped
to improve the quality of government schools. Moreover, the day-
to-day engagement of SMCs with school governance and quality
has also improved. Last year LAMP trained SMCs and other
community members to conduct community-led learning audit.
SMC members presented the outcomes of the audit at public
meetings. SMCs now have the capacity to make quality School
Development Plans (SDPs). Because our last few SDPs were
approved, the school got regular power and water supply, toilets,
additional classrooms, and teachers.
STAKEHOLDER: HEADMASTER
Improved Enrolment and
Attendance
IV
In Nuapada district,
more than 10,000 households
migrate to other states in
search of better employment
opportunities. Before LAMP,
there was no awareness on
the significance of education
among migrant parents.
School dropout rates were
very high and almost all
migrant children travelled with
their parents to worksites. Post
LAMP, all children (6–14 years)
are enrolled in schools and
school attendance is more than
90%.
—Sudarsan Panda
Headmaster Sudarsan Panda standing in front of his school in Dudkibahal village,
Nuapada, Odisha.
Photo Courtesy: Lokadrusti
“
”

5
Impact of LRCs: Results
from the Quantitative
and Qualitative Surveys
The quantitative survey based on
LATs and structured interviews was
augmented with qualitative research
based on in-depth interviews (IDIs) and
focus group discussions (FGDs) across
key stakeholders.
The qualitative study aimed to elicit stakeholder suggestions on areas
of improvement for the intervention.

IMPACT OF LRCs: RESULTS FROM THE QUANTITATIVE AND QUALITATIVE SURVEYS 45
The research revealed
that the SMC members
were aware of their roles
and responsibiliƟes, they
discussed performance and
progress of school children
during SMC meeƟngs;
they focused on the LRC
intervenƟon, and felt that
it helped retain children in
school.
“
As outlined in Chapter 2, the
quantitative survey based on LATs and
structured interviews was augmented
with qualitative research based on
in-depth interviews (IDIs) and focus
group discussions (FGDs) across key
stakeholders to evaluate the impact
of LRC on community engagement,
stakeholder engagement,
teacher training, and pedagogical
eﬀectiveness. The qualitative study
also aimed to elicit stakeholder
suggestions on areas of improvement
for the intervention.
5.1 Perspec ve of School
Management Commi ees
School management
committees (SMCs) are composed of
representatives from among parents,
students, teachers, school authorities,
and the community. The Right to
Education Act 2009 envisions the SMC
as the basic unit of a decentralized
model of governance with active
involvement of parents in the
school’s functioning. Active parental
participation has the potential to
improve the eﬃciency of a school as
parents have the highest incentive to
demand better quality of education
for their children.1 As part of their
job role, LRC facilitators hold monthly
meetings with SMCs and share
insights and feedback with the parents
and the community in a bid to keep
1 A. Banerjee, R. Banerji, E. Duflo, R.
Glennerster, S. Khemani (2010), ‘Pitfalls of
Participatory Programs: Evidence from a
Randomized Evaluation in Education in India’,
American Economic Journal: Economic Policy,
2:1, pp: 1–30. https://economics.mit.edu/
files/3117. Accessed on 23 May 2019.
them engaged and clued in on the
quality of education the children are
receiving.
As part of this study, 49 SMC
members were interviewed across
the five sample intervention blocks
of Nuapada using structured
questionnaires. The interviewees
included anganwadi workers
(AWWs), parents, SMC presidential
members, students, sarpanch/ward
members.
The research revealed that the
SMC members were aware of their
roles and responsibilities, they
discussed performance and progress
of school children during SMC
meetings; they focused on the LRC
intervention, and felt that it helped
retain children in school (Figure 5.1).
The community was clearly deeply
involved with the education of its
children.
Students who were members
of the SMC conveyed that they had
the freedom and opportunity to
voice their opinions and participate
in SMC meetings. Other children
of the school approached them
freely with issues and concerns

SMC chairperson Saibani Majhi presenting to the community, the performance chart of children in the
LRC at Pipelchhendi village, Nuapada, Odisha.

that needed to be raised in such
meetings. An enabling environment
on school premises, free of fear and
apprehension had instilled curiosity
in students and increased their
confidence.
5.2 Perspec ve of Head Teachers
of Hub Schools
The five head teachers of the
hub schools shared the following
observations during IDIs:
1. Improvement in learning
outcomes: LRC students fared much
better in terms of learning outcomes
when compared to non-LRC students.
The exposure to LRC’s pedagogical
methods and TLMs undoubtedly led
to this change.
2. Reduction in absenteeism:
Absenteeism and incidence of
dropping-out had both come down
with the LRC intervention.
3. Capacity building: The
LRC training of teachers was
regular and eﬀective. It focused
on lesson presentation, classroom
management, use of TLM,
motivation, questions, reinforcement
etc. The quality of training could be
rated as 8 on 10. It would be great
if education experts could impart
future training using TLM separately
designed for each topic.
4. Community engagement:
Interaction with community
members to understand their
perspective is both regular and
deep. With LRC, awareness of
parents regarding the education
Figure 5.1: Summary Responses of SMC Members during In-depth Interviews
0
20
40
60
80
100
Are aware of
roles and
responsibilities as
SMC members
adhering to
RTE Act 2009
91%
85%
75%
67.5%
Have discussed
performance of
children and
progress of
school and
other curricular
activities in past
SMC meetings
Have discussed
the LRC
intervention
in past SMC
meetings
Strongly
believe that
LRC helps in
retaining
children in
school
This is a positive achievement for the LRC wherein the community's responsibility
has extended to being involved in the child's progress at school.
‘Now, Ganiary village
is a 100% student
enrolment and drop-
out-free village.
Not a single child in
the village is out of
school.‘
— Head teacher of
government school,
Ganiary village

of their children has gone up. They
regularly send their children to
school and participate in monthly
meetings.
5.3 Perspec ve of Class I and II
Teachers from Hub and Spoke
Schools
In this section the term ‘teachers’
refers to those who teach classes
I and II in hub and spoke schools.
During IDIs, these teachers shared the
following insights:
1. Capacity building: Support and
handholding by the LRC facilitators
and the TLMs provided have improved
the quality of teaching and class
transactions. It would be great if LRC
experts could continue imparting
training in the future.
2. Community engagement:
Initially it was diﬃcult to convince
the parents about the eﬃcacy of the
LRC teaching methods (through
activities, art work, playing, dancing
etc.). However, as the children
began to show greater inclination to
participate in school-work and their
results improved, the parents too
changed their minds.
‘We greatly value the learning support provided in the
LRC to our students. It engages them in study beyond
school hours, which is fostering a positive attitude towards
learning. I have noticed a great change in the learning
behavior of students. This year, I asked the Xth Class
students about their performance in the Matric exam, and
found all of them to be confident of their grades. Now our
students do very well in music, dance, drawing, debate,
science fairs, competitions at the Block level and win prizes
for our school.’
— Head teacher of a hub school
‘Parents are seen to be leaving the children to continue
schooling while they are migrating for livelihood. Students
are not migrating now because they are doing well in
learning. Midday meal is also motivating them.’
— Teacher of government school
7 out of 10 parents reported
consulting with facilitators about the
management of the LRC.
Parents'
Engagement
Student's
Performance
All the 10 parents strongly felt
that the LRCs should continue
in their villages.
Sustainability
of LRC
All the 10 parents felt that the LRC was
contributing to learning improvement.
Activity-based teaching sessions and TLM
were instrumental in imparting
quality education.
"Without the LRC, the progress
in studies that the children are
making will halt. The LRC
inspires the children
to study well."

5.4 Perspec ve of Parents
During IDIs parents shared their
apprehensions about the impact
of the withdrawal of LRCs from the
village. They were very happy with the
progress of the children and keen for
the initiative to continue.
1. Parents’ engagement: LRC
facilitators engage with parents every
month to discuss management of
the LRC, progress of students, and
corrective course of action. Parents
also arrange for education fairs/ melas
in the village.
2. Students’ performance: The
LRC has made a huge contribution to
learning improvement of students.
Activity-based teaching sessions and
TLMs help impart quality education.
3. Sustainability of LRC: LRCs
should continue to operate so that
students can draw its benefits. If
they are discontinued all the good
work will be lost and the children will
regress to the old state poor learning
attainments and disinterest in school-
work.
5.5 Visits to the LRCs:
Perspec ve of LRC Facilitators
The LRC facilitators are
responsible for preparing schedules
and TLM prior to each class, conduct
classes pre- and post-school hours in
hub schools and visit spoke schools
during school hours to provide
technical assistance to government
school teachers. Facilitators also
meet parents and community
members formally or informally
for awareness building and to
discuss student enrolment, LRC
management, etc.
5.1.1 Way of Teaching
All facilitators have their unique
way to conducting a class. They
create their own recipe based on
the local context and culture, the
moods and vibes of the children,
the locally available material to
repurpose as TLM etc. Besides the
prescribed experiment kits and
TLM, the facilitators use practical
everyday objects to demonstrate
activities.
‘To teach “color” we ask children
to collect different types of
soil and use soil colors to draw
pictures on drawing sheets.’
‘We make dolls using mud and
soil colors. We also prepare
models of traditional farm
instruments by using the
branches of trees.’
‘Every object available in our
surrounding can be used as TLM.’
— LRC facilitators
‘This is a backward region. The education of our children will
be hampered if there are no LRCs. We will explore options
for support from community members to keep the LRCs
running.’
— Parents’ view on whether LRCs should be discontinued
‘The LRC provides TLMs and training through audios, videos,
and storytelling activities. We use these tools in teaching
Classes I and II.’
— Teacher of government school

5.1.2 Self -learning Environment
in LRCs
Each LRC has a Science corner,
Math corner, Language corner, and
a library for the kids. Displays of
diﬀerent activity material and danglers
prepared by LRC facilitators and
students are a common sight. LRC
facilitators are responsible for creating
an environment of self-learning for
students. This is done by adding
books and TLMs in the library space
and encouraging art work and group
work in classroom sessions. The set
up encourages students to go back
to the library, read the study material
and use readily available material to
come up with demonstrations. Each
LRC is full of sketches, craftwork,
demonstrations, and maps. Students
in the remediation class are
encouraged to explore and develop
their curiosity and ask questions.
5.1.3 Feedback from Facilitators
It might be useful to extend the
intervention from Mathematics,
Science, and Language to English and
Computers. Other improvements
could be an increase the number of
library books, more rigorous subject
training for the facilitator, and the
provision of a community hall in
each village so that LRCs always have
space to operate.
5.6 Sustainability of LRCs
Every stakeholder involved in the
LRC program either as a beneficiary
or a contributor understands its

pivotal role in improving students’
performance, enhancing their
confidence, and engaging the
community in the education of its
children. Some of the initiatives taken
up by the community to sustain LRCs
include:
1. Zero rent infrastructure: LRCs
operate from within the government
school premises or local community
halls so there is no need to buy or rent
real estate to run the LRC.
2. Ownership shift to youth:
Each LRC needs a dedicated facilitator
to conduct classes and engage with
the community. LRC involves the
educated youth of the village in
teaching, community engagement,
and organizing of community activities
such as education fairs. This provides
employment opportunities for the
village youth and ensures a longer
association with the program.
3. Community fund: The
community wants LRCs to continue
in the future and they realize the
relevance of fund flows to ensure
smooth functioning. Community
members have started collecting
funds to this end, though there are
no set rules and rates and the funds
raised depend on the community’s
will and capacity to contribute.
5.7 Classroom Observa ons
Apart from attending classes
held by LRC facilitators, regular
classes held at LRCs by government
school teachers were also observed
to understand the classroom
environment. For teachers to
establish a positive, encouraging,
and non-discriminating environment,
they should demonstrate
appropriate competencies that
represent a combination of traits,
Table 5.1: Classroom Observa on: Performance Parameters Studied Across 27 Classrooms (figures indicate number of
classrooms where said parameter was observed and categorized)
Always Sometimes Rarely Never
Teacher–Student Interaction
Teacher follows a prepared lesson plan 20 7 0 0
Teacher attends to all children and responds to the specific learning
needs of children
15 12 0 0
Lesson Delivery
Children frequently ask questions and respond to questions 15 12 0 0
Teacher enjoys the company of children 18 8 1 0
Classroom Environment
There is an indication of positive learning; children are motivated
frequently
20 7 0 0
Teacher promotes or teaches children to initiate and respond to
their peers
15 12 0 0
The child’s work is monitored and corrected by the teachers 16 10 1 0

abilities, knowledge, and skills such
communication skills, patience,
creativity, empathy, etc.2
Classroom observations
covered areas such as the
interaction between teachers and
2 Stojiljković, SnežanaandDjigic, Gordana
and Zlatković, Blagica (2012), ‘Empathy
and Teachers’ Roles’. Procedia - Social and
Behavioral Sciences, 69,pp: 960–66. 10.1016/j.
sbspro.2012.12.021.
students, lesson delivery by
teachers, and environment of the
classroom (Table 5.1). The classes
were found to be interactive and
the teachers seemed motivated to
conduct the classes and respond
to students’ queries properly.
However, there is a need to adjust
teaching methods and strategies to
support emotional, behavioral, and
social development of children from
the marginalized communities.

Nira Hans is an SMC member and resident of Dudkibahal village that
has around 180 households. Of these, 44% belong to ST, 7% to SC
and 48% to OBC communities. The SMC members meet every month
to discuss and take action on issues related to student attendance,
teacher attendance, learning outcomes, teaching–learning materials,
midday meals, school infrastructure, water, and hygiene. LAMP has
been working with the SMC since 2015 and has trained them on the
Right to Education (RTE) Act, roles and responsibilities, seasonal hostel
management and monitoring.
One of Nira’s children, who completed Grade VIII, also studied at the
LAMP LRC. Apart from receiving academic support, LRC children get
hands-on experience and exposure to art work, science experiments,
computers, internet, etc. One of the key changes among school
children, especially LRC children, is that they have become punctual
and dedicated to their studies. There is this renewed interest in
reading and children read even when they are back home.
This school has earned respect in the area for the quality of teaching at
the LRC and attracts children from nearby villages. Migrant parents in
and around the village enrol their children in the Dudkibahal Seasonal
Hostel and feel safe about their decision.
STAKEHOLDER: SMC MEMBER
Child-Migration-Free
Villages
V
In the last two
years, all children of the
primary and upper primary
schools including girls, have
transitioned to high school
and have completed studies
till Grade X. In 2019,
11 children including 7 girls
appeared for and passed
the Grade X board
examinations. Today, the village
is child-labor and child-
migration free.
— Nira Hans
Children from Sanmaheshwar LRC dance on the theme of RTE during the SMC meeting.
“
”

6
Conclusion
Findings of the evaluation study are
of interest to the central and state
governments, educationists, policy
makers, and practitioners as they
demonstrate the fact that continued
access to basic learning resources
and support along with community
participation leads to improvement
in public education provisioning and
learning outcomes.
Education is fundamental to the qualitative development of human
resources of a country.

EducaƟon is fundamental
to the qualitaƟve
development of human
resources of a country. This
is why the Government
of India amended the
ConsƟtuƟon to provide free
and compulsory educaƟon
for all children in the age
group 6 to 14 years, as a
fundamental right.
“ Education is fundamental to the
qualitative development of human
resources of a country. This is why
the Government of India amended
the Constitution to provide free and
compulsory education for all children
in the age group 6 to 14 years, as
a fundamental right. The Right of
Children to Free and Compulsory
Education (RTE) Act, 2009, means that
every child has a right to full-time
elementary education of satisfactory
and equitable quality in a formal
school which satisfies certain
essential norms and standards.
However, the quality elementary
education remains a serious concern
and is being discussed across
India today, with almost all state
governments striving to introduce
several measures to address this
issue.
AIF’s signature education program
LAMP is one such intervention which
works toward creating demand for
quality education in some of the most
underserved regions of India. A
third-party evaluation undertaken
in 2017–18 attempted to assess
the impact LRCs had on children,
communities, and government
schools. Based on a hub-and-spoke
model, the LRCs run in community-
owned spaces/government schools,
are equipped with innovative learning
materials and interactive technology
that address diﬀerent learning
needs of children across grades. The
overall aim of the LRC at the hub is
to demonstrate to spoke schools
as well as other schools in the area,
strategies to improve the quality
of education by bridging learning
deficits, strengthening STEM, improve
transition rates and promote school
completion.1
The LRC is a space where
children across ages and grades
can come and enrich their learning.
An LRC is typically set-up in a
room given by the Panchayat, or
any other space that is not in use.
Some LRCs have also been set-up
within government schools, at the
request, or with the permission, of
the SMC. In one village, Baddarlipada
(Nuapada, Odisha), where the village
infrastructure is really poor, the
local youth built the LRC themselves
with bamboo, mud, and straw.
Wherever LRCs are set up, they
are always zero-rent spaces. The
community also contributes towards
the refurbishment and repairs of the
LRC to make the spaces child-safe
and child-friendly. This demonstrates
that if the community sees benefit
and quality, the members go out of
the way to support education of their
children. LRCs also provide support
to the government school teachers
towards improving learning among
school children.
The LRC (at the hub) is supposed
to be a demonstration model for
quality education delivery aimed
to convince stakeholders including
the community, parents, and the
government of its teaching approach
and methods. The decision on the
number of spoke schools depends
on a variety of factors (distance from
1 STEM is a curriculum based on
the idea of educating students in four
specific disciplines — science, technology,
engineering and mathematics — in an
interdisciplinary and applied approach.

CONCLUSION 59
the hub, size of schools, topography
etc.). The LRC facilitators generally
take a demo session at the LRC hub,
and then visit the nearby spoke
schools during the day. This allows the
intervention to reach out to multiple
schools with the same resources.
The study reveals that LRC
intervention has been successful in
improving the learning outcomes
of students at the endline vis-à-vis
baseline across Math, Science, and
Bhasha (language). The results are
valid for learning outcomes intended
for each class.
Students of hub schools have
performed better than spoke schools
across all classes and subjects. This
is expected, given the intervention
design under which, students of
hub schools receive rigorous inputs
through daily LRC classes conducted
before and after school hours. In
comparison, LRC facilitators only
clarify doubts of spoke school
students of classes III to VIII once a
week.
Findings also reveal that LRC
students have performed significantly
better as compared to non-LRC
students across all classes and
subjects. On similar lines, students
of classes I and II of LRC schools are
more prepared/ ready for school in
comparison to non-LRC students. This
outcome may show larger impact on
the learning achievement in higher
classes as the learning deficit is
bridged early.
Community engagement of
LRCs is deep and sustained. After
the intervention, parents and other
stakeholders associated with the
SMCs were found to be more aware
of their role in the management of
the schools. Community members
including the parents, teachers,
sarpanch, AWWs, and the SMC
president discuss the management
of LRC in monthly meetings
organized with LRC facilitators and
block coordinators. They have a say
in the areas of improvement for
LRCs, organizing education fairs in
the village and in planning the LRC’s
future course of action. Communities
have started maintaining LRC
funds to tide over situations where
shortage of funds disrupts LRC
functioning. Such corpus may
not always be suﬃcient but it is
indeed a great source of motivation
indicating the genuine interest of
the community in the education of
its children. Parents discuss their
child’s progress in school with LRC
facilitators. They firmly believe in the
LRC way of teaching.
The engagement of
teachers in teaching
also increased with
demonstrated renewed
interest in learning
pedagogy.
“

The engagement of teachers
in teaching also increased with
demonstrated renewed interest in
learning pedagogy. Teachers, post
LRC inputs, acquiesced to being
able to better conduct classes than
earlier. Classroom observations
(Table 5.1) indicated that a majority
of the teachers came prepared for
the classroom teaching and were able
to create a positive environment to
motivate children. However, many
teachers were found to be spending
less time than required on classroom
management and organization,
improving participation of children,
providing feedback and indicating
corrective action. If the learning
capability of students as well as
their performance is to be improved
and enhanced then all the factors
of a teacher personality need to be
developed. Government teachers in
the school were found to be motivated
to implement the recommended
teaching–learning process with the
assistance of TLMs provided and
regular training conducted. Teachers
expressed interest in receiving regular
training in the future on specific
topics under subject experts. They
also expressed the desire to have
the LRC way of teaching extended to
other subjects such as English and
Computers.
The study indicated how timely
and age-specific intervention could
enhance the quality of education
and assure long-term sustainability,
improvement in teaching–learning
processes (pedagogy), school
leadership and community
participation.
However, there are some
challenges which need to be
addressed:
1. Shortage of teachers in
government schools, particularly in
rural areas, puts immense pressure
on LRC facilitators as schools expect
them to fill in for missing teachers,
apart from conducting the routine
LRC activities.
2. Engagement with students of
spoke schools is not as deep as that
in the hub school. As a result, the
bridging of learning deficits in spoke
schools is not as comprehensive as in
the hub LRC.
3. The LRC intervention has
created a sort of dependency among
the community given the shortage
of teachers and resources that they
face. While LRCs to a large extent
have been successful in bridging
learning deficit in intervention
schools, there is no clear-cut exit
plan.
Findings of the evaluation study
are of interest to the central and
state governments, educationists,
policy makers, and practitioners
as they demonstrate the fact that
continued access to basic learning
resources and support along with
community participation leads to
improvement in public education
provisioning and learning outcomes.
Investments, if made by the
government and other agencies
(under public–private partnership), in
initiating LRC will help in improving
learning outcomes and expanding
such eﬀorts across locations on a
large scale.

CONCLUSION 61
Seasonal hostels, a concept originally started by AIF and its partner
Lokadrusti in the year 2004 and adopted by the government
in 2008, are now managed by SMCs. In 2018–19, there were 60
seasonal hostels in Nuapada, with an intake of 3,150 students.
AIF through Lokadrusti, provides technical support, migration
mapping, community mobilization, counseling of parents, and
TLMs for students. They also carry out capacity building of SMC
members and caretakers running the hostels.
Hemanta Kumar Bhoi is in charge of the implementation of
inclusive education activities and community mobilization activities
of the Sarva Shiksha Abhiyan (SSA) in Nuapada. During his visits to
LAMP’s LRCs, Hemanta observed an improvement in the learning
levels of students, their engagement with co-curricular activities,
as well as significant personal development, wherein they
appeared as vocal, smart, and cheerful young people. “With
local educated youth recruited as LRC facilitators, there is no
language barrier, and a friendly teaching-learning environment is
ensured”, he adds.
STAKEHOLDER: GOVERNMENT OFFICIAL
Technical Support to
Seasonal Hostels
VI
Nuapada is a
migration sensitive district.
Seasonal migration is linked to
debt cycles and the need for
money to repay debts, which
compels the resident-villagers
into forced labor conditions
and exploitation. This often
results in children at risk of
dropping out from school, and
the only way out is to provide
safe spaces for them to stay
and learn. Seasonal Hostels
and LRCs have helped in
reducing drop-out rates.
— Hemanta Kumar Bhoi
Balmati Majhi (red sari), the cook, and her helper at the
Khamtarai Seasonal Hostel.
“
”

Annexes
Table A 4.1: Learning Outcomes: Mathema cs
Class III Class IV Class V
1. Identifying shapes
2. Counting numbers
3. Addition and subtraction
4. Measuring lengths and
weights
5. Daily tasks: days, time, and
recognizing coins
6. Financial calculation:
multiplication
1. Identifying shapes
2. Counting numbers
3. Working with 2-digit numbers: addition and
subtraction
4. Comparing objects as heavier/lighter, taller/
shorter, and capacity of different containers
5. Daily tasks: Days, time, and recognizing
coins, observing and studying the calendar
6. Financial calculation: addition, subtraction,
multiplication, and division
1. Acquiring understanding about shapes
2. Arithmetic operations in a equation
3. Comparison (>,=,<), increasing/decreasing
order
4. Comparing objects as heavier/lighter,
longer/shorter, and capacity of different
containers
5. Daily tasks: Studying the calendar, days and
time
Class VI Class VII Class VIII
1. Identify 2D and 3D shapes
2. Explore patterns in numbers
while doing various
operations
3. Comparison (more than,
equal to, farthest/closest)
4. Daily life situations related to
unit conversion, weight, time,
calendar involving four basic
arithmetic operations and
understanding directions
5. Financial calculation:
addition, subtraction,
1. Identify and count 2D and 3D shapes
2. Explore numbers while doing various
operations, usage of variables in equations
3. Use unitary method in solving various word
problems
4. Daily life situations related to unit, distance,
weight, currency, time, calendar involving
four basic arithmetic operations and
multiplication, division, chart, graph, page
number referencing
6. Work with fractions
7. Work with terms: equal to, greater than,
most, smallest, and between
1. Solve problem related to angles, sides, corners
of shapes
2. Explore numbers while doing various
operations, usage of variable in equation
3. Recognize pattern in operations, observe
reflection in mirror
4. Daily life situations related to unit, volume
involving four basic arithmetic operations and
5. Financial calculation: group distribution,
subtraction, multiplication, and division
6. Work with fractions and decimals
7. Work with terms: more, maximum, closest,
smallest, even, round-off
8. Apply HCF, LCM in a given situation
Table A4.2: Mean Difference in Mathema cs LO Scores: Endline versus Baseline
No. of students/respondents Average endline score/average
baseline score
Mean difference
p-value
Baseline Endline
Class III 539 408 2.26 1.77 0.40
Class IV 821 639 2.80 4.25 0.00*
Class V 1437 326 2.17 3.27 0.00*
Class VI 1566 817 3.21 4.04 0.00*
Class VII 1621 863 3.48 4.78 0.00*
Class VIII 1475 783 3.43 3.35 0.00*
Note: *Where p-value<=0.05, the mean difference is significant at 95% confidence interval

ANNEXES 63
Table A4.3: Mean Difference in Mathema cs LO Scores: Treatment versus Control Group
No. of students/respondents Average treatment score/
average control group score
Mean difference
p-value
Treatment (hub + spoke) Control
Class III 408 145 1.24 0.62 0.00*
Class IV 639 151 1.26 1.37 0.00*
Class V 326 156 1.28 1.32 0.00*
Class VI 817 156 1.18 0.91 0.00*
Class VII 863 155 1.33 1.66 0.00*
Class VIII 783 148 1.06 0.25 0.00*
Table A4.4: Mean Difference in Mathema cs LO Scores: Hub versus Spoke
No. of students/respondents Average hub score/
average spoke score
Mean difference
p-value
Hub Spoke
Class III 284 124 1.17 0.49 0.00*
Class IV 487 152 1.34 1.80 0.00*
Class V 541 142 1.25 1.26 0.00*
Class VI 671 146 1.24 1.35 0.00*
Class VII 712 151 1.24 1.35 0.00*
Class VIII 642 141 1.46 1.58 0.00*
Table A4.5: Learning Outcomes: Science
1. Identifying simple observable
features of the human body
2. Grouping objects based on
similarities and differences
3. Identifying relationships in the
family
4. Voicing opinions on good/bad
behavior
5. Recording observations,
experiences, information on
objects/activities/places
6. Identifying professions
1. Recording observations regarding the
neighborhood and identifying India’s festivals/
national bird/flower/animal
2. Identifying important days based on historic
significance
3. Identifying simple features of objects
4. Identifying vegetables, fruits, shrubs, trees
5. Explaining the process of producing and
procuring objects of daily needs
6. Voicing opinions on issues observed/
experienced in family/neighborhood
7. Explaining herd/group behavior
8. Identifying relationships in the family
1. Grouping objects and materials
according to various features/
properties
2. Recording observations in an organized
manner
3. Explaining unusual features of animals
and their responses
4. Voicing opinion on issues observed
5. Guessing conditions of phenomena
6. Identifying signs and directions of
different objects
1. Identifying materials and
organisms on the basis of
observable features
2. Explaining processes and
phenomena
3. Classifying materials, organisms,
and processes based on
observable properties
4. Applying learning of scientific
concepts in day-to-day life
1. Identifying materials and organisms on the
basis of observable features
2. Differentiating materials and organisms on the
basis of their properties
3. Explaining processes and phenomena
4. Measuring and calculating
1. Differentiating materials and organisms
on the basis of their properties
2. Explaining processes and phenomena
3. Understanding shapes, volumes,
weights, and mixtures
4. Classifying materials and organisms
based on properties
5. Applying learning of scientific concepts
in day-to-day life
6. Conducting simple investigations to
seek answers to queries

Table A4.6: Mean Difference in Science LO Scores: Endline versus Baseline
No. of students/respondents Average endline score/
average baseline score
Mean difference
p-value
Baseline Endline
Class III 488 404 2.23 3.26 0.00*
Class IV 808 630 3.35 3.31 0.00*
Class V 1460 675 2.85 0.98 0.00*
Class VI 1553 812 1.60 0.63 0.00*
Class VII 1614 855 3.20 2.93 0.00*
Class VIII 1502 772 2.61 6.77 0.00*
Table A4.7: Mean Difference in Science LO Scores: Treatment versus Control Group
average control group
score
Mean difference
p-value
Treatment
(hub + spoke) Control
Class III 404 138 1.48 1.92 0.00*
Class IV 630 151 1.35 1.23 0.00*
Class V 675 155 1.53 0.52 0.00*
Class VI 812 161 1.19 0.27 0.00*
Class VII 855 144 1.38 1.19 0.00*
Class VIII 772 148 1.28 2.43 0.00*
Table A4.8: Mean Difference in Science LO Scores: Hub versus Spoke
No. of students/respondents Average hub score/
average spoke score
Mean difference
p-value
Hub Spoke
Class III 292 112 1.57 2.39 0.00*
Class IV 496 134 1.17 0.70 0.00*
Class V 540 135 1.33 0.40 0.00*
Class VI 677 135 1.10 0.14 0.005*
Class VII 717 138 1.36 1.18 0.00*
Class VIII 637 135 1.23 2.10 0.00*

ANNEXES 65
Table A4.9: Learning Outcomes: Language (Odia)
1. Communicate in own bhasha (oral
competency)
2. Respond to comprehension questions
related to stories and poems orally
and in writing
3. Word meaning
4. Identify main ideas and draw
conclusions
5. Synonyms and antonyms
6. Correct the spelling/sentence
1. Communicate in own bhasha
(oral competency), bird/flower/
animal
2. Respond to comprehension
questions related to stories and
poems orally and in writing
3. Infer meaning of word
competency)
3. Infer meaning of word
competency)
2. Respond to comprehension questions
related to stories and poems orally
and in writing
3. Identify details, characters, main
ideas, and sequence of ideas and
events
4. Infer meaning of words
1. Communicate in own bhasha
(oral competency)
events
competency)
events
Table A4.10: Mean Diﬀerence in Language (Odia) LO Scores: Endline versus Baseline
Mean diﬀerence
p-value
Baseline Endline
Class III 1346 402 2.34 0.59 0.003*
Class IV 1346 623 3.28 1.85 0.012*
Class V 1459 682 2.83 4.95 0.000*
Class VI 1584 821 3.09 4.00 0.000*
Class VII 1637 860 1.72 1.86 0.000*
Class VIII 1481 769 2.54 7.19 0.000*

Table A4.11: Mean Difference in Language (Odia) LO Scores: Treatment versus Control Group
average control group score
Mean difference
p-value
Treatment (hub + spoke) Control
Class III 402 137 1.24 0.20 0.007*
Class IV 623 150 1.25 0.53 0.308
Class V 682 156 1.36 2.03 0.898
Class VI 821 155 1.50 1.98 0.519
Class VII 860 154 1.20 0.75 0.000*
Class VIII 769 150 1.28 2.61 0.110
Table A4.12: Mean Difference in Language (Odia) LO Scores: Hub versus Spoke
No. of students/respondents Average hub score/average
spoke score
Mean difference
p-value
Hub Spoke
Class III 284 118 1.14 0.12 0.001*
Class IV 483 140 1.15 0.34 0.000*
Class V 542 140 1.22 1.47 0.000*
Class VI 673 148 1.24 1.19 0.000*
Class VII 722 138 1.17 0.65 0.000*
Class VIII 629 140 1.25 2.44 0.000*
Table A4.13: Learning Outcomes: School Readiness
Classes I and II
1. Pre-number concepts
2. Space concept
3. Classification of fruits and vegetables
4. Following instructions
5. Number/object matching
6. Pattern making
7. Relative comparisons: number – greater/ lesser
Table A4.14: Mean Difference in School Readiness LO Scores: Endline versus Baseline
Mean difference
p-value
Baseline Endline
Classes I
and II
1377 1176 2.12 9.10 0.006*

ANNEXES 67
Table A4.15: Percentage of Students (Classes I and II)Achieving School Readiness Learning Outcomes (Endline versus
Baseline)
Learning Outcome Percentage of Students Who Answered Correctly
Baseline Endline
1 Pre-number concepts 17.6 93.4
2 Space concept 29.3 89
3 Classification of fruits and vegetables 56.7 90.9
4 Following instructions 34.9 94.9
5 Number/object matching 81.7 91.2
6 Pattern making 28.1 71.3
7 Relative comparisons: number – greater/ lesser 50.3 63.9
Table A4.16: Weighted School Readiness Mean Scores: Treatment versus Control Group
Treatment Control Treatment mean
score/Control mean
score
Mean
diﬀerence
p-value
No. of
respondents
Mean
score
No. of
respondents
Mean
score
Classes I
and II
1176 17.20 306 16.68 1.03 0.52 0.00*

Strengthening Learning Outcomes for Migrant Children: An Evaluation of the Learning and Migration Program (LAMP)

More Related Content

Similar to Strengthening Learning Outcomes for Migrant Children: An Evaluation of the Learning and Migration Program (LAMP) (20)

More from jonesmaryy024 (20)

Recently uploaded (20)

Strengthening Learning Outcomes for Migrant Children: An Evaluation of the Learning and Migration Program (LAMP)