C4D VAC Systematic Review Report

C4D WORKING
PAPER SERIES
Communication for Development

Approaches to Address Violence
Against Children:
A Systematic Review
tle: 3: C4D and Violence Against Children - working paper

tle: 4: C4D and Violence Against Children - full report
Communication for Development (C4D)
Working Paper Series
Communication for Development Approaches

to Address Violence Against Children:
A Systematic Review
Authors: Suruchi Sood and Carmen Cronin
© United Nations Children’s Fund (UNICEF),

New York, 2019
Suggested citation: Sood, S. and C. Cronin,

Communication for Development Approaches to
Address Violence Against Children: A Systematic
Review, UNICEF, New York, 2019.
The designations in this publication do not imply

an opinion on legal status of any country or
territory, or of its authorities, or the delimitation
of frontiers.
Cover photo: © UNICEF/UNI159383/Pirozzi
For further information, contact:
Communication for Development and Child

Protection Sections
Programme Division
United Nations Children’s Fund
3 United Nations Plaza, New York, NY 10017
Email: C4DHQ@unicef.org
Website: www.unicef.org/cbsc
ISBN Number 978-0-9911961-6-6

Communication for Development
Approaches to Address Violence
Against Children:
A Systematic Review
In the State of Palestine, five-

year-old triplets (left-right)
Alma, Maryam and Lareen
have taken shelter
in a mental health clinic,
where their father is a nurse,
in Gaza City.
© UNICEF/UNI167953/D’AKI
FOREWORD
D
espite a clear legal and ethical imperative to protect children, Violence
against Children (VAC) remains pervasive. It affects millions of children
every year, in their homes, communities, schools, workplaces, in detention
centres and childcare institutions, and online. It can have long lasting, and often
lifelong, negative effects. The vast majority of children never speak out about
their experiences and even fewer receive the services they need to recover.
When the global community adopted the Sustainable De- novation and best practice in shifting social norms related to
velopment Goals (SDGs) in 2015, targets were set to eradi- VAC. C4D is indeed integral to UNICEF’s VAC programmatic
cate all forms of violence, abuse, exploitation and trafficking approach under the Strategic Plan 2018-2021. Changing social
against boys and girls, as well as harmful practices by 2030. norms is also embedded in the VAC Theory of Change, as a
There is increasing recognition that VAC places a long-term stand-alone outcome, and as a cross-cutting strategy that influ-
burden on health and social services, undermines invest- ences all other outcomes.
ment and development in other sectors (such as health, early
childhood development, nutrition and education) and con- Even so, the evidence and documentation of C4D best practices
strains economic growth. Conversely, investment in ending to address VAC is limited. There is an urgent need to improve
violence against children and women can accelerate develop- access, dissemination, systematization, and use of data and evi-
ment across all the SDGs. dence on social and behavior change and community engage-
ment to promote protective practices. In recent years, several
Violence against children encompasses efforts have been undertaken to strengthen this. Yet there are
few such efforts focused on VAC.
“all forms of physical or mental
violence, injury and abuse, neglect or This report is part of a package of evidence and tools that
negligent treatment, maltreatment or includes this systematic review of C4D interventions to ad-
dress VAC, an evidence review of randomized controlled
exploitation, including sexual abuse"
(Article 19, United Nations Convention on the trials, a Technical Guidance for C4D programmes address-
Rights of the Child) ing VAC, and training materials related to the Guidance.
This suite of materials is a result of UNICEF’s investments
Transforming social norms that condone or facilitate VAC or in C4D capacity and its collaboration with Drexel Univer-
prevent child victims from accessing support, and reinforcing sity to strengthening the evidence-base on what works for
positive norms that protect children from violence, has been preventing and responding to VAC through C4D. Several
increasingly recognized as crucial by the global community – UNICEF offices are using these resources to develop coun-
it is, for example, one key strategy of the INSPIRE package for try roadmaps and strengthen their strategies to address VAC
ending violence against children (WHO, 2016). through robust C4D programming. We hope that more
UNICEF offices, governments, partner agencies, and others
UNICEF has a critical role to play in ending VAC and achieving working in this area will find this package useful. We hope
the SDGs. It is a leading global agency with the profile, reach, that it will also provide the hard evidence that governments
expertise and mandate to achieve impact at scale. Building on and development agencies need for scaling up efforts to re-
its expertise and technical leadership in Communication for duce violence against children.
Development (C4D), UNICEF is also well placed to lead on in-
RAFAEL OBREGÓN STEPHEN BLIGHT

Chief, Communication for Senior Advisor, Child Protection
Development Programme Division, UNICEF Programme Division, UNICEF
2 Communication for Development Approaches to Address Violence Against Children: A Systematic Review
Contents
Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
List of tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
List of figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1 Executive summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3 Structure of this report. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Manuscript selection and sampling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5 Limitations of this systematic review ..................................................... 20

6 Key findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
SECTION 1: Sampling information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
SECTION 2: Programme design elements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Assessment of conceptual (theoretical) frameworks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Assessment of programme objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Assessment of communication objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Assessment of the level of influence of the interventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Assessment of programme implementing partners. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
SECTION 3: Programme implementation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Assessment of the intended audiences for interventions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Assessment of the strategic communication approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Assessment of communication channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
SECTION 4: Overall programme evaluation processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Assessment of the overall evaluation designs.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Assessment of research methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Assessment of analysis frameworks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Assessment of sampling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Assessment of indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
SECTION 5: Thematic analysis of key results by type of research. . . . . . . . . . . . . . . . . . . . . . 44

Formative research results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Process evaluation results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Impact evaluation results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7 Overall recommendations 62
.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
End notes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Tables
TABLE 1: Key terms for the systematic review.....................................................................................................................................13
TABLE 2: Inclusion and exclusion criteria for the systematic review..............................................17
TABLE 3: Search engines utilized for the systematic review..........................................................................17
TABLE 4: Location of interventions for manuscripts in the

systematic review........................................................................................................................................................................................................ 25
TABLE 5: Conceptual frameworks in the manuscripts included in the

systematic review........................................................................................................................................................................................................ 27
TABLE 6: Implementing agencies and partners for manuscripts in the

systematic review........................................................................................................................................................................................................33
TABLE 7: Intended audiences for manuscripts in the systematic review.............................. 35
TABLE 8: Evaluation design for manuscripts in the systematic review...................................... 38
TABLE 9: Research methods for manuscripts in the systematic review................................. 39
TABLE 10: Analysis methods for manuscripts in systematic review................................................ 39
Figures
FIGURE 1: Overall recommendations based on the systematic review ................................... 11
FIGURE 2: Structure of this report’s findings section .........................................................................................................15
FIGURE 3: Summary of manuscripts in the systematic review......................................................................18

FIGURE 4: Coding criteria by thematic area of the systematic review�� 19
FIGURE 5: Year of publication for manuscripts included in the

systematic review�� 25
FIGURE 6: Summary of sampling findings....................................................................................................................................................... 26
FIGURE 7: Example of conceptual model from KiVA intervention.......................................................... 28
FIGURE 8: Levels of influence included in the systematic review database.................. 30
FIGURE 9: Results on levels of influence from the manuscripts included in the

systematic review database .............................................................................................................................................................31
FIGURE 10: Examples of C4D interventions influencing policy .................................................................... 32
FIGURE 11: Summary of programme design findings............................................................................................................ 34
FIGURE 12: Results on the strategic communication

approaches from the systematic review............................................................................................................... 36
FIGURE 13: Results on the communication channels used in the

systematic review .................................................................................................................................................................................................... 36
FIGURE 14: Summary of programme implementation findings....................................................................... 37

FIGURE 15: Indicators from the manuscripts included in the
systematic review .....................................................................................................................................................................................................41
FIGURE 16: Summary of overall evaluation processes findings ................................................................. 43
FIGURE 17: Summary of findings from formative research results .................................................... 48
FIGURE 18: Summary of findings from process evaluation results .................................................... 52
FIGURE 19: Summary of findings from impact evaluation results ...........................................................61
FIGURE 20: Potential directions for future research and study.................................................................... 68
Acknowledgements
RESEARCH TEAM
PRINCIPAL INVESTIGATOR: Suruchi Sood, Associate Professor, Dornsife School of Public Health,
Drexel University
CO-PRINCIPAL INVESTIGATOR: Günay Salazar
CO-PRINCIPAL INVESTIGATOR: Ami Sengupta
RESEARCH STAFF: Carmen Cronin, Michelle Gordon, Jessica Lopez, Kelli Kostizak, Sarah Stevens
STUDENT ASSISTANTS: Joëlla Adams, Srinidhi Bhatt, Janay Brandon, Mena El Turky, Nina
Figueroa, Andrew Issa, Priyanka Padidam, Nikhil Shah, Anuja Thatte
UNICEF
MANAGEMENT TEAM:
• Neha Kapil, C4D Specialist, UNICEF Headquarters
• Charlotte Lapsansky, C4D Specialist, UNICEF Headquarters
• Clarice Da Silva e Paula, Child Protection Specialist, UNICEF Headquarters
• Marina Komarecki, IKM Specialist, UNICEF Headquarters
• Sophie Flynn, Child Protection Consultant, UNICEF Headquarters
• RyAnn Babcock Waldemarsen, Programme Support, UNICEF Headquarters
• Mercy Abgai, Programme Support, UNICEF Headquarters
OVERALL GUIDANCE:
• Theresa Kilbane, Senior Advisor, Child Protection, UNICEF Headquarters
• Rafael Obregon, Chief, C4D, UNICEF Headquarters
• Stephen Blight, Senior Advisor, Child Protection, UNICEF Headquarters
• Susan Bissell, Chief, Child Protection, UNICEF Headquarters
• Cornelius Williams, Chief, Child Protection, UNICEF Headquarters
INPUTS AND CONTRIBUTIONS:

• Susana Sottoli, Associate Director, UNICEF Headquarters
• Patricia Portela Souza, C4D Specialist, UNICEF Headquarters
UNICEF
REVIEWERS:
• Ketan Chitnis, C4D Chief, UNICEF Vietnam
• Rudrajit Das, C4D Specialist, UNICEF
• India Ayda Eke, Child Protection Specialist, UNICEF Headquarters
• Fabio Friscia, C4D Specialist, UNICEF WCARO
• Karin Heissler, Child Protection Specialist, UNICEF Headquarters
• Mika Kunida, C4D Specialist, UNICEF WCARO
• Paolo Mefalopulos, C4D Chief, UNICEF India
• Mario Mosquera, C4D Specialist, UNICEF India
• Joachim Theis, Child Protection Chief, UNICEF India
• Cristina del Valle, Child Protection Specialist, UNICEF WCARO
• Nance Webber, C4D Chief, UNICEF Bangladesh
• Clara Sommarin, Child Protection Specialist, UNICEF Headquarters
• Caroline Bacquet, Child Protection Specialist, UNICEF Headquarters
• Jennifer Keane, Child Protection Specialist, UNICEF Headquarters
• Child Protection Regional Advisors
EXTERNAL REVIEWERS
• Sue Goldstein, Programme Director at Soul City, Johannesburg Area, South Africa
• Lisa M. Jones, Ph.D. Research Associate Professor of Psychology, Crimes against
Children Research Center (CCRC), University of New Hampshire
• Catherine L. Ward, Ph.D. Associate Professor, Department of Psychology, University of
Cape Town
EDITING AND PROOFREADING

Robert David Cohen and Teresa Stuart, Rainbarrel Communications
Lisa Hiller-Garvey, Small World Stories
DESIGN AND LAYOUT

Big Yellow Taxi, New York
1 Executive summary
Children and their teacher

play a game outdoors in a
circle, during a recreation
period at a UNICEF-supported
school in the Boy Rabe
Monastery displacement
camp in Bangui, the Central
African Republic.
T
“Violence against children cuts across he Programme Division at
boundaries of geography, race, class, UNICEF Headquarters in New
York initiated a research study
religion, and culture. [...] Violence against
to analyse the effectiveness of
children is never justifiable. Nor is it
© UNI CE F/NYHQ2 014 -0 328 /G RARUP
Communication for Development (C4D)1

inevitable. If its underlying causes are approaches to address Violence Against
identified and addressed, violence against Children (VAC). 2 The overall study aims
children is entirely preventable.” to develop guidance and recommenda-
— World report on violence against children, tions for future programming and re-
United Nations, 2006 search. This report presents the findings
from a systematic literature review.
EXECUTIVE SUMMARY
VAC is ubiquitous and manifests itself in differ-

ent forms. The term ‘violence’ is often used to
encompass the complexity and range of issues
children face -- from trafficking to bullying
to child neglect.3 The many manifestations of
VAC explains the over 80,000 initial total hits
yielded by the review. Moreover, each form of
violence is influenced by a unique set of politi-
cal, economic, cultural, and social factors. As
such, it is critical to examine VAC along a con-
tinuum, rather than addressing it uniformly.
Overlapping results for different searches re-

vealed that VAC cannot be easily pigeonholed
by topic. Rather, VAC must be conceptualized
as multiple forms or facets of violence. Linkages
between different forms of VAC and other forms
of violence, such as violence against women, and
intimate partner violence, must be explored. No
doubt this overlap presents itself as a challenge
when implementing and evaluating interven-
tions. However, it is imperative that the interrela- that individuals are embedded within a larger 10-year-old Vova, who has
been displaced from the
tionships and complexities between these issues social system. Effective interventions must keep city of Bryanka in Luhansk
be recognized and addressed holistically. in mind the interactions between levels in order Oblast (Region), sits on
his bed in a dormitory at
to effectuate sustainable change. The social eco- an accommodation centre
for people displaced by
The number of manuscripts (peer-reviewed logical model provides a framework to address
the country’s continuing
articles and grey literature) related to the use the interactions between levels. Interventions conflict, in the Pushcha
Vodytsia neighbourhood
of C4D approaches to address VAC has steadi- that cut across the levels of the social ecological in Kyiv, Ukraine.
ly increased each year since 2000. Of the 302 model should work towards addressing social,
manuscripts that were coded, 44 per cent dis- emotional, and behavioural skills (for example,
cuss an intervention implemented in a devel- self-efficacy) of individuals and groups, as well
oping country, which speaks to the geographic as norms, instead of only addressing individual
robustness of this review. A greater proportion knowledge and attitudes.
of manuscripts discuss interventions in urban
contexts as compared to rural contexts. Manuscripts reviewed did not necessarily explic-
itly state the use of C4D approaches. However,
Roughly half of the interventions reviewed do not upon closer examination, it became apparent
explicitly reference a conceptual model to under- that the majority of responses to VAC were in-
pin the interventions. Those that do, typically cite herently communicative. Programmes address-
individual or cognitive conceptual models and a ing VAC often use C4D approaches to reduce
majority (over 80 per cent) focus on the individu- harmful practices using a ‘harm reduction’
© UNI CE F/NYHQ2 014 -1 903 /K REP KIH
al level of change. About 11 per cent use commu- framework. Often in these cases, programme
nity approaches and slightly less than 10 per cent objectives focus on the negative, whereas C4D
report using an ecological approach. messages for the same intervention focus on
positive changes. Overall programme objectives
While cognitive and individual-based behav- should be linked to communication objectives,
iour change approaches are valid and useful in which in turn yield C4D messages. The links
certain contexts, there is a growing realization between these overall programme objectives and
EXECUTIVE SUMMARY
by university-based researchers) tended to use

training/ capacity building trajectories.
Many of the interventions focus on interven-

ing in the setting where violence occurs, such as
the home or school. However, there is a need to
look beyond settings, and take into consideration
the forms of violence that occur in less concrete
or discrete places, such as verbal or sexual ha-
rassment of adolescent girls in public spaces. A
norms-based approach has the potential to ad-
dress violence more holistically and at a broader
level. It can also allow for the integration of inno-
vative communication technologies as both chan-
nels for dissemination and mechanisms to help
track intervention implementation and success.
The paucity of robust and rigorous evaluations in

this review underscores the need to invest more
heavily in research. Few manuscripts described
formative research processes (17 per cent) or pro-
(Standing, front row, subsequent communication objectives and mes- cess evaluations (14 per cent). Significantly more
second from left) Princess
Johnson, eight, prays
sages should be clarified in order to achieve the manuscripts reported on impact evaluations
with her schoolmates desired outcomes in relation to behaviour and (around 16 per cent of manuscripts for random-
before the start of
class, in Kpallah Town, social change. ized controlled trials and 15 per cent on a pre-
a community in the post case-control design without randomization).
city of Brewerville in
Montserrado County, At the same time, it was difficult to distinguish Qualitative observational data (reported in 28
Liberia. between overall Child Protection (CP) proper cent of manuscripts) is a commonly utilized
gramme objectives that considered the spectrum evaluation methodology. This type of data, how-
of initiatives for prevention and response and ever, makes a weak case for C4D attribution to
C4D objectives that focused more specifically outcomes. Overall, there was a serious underuti-
on behavioural and social change outcomes. lization of participatory research methods for any
There was little to no evidence of the utilization and all types of evaluations.
of SMART or SPICED criteria when describing
programme objectives. The majority of objec- There is a serious lack of evaluation data on
tives relied on changing knowledge or improv- the effectiveness of interventions and strong
ing comprehension, instead of activating higher evidence of the need for additional effective-
levels of cognitive and affective change such as ness evaluations, specifically for low-and
assimilating or evaluating information. middle-income countries. The focus on the
individual, as seen in the selection of concep-
Implementation related findings demonstrated tual models, the wording of objectives, and
that the intended audience for the majority of in- the level of influence, persists with regards to
© UNI CE F/NYHQ2 015 -1 210 /G RILE
terventions were individuals and did not report measurement indicators, which are further
segmenting audiences into primary, secondary, centred on low-order individual cognitive
or tertiary groups. Developing countries tended constructs. Lack of information on sampling
to utilize campaigns with mass media chan- frameworks, small sample sizes, and other
nels such as television, radio, and print, where- methodological issues bring to light the lack
as United States-based interventions (often led of specificity and sophistication in the evalu-
EXECUTIVE SUMMARY
ation data being collected. These issues ul- Moving forward, future research and practice
timately raise doubts about the quality of should consider the following 15 overall rec-
information utilized to examine effectiveness ommendations, categorised under three broad
of interventions. themes relevant to contextualisation and fram-
ing of the issue, programme design, implementa-
Local indicators (incidence and prevalence) on tion and evaluation, which are elaborated upon
VAC are not always readily available or acces- later on in the report.
sible. This being the case, an essential first step
to identifying the magnitude of VAC as a set of FIGURE 1: Overall recommendations based on the Systematic
issues deserving of global advocacy and ad- Review
equately resourced interventions would be the
creation of standardized definitions and mea-
Contextualisation and framing of the issue
surements of incidence and prevalence.
The review shows that prevention efforts can

• Specifically address children within other forms of violence
contextualize violence as both a cause and an
• Explore linkages between different forms of VAC
outcome. Preventing and responding to VAC
• Explore VAC along a spectrum
can be an end in and of itself. Or, VAC can be
• Contextualize VAC both as a cause and an outcome
embedded as a causal factor within interventions
• Start early and continue into adulthood
designed to promote health or education. Both
• Move beyond a place-based approach to a wider norms-based approach
methods of understanding VAC are valid.
to incorporate innovative communication channels and tackle the culture
of violence
There was strong evidence in support of early
childhood programmes (e.g. positive parenting
and early childhood development). These types Programme design
of programmes can lay the groundwork for gen-
erational change. By raising boys and girls that
reject violence, new norms around masculinity • Embrace the social ecological model for behaviour and social change
and femininity can be created. • Broaden conceptualizations of C4D approaches to encompass efforts that
involve any form of communication/ communicative action and cut across all
Finally, the role of gender cannot be overlooked levels in the social ecological model
when discussing VAC. Girls are disproportion- • Segment audiences by gender and address gender specific needs and
ately at risk for most types of violence and their differences
specific needs must be addressed. Female em- • Clarify direct and indirect linkages between stated programme outcomes
powerment has proven to be a successful and and C4D or communication objectives and messages
integral component of prevention efforts. The • Reconfigure programme and C4D/ communication objectives of VAC interventions
review pointed out the need to engage men and • Move beyond individually-focused knowledge, attitudes, and practices by
addressing social, emotional, and behavioural competencies
boys in gender transformative programmes.
While focusing on girls is important, it should
not be at the expense of boys, who are also vic- Programme implementation and evaluation
tims of violence.
In summary, it is essential to expand outcome • Position VAC as a 'glocal' issue through qualitative and
evaluation studies so as to enhance our under- quantitative measurement
standing of global best practices, which, in turn, • Enhance investment in research
need to be contextualized based on what works • Scale up promising interventions
at a local level to address VAC.
2 Background
A group of children play ëis

ëjam mu mee moní or Eenie,
Meenie, Miney, Mo in Eng-
lish in Adone village of Ta Oi
district in Saravane province,
Lao PDR.
T
This research study, initiated his study focuses specifically on the
© UNI CE F/LAO-20 15-NOORAN I- 0238/
by the Programme Division at effectiveness of C4D approaches in

UNICEF headquarters in New York, order to better understand what works
vis-a-vis prevention and response. It
systematically analyses interventions
sought to answer the following research ques-
addressing violence against children tion: ‘What are the effects of communication
through the use of communication for for development approaches to address violence
development (C4D) approaches. against children?’.
BACKGROUND
This systematic review examines available There are differences among practitioners and
peer-reviewed publications and grey literature scholars on all the communication approach-
(reports and project documents) on C4D in- es and terms that constitute C4D. Therefore,
terventions addressing VAC within a 13-year the systematic review included several com-
time period (from 2000 to 2013). The research plementary terms such as behaviour change
question posed above requires a clear under- communication, social mobilization, media
standing of the population (children), issue campaigns, and advocacy. With respect to
of concern (VAC), and intervention approach forms of VAC, a mix of broad and specific is-
(C4D). Definitions for these terms are provid- sues is included (the complete list of key terms
ed below: used for this systematic review is presented in
Table 1).
THE POPULATION FOCUS FOR THIS RE-
VIEW IS ON CHILDREN. The United Nations
Convention on the Rights of the Child (CRC)
defines a child as “a human being below the age
of 18 years, unless under the law applicable to
child, majority is attained earlier”.4
TABLE 1: Key terms for the systematic review
THE ISSUE FOCUS OF THIS REVIEW IS
Communication (for/ and) development
ON VAC. Article 19 of the CRC defines vi-
Behaviour change communication
olence as: “all forms of physical or mental
Social change communication
violence, injury or abuse, neglect or negli- Advocacy
gent treatment, maltreatment or exploitation, Social mobilization
including sexual abuse”. General Comment Community mobilization
no. 13 seeks to guide State Parties in un- Approaches Participatory communication
derstanding their obligations under article Communication campaigns
19 of the CRC to prohibit, prevent and re- Strategic communication
Media campaigns
spond to all forms of violence. It is based on
Social media
the fundamental assumption that: “No vio-
Interactive communication technologies
lence against children is justifiable; all vio-
lence against children is preventable”, and it Children
builds on existing guidance provided by the Minors
CRC Committee and on recommendations Girls
Populations Boys
coming from various United Nations mecha-
Adolescents
nisms and reports, including the Secretary-
Youth
General’s Study on Violence against Children
(United Nations, 2006). Child protection
Violence
THE INTERVENTION FOCUS OF THIS RE- Gender-based violence
VIEW IS ON COMMUNICATION FOR DE- Sexual abuse/ sexual violence/ sexual exploitation
Trafficking
VELOPMENT, which UNICEF broadly defines
Child Labour
as “an evidence-based and participatory pro-
Issues Child marriage/ early marriage
cess that facilitates the engagement of children, Honour killings
families, communities, the public and deci- Female genital mutilation/ cutting
sion makers for positive social and behavioural Discrimination
change in both development and humanitarian Corporal punishment
contexts through a mix of available communi- Bullying
Gang violence
cation platforms and tools” (UNICEF, 2019).
3 Structure of this report
Children hold books and other

supplies from their new UNI-
CEF school kits at the accom-
modation centre where they
now live, on the outskirts of
the city of Kharkiv in Kharkiv
Oblast Province, Ukraine.
The following chapters summarize the methodology

utilized for conducting the systematic review, including
an acknowledgement of the limitiations of a systematic
review methodology, and of this review in particular.
© UNICEF/NY HQ2014 -1939 /ZMEY
The central section presents the key findings from the

systematic review, organized into five subsections (see
Figure 2). The final section provides a summary of
overarching findings and recommendations for future
research and practice.
STRUCTURE OF THIS REPORT
The key findings of this systematic review are 4. OVERALL PROGRAMME EVALUA-
organized into the five subsections below: TION PROCESSES: includes informa-
tion on overall programme evaluation
1. SAMPLING INFORMATION: includes processes such as evaluation design,
information about sampling for the sys- research methods, analysis frame-
tematic review. works, sampling, and indicators.
2. PROGRAMME DESIGN ELEMENTS: 5. THEMATIC ANALYSIS OF THE KEY

provides information on the con- RESULTS BY TYPE OF RESEARCH:
ceptual frameworks reported in the provides a thematic analysis of the
manuscripts, programme and commu- key results and discussion sections
nication objectives, level of influence of the manuscripts included in the
that the interventions focus on, and the systematic review. The analysis cat-
implementing agencies and partner- egorizes the results by formative re-
ships involved in the interventions as search, process evaluation and impact
reported in the manuscripts included in assessment/ evaluation.
the systematic review.
The key findings from each of the five subsec-
3. PROGRAMME IMPLEMENTATION: tions are summarized within the individual
contains a summary of the intended subsections to provide an overall picture of key
audiences, strategic approaches, and learnings from the systematic review.
communication channels utilized by
the interventions reported in the manu- A visual map of the findings is presented in Fig-
scripts included in this systematic re- ure 2. This map is intended to help the reader
view. visualize the flow of the findings, see how each
section builds upon the previous ones, and find a
specific section of interest.
FIGURE 2: Structure of this report’s findings section
5.
4.
2. 3. Thematic
1. Overall
Programme Programme analysis of the
Sampling programme
design elements implementation key results by
evaluation
type of research
Year Conceptual frameworks Intended audiences Evaluation design Formative research
Location Programme objectives Strategic approaches Research methods Process evaluation
Impact assessment/
Summary Communication objectives Communication channels Analysis frameworks
evaluation
Level of influence Summary Sampling Summary
Implementer and partners Indicators
Summary Summary
4 Methods
Girls at Bhagyanagar Chil-

dren’s Home in India: two near-
by children’s homes separately
house girls and boys mostly
between 6-14 years who are
orphans, children of migrant
labourers and potential child
labourers.
A critical first step in designing a robust systematic

review is reaching overarching agreement on the
inclusion and exclusion criteria for studies to be
© UNI CE F/INDA201 4-000 40/BISWAS
reviewed. Such criteria ensure that all relevant literature

is included while winnowing out those not pertinent
to the question at hand. The inclusion and exclusion
criteria used for this secondary analysis are summarized
in Table 2.
METHODS
TABLE 2: Inclusion and exclusion criteria for the systematic review

• A 2000 onwards time frame for intervention reviews (2000-2013)
• Reference to a specific programme or intervention
Inclusion criteria • The programmes or interventions include children as audiences or beneficiaries
• The programmes or interventions address key ‘violence’ issues
• The programmes or interventions include C4D approaches
• Manuscripts written or published before 2000
• Manuscripts solely on policies, editorials and commentaries
Exclusion criteria • Manuscripts on violence or C4D approaches but not specific to children
• Manuscripts on VAC but not utilizing C4D approaches
• Manuscripts on C4D approaches but not specific to violence
The list of databases included in this review is as well as the Communication Initiative data-
summarized below in Table 3. A majority of these bases, were all selected in order to include grey
databases include peer-reviewed literature, while literature.
others such as Google Scholar, ELDS and BLDS,
TABLE 3: Search engines utilized for the systematic review

SEARCH ENGINE Description
PUBMED CENTRAL United States National Library of Medicine
ACADEMIC SEARCH COMPLETE Journals and periodicals across disciplines
PSYCINFO Behavioural sciences and mental health
SOCIOLOGICAL ABSTRACTS Sociology and other behavioural and social sciences
GOOGLE SCHOLAR Scholarly literature including theses, books, abstracts and articles
Boys queue to kick a foot-
ball during a recreation
ELDIS AND BLDS Institute of Development Studies, Sussex, United Kingdom
period at Sheikh Radwan
School in Gaza City on the
COMMUNICATION INITIATIVE Database of communication strategies and media for development
first day of the new school
year in the Gaza Strip.
© UNI CE F/NYHQ2 014 -1 508 /E L BABA
METHODS
Manuscript selection The staffing to conduct the systematic review in-

and sampling cluded seven Masters of Public Health students at
Drexel University. A series of quality control mech-
I
n order to systematize the review process, anisms were employed at each stage. After data
the shortlisting of manuscripts involved entry was completed, the Principle Investigator on
three stages of review: the study identified a systematic random sample of
30 manuscripts (ensuring that at least two articles
1. TITLE REVIEW: Titles deemed relevant per student involved in data entry were included)
were shortlisted for abstract review. to calculate inter-rater reliability, which was manu-
2. ABSTRACT REVIEW: Shortlisted ab- ally calculated at 84 per cent, indicating statistically
stracts were scanned for relevance based substantial agreement. The kappa statistic was uti-
on their adherence to the inclusion criteria. lized to examine inter-rater agreement and yield-
ed a value of 0.69. According to Landis and Koch
3. FULL TEXT REVIEW: Full text manu- (1977), a value of 0.61 - 0.80 can be considered as
scripts of the relevant abstracts were read substantial agreement.
and entered into a database.
The analysis of data presented in this report uti-

Figure 3 below summarizes the sampling process lized a mixed methods approach. The quantita-
and sample size information. While the total tive components of the coded data were analysed
number of hits based on the key search criteria through frequencies and cross-tabulations in
was 80,532, only 302 manuscripts made the final STATA 12. The text data was analysed by using
cut. a grounded theory approach, in which the coded
data was categorized into themes, which were fur-
The 302 manuscripts were coded into an ex- ther refined and expanded during the analysis pro-
cel database that included 96 columns of in- cess (Charmaz, 2006).
formation categorized into five thematic areas:
background information, programme design,
programme implementation, programme evalu-
ation, and finally limitations and recommen-
dations, as outlined in the manuscript by the
author(s). Figure 4 provides a breakdown of in-
formation collected for each thematic area.
FIGURE 3: Summary of manuscripts included in the systematic review
Total hits: 80,532
Selected for abstract review: 1,909
Selected for full text review: 414 articles; 119 reviews (total:533)
Selected for coding: 235 articles; 68 reviews (total: 302)
METHODS
FIGURE 4: Coding criteria by thematic area of the systematic review
2. 3. 4. 5.
1.
Programme Programme Programme Limitations and
Background
design implementation evaluation recommendations
Citation Objectives Communication approach Design Theoretical
Channels Evaluation type Operational

Type of publication Conceptual framework (formative, process,
Audiences impact) Legal/ ethical
Region Level of influence Partnerships Results Methodological
10-year-old Melina and her

© UNI CE F/NYHQ2 015 -1 157 /N EWAR
eight-year-old sister, Monica,

play with a chick in remote
Gerkhutar Village Development
Committee in Nuwakot District,
1 of 12 districts that have been
severely affected by the massive
earthquake in Nepal.
5 Limitations of
this systematic review
Ronaldo Durant plays with

his siblings at their home in
Linden, Guyana.
There are several limitations inherent in conducting any systematic

review, and the following are pertinent to this specific review and need to
be mentioned up front.
• Much of the analysis and reporting in any system- implementation and evaluation processes for any
atic review is dependent on the focus of the select- given intervention.
© UNI CE F/UKL A2 014 - 1 179 /MILTON
ed manuscript, and in the case of peer-reviewed

articles, constraints of the publication process.5 • The inclusion of grey literature in this report is in-
dicative of the need to conduct more robust eval-
• Not all of the information about an intervention can uations including at the design (formative) and
be reported in one manuscript. This is likely the implementation (process) stages.
reason for missing information regarding design,
LIMITATIONS OF THIS SYSTEMATIC REVIEW
T
he analysis presented in this system- • An intervention aimed at educating
atic review is based on a grounded adolescent girls on family life education
theory approach. In grounded the- issues in a village of Uttrakhand, India,
ory, categories and themes emerge was coded as part of the searches on
or are generated from the data and are further ‘social change approaches to address
refined and expanded throughout the data discrimination’ as well as ‘communica-
analysis process. This approach differs from a tion and development interventions to
quantitative categorization of the manuscripts. address child marriage’ (Saxena, Sriv-
Thus, it is hard to apply ‘rule-of-thumb’ criteria astava and Ahuja, 2009).
and classify manuscripts based on the strength
of the evidence presented. In order to quanti- The inherent complexity of C4D approaches and
tatively compare effectiveness, one of two cri- VAC makes it difficult to create mutually ex-
teria needs to be met: either specific common clusive categories for the reviewed manuscripts
outcomes need to be identified or approaches based on either approaches or issues. There-
need to be limited to examine their impact on fore, one cannot make broad generalizations of
a variety of outcomes.6 Otherwise there is no the terms C4D and VAC, but rather one needs
common ground upon which to make judg- to examine specific issues and approaches. At
ments about effectiveness. VAC encompasses the same time, this complexity is a key learning
a range of issues and C4D includes an array of upon which broad-based guidance on the con-
approaches. This systematic review combines ceptualization of VAC programmes can be built.
both VAC and C4D as an overall framework;
therefore the validity of such aggregation based The systematic review attempted to cast a wide
on strength of evidence is limited. On the one net and gather information from as many data
hand it is impossible to compare the effective- sources as possible by using multiple and dif-
ness of interventions addressing different is- ferent types of databases. Nevertheless, some
sues and using multiple approaches. At the relevant databases were inadvertently excluded
same time it is hard to take the interventions such as ERIC (Education Resources Informa-
included in the review and fit them into mu- tion Centre), while others were deliberately
tually exclusive categories based upon either omitted, for example, student theses and dis-
approach or issue. Sometimes interventions ad- sertations. The selection of key terms for the
dress multiple violence issues and/ or use sev- review was based on a review of several back-
eral approaches. ground documents and consensus among the
research team, as well as C4D and Child Pro-
Apart from the large number of approaches and tection teams at UNICEF Headquarters. This
issues being covered in this review, there were process resulted in a well-defined yet broad
overlaps between C4D approaches and different set of approaches and forms of violence. How-
forms of VAC as evidenced by the fact that the ever, some commonly used approaches such as
literature search revealed the same manuscripts information campaigns, IEC approaches, and
even after the related key terms were varied. For educational campaigns were not included in
example: the search terms because of their limited use in
academia.
• The evaluation of the first five years of
the California Wellness Foundation’s Vio- A key limitation of including a variety of key
lence Prevention Initiative was coded as terms to describe both C4D and child protection
an example of community mobilization, meant that in some cases it was difficult to parse
as well as a communication for develop- out the communication aspect of some of the
ment approach (Greenwood et al., 2001). manuscripts that were included in the review,
for example, counsellor training to address VAC
(Left to right) Nancy, a
therapist, uses a puppet or dance therapy for violence prevention. Simi- reported interventions are from industrialized
as an aid during a ses-
sion with nine-year-old
larly, in some instances the extent to which C4D countries. In some ways this is a function of the
Natalia (name changed), interventions were designed to address VAC academic nature of this systematic review and
at the CEPAT (Centro de
PrevenciÛn y AtenciÛn
was difficult to determine. An example of this reflects the biases associated with searching in
Terapéutica) therapeutic are programmes addressing HIV-related stigma United States-based databases and peer-review
care centre, in the city of
PotosÌ, capital of PotosÌ and discrimination, where VAC related preven- publications.7 At the same time it is important
Department, in Bolivia. tion and response is related to programmes de- to situate this systematic review in the context
© UNI CE F/NYHQ2 013 -1 500 /P IROZZI
signed to address issues pertinent to health and of other reviews of this nature. For example,
education. Mikton and Butchart (2009) while reviewing
child maltreatment programmes, noted in their
Another key limitation of this review could be results a woeful imbalance in the geographic
the fact that more than half of the information distribution of child maltreatment prevention
presented here is based on data from the devel- research with over 99 per cent of the publica-
oped world. Approximately 56 per cent of the tions coming from high-income countries.
LIMITATIONS OF THIS SYSTEMATIC REVIEW
These authors contend that their findings par- sufficient programmatic details as required by
allel the 10/90 gap in areas of health research.8 this systematic review database. It is important to
This review, in contrast, is therefore geographi- mention that the inclusion of both types of man-
cally robust given that 44 per cent of the manu- uscripts constitutes a limitation because exist-
scripts comes from developing countries. It is ing reviews/ analyses yield different results and
possible that the inclusion of multiple search have to be interpreted differently from individual
terms to define C4D and VAC, as well as the interventions.
use of multiple databases, resulted in the over-
all robustness of the sample. A final key limitation of this review was that it
only included peer-reviewed journal articles and
In addition, it is important to note that while the reports published in English between 2000 and
United Nations Secretary-General’s Study not- 2013. The 2000 onwards time frame was con-
ed that VAC happens everywhere and across all sidered appropriate as research associated with
social groups, specific types of VAC are related examining effectiveness of communication pro-
to social and cultural norms of a given society. grammes is relatively new. Therefore, looking at
According to UNICEF, the practice of female interventions and evaluations preceding the 13-
genital mutilation/ cutting (FGM/ C) is concen- year time frame was not likely to yield pertinent
trated in 29 countries in Africa and the Middle results. Finally, the utilization of communication
East.9 Any literature associated with FGM/ C is as a mechanism to address VAC has not been
therefore likely to come from these specific coun- conducted in a systematic manner focused on
tries. While the information from interventions evidence. While examining interventions previ-
implemented in industrialized countries might ous to this time frame may have been important
not be directly applicable to developing country from a historical perspective, it was considered
contexts, the information and data remain valid
sources from which to draw conclusions and rec-
ommendations regarding C4D approaches to ad- “While the United Nations Secretary-
dress VAC. A related limitation is that this review
General’s Study noted that VAC happens
focused specifically on VAC and did not attempt
to examine causal or contributing factors or out- everywhere and across all social groups,
comes associated with VAC such as alcohol abuse, specific types of VAC are related to social
or study the relationship of VAC with environ- and cultural norms of a given society.”
mental factors such as urbanization, migration,
and displacement.
This systematic review involved a fairly large

team to complete the tasks in the allocated time. unlikely to add substantially to the research
While the inter-rater reliability is within the sub- question at hand. It is entirely possible that this
stantial range and despite efforts to ensure that review was unable to record manuscripts that
the data selection and entry procedures were were published exclusively in other languages.
standardized, variations by coder were detected. For example, evaluations published exclusively
in Spanish are noticeably absent. Some attempts
This systematic review included existing reviews were made to examine reports available only in
and meta-analyses, as well as manuscripts deal- Spanish through a search of the Communica-
ing with individual interventions. The reason for tion Initiative Spanish language website. While
including both types of manuscripts was twofold. the search yielded a large number of initial hits,
First, reviewing existing reviews/ analyses would evaluation information was noticeably absent in
prove to be cumbersome and repetitive. Second, these publications.
existing reviews/ meta-analyses seldom include
6 Key findings
Two girls read together from
a book during a class in the
UNICEF-supported Deari
Elementary School in the city
of Keren, capital of Anseba
Region, in Eritrea.
The key findings from this systematic SECTION 1:

review are organized into five sections. Sampling information
T
• SECTION 1: Sampling information he systematic review database in-
• SECTION 2: Programme design elements cluded background information
3: Programme implementation
© UNI CE F/NYHQ2 014 -3 552 /P IROZZI
• SECTION
about each manuscript such as: com-
• SECTION 4: Overall programme evaluation processes
plete citation, type, and year of pub-
• SECTION 5: Thematic analysis of the key results by
type of research lication. In addition, geographic location was
noted at increasing levels of specificity starting
with region, then country, and finally sub-na-
tional level. Information on year of publication
indicates that approximately half of the articles
KEY FINDINGS
FIGURE 5: Year of publication for manuscripts included in the systematic review

25%
20%
15%
10%
5%
0%
=8) n=13) (n=7) n=11) n=12) n=18) n=16) n=24) n=33) n=34) n=29) n=46) n=37) n=14)
0 (n 1( 2 3( 4( 5( 6( 7( 8( 9( 0( 1( 2( er
200 200 200 200 200 200 200 200 200 200 201 201 201 quart
s t
3 (1
201
reviewed (53 per cent) were published between scripts, 168 (over 50 per cent) came from in-
2008 and 2013 (see Figure 5). dustrialized countries with the United States of
America leading the list by a large margin with
Data on the geographic location of implemented 118 manuscripts (about 39 per cent). Some 31
interventions was categorized according to the manuscripts (10 per cent) were classified as being
UNICEF regional office classification. The re- global. In most cases these global manuscripts
sults indicate that 30 manuscripts (about 10 per reflected existing reviews and meta-analyses of
cent) were from the South Asian region – with a VAC programmes (Table 4).
majority coming from India (53 per cent). Some
18 manuscripts (6 per cent) were from Eastern Information on the location of the interventions
and Southern Africa. East Asia and the Pacific was also classified by geography ranging from
and the Middle East and North Africa each con- regional or province level to sub-regions. For
tributed approximately 15 manuscripts (roughly some countries (United States of America and
5 per cent). By far the largest number of manu- India), states were mentioned as being the focus
TABLE 4: Location of interventions for manuscripts in the systematic review

NUMBER OF MANUSCRIPTS PROPORTION IN DATABASE (N = 302)
Central and Eastern Europe 3 1.0

Western and Central Africa 10 3.3
Eastern and Southern Africa 18 6.3
The Americas and Caribbean 12 4.0
LOCATION
South Asia 30 9.9

East Asia and Pacific 15 5.0
Middle East and North Africa 15 4.6
Industrialized countries 168 55.6
Global 31 10.3
Region/ province 36 11.9
Sub-Region/ province 3 1.0
SUB-LOCATION
District 13 4.3
State 24 7.9
Urban centre 81 26.8
Rural areas 12 4.0
Multiple 80 26.5
KEY FINDINGS
Children play in a group

in Pakihchala, Bhaluka, in of the interventions. In addition, whenever pos- was relatively smaller: 12 manuscripts (4 per
the Mymensingh district
of Bangladesh. sible urban and/ or rural locations were noted. cent). This information sheds some light on the
The data shows that 80 interventions (over 25 nature and scope of the interventions. However,
per cent) were implemented in multiple locations this data is based on what was reported in pub-
covering, for example, both urban and rural lished manuscripts and therefore may not reflect
areas. A large proportion of the interventions, a complete picture of intervention locations.
81 manuscripts (27 per cent), focused on urban
locations while the proportion of rural locations
FIGURE 6: Summary of sampling information
• Peer-reviewed publications made up oped/ industrialized countries (a third

25 per cent of the sample and project coming from the United States alone),
reports made up the other 75 per cent. 44 per cent came from the developing
• The general trend in publications world.
related to the utilization of C4D ap- • A greater proportion of manuscripts
© UNI CE F/BANA201 4-006 78/PAU L
proaches for VAC was promising, with are from urban rather than rural
a steady increase in the numbers of locations.
manuscripts published every year • 10 per cent of the manuscript repre-
since 2000. sent ‘global’ experiences, mainly ex-
• While the sample of manuscripts in- isting reviews and meta-analyses of
cluded in this review favoured devel- interventions addressing key VAC.
KEY FINDINGS
SECTION 2: models and frameworks in the manuscript. The

Programme design elements data entry tool permitted the inclusion of sev-
eral models, and students were asked to include/
T
he information associated with pro- refer to other models as and when they came up
gramme design elements gleaned from during the course of data entry. The list of mod-
the systematic review includes the fol- els is summarized in Table 5.
lowing information:
There was limited evidence that manuscripts
• Conceptual (theoretical) frameworks included an explicit statement of conceptual
models to underpin described interventions. In
• Programme objectives fact, 129 (43 per cent) of the manuscripts did not
clearly reference any conceptual model.
• Communication objectives
Cognitive models, reported in 51 manuscripts
• Level of influence (17 per cent), were among the most commonly
mentioned. In some cases, interventions cre-
• Implementing agencies and partners ated their own frameworks/ models by building
upon existing theories of change. For example,
ASSESSMENT OF CONCEPTUAL two of the manuscripts from Finland’s KiVa
(THEORETICAL) FRAMEWORKS10 intervention employed the participant role ap-
Two types of analyses were undertaken to exam- proach to school bullying (Karna et al, 2011a;
ine the overall conceptual (theoretical) frame- Karna et al., 2011b). This approach (see Figure
works specified in the manuscripts. First, all of 7) builds upon constructs of social learning
the frameworks or models explicitly mentioned theory11 and stages of change.12 The review indi-
in the manuscripts were included in the data- cated that an additional 18 manuscripts focused
base. If more than one model/ framework was on individual steps and stages of change mod-
described, then the database included multiple els. Therefore, close to a quarter of the manu-
columns to note all of them. A second set of scripts explicitly relied on individually based
analyses pertained to the implicit inclusion of approaches to behaviour change.
TABLE 5: Conceptual frameworks in the manuscripts included in the systematic review

MODELS NUMBER OF PROPORTION NUMBER OF PROPORTION
MANUSCRIPTS IN DATABASE MANUSCRIPTS IN DATABASE
(N = 302) (N = 302)
EXPLICIT MODEL IMPLICIT MODEL

Cognitive approaches 51 16.9 51 16.9
Community approaches 34 11.2 35 11.6
Social ecological model 24 7.9 32 10.6
Steps/ stages approach to behaviour change 18 5.9 28 9.3
Social marketing 15 5.0 -
Social norms 13 4.3 9 3.0
Systems level addressing policy 11 3.6 19 6.3
Human rights, gender and multicultural 8 2.6 -
approaches
Steps/ stages approach to social change 5 1.7 12 4.0
Narrative approaches 3 1.0 9 3.0
Social network models 3 1.0 -
Planning models (COMBI, P process) 3 1.0 1 0.3
KEY FINDINGS
Community approaches, on the other hand, interpreted given the subjective nature of this cod-
were explicitly mentioned in 34 manuscripts ing. However, it is interesting to note while only 24
(11 per cent). Twenty-four manuscripts (slightly of the manuscripts (8 per cent) explicitly stated us-
less than 10 per cent) reported utilizing an eco- ing an ecological approach, the coders found evi-
logical approach.13 One word of caution when dence of an ecological perspective in an additional
interpreting these results is that comprehensive 32 manuscripts (11 per cent).14 In another example,
models such as the ‘social ecological frame- only 11 manuscripts (4 per cent) made an explicit
work’ encompass various levels of influence reference to changes at the system or policy-level as
ranging from the individual level to the policy- their conceptual framework, but an additional 19
level. It is likely that the 24 manuscripts (8 per manuscripts (6 per cent) were thought to be im-
cent) that reported relying on the social ecolog- plicitly attempting to address policy-level changes
ical approach included elements from cognitive primarily through advocacy.
approaches stages of behaviour change, social
FIGURE 7: Example of conceptual model from KiVa intervention
Participant role approach to school bullying
Finland’s KiVa intervention employed the participant a “positive change in the behaviours in classmates can
role approach to school bullying. According to this ap- reduce the rewards gained by bullies and consequent-
proach, bullying is perceived as a group phenomenon ly their motivation to bully in the first place” (Kärnä,
largely enabled and maintained by class members tak- 313). This is accomplished through skill-building exer-
ing on different participant roles. These roles include: cises emphasizing self-efficacy, anti-bullying attitudes,
assistants, reinforcers of the bully, outsiders or defend- and empathy as well as the establishment of a school
ers of the victim. These roles moderate behaviour with- mechanism to address instances of bullying immedi-
in the class social group by ‘peer group power’. The ately through counselling and small group discussions.
KiVa interventions encourage participants to take ac- In some ways the KiVa model combines both cognitive
tion against bullying by becoming defenders or helpers elements (role playing to build efficacy) and stages of
of victimized classmates. KiVa is based on the idea that change (moving individuals to serve as defenders).
networks, community-based approaches and ASSESSMENT OF PROGRAMME

social norms. There was evidence of additional OBJECTIVES
models as well, for example social marketing The programme design analysis of the manu-
was referenced in 15 manuscripts (5 per cent), scripts included an assessment of the overall
and some 13 manuscripts (4 per cent) reported programme objectives. Thematic categorization
addressing social norms. In keeping with UNI- using a grounded theory approach yielded the
CEF’s strong focus on human rights-based ap- following broad categories:
proaches, eight of the manuscripts (3 per cent)
specifically referenced a human rights, gender • Policy frameworks;
or multicultural focus. • Overall impact or change over time in
incidence and prevalence;
Subjective interpretation of theoretical constructs • Social change (e.g. community empow-
by the coders indicated that over 50 per cent of the erment, social norms, and social move-
manuscripts implicitly made reference to one or ments); and
more of the models listed above. Information on
• Individual change (e.g. behaviours, self-
implicit referencing of models has to be cautiously
efficacy, attitudes and knowledge).
KEY FINDINGS
The objectives related to policy frameworks were communication messages for the same inter-
written in broad and generalizable terms and ventions focused on raising the value of the girl
typically sought broad-based social and struc- child and changing gender-related attitudes. An-
tural change. This was especially true of child ti-bullying interventions worked in a very simi-
labour reduction and elimination programmes. lar fashion. Programme objectives were couched
The second set of broad programme objectives in terms of reducing bullying in schools, but the
addressed overall impact (measured in terms interventions themselves worked on issues of
of prevalence and incidence of violence against social support and peer relations, issues whose
children), and was most commonly designed linkages to bullying are not always readily ap-
to address issues such as bullying and/ or gang parent. Clarifying the links between overarching
violence. programme objectives and communication mes-
sages by establishing intermediate communica-
The specific programme objectives geared to- tion objectives would help determine how the
wards social change dealt with changes in at- communication messages contribute to achiev-
titudes and norms associated with harmful ing overall programme objectives.
practices such as FGM/ C and child marriage.
A large subcomponent of the social change ob- Of the 302 manuscripts, only one did not specify
jectives were written in ‘empowerment’ terms the programme objectives associated with the
and aimed to bring about changes in efficacy interventions being described. Interestingly, the
(both at individual and collective levels). analysis showed little to no evidence of the utili-
zation of SMART or SPICED15 criteria when de-
The final category of programme objectives fo- scribing programme objectives and subsequently
cused on individual level changes. The pro- indicators designed to measure these objectives.
gramme objectives geared towards individual
level changes comprised the single largest group ASSESSMENT OF COMMUNICATION
in terms of stated programme objectives. A large OBJECTIVES
proportion of these individually focused behav- Programme objectives are often broad and in-
iour change interventions appeared to address terventions utilizing C4D often have within
gaps in awareness and knowledge and therefore them communication objectives that specify
aimed at a fairly low level along an individual be- what the communication components of a given
haviour change continuum. project are designed to achieve. The communi-
cation objectives included in the manuscripts
Interestingly, the thematic analysis indicates were entered verbatim into the systematic re-
that the stated objectives of the individual be- view database. In order to analyse the content
haviour change interventions were universally of the communication objectives, elements were
designed to reduce or eliminate harmful prac- adapted from Bloom’s Taxonomy for cogni-
tices. Their objectives made little stated effort tive and affective learning objectives that were
to focus on positive changes, which individu- originally used in the field of education. The
als could undertake, yet the messaging for such cognitive domain has been defined as the area
interventions was positively framed. This find- of learning devoted to acquiring information,
ing points to a potential misalignment between knowledge and intellectual abilities (Morrison,
the negatively couched objectives and the posi- Ross and Kemp, 2007). Skills in the cognitive
tively framed messaging derived from those domain consist of six levels, from lowest order
negatively couched programme objectives. For processes to highest: knowledge, comprehension,
example, the programme objectives of interven- application, analysis, synthesis and evaluation.
tions addressing child marriage were couched in These skills are associated with qualitatively
terms of raising a community’s awareness of the measurable verbs that allow curriculum design-
harmful effects of child marriage or to reduce ers to explicitly indicate what the student must
the incidence of child marriage. Meanwhile, the do in order to demonstrate learning (Anderson
KEY FINDINGS
and Krathwohl, 2001; Krathwohl, 2002). For ex- ing’, and ‘evaluating’. Evidence has shown that
ample, knowledge is associated with verbs such only addressing the cognitive domain will not
as ‘remember’, ‘recognize’ and ‘recall’. Similarly, achieve sustainable change.
comprehension is associated with verbs such as
‘distinguish’, ‘estimate’, ‘explain’, ‘generalize’, ASSESSMENT OF THE LEVEL OF
‘infer’, ‘interpret’, ‘paraphrase’, ‘rewrite’, ‘sum- INFLUENCE OF THE INTERVENTIONS
marize’, and ‘translate’. The affective domain has The systematic review database included a vari-
been defined as the area of learning devoted to able titled: ‘Level of influence’. This variable as-
attitudes and values (Morrison, Ross and Kemp, sessed the focus of the intervention in terms of
2007). Skills in the affective domain describe seven levels ranging from the individual to the
the way people react emotionally and their level global (see Figure 8).
of empathy. There are five levels in the affective
domain, from lowest order processes to highest: The results from the analysis of the level of influ-
receiving, responding, valuing, organizing and ence (see Figure 9) indicate that overall, a major-
characterizing. ity of the interventions, 253 manuscripts (over 80
per cent), focused on the individual level.16
Of the 302 manuscripts, 28 (9 per cent) did not
specifically provide any C4D objectives guiding Slightly less than two thirds, 186 manuscripts
the intervention. For multifaceted programmes, (62 per cent), addressed the interpersonal level
all of the objectives are rarely included. In many (for example, communication between parents
instances, the communication objectives noted in and children). Approximately half of the in-
a given manuscript related only to the specific interventions, 162 manuscripts (54 per cent), in-
tervention component being discussed. Another cluded organizational and community elements.
shortcoming that emerged while analysing com- The social/ environmental realm was included
munication objectives was that interventions did in approximately 36 manuscripts (12 per cent).
not distinguish between overall programme ob- A few of the manuscripts, especially those that
jectives and specific C4D objectives. In addition, consisted of reviews or analyses of individual
it is important to note that all the C4D objectives interventions, were categorized as belonging to
were written in ‘cognitive’ terms with no reference the ‘global’ realm as they attempted to summa-
to addressing the affective domain. As far as the rize and synthesize evidence from a wide range
cognitive objectives were concerned, the empha- of individual interventions. It is important to
sis was on lower-level domains such as knowledge note that although these are categorized as being
and comprehension. For example, verbs such as global in nature, the individual interventions be-
‘awareness’, ‘understanding’, and ‘increasing’ were ing reviewed were likely to belong to one of the
more commonly utilized as compared to higher other ‘levels of influence’.
order cognitive skills such as ‘examining’, ‘assess-
FIGURE 8: Levels of influence included in the systematic review database
• Individual/ intrapersonal (one person)

• Interpersonal (couples, family)
• Community (neighbourhoods, villages, localities)
• Organizational (schools, law enforcement, workplace)
• Social/ environmental (grassroots activism)
• Policy (government – executive, legislative, judicial)
• Global level
30 Communication for development approaches to address violence against children: A systematic review
KEY FINDINGS
FIGURE 9: Results on levels of influence from the manuscripts included in the

systematic review database
100%
80%
60%
40%
20%
In Guinea, a girl looks up

0% from writing in a notebook
Individual Interpersonal Community Organizational Social Policy Global while attending class at
(n=253) (n=186) (n=162) (n=151) (n=36) (n=69) (n=11) the Mangalla school, in
the town of Guéckédou,
Guéckédou Prefecture.
© UNI CE F/NYHQ2 015 -0 570 /D E MUN
KEY FINDINGS
FIGURE 10: Examples of C4D interventions influencing policy
Inter-American Development Bank efforts in Colombia and Uruguay
The Inter-American Development Bank’s efforts in victims. There was no evaluation data to correlate
Colombia and Uruguay from 1996 to 2006 aimed reductions in violent behaviour with interventions
at rebuilding trust in criminal justice institutions working with at-risk youth, although the results sug-
and sought to: improve police performance; to im- gest moderate success in interventions that sought
prove the treatment of domestic violence victims; to to place youth, in the labour market and schooling.
strengthen civil society organizations; to reduce the The strategy to promote civil society’s involvement
potential for youth to become involved in crime; and included the creation of centres providing various
to raise social awareness. Results on strengthening legal services in both countries that met a large need
criminal justice systems, police reform and working as indicated by the significant cases handled by the
with victims of domestic violence in both countries centres. The results from the public awareness cam-
were positive. Police stations were effective in pro- paigns were mixed. Some campaigns changed at-
viding protection to victims, but were less adept at titudes towards violence while others were less
mediating cases of physical aggression against wom- successful. The main limitations noted by the authors
en. In Uruguay, project activities, including a public was that complex problems require multifaceted so-
awareness campaign and training of public officials, lutions, which can lead to the creation of ambiguous
resulted in a decrease in the number of households programs and less success. Moreover, these types of
reporting domestic violence incidents associated interventions can be less cost-effective than small-
with nine projects designed to improve services for scale and short-term interventions.
Communication campaign to change policy: Eastern Nigeria
This campaign used a ‘Community Action Cycle’ comparison group design, indicated that programme
approach to contribute to the elimination of female exposure was associated with the expected improve-
genital cutting (FGC) in the project communities by ments in all the pertinent indicators and that the mul-
challenging individuals and communities to examine timedia communication programme was effective in
their beliefs and values around FGC and encourag- changing FGC-related attitudes and promoting the in-
ing action towards eliminating the practice. A robust tention not to perform it. The authors reported some
evaluation, using cross-sectional surveys from ran- policy-level impact, with the Enugu State House of
domly selected households based on an intervention- Assembly passing a bill in 2005 to abolish FGC.
Some 69 manuscripts (23 per cent) reported that Bank’s violence reduction programs in Colom-
they had a policy-level component and described bia and Uruguay from Alda, Buvinic and Lamas
communication efforts that had been utilized in (2006). In this example, the overall country pro-
changing policies either through the advocacy gramme was designed to reduce violence using
focus of the intervention itself or as a result of a multi-pronged approach. The second example
the success of communication campaigns. Two comes from Eastern Nigeria on a multimedia
pertinent examples are noted in Figure 10. The campaign contributed to policy change in Enugu
first describes comprehensive country-level ef- State (Babalola, Brasington, Agbasimalo, Hel-
forts in an overview of Neighbourhood Peace- land, Nwaguma and Onah, 2006).
keeping: The Inter-American Development
KEY FINDINGS
ASSESSMENT OF PROGRAMME As the data presented in Table 6 indicates, 24 of

IMPLEMENTERS AND PARTNERS the 302 manuscripts (just under 8 per cent) did
The final component of programme design ex- not provide any information on the implementers,
amined the organizations involved in programme while some 50 manuscripts (17 per cent) noted
implementation and the various programme multiple implementing partners. Of the 302 man-
partnerships. One key limitation in the analysis of uscripts, 30 (approximately 10 per cent) indicated
this component is that it is based on a secondary a governmental agency as the key implementer,
review of manuscripts. Information on the imple- and an additional 30 (10 per cent) relied on in-
menting agencies and partnerships is limited to ternational non-governmental organizations. In
what was available in the manuscripts or what some 13 manuscripts (4 per cent) were led by local
could be deduced from reading them. In some research agencies. Given the vast number of peer-
cases, the manuscripts were devoted to describing reviewed journal articles included in this system-
the ‘processes’ involved in the design and imple- atic review, it is not surprising that the largest
mentation of interventions. This is especially true proportion of interventions, 99 manuscripts (close
of interventions that relied on community or- to 33 per cent), were implemented by universities.
ganizing and community-building approaches.
In these cases, the manuscript authors explicitly In 151 manuscripts (50 per cent) no partner-
provided information on the various implement- ships were mentioned. As noted before, this
ers and partnerships involved in the intervention. data is based on a review of published manu-
In other cases, the manuscripts focused on a dif- scripts and it is possible that the manuscripts
ferent aspect of an intervention, for example, les- did not mention specific partnerships, which
sons learned or evidence of effectiveness. In these may nevertheless have existed at the ground
instances, it is possible that not all of the interven- level. Partnership models comprised of mul-
tion implementers and partners were mentioned. tiple organizations appears to be the norm as
The codes employed in the systematic review demonstrated in 63 manuscripts (almost 21 per
database for the implementing agencies and part- cent) and of these partnerships, local and inter-
nerships were very similar and included: gov- national non-governmental organizations col-
ernment, private sector, local non-governmental laborating with government partners appear to
organizations, international non-governmental be the most common combination. Other com-
organizations, grassroots organizations, local and mon partners appear to be research organiza-
external research agencies, and universities. tions and universities.
TABLE 6: Implementing agencies and partners for manuscripts in the systematic

review
ORGANIZATIONS INVOLVED NUMBER OF PROPORTION IN NUMBER OF PROPORTION IN
MANUSCRIPTS DATABASE MANUSCRIPTS DATABASE
(N = 302) (N = 302)
IMPLEMENTING AGENCIES PARTNERS

Government 30 9.9 29 9.6
Private sector 3 1.0 3 1.0
Local NGOs 44 14.6 14 4.6
International NGOs 33 10.9 15 5.0
Grassroots initiatives 6 2.0 0 -
Research organizations 13 4.3 14 4.3
Universities/ schools 99 32.8 13 4.3
Multiple organizations 50 16.6 63 20.9
Missing 24 7.9 151 50
KEY FINDINGS
FIGURE 11: Summary of programme design elements
• Close to half of the manuscripts did not explicitly reference a conceptual model.
• Social ecological models were largely implicit. It is possible that conceptual models
are implicitly followed but not explicitly stated.
• When mentioned, individual and cognitive conceptual models were most com-
monly utilized.
• Programme objectives did not follow measurable criteria.
• Programme objectives were universally designed to reduce or eliminate harmful
practices and traditions, yet the C4D messaging often focused on positive changes
individuals could undertake. Thus, this creates a potential misalignment. Without
clearly linking programme objectives and C4D messaging, it is hard to understand
what the role of C4D really is in promoting social and behaviour change.
• Communication objectives were not often specified and when mentioned focused on
individual and cognitive dimensions.
• The level of influence in a majority of interventions is individually focused.
• Collaboration and partnerships for implementation with international and national
non-governmental organizations working closely with government were reported
as the norm. The academic nature of a systematic review likely explains the large
(Right) a child at the front number of university-supported interventions.
of a queue of young chil-
dren mimics the actions
of a woman showing him
how to wash his hands
properly in a village in
Forécariah Prefecture, in
Guinea.
© UNI CE F/NYHQ2 015 -1 285 /L A ROS E
KEY FINDINGS
SECTION 3:
Programme implementation
T
he analysis of programme implementa-
tion focused on the intended audienc-
es for the interventions, the strategic
communication approach noted in the inter-
vention, as well as the specific communication
channels used for interventions.
ASSESSMENT OF THE INTENDED

AUDIENCES FOR INTERVENTIONS
The code sheet for the systematic review listed a
variety of different intended audiences/ partici-
pant groups for interventions. Intended audience/
participant groups refers to the audience or indi-
viduals an intervention is trying to reach and does
not necessarily refer to the individuals who hold
the decision-making power to change a certain
behaviour.
A boy sits at his desk in

Three core groups of audiences emerged from 2. Another audience category included a classroom at Robertito
the review: professionals such as service provid- School at the Cerro Rico
Mines in the city of PotosÌ,
ers, teachers, law enforcement officials Bolivia.
1. The intended audiences included in the and healthcare workers.
reviewed interventions encompassed
general audiences such as children, 3. A final group of intended audiences in-
parents/ caregivers, the general public, cluded opinion leaders such as commu-
males, females, and in a few isolated nity influentials and policymakers.
cases, perpetrators. 17
TABLE 7: Intended audiences for manuscripts in the systematic review

INTENDED AUDIENCES NUMBER OF MANUSCRIPTS PROPORTION IN DATABASE (N = 302)
Children 215 71.2

Parents 178 58.9
Teachers 119 39.4
Key influentials 84 27.8
Service providers (social workers) 77 25.5
© UNI CE F/NYHQ2 014 -0 496 /MARKIS Z
Policymakers 75 24.8
Law enforcement 64 21.2
Caregivers 63 20.9
Health-care providers 46 15.2
General public 36 12.9
Perpetrators 2 0.7
KEY FINDINGS
FIGURE 12: Results on the strategic communication approaches from the systematic
review
100%
80%
60%
40%
20%
0%
Advocacy Mass media Community Interpersonal Capacity building Intermediated
(n=84) (n=93) (n=143) (n=181) (n=230) (n=6)
In the event that interventions targeted mul- ASSESSMENT OF THE STRATEGIC

tiple intended audiences, the database included COMMUNICATION APPROACH
multiple columns to note all of the mentioned The systematic analysis database focused on six
audiences. Table 7 provides information different types of strategic communication ap-
on the individual audiences included in the proaches. The database included multiple col-
interventions. umns to note all of the mentioned approaches.
Analysis of the reported data by approach is pro-
The review found that 71 manuscripts (24 per vided in Figure 12. The results indicate that 254
cent) described interventions targeting one type of the manuscripts (84 per cent) reported the use
of audience. An additional 63 manuscripts (21 of more than one approach to meet the interven-
per cent) focused their efforts on two audience tion’s objectives.
categories. What stands out is that most inter-
ventions are directed towards one group and do Information on the strategic approach shows
not necessarily consider primary, secondary and that 84 interventions (less than 33 per cent)
tertiary audience segments. used advocacy and 93 interventions (less than
33 per cent) used mass media as their main
communication vehicle. In all, 143 manuscripts
FIGURE 13: Results on the communication channels used in the systematic review
100%
80%
60%
40%
20%
0%
Television/ film Radio Print Internet Mobile Local media Counselling Public
(n=84) (n=39) (n=179) (n=48) technologies (n=48) (n=223) forums
(n=12) (n=63)
KEY FINDINGS
(47 per cent) were based on community-based of communication. An additional 96 manu-

approaches. A majority of the interventions, scripts (about 31 per cent) described interven-
230 manuscripts (over 76 per cent), relied on tions using two channels of communication.
interpersonal training as a means to commu- This data potentially shows that almost two
nicate their information regardless of their thirds of interventions do not utilize multiple
intended audiences/ beneficiaries. The high channels to meet the needs of diverse and seg-
number of interventions focusing on capacity mented audiences.
building/ training might be related to the fact
that many of the reported activities were based Figure 13 provides information on the spe-
out of universities in the United States of Amer- cific channels of communication utilized. As
ica, which are oriented more towards training the data indicate, a majority of the interven-
and education, and have a more circumscribed tions, 223 manuscripts (74 per cent), employed
focus on large-scale mediated campaigns. The interpersonal communication and counsel-
mediated campaigns included in the review ap- ling as their primary approach, and in most
peared to be from other parts of the world. cases counselling was accompanied by the use
of printed materials (179 manuscripts or 59
ASSESSMENT OF COMMUNICATION per cent use print materials). Mass media ap-
CHANNELS proaches such as television and film were uti-
Communication channels utilized by the inter- lized in 84 manuscripts (28 per cent), and radio
ventions within the systematic review ranged in 39 manuscripts (13 per cent). Community-
from mass media sources such as television/ based local media appeared in 48 manuscripts
film, radio, print; new communication tech- (16 per cent), and public forums as advocacy
nologies like Internet and mobile phones; lo- tools appeared in 63 manuscripts (22 per cent).
cal media such as street theatre performances; New technologies and mobile forms of commu-
counselling, either expert or peer, and advocacy nication such as cellular phones were used to a
interventions such as public forums and com- lesser degree, although cross-tabulations with
munity dialogue. A review of the information the year of publication show a growing reliance
indicates that 82 manuscripts (27 per cent) de- on new technologies in later years.
scribed interventions utilizing only one channel
FIGURE 14: Summary of programme implementation findings
• In line with the programme design-related findings that interventions utilize individ-
ual conceptual/ theoretical models, the intended audience for most interventions is
directed towards individuals rather than dyads or groups.
• Interventions commonly targeted children and parents and did not report primary,
secondary and tertiary audience segments.
• Campaigns using media channels such as television, radio and print appear to be
more commonly utilized in developing countries – United States-based interventions
follow training/ capacity building trajectories led by university-based researchers fo-
cusing on violence in institutionalized settings, for example in schools or within small
community settings.
• The focused nature of and academic involvement in the interventions from industrial-
ized countries allows for higher levels of systematic evaluation and documentation.
• The most commonly utilized C4D approaches are interpersonal and capacity build-
ing. The communication channels most frequently utilized are also interpersonal.
• Interventions tend to use one or two channels of communication contrary to a multi-
ple-channel approach designed to reach a wider range of audiences.
• Utilization of mediated technologies show an increase over time.
KEY FINDINGS
SECTION 4: The data on evaluation design (Table 8) shows

Overall programme evaluation that 48 manuscripts (16 per cent) utilized the
processes most widely accepted and most robust evalua-
tion design (randomized controlled trials). An
T
he primary research question this re- additional 45 manuscripts (15 per cent) relied
view was designed to answer pertained on the most feasible and still robust programme
to effectiveness. Therefore, this sys- evaluation methodology: the pre and post with
tematic review placed substantial attention case-control design. An additional 41 manu-
on the evaluation components of the manuscripts (14 per cent) used only a pre and post test
scripts. Specifically, it examined the following design and an even smaller number, 13 manu-
evaluation aspects of programmes: evaluation scripts (4 per cent), utilized a case control evalua-
designs, research methods, analysis frame- tion design without pre and post measurements.
works, sampling related information, and in- The most common evaluation design found in
dicators as reported in the manuscripts. 85 manuscripts (28 per cent) consisted of obser-
vational studies featuring qualitative data with
The evaluation information was broadly catego- an additional 12 manuscripts (4 per cent) includ-
rized as formative, process or impact.18 Of the 302 ing a combination of qualitative and quantitative
manuscripts included in the database, 51 (17 per methods.
cent) discussed formative research, 43 (14 per cent)
included information on process evaluation, and ASSESSMENT OF RESEARCH
227 (75 per cent) included impact evaluation infor- METHODS
mation. It is evident that there is little information The review also included a broad categoriza-
on formative and process evaluation utilized in the tion of different research methods: quantitative,
interventions being examined. Despite this limi- qualitative, mixed (combination of qualitative and
tation, wherever feasible, the information in this quantitative techniques) and, finally participatory
section is disaggregated by formative, process and methods. Methodological information was catego-
impact evaluation. rized by formative, process and impact assessment
(Table 9).
ASSESSMENT OF THE OVERALL
EVALUATION DESIGNS Interestingly, the utilization of qualitative meth-
To assess the overall evaluation designs, manu- ods was more common at the formative and pro-
scripts were categorized based on the following cess phases than the use of quantitative methods
evaluation designs: randomized controlled trial; alone. Mixed methods are also utilized to a fair
pre- and post-test design; case control design; degree in these two evaluation phases. When it
observational study (quantitative); observational comes to impact evaluations, more manuscripts
study (qualitative). report using quantitative methods (116 manu-
scripts or 38 per cent) as compared to qualita-
TABLE 8: Evaluation design for manuscripts in the systematic review

EVALUATION DESIGN NUMBER OF MANUSCRIPTS PROPORTION IN DATABASE (N = 302)
Randomized controlled trial 48 15.9

Pre and post 41 13.6
Case control 13 4.3
Observational quantitative 37 12.3
Observation qualitative 85 28.1
Policy review 7 2.3
Observational both 12 4.0
Pre post case control 45 14.9
KEY FINDINGS
TABLE 9: Research methods for manuscripts in the systematic review

FORMATIVE EVALUATION PROCESS EVALUATION IMPACT EVALUATION
Number of Proportion Number of Proportion Number of Proportion

manuscripts in database manuscripts in database manuscripts in database
(n = 302) (n = 302) (n = 302)
Quantitative 13 4.3 15 5.0 116 38.4
Qualitative 24 7.9 18 6.0 80 26.5
Mixed methods 12 4.0 10 3.3 41 13.6
Participatory 11 3.6 4 1.3 9 3.0
tive methods (80 manuscripts or 27 per cent). To The results in Table 10 indicate that formative re-
some extent this might be a function of the fact search analysis tends to focus on thematic reviews
that specific journals prefer manuscripts uti- and involves univariate analyses, i.e. the descrip-
lizing quantitative methods. The utilization of tion of individual variables such as knowledge and
participatory methods was very low. Of the 302 attitudes. The high use of thematic analysis meth-
manuscripts analysed, only 24 (8 per cent) men- ods is not surprising given the high proportion of
tioned the use of participatory methods. Broken qualitative methodologies used during formative
down by type of evaluation: only 11 manuscripts research to uncover critical themes to be included
cited use of participatory methods during for- in interventions. However, the lack of comparative
mative evaluation, four during process evalua- bivariate or multivariate analysis indicates a lack
tions, and nine during impact evaluations. of specificity and sophistication in terms of under-
taking any causal or behavioural analysis. Process
ASSESSMENT OF ANALYSIS evaluation analysis methods focus on the use of
FRAMEWORKS univariate and comparative techniques. It appears
The systematic review database categorized that most process evaluations involve small sam-
commonly used statistical techniques for quanti- ple sizes and focus on descriptions of quantitative
tative studies as follows: univariate (frequencies, information and comparisons of qualitative data.
counts), bivariate (comparisons utilizing two Close to a third of the impact evaluations utilize
variables) and multivariate (three or more vari- multivariate techniques, indicating that mostly
ables controlling for background and confound- quantitative data is being analysed while control-
ing variables) techniques. Qualitative analysis ling for background and confounding variables.
frameworks included the following categories: Four out of 10 impact evaluation manuscripts (40
content analysis, thematic analysis and compara- per cent) relied on simple descriptive information
tive analysis methods. to make a case for effectiveness; this is worrisome
considering the complexity inherent in implement-
ing and evaluating interventions.
TABLE 10: Analysis methods for manuscripts in systematic review

FORMATIVE EVALUATION PROCESS EVALUATION IMPACT EVALUATION
Number of Proportion in Number of Proportion in Number of Proportion in

manuscripts database manuscripts database manuscripts database
(n = 302) (n = 302) (n = 302)
Univariate 11 3.6 17 5.6 56 18.5
Bivariate 6 2.0 5 1.7 68 22.5
Multivariate 6 2.0 5 1.7 99 32.8
Content 9 3.0 7 2.3 30 9.9
Thematic 14 4.6 7 2.3 44 14.6
Comparison 4 1.3 30 9.9 35 11.6
KEY FINDINGS
from output mapping by large-scale programmes

to conducting a single focus group discussion
with six participants. Most of the process evalua-
tion results are reported from within structured
school-based programmes. In part this could be
because such programmes involve training and
educational components that are relatively easy
to monitor. Of greater importance perhaps is the
lack of information on process evaluation. The
information on sampling frameworks for the
process evaluation did not yield any generalizable
information as these frameworks appeared mostly
to be derived from the individuals included in a
given intervention and/ or involved the tracking
of routine data pertinent to the intervention.
A review of the sampling information among the

277 manuscripts (75 per cent) that conducted im-
pact evaluations showed that over eight out of 10
manuscripts (80 per cent) focusing on individual
interventions relied on primary data collection to
A boy plays with a dog
on the porch of Kharik- ASSESSMENT OF SAMPLING make a case for effectiveness. The manuscripts,
khyong Government
A review of the sampling information for the which were reviews or meta-analyses of existing
Primary School in Ran-
gamati, Bangladesh manuscripts that included formative research interventions, utilized secondary sources. Only two
information revealed that a majority, 51 manu- interventions reported relying on external data: one
scripts (17 per cent), sampled youth. In some intervention used tracking crime statistics as an
cases, samples involved children of specific ages evaluation mechanism (McGarrell et al., 2103) and
often denoted by school year and corresponding another project used programme records as a part
intended audiences for programmes. Very few of a terminal assessment (Pepler and Craig, 2011).
cases of multiple respondents/ participants were
evident. Only a handful of the studies that de- A majority of the impact assessments relied on data
scribed their formative research processes iden- collection from a single source and disaggregation
tified influentials or professionals like teachers if done was on the basis of gender (i.e. data collec-
as being a part of their sample. Very little infor- tion among boys and girls). Less than a quarter of
mation was available on the formative research the assessments reported involving multiple re-
sampling framework. When provided, the infor- spondents such as children, parents and commu-
mation included the utilization of community- nity leaders, or students and teachers. In line with
based focus group discussions, schools, work data collection techniques, the sampling frame-
sites, and the use of random digit dialling. Some work was often derived at the individual level; for
evidence of purposive sampling strategies was community-based programmes households were
evident, for example, organizations involved in used as a sampling frame, and for school-based
interventions that provided access to their ben- programmes classrooms served as the criteria for
© UNI CE F/BANA201 4-015 92/MAWA
eficiaries and child welfare staff providing access selection. In less than 10 per cent of the cases, the
to foster parents and children. sampling unit was conceptualized at a larger level,
and units like a school or a community were uti-
Among the 43 manuscripts (14 per cent) that in- lized for sampling and subsequent analysis. One
cluded process information, a wide variation in notable exception was randomized controlled tri-
sampling information was observed. Addition- als based in schools, which were able to randomize
ally, a range of methods was used to gather data, sampling at the school or classroom level.
KEY FINDINGS
FIGURE 15: Indicators from the manuscripts included in the systemic review
100%
80%
60%
40%
20%
0%
Individual Interpersonal Community Structural Output
(n=228) (n=43) (n=56) (n=33) (n=45)
ASSESSMENT OF INDICATORS Interpersonal level indicators included measure-

Another key topic covered by the evaluation comments of dialogue, discussion and conflict resolu-
ponent of this systematic review focused on the tion. For example, indicators that measured the
type of indicators being examined in the inter- proportion of individuals who engage in inter-
ventions (see Figure 15). The results indicate that personal communication and initiate discussion.
228 manuscripts (76 per cent) included indicators Interventions focused on changing issues with a
based on data collected from individuals, with- strong interpersonal foundation, such as gender-
out any attempt to aggregate at the family, school based violence, child abuse, and bullying pro-
or community level. All the other indicator types grammes, frequently employed these indicators.
were reported in anywhere from 33 to 56 manu- Several anti-bullying interventions used reports
scripts (anywhere from 10 to 20 per cent). from multiple sources including students, par-
ents, teachers, or independent observers to record
Qualitative analysis of the individual level indica- and validate data.19 Another intervention, Expect
tors revealed that the majority sought to reflect Respect, which focused on decreasing teen dating
change in the knowledge, attitudes, beliefs, values violence and sexual assault by improving healthy
and self-efficacy of the intervention’s target au- conflict resolution skills in a south central urban
dience as opposed to changes in behaviour. For area in the United States., measured the frequency
example, most of the anti-bullying programmes of perpetration and victimization in dating rela-
included a measurement of students’ knowledge tionships (Ball et al., 2012).
of bullying and awareness of different types of
bullying. Many interventions also performed a Community level indicators were less common
pre- and post-test questionnaire measuring stu- than individual level indicators and were typi-
dents’ self-efficacy in handling hypothetical bul- cally utilized for programmes focused on ad-
lying scenarios. Other individual level indicators vocacy or changing social norms. For example,
included measurements of an individual’s atti- indicators that focused on public denouncement
tudes toward early marriage and family planning, of female genital mutilation/ cutting (FGM/ C)
teacher’s perceptions of class social dynamics, and as an intermediate measure to examine decrease
an instructor’s evaluation of individual students’ the incidence of FGM/ C.20 A 2007 report by the
on their acquisition of taught skills. As many of International Centre for Research on Women
the interventions were informed by cognitive-be- focused on the following indicators of success:
havioural theories, changes in knowledge, beliefs transformation of harmful social norms, support-
and attitudes were often the primary indicators of ing and scaling up of community programmes,
intervention success. and increased access to girls’ education. Perkins,
Craig and Perkins (2011) describe a social norms-
KEY FINDINGS
The fourth type of indicator, structural factors,

was the least commonly used. Programmes us-
ing structural factors were often focused on
advocacy and measured change in local, state or
national policy regarding their area of interest.
Boothby and Stark (2011) examined Surveillance
in Child Protection Systems Development data
through a case study in Indonesia to identify a
‘road map’ of a national child protection infor-
mation system in Indonesia and included advo-
cacy to increase collaboration as a key indicator.
The authors indicated that the creation of a sin-
gle information system (instead of several differ-
ent branches and different databases), a common
language, and partnerships were critical.
The final indicator, outputs, typically reflected

intervention fidelity or fulfilment of specific in-
tervention benchmarks. Some examples of output
level indicators include: the number of training
workshops offered, the number of participants,
among others. Some specific examples of inter-
ventions describing their utilization of interven-
tion fidelity indicators are provided below:
• An evaluation of a university-commu-
nity partnership to promote a safer
school environment through the estab-
lishment of a Safe School Task Force,
measured the accomplishments of this
task force through mapping solutions,
outreach activities, and a student-
led movie and message night (Adler,
Chung-Do, and Ongalibang, 2008).
• Barron and Topping (2011) used video

Teacher Antonio Mendoça
engages Usher Sanca,
based intervention among adolescents in five analysis of interactions to study the
seven, in learning about middle schools to reduce bullying. The interven- fidelity of a sexual abuse prevention in-
letters at the UNICEF-sup-
ported Ponta Nova Unified
tion used print media posters to display accurate tervention. They specifically examined
School, in the village of norms regarding bullying attitudes and behav- indicators intervention fidelity such
Ponta Nova, Oio Region,
© UNI CE F/P FPG20 14P- 0622/ LY NCH
Guinea Bissau. iours from baseline survey results. This interven- as: presenter body language and voice
tion utilized pre- and post-measures to compare tone, presenter behaviour manage-
the following indicators: perceptions of peer bul- ment, presenter explanation behaviour
lying, pro-bullying attitudes, personal bullying of and the giving of instructions, length
others, victimization as well as support for report- of turn-giving, listening and respond-
ing bullying to adults. ing, responding to and affirming disclo-
sures and adherence to content.
KEY FINDINGS
FIGURE 16: Summary of overall programme evaluation processes findings
• Interventions do not report on forma- utilization of robust evaluation designs

tive research processes in detail, with (randomization at the school or class-
only 51 manuscripts (17 per cent) re- room level), checking for implementa-
porting on formative research. tion fidelity.
• There is a marked paucity of informa- • While there may be a bias on the part
tion on how formative research is uti- of journals to accept more unconven-
tional ‘participatory’ methodological
lized for programme design.
studies, there is nevertheless seri-
• Process evaluations (monitoring) ous under-utilization of participatory
are uncommon with only 43 manu- methods with only 24 of the manu-
scripts (14 per cent) reporting process scripts (under 4 per cent) reporting
evaluations. the use of participatory evaluation
• When conducted, process evaluations methodologies.
(monitoring) are narrowly defined as • Lack of information on sampling

a means to measure intervention out- frameworks and small sample sizes
puts and not as a vital mechanism for combined with the most basic anal-
mid-course corrections or a method of ysis techniques reported in most
tracking outcomes. manuscripts bring to light the lack of
• Reliance on qualitative observational specificity and sophistication in the
data as reported by 85 manuscripts (28 evaluation data being collected and
per cent) makes an overall weak case leads to questions about quality and
for attribution. effectiveness.
• Randomized controlled trials were re- • In line with the individual level objec-
ported in 48 manuscripts (16 per cent). tives and audience segments, inter-
An additional 45 manuscripts (15 per vention evaluation indicators mostly
cent) reported a pre-post case control focus on data collected from individu-
design (without randomization). The als without attempts at aggregation.
lack of randomization is not surprising • In line with the individual-level cogni-
given the field-based nature of inter- tive conceptual basis of most interven-
vention evaluations. tion, indicators are mostly individually
• School-based intervention designs based knowledge, attitudes and prac-
lend themselves more easily to the tice-related measures.
• A randomized controlled trial to estab- In an attempt to redress structural

lish a ‘Communities that Care’ preven- inequality and institutionalized vio-
tion system, designed to reduce levels lence, the ‘Creating New Choices:
of risk associated with problem behav- Violence Prevention Project in Aus-
iours, such as adolescent drug use and tralia’ (Sidey and Lynch, 2001) spe-
delinquency, used participant attendance cifically examined satisfaction with
records, changes in participant attitudes, the intervention model as a means to
and knowledge to establish benchmarks examine initial effectiveness.
and milestones to measure intervention
outputs (Quinby et al., 2008).
KEY FINDINGS
SECTION 5: • A child abuse intervention in Massa-

Thematic analysis of key chusetts, USA, showed clear evidence
results by type of research of concern for child abuse as a serious
problem and community-level desire to
T
his section of the report provides infor- be involved in addressing the issue of
mation from a detailed thematic analy- child abuse (Schober, Fawcett and Ber-
sis of the key results and discussion nier, 2012).
sections of the manuscripts included in the sys-
tematic review. The information is presented in • Formative research on a proposed in-
the following categories: formative, process and tervention related to child labour and
impact assessment results. Again, key themes sexual exploitation of children in Thai-
were derived using a grounded theory approach land revealed that contrary to common
in which categories and themes emerge from misconception, birth position, parental
the coded textual data from each individual marital instability and educational at-
manuscript included in the systematic review. tainment were significant predictors of
a girl entering hazardous labour rather
Given the large number of themes that emerged, than parental wealth (Taylor, 2005).
the themes were further clubbed into theoretical,
programmatic and methodological domains. It is Unfortunately, there is little reported effort to
important to note that the categorization of themes understand social and structural determinants
into theoretical, operational and methodologi- of change. Research data clearly shows the nega-
cal domains is not meant to imply that these are tive consequences of leaving broader concerns
discrete categories. There is a great deal of overlap unaddressed. A review of 87 child sexual abuse
within the themes in each domain and it is evident programmes notes that failures in prevention in-
that themes have implications both for program- terventions are often due to a lack of attention to
ming and evaluation. These divisions are simply social and political realities rather than a measure
meant to provide an overall structure and guid- of intervention effectiveness (Plummer, 2001).
ance for future programming and research. These
themes are highlighted first and then followed by Interestingly, there is adequate evidence that
examples from the systematic review database that formative research focuses on barriers and so-
further illustrate key points. To the extent possible, lutions. Such examinations, however, are often
relevant examples both from industrialized and de- couched in terms of individual level barriers and
veloping countries have been provided to help the solutions rather than social and structural deter-
reader make contextual connections. minants. Some notable exceptions are:
FORMATIVE RESEARCH RESULTS • Recent research on child marriage in

Of the 302 manuscripts included in the database, India identified a combination of indi-
51 (17 per cent) discussed formative research. vidual (e.g. lack of awareness), as well as
The key themes noted below are based on the re- social and structural factors (e.g. gender
sults from the specific manuscripts that included norms affecting girls’ value/ role in the
information on formative research. community, economic considerations
associated with poverty and dowry, and
SUMMARY OF FORMATIVE RESEARCH weak enforcement of the Prohibition of
RESULTS: THEORETICAL Child Marriage Act 2006 as perpetuating
Key results from formative research were noted the practice of child marriage.
in terms of individual levels of knowledge, at-
titudes and practices. The following examples • Additional research on the same topic
provide evidence demonstrating the impor- has shown a societal preference for
tance of formative research: male progeny, and the perception that
KEY FINDINGS
child marriage protects the virtue of girls • Formative research from an anti-bul-
reinforces the practice (ICRW, 2011). lying intervention revealed that verbal
harassment and the use of derogatory
• In another example, formative research language is a pervasive problem in
from South Africa on perceptions of United States schools leaving students
gender-based violence revealed that withdrawn, distracted, wounded and
poverty, ubiquitous gendered violence, even ready to turn to physical violence
sexually exploited children and unsafe (Wessler and DeAndrede, 2006).
recreational spaces were major themes.
Youth were identified as being con- • In Australia, formative research on an
sumed by issues of safety rather than intervention designed to implement
pursuing other developmentally appro- community-based responses to child
priate markers (Mitchell, 2003). These abuse showed that while child abuse
authors note that formative research of- is a serious social problem, it is poorly
ten does not focus on underlying issues. understood by the public on a number
of levels including: its true extent and
SUMMARY OF FORMATIVE RESEARCH nature; the short- and long-term social
RESULTS: OPERATIONAL and financial costs of child abuse to
Formative research gives interventionists the children, families, and the community;
ability to understand intended audiences. This as well as knowledge of common per-
issue is critical for VAC interventions that ad- petrators of child abuse. In addition,
dress ‘perpetrators’ who may otherwise be sub- there was a significant level of anxiety
ject to unfair labels and blame. on the part of adults in accepting the
legitimacy of children’s rights. The au-
• A research study with United States mid- thors concluded that clear social and
dle school youth found that these youth political commitments to children are
have specific and strongly-held percep- necessary to prevent child abuse (Tuc-
tions about violence and perpetrators of ci, Mitchell and Goddard, 2001).
violence. However, despite acknowledg-
ing acts deemed violent by their own The literature suggests that formative research
definitions, these youth did not internal- can play an important role in terms of design-
ize these actions and did not view them- ing programmes that are community owned.
selves as ‘violent’. The authors conclude For example:
that this finding has significant implica-
tions for behaviour change, especially in • A comprehensive child protection forma-
a social marketing context (Quinn, Bell- tive research project in Northern Thai-
Ellison, Loomis and Tucci, 2007). land resulted in raising awareness of
child abuse, the creation of a volunteer-
• A South African intervention addressing led community child protection team,
male involvement in gender-based vio- and the development of a participatory
lence found that barriers to participation child protection model focusing on child
in the proposed intervention included protection at individual, family and com-
denial or minimization of gender-based munity levels. The model emphasized
violence (Ditlopo et al., 2007). participation from a variety of commu-
nity members and local organizations
Formative research can serve as a means to iden- (Auemaneekul, Senaratana, Juntarawijit,
tify key issues that must be addressed. Some ex- Sripichyakan and Ensign, 2009).
amples include:
KEY FINDINGS
• Formative research to help design a In addition, results from the systematic review
comprehensive violence prevention in- database demonstrate that formative research is
tervention at a city level found general especially important for programmes that wish
perceptions to be that youth violence is to utilize new communication technologies.
a complex problem with many risk fac-
tors. Many of the young people felt help- • Research by Constantino, Crane, Noll,
less to change the environment and were Doswell and Braxter (2007) on the effica-
doubtful that things would ever change. cy of utilizing email mediated interactions
This research resulted in the identifica- revealed that email is a feasible and ac-
tion of five areas of collaboration: create ceptable way to provide support and in-
after-school and summer programmes, formation to abused women (interviews
increase knowledge of existing resources with children also yielded similar results).
for violence prevention, promote positive Results showed that participants were
involvement of police with children, pro- interested in the device, compliant with
vide parent education to reduce violence, learning and using it, and responded to
and develop strategies for influencing interventionists’ email questions.
media portrayal and use of violence (Mey-
er, Cohen, Edmonds and Masho, 2008). • Recent survey research associated with
cyberbullying revealed that effective in-
Formative research focusing on audience prefer- terventions could involve taking away an
ences emerged as an important method of iden- offender’s computers and cell-phones
tifying communication channels that might not and restricting access to social network-
have been previously considered. For example: ing sites (Kraft and Wang, 2009).
• Research by Self-Brown et al. (2008) to Formative research can feed directly into mes-
assess participants’ comfort level with sage and material design. Analysis of the manu-
materials associated with the Darkness to scripts that include formative research report
Light (D2L) child sexual abuse prevention that pre-testing of pilot messages and materials
campaign found video-based informa- is a critical part of the formative research pro-
tion to be especially useful as a way to in- cess. Several examples from the systematic re-
crease knowledge about the issue; video view provide information to reinforce this point:
and print materials were seen as a viable
way to make the topic more approach- • Formative research designed to identify
able. At the same time participants ‘ideal’ foster parenting qualities was
wanted more information, specifically used to develop media messages cater-
on ways to talk to and educate children ing to identified characteristics of the
about the topic as well as culturally-ap- high-quality foster parents (Duerr Ber-
propriate materials in local languages. rick, Shauffer and Rodriguez, 2011).
• Survey-based formative research on a • An evaluation of specific video mes-

South African intervention aimed at in- saging demonstrated that mothers
creasing male involvement in address- easily grasped the scope of childhood
ing gender-based violence indicated injuries through this intervention tech-
that respondents recommended edu- nique. Graphic images, mood-inducing
cational workshops and entertainment sound effects, and testimonials by other
education drama to help change atti- mothers all helped to achieve project
tudes towards gender-based violence objectives (Morrongiello, Zdzieborski,
(Ditlopo et al., 2007). Sandomierski and Lasenby-Lessard,
2009).
KEY FINDINGS
• Formative research results from a supports exist to help address VAC

media-based intervention to address (Payne and Williams, 2008).
sexual abuse showed that potential
participants indicated no discomfort The systematic review confirms the crucial role of
in viewing the media materials or the formative research in the design of culturally sen-
booklet, and there was overwhelming sitive interventions. Findings from the systematic
agreement that media is an effective review reveal a need for more specific information
means to raise society’s awareness, around communication related barriers and bot-
increase people’s knowledge of child tlenecks to desired change. Language, education,
sexual abuse, and make the topic of and cultural barriers may hinder the effectiveness
child sexual abuse more approachable. of C4D approaches addressing VAC, according to
At the same time, individuals noted that the data. The review highlighted a few examples
knowledge is not enough to prevent of formative research that reportedly contributed
child sexual abuse. A combination of to making interventions more culturally-sensitive.
mass media with broad-based commu-
nity involvement along with parental • The authors of a dating violence inter-
education and involvement was viewed vention for urban youth linked its suc-
as necessary to curb child sexual abuse cess to formative research that allowed
(Self-Brown, Rheingold, Campbell and it to be culturally sensitive to audience
de Arellano, 2008). needs (Akeo et al., 2008).
Formative research has proven effective in deter- • An evaluation of the UNICEF-led Mee-
mining key audiences, as well as partners who na Programme in South Asia indicated
should be involved in interventions addressing that one key factor in its success was its
VAC. For example: underpinning in research and develop-
ment processes (Chesterton, 2004).
• Participants in formative research on the
feasibility of a mental health promotion SUMMARY OF FORMATIVE RESEARCH
intervention in American schools indi- RESULTS: METHODOLOGICAL
cated that participants ranked teachers Though rarely utilized in the literature reviewed,
as the key stakeholder group that would community-based participatory research (CBPR)
be most likely to support and implement has proven effective in obtaining a comprehen-
interventions (Evans, Mullett, Weist and sive picture of barriers and motivators when de-
Franz, 2005). signing interventions. Of note are the following
manuscripts:
• Research on ecological approaches to
designing interventions associated with • Leff et al. (2010) report on the results
youth violence showed that coalition- from primary and secondary data using
building in community-based interven- CBPR to develop a youth violence pre-
tions has a better chance of success vention intervention. The formative re-
when careful formative research is con- search consisted of focus groups to gain
ducted to understand social as well as a better understanding of the strengths
inter- and intra- organizational networks and challenges within the local commu-
(Bess, Speer and Perkins, 2012). nity and experiences with prior violence
prevention interventions, as well as to
• For community-based programmes, elicit specific indicators of programmatic
formative research in the form of com- and/ or community success in regard to
munity resource mapping is also criti- violence prevention. Literature reviews
cal to understand where and what and pilot testing of the initial interven-
KEY FINDINGS
tion revealed that most youth thought • Another noteworthy example focused
it would be easy to implement the skills on building social capital through neigh-
and strategies they had learned in the bourhood mobilization. Formative re-
intervention. Then, prior to implemen- search involved an asset mapping
tation, site partners and community component and service delivery compo-
partners along with the academic team nent, as well as institutional and organi-
members worked together to finalize the zational mapping. The data from these
intervention design and all additional ma- efforts was collected by community
terials. Large-scale implementation was members and housed in the centre to be
discussed at a community symposium. used by grassroots groups. Based on the
The academics wanted to conduct a ran- information generated from the forma-
domized controlled trial, but community tive research, intervention implemen-
participants expressed a strong desire tation involved a multi-pronged effort
that all sites should receive the full inter- including a dedicated space for social
vention. Thus, a staged cluster random- and community services and activities,
ized trial was developed instead. As this as well as educational initiatives for mar-
example shows, not just the intervention ginalized populations to improve collabo-
designs, but also the evaluation proce- ration between parents and schools in
dures were based on collaboration. order to best meet the needs of chil-
dren (Payne and Williams, 2008).
FIGURE 17: Summary of findings from formative research results
THEORETICAL • Plays a crucial role in designing interventions that

Key results from formative research were noted in are community-owned and culturally sensitive.
terms of individual levels of knowledge, attitudes and • Constitutes an important method of identifying
practices, and there is considerable evidence of the communication channels that might not have
importance of such research. been previously considered. This is especially
important for interventions that wish to utilize
There is adequate evidence that formative research new communication technologies.
focuses on barriers and solutions. However, such ex- • Feeds directly into message and material design.
aminations are often couched in terms of individual • Proves effective in determining key audiences, as
level barriers and solutions. Unfortunately, there is well as partners who should be involved in pro-
little reported effort to understand social and struc- grammes addressing VAC.
tural determinants of change.
METHODOLOGICAL
OPERATIONAL Although rarely utilized, community-based participa-
The literature reinforces the importance of conduct- tory research (CBPR) has proven effective in obtaining
ing formative research. Formative research: a comprehensive and complete picture of barriers and
• Gives programmers the ability to understand in- motivators when designing interventions and should be
tended audiences. considered as part of any formative research.
• Serves as a means to identify key issues that
must be addressed.
KEY FINDINGS
PROCESS EVALUATION RESULTS SUMMARY OF PROCESS EVALUATION

Of the 302 manuscripts included in the da- RESULTS: OPERATIONAL
tabase, 43 (14 per cent) included information Process evaluation is especially critical for
on process evaluation or monitoring. The key large-scale programmes where implementation
themes noted below are based on thematic may vary due to external reasons. For exam-
analysis of these manuscripts. ple, Rijsdijk et al. (2011) describe the process
evaluation for a comprehensive sex education
SUMMARY OF PROCESS EVALUATION intervention targeting adolescents in Uganda.
RESULTS: THEORETICAL This intervention was designed to empower
Process evaluation allows interventions to make and support young people in making their
a link between the implementation of specific own informed decisions about sex by chang-
intervention activities and subsequent impact, ing attitudes about sexual coercion, to increase
which in turn makes a stronger case for employ- self-efficacy in dealing with situations where
ing C4D approaches. For example: unwanted sex could happen, and to increase the
intention to deal with unwanted sex and force.
• The ‘Communities that Care (CTC) Pre- The process evaluation data from this interven-
vention System’ established in 12 com- tion showed that full implementation was more
munities reported that where process effective than partial implementation (which
evaluation was conducted, on average 90 resulted from not having enough computers).
per cent of the core components achieved
high implementation fidelity. These held Process evaluation can disaggregate measure-
over time and quality was maintained ment of implementation among diverse audienc-
over five years (Quinby et al., 2008). es. For example:
• In another example, process evaluation • Results from a foster care interven-

of a positive youth development inter- tion for children with clinically signifi-
vention in Hong Kong utilizing different cant behaviour problems supports the
cohorts by Law and Shek (2012) indicat- efficacy of an enhanced intervention
ed high overall intervention adherence among urban, primarily African-Amer-
and implementation quality. In addi- ican foster parents. The findings sug-
tion, the authors were able to correlate gest that specific enhanced elements
intervention implementation with qual- of standard protocol (including weekly
ity and success. Multiple regression consultations by a practitioner trained
analyses further showed that both im- in behavioural interventions) can po-
plementation process and intervention tentially reduce the level of behaviour
adherence were significant predictors problems in this high-risk population
of intervention quality and success. of foster children (Leathers, Spielfogel,
McMeel and Atkins, 2011).
• The process evaluation of an FGM/ C
intervention in Egypt found that men • Another process evaluation of a cogni-
and women exposed to the intervention tive-behavioural therapy approach also
retained more information regarding found that the use of the intervention
the negative health consequences of and abuse-specific content was consis-
the traditional practice when compared tently applied across a range of diverse
with those who were not exposed to client characteristics, thereby building
the intervention, thereby establishing a case for the approach’s cross-cultural
links between intervention implementa- relevance (Kolko, Iselin and Gully, 2011).
tion and effectiveness (Barsoum, Rifaat,
El-Gibaly, Elwan and Forcier, 2009).
KEY FINDINGS
• In an intervention designed to study re- checklist (Hahlweg, Heinrich, Kuschel,

lational aggression in schools, process Bertram and Naumann, 2010).
results suggested strong acceptability
and feasibility of the intervention, espe- • Methodological variation is evident
cially for girls (James et al., 2011). in the use of self-reporting to exam-
ine initial outcomes of PeaceBuild-
Process evaluation can indicate initial success ers, a universal school-based violence
and serve as a means to validate and expand in- prevention intervention. In this case,
terventions. For example, the results of an inter- teacher interviews report on their use
vention for maltreating fathers showed that in of the PeaceBuilders curriculum and
the first year, 105 men were referred to the in- expressed agreement or disagreement
tervention, greatly outnumbering the number of with the quality of implementation, as
treatment slots available during the same period well as hypothesized effectiveness, us-
(Scott and Crooks, 2007). The process evaluation ing Likert scaled measures. Overall, the
data tracked the source of referrals and found authors found consistent behaviour ef-
that the intervention even included a few self- fects when comparing teacher self-re-
referrals. Tracking of the numerous requests for ports between intervention and control
training and implementation from other com- schools (Flannery et al., 2003).
munities suggested that the need for similar
programmes was not unique to the initial project • In a sexual abuse prevention interven-
community. tion, fidelity and intervention effec-
tiveness was measured through video
SUMMARY OF PROCESS EVALUATION analysis of interactions between par-
RESULTS: METHODOLOGICAL ticipants and facilitators (Barron and
Process evaluation is commonly used as a way Topping, 2010).
to examine fidelity, i.e. the extent to which inter-
ventions are implemented according to plan. The Process evaluations can be conducted regard-
examples of fidelity included in the systematic less of the nature and scope of the intervention.
review varied widely in terms of methods (quali- While there were relatively few examples of pro-
tative, quantitative), techniques (interviews, ob- cess evaluations, the variety of projects ranged
servations, video recordings) and the extent of from one-on-one counselling efforts to small
process monitoring. For example: school-based interventions to large-scale media
efforts.
• One study aimed at increasing knowl-
edge of sexual abuse among elemen- • In a process evaluation of a one-on-
tary schoolchildren in Hawaii included one counselling intervention utilizing a
observations for each individual les- cognitive-behavioural therapy model,
son in order to ensure that school staff clinicians in a child protection centre
were appropriately trained and sup- reported high levels of use for all in-
ported to teach the sensitive curricu- tervention content items. The authors
lum (Baker, Gleason, Naai, Mitchell conclude that the intervention had been
and Trecker, 2013). incorporated and reasonably main-
tained in the practitioners’ repertoires
• In a randomized controlled universal well after training ended (Kolko, Iselin
prevention trial assessing long-term and Gully, 2011).
outcomes of a positive prevention in-
tervention found that 50 per cent of • The process evaluation for a interven-
sessions consistently performed fideli- tion titled V.I.K. (Very Important Kids), a
ty checks through a protocol adherence school-based intervention designed to
reduce teasing and unhealthy weight-
control behaviours, demonstrated that
outreach can be feasibly done, at a rela-
tively low cost, and can engage students,
parents and staff (Haines, Newmark-Sz-
tainer, Perry, Hannan and Levine, 2006).
• One unique process evaluation model

involved the evaluation of a ‘place-
based strategy’ via the creation of a
Youth Violence Prevention Centre in
Flint, Michigan in the United States.
This centre created infrastructure, net-
working opportunities and intervention
strategies that facilitated and organized
efforts to address youth violence. The
process evaluation model allowed for
the measurement of organizational em-
powerment in community mobilization
as a means to address youth violence
Nine-year-old Nashala
(Griffith et al., 2008). • A review by Cornell (2011) of a student Maharjan, holding a
schoolbook, sits on a
threat assessment guideline interven-
piece of timber in the
Process evaluation helps identify Type 3 error tion designed to prevent school-based rubble and ruins of a
destroyed home, in the
(i.e. when interventions are not implemented violence in the United States compared Khokana neighbourhood
according to plan).21 In updating a meta-anal- intervention and control students to in Lalitpur District, in
Kathmandu Valley, one of
ysis of school-based programmes designed to discover that students in intervention the areas hardest hit by
address aggressive and disruptive behaviours, schools were approximately four times the massive earthquake
in Nepal.
Wilson and Lipsey (2007) concluded that more likely to receive counselling ser-
school-based programmes had positive ef- vices; two-and-a-half times more likely
fects on relevant outcomes. However, different to receive a parent conference; one-
treatment modalities, for example, universal third as likely to receive long-term sus-
interventions versus those for specific sub- pension; and one-eighth as likely to
populations, produced largely similar effects receive an alternative school placement
and the authors concluded that effects were indicating that the intervention was
larger for interventions that were implemented effective.
according to plan. This underscores the im-
portance of intervention implementation as • Safe Neighbourhoods data on gun
the key to success and, in turn, allows us to offenses between 2000-2005 indicat-
build a case for process evaluation. ed a statistically significant decline
in firearm offenses in three of the
Process evaluation is useful in tracking chang- five programme sites (Kroovaud Hip-
es in outcomes. Traditionally, process evalu- ple, Frabutt, Corsaro, McGarrell and
© UNI CE F/NYHQ2 015 -1 121 /PANDAY
ations are considered tools to track outputs, Gathings, 2007).

and this was the most common use of process
evaluation in the reviewed manuscripts. How-
ever, we found a few notable exceptions that
moved beyond outputs to tracking outcomes.
These include the following:
KEY FINDINGS
FIGURE 18: Summary of findings from process evaluation results
THEORETICAL METHODOLOGICAL
Process evaluation allows interventions to make a • Process evaluation is commonly used as a way to
clear link between the intervention activities and im-
examine fidelity of implementation.
pact, which in turn makes a stronger case for employ-
ing C4D approaches. • Process evaluations can be conducted regardless
of the nature and scope of the intervention.
OPERATIONAL • Process evaluation helps identify Type 3 error
• Process evaluation is especially critical for large- (when interventions are not implemented accord-
scale interventions where implementation may ing to plan).
vary due to external reasons. • Process evaluation is useful in tracking changes in
• Process evaluation can disaggregate measurement outcomes and outputs.
of implementation among diverse audiences.
• Process evaluation can indicate initial success
and serve as a means to validate and expand
interventions.
© UNI CE F/UNI18 0551/ TINGT ING
Early childhood development

project in Chongqing, China.
KEY FINDINGS
IMPACT EVALUATION RESULTS • Efficacy of child abuse and neglect pre-

Of the 302 manuscripts included in the database, vention messages in a multi-pronged ef-
227 (75 per cent) included impact evaluation in- fort called the Florida Winds of Change
formation. The key themes from each group are reported that exposure was significantly
presented below, each with supporting examples related to measures of five out of six
from the database. outcomes: knowledge of child devel-
opment, knowledge of community re-
SUMMARY OF IMPACT EVALUATION sources for parents, attitudes towards
RESULTS: THEORETICAL prevention, motivation to prevent child
Impact evaluation data highlights the impor- abuse and neglect, action or behaviour
tance of utilizing an ideation model22 at the indi- to prevent child abuse and neglect (Ev-
vidual level and community level while looking ans, Falconer, Khan and Ferris, 2012).
to the social ecological model as a holistic frame-
work. There is ample evidence that C4D inter- Evidence of social change attributed to pro-
ventions are successful in promoting awareness grammes is limited, while the need for interven-
and knowledge. In addition, there are examples tions to address normative factors is imperative. As
in the systematic review database of interven- previously noted, a majority of the objectives for
tions that had positive outcomes but may have interventions are couched in terms of individual
been considered as failures for not achieving behaviour change. At the same time, the evidence
their narrowly defined ‘behavioural’ programme points to the complexity of addressing VAC and
objectives. Examining C4D approaches using an the need to tackle these issues across all levels of
ideation perspective is therefore likely to yield the social ecological model, specifically focusing on
critical information on changes in intermediate norms. The following examples showcase interven-
factors resulting in behaviour and social change. tions that addressed norms and social change.
Some examples from the database include:
• A domestic violence prevention interven-
• The school-based Gang Resistance tion among female Iranian high school
Education and Training (G.R.E.A.T.) students based on the PRECEDE-PRO-
intervention resulted in significant CEED Model23 indicated that this model
differences between G.R.E.A.T. and was effective since it focused on building
non-G.R.E.A.T. students on five out of life skills through education. (Soleiman-
32 outcomes, of which four were attitu- Ekhaiari, Shojaeizadeh, Foroushani, Gho-
dinal (Peterson and Esbensen, 2004). franipour and Ahmadi, 2013).
• A report by Brady (2007) examined the • A study involving men in research on

Ishraq Programme in rural West Egypt, gender norms reveals that a gender-
an intervention designed to provide transformative approach and promotion
new opportunities to adolescent girls in of gender-equitable relationships be-
socially conservative settings, and re- tween men and women are more effec-
ported that Ishraq girls expressed a de- tive in producing behaviour change than
sire to marry at older ages and a desire narrowly focused interventions, as are
to have a say in choosing a husband. interventions that reach beyond the indi-
vidual level to the social context (Barker,
• Walker’s (2012) review of early marriage Ricardo, Nascimento, Olukoya and San-
in Africa uncovered the need for a sub- tos, 2010).
regional strategy to reduce poverty and
address societal inequality in general, • A review of the evidence of the effec-
and more specifically, gender inequality. tiveness of programmes to change gen-
KEY FINDINGS
der norms across the different levels of • An innovative intervention described

the social ecological model including by Reich, Penner, Duncan and Auger
household and community levels report- (2012) used children’s books as a way to
ed by Keleher and Franklin (2008) indi- change new mothers’ attitudes about
cates that gender norm change requires corporal punishment. They found the
simultaneous multilevel interventions books to be an effective and low-cost
designed to influence the underlying way to teach low-income, new mothers
determinants. The status of women and about typical child development and
girls and their opportunities are depen- effective parenting. Results from the
dent on protective upstream legislation evaluation indicated that intervention
among other factors. Strategies for in- differentially impacted women across
creasing the levels of education of girls different demographic segments, in-
or raising their access to health services dicating the need to design and imple-
will have little or no effect on lessening ment discrete strategies to address
the gender gap between men and wom- specific population needs.
en, rich or poor, if they are not embed-
ded in human rights frameworks that Evidence exists regarding the success of inter-
affirm, guide, and monitor violations of personal and community programmes targeting
equal and universal rights. at-risk populations. Apart from the point made
about tailoring programmes, this systematic
review highlights the importance of specifically
“The evidence points to the complexity of
targeting at-risk subpopulations.
addressing VAC and the need to tackle these
issues across all levels of the social ecological • The evaluation of PeaceBuilders, a
model, specifically focusing on norms.” school-based, universal violence pre-
vention intervention that disaggregated
low, medium, and high-risk children,
SUMMARY OF IMPACT EVALUATION showed that changes attributable to the
RESULTS: OPERATIONAL intervention were not uniform across
Impact evaluation data suggests the importance risk categories. Significant behaviour
of contextualizing interventions based on indi- changes were found for children clas-
vidual and cultural needs. sified as high-risk for future violence
between baseline and end-line (as mea-
• A systematic review of 87 programmes sured by teacher-reported aggression,
to address child maltreatment indicated which decreased, and teacher-rated
that intervention success is grounded social competence, which increased)
in the acknowledgement of cultural dif- (Flannery et al., 2003) as compared to
ferences (Plummer, 2001). Conversely, students in other categories.
the author identified the failure to adapt
programmes to specific sub-popula- • The impact evaluation of a universal
tions (lack of audience segmentation) school-based violence prevention in-
as a key barrier to effectiveness. tervention on social-cognitive out-
comes presented data showing that
• Evaluations of the Finnish KiVa anti-bul- high-risk students benefited from the
lying intervention for grades 1-9, which intervention. High-risk students showed
has gone to scale based on several decreases in beliefs and attitudes sup-
evaluations of effectiveness, also high- porting aggression and increases in self-
light the importance of tailoring inter- efficacy, as well as beliefs supporting
ventions for subpopulations. non-violent behaviour. Effects on low-
KEY FINDINGS
risk students were in the opposite direc- individual level to the social context.
tion, thus suggesting a different pattern
of intervention is needed depending on Some specific examples of male involvement
risk. For example, combining school- from the systematic review include:
based programmes with family-based
programmes and tailoring interventions • Findings from the Men as Partners
based on risk status (low versus high) of (MAP) Programme in Soweto, South Af-
students (Simon et al., 2008). rica, indicated that the intervention was
successful in improving norms and at-
It is important to keep gender in mind as a key titudes supportive of gender egalitarian
variable in and of itself when designing pro- relations, among other positive chang-
grammes specifically for men and boys and as es (Ditlopo et al., 2007).
a potential evaluation confounder for research.
Gender focus in the manuscripts centered • Barker’s (2006) reflections on the im-
mainly on issues pertinent to girls and women. pact of engaging boys and men to
This was obviously true for FGM/ C, but also empower girls concluded that pro-
for global programmes dealing with child mar- grammes can be effective in changing
riage.24 A similar trend was noticed with regard attitudes and behaviours of men and
to descriptions of child trafficking interven- boys, which is positive for well-being
tions.25 Interestingly, the focus on girls was also of women and girls. He also noted
evident in programmes designed to promote that comprehensive, multi-theme pro-
empowerment.26 grammes focusing on norm transfor-
mation were the most effective.
The systematic review revealed only a few ex-
amples of interventions that specifically sought • Casey et al. (2013) more recently re-
to involve boys and men as key audiences to viewed the challenges associated with
support change. Nonetheless, the importance of global efforts to engage men in the pre-
such approaches has been highlighted. In a re- vention of violence against women us-
view of 58 programmes by Barker, Ricardo, Nas- ing an ecological perspective and came
cimento, Olukoya and Santos (2010), the authors to a similar conclusion.
categorized interventions working with men and
boys to improve health and gender roles as using • The review revealed some disaggrega-
a 1) gender-neutral approach, 2) gender-sensitive tion of reporting of differential impact on
approach, or 3) gender-transformative approach boys and girls. Haner, Pepler, Cummings
and concluded that few interventions go beyond and Rubin-Vaughn (2010) reported that in
a pilot stage or short-term. But there is compel- response to an arts-based curriculum on
ling evidence that well-designed interventions bullying prevention, boys reported less
with men and boys can lead to positive changes bullying than girls over time.
in their behaviours and attitudes related to sex-
ual and reproductive health; maternal, newborn We found only one reference to the need to ex-
and child health; their interaction with their pand awareness of boys’ issues and male gen-
children; their use of violence against women; der issues, expand boys’ participation, establish
their questioning of violence with other men; interventions and information for boys, or de-
and their health-seeking behaviour. Gender- velop outreach for boys (Frederick, 2010). The
transformative approaches and the promotion focus on VAC specific to girls is understandable,
of gender-equitable relationships between men given the higher levels of VAC that girls experi-
and women are more effective in producing be- ence. However, boys often get left out, which can
haviour change than narrowly focused interven- mean glossing over the forms of violence most
tions, as are interventions that reach beyond the pertinent to boys in terms of severity (e.g. child
KEY FINDINGS
labour and corporal punishment), and their spe- cessful school-based violence prevention pro-
cific vulnerabilities, as well as unique needs. gramme found that most of the violence was, in
fact, taking place in the family. An evaluation
Place-based programmes need to consider the of Kornblum’s body-based violence prevention
in-between spaces where VAC occurs. On the one curriculum using dance/ movement therapy for
hand, there is sufficient data to show that schoolchildren showed a reduction in problematic be-
based programmes are successful in combating haviours (reported by teachers post-intervention),
some forms of VAC, especially bullying. On the but also revealed that the primary place where
other, place-based programmes tend to disre- children were facing problems was at home with
gard other non-discrete locations where bullying parents and siblings (Hervey and Kornblum,
may take place (e.g. social media sites). Smothers 2006). These collective findings therefore under-
and Smothers (2011) developed and evaluated a score the importance of addressing VAC in dis-
social ecological model of sexual abuse preven- crete and less discrete spaces.
tion that was implemented in a public school with
diverse urban youth. The results showed signifi- Several impact assessments on various issues
cant gains in knowledge in all areas assessed and relating to VAC highlight the importance of ad-
indicated that children and adolescents can be dressing VAC as part of early childhood devel-
Girls at Bhagyanagar taught healthy relationship skills that might pro- opment programmes. The systematic review
Children’s Home in India:
two nearby children’s tect them from predatory behaviours or maladap- yielded data on interventions that address is-
homes separately house tive peer relationships. Similarly, classroom-based sues pertinent to infants and young children
girls and boys mostly
between 6-14 years who instruction can reach both potential victims and such as neglect, maltreatment and abuse.
are orphans, children of
offenders while simultaneously training school
migrant labourers and
potential child labourers. staff and faculty. But in another example, a suc- • An analysis of child maltreatment pro-
grammes indicated that parent educa-
tion with a home-visiting component
show promising results. Factors lead-
ing to intervention success included:
the receipt of services before or as
close to birth of the first child as pos-
sible; services that focus on the child’s
particular development level; opportu-
nities for parents to model the behav-
iours being promoted; sufficient time
commitments; an emphasis on social
supports and the skills needed to ac-
cess these supports; a balance of home
and group-based alternatives; and rec-
ognition of cultural differences (Port-
wood, 2006).
• Russell, Trudeau and Britner (2008)

provide information on an intervention
© UNI CE F/INDA2014 -000 42/BISWAS
designed to increase public awareness

of the caregiving practices connected
to Shaken Baby Syndrome.
• Bugental and Schwartz (2009) examine

the efficacy of a cognitive intervention
– designed to improve a parent’s abil-
KEY FINDINGS
ity to interpret a child’s motivations and ated with genital terminology (Kenny,
behaviours – as part of a programme Wurtele and Alonso, 2012).
to prevent child mistreatment among
medically at-risk infants. • The Pathways to Prevention Preschool
Intervention Program appeared to be
In addition, several examples of successful in- effective in improving levels of chil-
terventions specifically targeting preschool and dren’s behaviour over and above the
kindergarten children are provided below. effect of the regular preschool curricu-
lum. Case studies developed through
• An impact assessment among elemen- staff and participant interviews suggest
tary schoolchildren in Hawaii, USA, positive outcomes. Cost analysis re-
indicated that intervention participants vealed that the preventive project was
scored significantly higher on post- cheaper than remedial programmes
test measurements on knowledge of analysed, as well as more cost-effective
appropriate and inappropriate touch- (Leech, 2005).
ing as compared to controls. Chil-
dren participating in the intervention • There is some evidence that bullying
were also able to identify a response prevention programmes conducted as
to unwanted touching that many had early as in kindergarten can change
not known at pre-test (Baker, Gleason, teachers’ attitudes towards victimized
Naai, Mitchell and Trecker, 2013). children, increase teachers’ self-effi-
cacy in handling bullying, and prevent
• An examination of the long-term out- bullying from occurring (Alsaker and
comes of a randomized controlled trial Valkanover, 2012).
of Triple P, a universal prevention pro-
gramme associated with dysfunctional
“Impact assessments on various issues
parenting behaviours, showed that at
the two-year follow up, both mother relating to VAC highlight the importance of
and father participants reported sig- addressing VAC as part of early childhood
nificant reductions in such parenting. development programmes.”
Additionally, mothers also reported
an increase in positive parenting be-
haviour and significant reductions in
internalizing and externalizing child Impact assessment results showcased the im-
behaviour (Hahlweg, Heinrich, Kuschel, portance of long-running interventions. How-
Bertram and Naumann, 2010). ever, there was little emphasis on cost benefit
or effectiveness comparisons of interventions
• An evaluation of a personal safety based on duration. For example:
intervention with Latino preschool-
ers in the United States showed that • An evaluation of a bullying prevention
participants made greater knowledge intervention for elementary schools com-
gains than controls with regard to pared three schools with differing inter-
genital terminology, recognizing the vention lengths (one year versus two
inappropriateness of touch requests years versus three months). The authors
even when coming from ‘good’ peo- found that students in the two-year inter-
ple, and learning to recognize, resist vention reported more positive attitudes
and report inappropriate touching. towards bullying victims compared to
Children in the intervention retained students in the three-month intervention
knowledge gains except those associ- (Beran, Tutty and Steinrath, 2004).
KEY FINDINGS
• An evaluation of a sexual abuse preven- comments, suggestions and other in-

tion workshop in a multicultural, impov- formation. Initial results showed that
erished urban area indicated that there parents in the cell phone enhanced
were no statistically significant differ- intervention were more likely to re-
ences between intervention and control schedule appointments with their so-
groups at year one, but at year two there cial workers but less likely to miss an
was an increase in knowledge of inappro- appointment, compared to the pre-ex-
priate touching among intervention par- isting programme without the texting
ticipants and self-reported experiences component. The authors report that
of victimization decreased (Daigneault, cell phones may be a better means of
Hébert, McDuff and Frappier, 2011). communicating with young parents
with multiple risks (Bigelow, Carta, and
• A preliminary evaluation of Expect Burke, 2008) than through traditional
Respect, a dating violence prevention channels.
intervention for at-risk youth using
support groups, indicated that support • Love, Sanders, Metzler, Prinz and Kast
group participants reported using sig- (2013) assessed parents’ perceptions of
nificantly more healthy conflict resolu- an online parenting intervention deliv-
tion skills at the end of the intervention. ered through a social networking site
The authors contend that the support using interviews with parents at high
group format, coupled with the extend- risk for child maltreatment. The study
ed intervention duration, was success- found that social networking sites were
ful in increasing the healthy conflict a feasible mechanism for reaching a
resolution skills of these at-risk stu- low-income and ethnically diverse pop-
dents (Ball et al., 2012). ulation in Los Angeles County, Califor-
nia. Parents watching the online videos
There is a growing body of data on the efficacy were enthusiastic, but also discussed
of new communication technologies as a mecha- the value of learning through shared
nism for intervention implementation, as well experiences and having a moderator.
as for evaluation. A few interventions that were
included in the systematic review reported ex- • Compared to a control group, partici-
amining the efficacy of new communication pants in a cyberbullying prevention
technologies for implementation and monitor- intervention reported gaining greater
ing. For example: control over personal information as
well as improved interactions with oth-
• Bailey and Ngwenyama (2010) exam- ers on social networks (Ortega-Ruiz, Del
ined the impact of increased exposure Rey, and Casas, 2012).
to violence and community conflict by
improving computer literacy through The use of new communication technologies as
telecentre usage. an effective means to screen and follow up with
participants was also evident in the results from
• The addition of a mobile phone-based the systematic review. For example:
texting component for a parenting in-
tervention for families at risk for child • SafeCare, an evidence-based child
neglect showed promising results. maltreatment prevention intervention,
Families in the cell-phone-enhanced used Apple iPhones to enhance and
intervention received text messages reduce face-to-face home safety ses-
with prompts and questions relating to sions. This intervention found that the
the intervention, as well as supportive SafeCare Safety module yielded sub-
INTRODUCTION
stantial reductions in household hazards

(including poisons, electrical, choking
and suffocation hazards) across all par-
ticipants, which were maintained at a
one-month follow up post-intervention.
The iPhone greatly enhanced commu-
nication between visits through texting,
email, and phone or voicemail messag-
es while also simplifying intervention
logistics. In addition, the iPhone reduced
face-to-face time data collection without
compromising results (Jabaley, Lutzker,
Whitaker and Self-Brown, 2011).
• Constantino, Crane, Noll, Doswell and

Braxter (2007) explored the feasibility
of email-mediated interaction in moth-
ers and children who had survived
abuse. The intervention was designed
specifically for mothers and children
who had obtained a Protection From
Abuse order. Results indicated that
email was a feasible and acceptable
way to provide support and informa- cally abusive parents (Johnston, Kend- A young girl washes her
hands with soap included
tion. Participants were interested in the rick, Polnay, and Stewart-Brown, 2008). in UNICEF-provided hy-
giene kit in Baluwa village
device, compliant with learning and us-
in Gorkha district, Nepal.
ing it, and responded to interventionist • The evaluation of a community media
email questions. campaign on the prevention of child
sexual abuse indicated that the media
Impact evaluations suggest that integrating dif- campaign had significant impact on at-
ferent communication channels and contextual- titudes and significant impact on primary
izing C4D within interventions geared towards prevention responses to hypothetical
policy and structural change is required to bring vignettes. Exposure to the intervention
about sustainable change. The following examples booklet positively affected knowledge
showcase how C4D interfaces with other inter- as compared to no campaign exposure.
ventions to bring about sustainable change and Participants retained knowledge relevant
can serve as an essential part of larger initiatives. to child sexual abuse prevention imme-
diately following exposure to the materi-
• A review of individual and group-based als. However, the results show that these
parenting programmes for the treat- knowledge gains were not maintained
ment of child abuse and neglect sug- (Rheingold et al., 2007).
gests -- through comparative analysis
© UNI CE F/P FPG20 15-29 17/PAN DE Y
-- that programmes directly addressing • A mediated intervention on child abuse

abusive parenting may be more effec- in the Oredo Local Government Area of
tive than those that do not directly reach Edo State in Nigeria revealed that tele-
out to abusive parents. However, the au- vision programmes addressing child
thors feel that further research is need- abuse were able to place the issue in the
ed to assess what should be provided as minds of the public. They also raised
core programme components for physi- awareness about the dangers of child
KEY FINDINGS
abuse in its many forms: child labour, tations arising from the method of data collec-
child battering or sexual exploitation. tion. Examples of the latter included the use of
However, this awareness has not trans- written questionnaires that children with low
lated into behaviour and social change, literacy and poor writing skills had difficulty
judging by continued widespread inci- filling out, or problems that arose due to the use
dence of child battering, child traffick- of translators.
ing, street hawking and begging. The
results also showed that addressing Of greater concern is the overwhelming majority
child abuse through television is a suc- of manuscripts lamenting the paucity of adequate
cessful advocacy tool, which resulted data. For instance, a review by Wessells (2009)
in many governmental and non-gov- concluded that the state of the evidence regard-
ernmental agencies becoming engaged ing child-focused community groups is anecdotal,
in activities to reduce the problem. In impressionistic, unsystematic and underdeveloped,
fact, the Edo State government passed and specifically pointed to the paucity of informa-
into law bills on human trafficking and tion from low and middle-income countries.
sexual exploitation. However, as the re-
port concludes, these efforts have not Interventions driven by methodological rigor alone
decreased the incidence of child abuse needed to contend with ethical concerns. It is im-
(Osakue Stevenson and Awosola, 2008). portant to consider the responsibilities associated
with withholding interventions that may be useful
SUMMARY OF IMPACT EVALUATION for control group participants. For example:
RESULTS: METHODOLOGICAL
There is a serious lack of evaluation data on the • A theory-driven alternative approach
effectiveness of interventions and strong evi- to school bullying that focused on so-
dence of the need for additional effectiveness cial emotional learning and positive
evaluations, specifically those from low and youth development was associated
middle-income countries. The systematic re- with significant reductions in bullying
view found that close to two thirds of the studies and victimization. However, even after
in the database noted at least one methodologi- controlling for gender, the researchers
cal limitation in the intervention. Lack of ran- found that the control group experi-
domization, the absence of a comparison or enced increases in bully and victim be-
control group, short observation periods and haviours (Domino, 2013).
lack of follow up were the most commonly listed
weaknesses concerning methodological design. • From an ethics standpoint, a review of
With regards to indicators, a large proportion sexual abuse and exploitation of boys in
of studies acknowledged problems of reliability South Asia noted the alarming absence
and validity associated with using self-report- of informed consent procedures, con-
ed data. They underscored the absence of vali- fidentiality protocols and other mecha-
dated measures, and often resorted to proxy nisms to ensure responsible and ethical
measures to assess intervention effectiveness, conduct in research (Frederick, 2010).
despite not being able to link improvements to
the intervention itself. Sampling issues included • Other ethical issues in the systematic
small sample sizes resulting in limited statisti- review concerned privacy and confi-
cal power, limited ability to generalize findings, dentiality issues when using technol-
and sampling bias, especially for voluntary inter- ogy-based or enhanced programmes,
ventions or when purposive sampling was used. especially with survivors of abuse
Several studies cited barriers relating to data (Constantino, Crane, Noll, Doswell and
collection, such as missing data due to untrained Braxter, 2007).
implementers; use of leading questions, or limi-
KEY FINDINGS
The scope of effectiveness evaluations can be ex- happy, with improved achievement mo-
panded to include both qualitative and participa- tivation and assertiveness. They had an
tory evaluation methods. This systematic review improved understanding of setting real-
yielded limited data on qualitative evaluations. istic goals, listened and reflected better,
and had improved insights about them-
• One of the few examples of a mixed selves that culminated in an improved
methods approach that has proven effec- perception and self-awareness.
tive is a study by John (2009) examining
the impact of a life skills development in- • The evaluation of Soul Buddyz Clubs
tervention on the behavioural aspects of in South Africa indicated that success-
children in need of care and protection. ful clubs actively realized the vision and
Statistically significant positive changes objectives of Soul Buddyz by mobiliz-
occurred in the intervention group, and ing children as agents of change in their
focus groups revealed that juveniles in own lives and that of the community
the intervention felt more relaxed and (Schmid, Wilson and Taback, 2011).
FIGURE 19: Summary of findings from impact evaluation results
THEORETICAL • There is increasing evidence on the efficacy of

• Impact evaluation data highlights the importance new communication technologies as a mecha-
of utilizing an ideation model. nism for intervention implementation, as well as
• Evidence of social change by interventions is lim- for evaluation.
ited, while the need for interventions to address • Several impact assessments on various issues
normative factors is imperative. relating to VAC highlight the importance of ad-
dressing VAC as part of early childhood develop-
OPERATIONAL ment interventions.
• Impact evaluation data suggests the importance • Impact assessment results showcased the im-
of contextualizing interventions based on individ- portance of long-running interventions.
ual and cultural needs.
• Evidence exists regarding the success of inter- METHODOLOGICAL
personal and community interventions targeting • There is a serious lack of evaluation data on ef-
at-risk populations. fectiveness of interventions and strong evidence
• It is important to keep gender in mind as a key of the need for additional effectiveness evalua-
variable, designing interventions specifically for tions, specifically those from low and middle-in-
men and also keeping gender in mind as a poten- come countries.
tial evaluation confounder for research. • Interventions driven by methodological rigor
• Place-based interventions need to consider in- alone need to contend with ethical concerns.
between spaces where VAC occurs. • The scope of effectiveness evaluation can be ex-
• Impact evaluations suggest that C4D works in panded to include both qualitative and participa-
conjunction with other interventions to bring tory evaluation.
about sustainable change.
7 Overall
recommendations
Boys at Bhagyanagar Children’s

Home in India: two nearby chil-
dren’s homes separately house
girls and boys mostly between
6-14 years who are orphans,
children of migrant labourers
and potential child labourers.
“Today, violence results in more than 1.5 million people

being killed each year, and many more suffer non-fatal
injuries and chronic, non-injury health consequences.
Despite the fact that violence has always been present,
the world does not have to accept it as an inevitable
part of the human condition. Violence can be prevented.
This is not an article of faith, but a statement based on

evidence.”
Violence prevention: The evidence.
World Health Organization, 2010
OVERALL RECOMMENDATIONS
I
n an overview summarizing the major forms of violence mostly acknowledged the
achievements and remaining challenges nexus of VAC with violence against women
of the Convention on the Rights of the (VAW). There is overwhelming agreement that
Child, Doek (2009) reported that na- individuals directly engaged in or responsible
tional efforts addressing VAC have often been for violence at some point in time have been
hampered by a lack of financial and human victims themselves and that the negative rami-
resources or political will. In turn, this has fications of VAW are multiplied manifold if and
impeded the proper infrastructure to support when children are involved. However, strategies
children’s rights from forming. He concluded dealing with VAW do not often segment by age
that addressing VAC takes more than having or address VAC as a correlated issue. For exam-
the proper infrastructure, it requires having a ple, there is an absence of age-based disaggre-
culture that supports the rights of the child. gation in programmes addressing human
trafficking. Disaggregation by age would allow
C4D approaches can challenge social and cul- child trafficking to emerge as a separate issue
tural norms that perpetuate and condone vio- deserving focused attention.
lence. Approaches to tackle social norms have
been successfully used to reduce alcohol mis- EXPLORE LINKAGES BETWEEN
use and smoking behaviours; also, to change
attitudes of young males towards risky sexual
2 DIFFERENT FORMS OF VAC. The
overlap of manuscripts shortlisted
behaviours and bullying (WHO, 2009). Within across different search criteria highlights that the
UNICEF, C4D is integrated as a key cross-cut- different types of VAC cannot be pigeonholed by
ting strategy that promotes long-term behav- topic. Rather, they must be conceptualized as
iour and social change and is defined as: “an multiple forms or facets of violence. For example,
evidence-based and participatory process that child marriage and female genital mutilation/
facilitates the engagement of children, families, cutting and gender-based violence are intrinsical-
communities, the public and decision makers ly linked, as are corporal punishment and child
for positive social and behavioural change in abuse/ maltreatment. Similarly, child trafficking
both development and humanitarian contexts and sexual exploitation co-occur. No doubt this
through a mix of available communication overlap complicates both implementation and
platforms and tools” (UNICEF, 2019). Herein evaluation of programmes; however, it is impor-
lies the critical importance of exploring the tant that the interrelationships and complexities
role that a range of communication approaches between these issues be recognized and addressed
underlining C4D can play in addressing VAC. holistically. This is especially important given the
Key recommendations from this systematic re- magnitude of their prevalence, the plethora of
view that apply specifically to child protection mediating and causal factors involved, and the
programmes as a whole, and their C4D com- multitude of C4D approaches we have at our
ponents in particular, are categorised under the disposal to address VAC.
three broad themes below.
EXAMINE VAC ALONG A SPEC-
I. CONTEXTUALISING AND
FRAMING VAC
3 TRUM. THIS RECOMMENDATION
RELATES TO THE PREVIOUS ONE
SPECIFICALLY ADDRESS CHIL- REGARDING THE LINKAGES BETWEEN DIF-
1 DREN WITHIN THE CONTEXT OF

OTHER FORMS OF VIOLENCE. The
FERENT TYPES OF VAC. Associated topics such
as VAW yield more consistent research and data
search term combinations yielded over 80,000 than when searching for VAC. One potential
hits. Most of the manuscripts discarded by this explanation may be that VAW has been mea-
systematic review dealt with violence but lacked sured in discrete, verifiable forms – physical,
specificity in terms of their focus on children. emotional and sexual -- whereas VAC is often
For example, interventions addressing other measured across broad conceptualizations.
These can range from discriminatory nutrition MOVE BEYOND A PLACE-BASED

practices associated with girl children, all the way
to interventions addressing female foeticide.
6 APPROACH TO A NORMS-BASED
APPROACH TO INCORPORATE
Further, the review shows that the response to INNOVATIVE COMMUNICATION CHANNELS
child labour has largely been in the area of policy AND TO TACKLE THE CULTURE OF VIOLENCE.
change, while at the community and individual Most recent programmes have drawn on the
levels child labour might be considered a social United Nations Secretary-General’s Study on
good (helping poor children in need) rather than VAC and focused on the settings prioritised in the
as a facet of VAC. Hence, child labour persists on study. These are: the home and family, schools
a massive scale but remains virtually unseen. On and educational settings, care and justice systems,
the other hand, sexual abuse is clearly typified as work setting, and the community. Many of the
criminal conduct, and although reporting has programmes associated with these settings rely on
become more commonplace, it continues to be techniques that are associated with the location.
considered a deviant behaviour. It is important, Therefore, programmes in school and educational
therefore, to consider the full spectrum of poten- settings usually involved elements of teacher
tial causes and impacts when addressing VAC. (expert or peer) based training. Meanwhile, pro-
grammes in the home and family-level often fell
CONTEXTUALIZE VAC BOTH AS A within the social work realm. Such narrow con-
4 CAUSE AND AN OUTCOME. One of

the key difficulties during the
ceptualizations result in the replication of ‘tried
and true’ approaches for communication that also
course of the literature search related to the have a narrowly defined focus. It is important to
complexity of understanding the nature of VAC, consider innovative ways to communicate about
as it encompasses such a wide and complex VAC. There is some evidence of new communica-
range of issues. At one end of the spectrum, the tion technologies being utilized as a channel for
systematic review revealed that programmes are dissemination and also for tracking programme
based on a wide range of causal factors, ranging implementation and success in this regard.
from the social to the psychological that result in
VAC. In this regard, a majority of the manu- Place-based interventions can be implemented
scripts portrayed VAC as an outcome. However, and evaluated in a streamlined fashion. For ex-
at the other end, violence emerges as a causal ample, conducting a randomized controlled trial
factor for adverse educational outcomes and and monitoring fidelity of implementation can
long-term morbidity and mortality. Both ways of all take place in an institutional environment
understanding VAC are valid, although strate- such as a school. However, while it is important
gies to address it may differ. to begin interventions at one place, it is equally
important to branch out to address more com-
START EARLY AND CONTINUE plex social contexts. One must consider issues
5 INTO ADULTHOOD. There is evi-

dence of long-term success for early
relating to VAC that occur in less concrete or
discrete places (such as verbal and sexual harass-
childhood programmes. Both family-based ment that adolescent girls face in public spaces)
programmes promoting positive parenting and which are harder to tackle and address. Given
early childhood development programmes can the ubiquitous nature of VAC, it is also impor-
address child maltreatment and neglect starting tant to address it at a broader level. One possibil-
in infancy. These have the potential to sustain ity could be to address VAC through a normative
positive effects well into adulthood. Primary lens and tackle what can be considered a ‘cul-
prevention efforts fostering generational change ture of violence’. One relevant example might be
are instrumental for creating new norms of child labour. A shift from the traditional policy
masculinity and raising a generation of men who perspective to a normative approach would help
do not accept violence as a norm and girls who frame this issue as one of VAC and may help
refuse to accept violence silently. address both the supply and demand for child
INTRODUCTION
labour. Linking school violence to the normative SEGMENT AUDIENCES BY GENDER

violence happening outside schools would be an-
other way of broadening the perspective.
9 AND ADDRESS GENDER-SPECIFIC
NEEDS AND DIFFERENCES. A num-
ber of points highlighted the importance of gender.
II. DESIGNING THE
PROGRAMME 1. Female empowerment is necessary and
EMBRACE THE SOCIAL ECOLOGI- successful in addressing violence, and
7 CAL MODEL FOR BEHAVIOUR AND

SOCIAL CHANGE. Much of the
it is therefore essential to implement
interventions that specifically meet the
published literature on VAC programmes utilized needs of girls.
a cognitive and individually-based behaviour 2. The need to involve men and boys
change approach. While such approaches are more often in gender transformative in-
useful and valid in some instances, there is a terventions was highlighted.
growing realization that individuals are embed- 3. A focus on females must not exclude
ded within a larger social system. Thus, effective young boys from VAC issues as they
programmes have to consider the interpersonal, are also affected.
family and community dimensions to generate
4. There is a need to address both men
and measure change. At the same time, a concert-
and boys in terms of their roles both as
ed institutional, policy and overall national level
victims and perpetrators. For example,
response is also critical to address VAC and create
anti-bullying programmes have been
an enabling environment for change. This has
designed to address the needs of boys
proven effective especially in addressing female
by examining ‘machismo’ and its role
genital mutilation/ cutting in many countries,
in VAC.
where a combination of challenging harmful
social norms together with legislative and policy 5. There is also a growing demand to ad-
level approaches has yielded positive results. dress the specific needs of marginalized
groups such as gay, bisexual and trans-
BROADEN CONCEPTUALIZATIONS gender youth who are often subjected to
8 OF C4D APPROACHES TO ENCOM-

PASS EFFORTS THAT INVOLVE
widespread discrimination and violence.
CLARIFY THE DIRECT AND INDI-

ANY FORM OF COMMUNICATION/ COMMUNI-
CATIVE ACTION AND CUT ACROSS ALL LEV-
10 RECT LINKAGES BETWEEN STAT-
ED PROGRAMME AND
ELS OF THE SOCIAL ECOLOGICAL MODEL. COMMUNICATION OBJECTIVES AND MES-
This review included many approaches that inter- SAGES. Programmes addressing VAC have often
sect with a broad understanding of what C4D is been couched within a ‘harm reduction’ frame-
and its integral role in the uptake and mainte- work and both programme and communication
nance of individual and social change. Much of objectives have been listed in negative terms (e.g.
the information on programmatic responses for reduction of harmful practices). Communication
VAC reveals that these are inherently communi- messages, on the other hand, have focused on the
cative: counselling, awareness raising, disclosure, positive, for example, through the frequent use of
negotiation, public denouncement, resistance and role models for desired behaviours. Clearer link-
confrontation. Indeed, in the preliminary review ages and flow of logic between the articulation of
of manuscripts, it was often difficult to distin- objectives and messaging that is designed to ad-
guish between overall programme approaches dress or meet those objectives is needed. Estab-
and discrete C4D components. Interventions were lishing intermediate communication objectives
not necessarily described as utilizing C4D ap- would help determine how communication mes-
proaches, even when they clearly did. Neverthe- sages contribute to achieving overall programme
less, upon closer examination, C4D techniques objectives. A related issue deals with labelling all
dominated in efforts to reduce harmful practices. traditions as being ‘harmful’. This can lead to
negative unintended consequences, most notably Promoting a core understanding of what VAC is
by driving a practice underground, rather than in the form of a standardized definition and
creating sustained behaviour and social change. measurements of incidence and prevalence is an
essential first step to gauging the magnitude of
RECONFIGURE PROGRAMME AND VAC as an issue deserving of global advocacy
11 COMMUNICATION OBJECTIVES
OF VAC INTERVENTIONS. Many of
and concerted action. The United Nations Secre-
tary-General’s Study on VAC is a seminal effort
the interventions did not use the basic SMART to provide a detailed global overview of VAC,
and SPICED criteria when describing programme where it occurs, and ways to combat it. However,
objectives. It was also challenging to distinguish local indicators of VAC are not routinely acces-
between overall programme and C4D objectives. sible. This hinders the data comparison at global
VAC is an emotionally charged issue and yet levels and renders situation analyses for behav-
programme and communication objectives were iour and social change interventions difficult to
mainly couched in cognitive terms with little undertake.
reference to the power of emotions to promote
behaviour and social change. Cognitive objectives ENHANCE INVESTMENT IN RE-
could move beyond enhancing knowledge and
comprehension to ensuring the activation of
14 SEARCH. The grey literature con-
tains many examples of creative
higher-level cognitive changes associated with interventions addressing VAC, especially imple-
assimilation and evaluation of information. mented in low and middle-income countries.
However, much of this creative programming is
MOVE BEYOND INDIVIDUALLY unaccompanied by information on the effective-
12 FOCUSED KNOWLEDGE, ATTI-

TUDES AND PRACTICE BY AD-
ness of these strategies. Much of the published
literature, especially when grounded in commu-
DRESSING SOCIAL, EMOTIONAL AND nity-based approaches, tended to focus on the
BEHAVIOURAL COMPETENCIES. Interventions ‘process’ of implementation rather than effec-
that encompass training, capacity-building, and tiveness. It is essential to expand the number of
efficacy approaches -- either as the end result or as outcome evaluation studies to enhance our un-
essential building blocks – have reported positive derstanding of global best practices. This needs
outcomes. This was especially true of anti-bully- to be contextualized based on what works at a
ing programmes for school children, youth proj- local level.
ects to reduce gang violence, or child marriage
initiatives that provide economic opportunities SCALE UP PROMISING INTERVEN-
for adolescent girls. At the same time, enrichment
programmes providing preschool children with
15 TIONS. Much of the robust evi-
dence for effectiveness is based on
both academic and social skills showed promise small-scale pilot projects employing randomized
in addressing child maltreatment and neglect. controlled trial designs. This is especially true of
Even within interventions with narrowly focused programmes dealing with corporal punishment,
individual level outcomes, it is important to step bullying and gang violence among youth. How-
beyond the cognitive knowledge/ awareness di- ever, there is scant information on the potential
mensions and address emotional, relational and for scale up of promising interventions. A practi-
behavioural competencies to effect change. cal way forward might be to develop a road map
of such pilot interventions along with scaling up
III. IMPLEMENTING guidelines. After evaluating progress over a
AND EVALUATING THE reasonable time period, programme planners
PROGRAMME and implementers can determine whether to
POSITION VAC AS A ‘GLOCAL’ IS- continue, adapt or discard the intervention.
13 SUE THROUGH QUALITATIVE AND

QUANTITATIVE MEASUREMENT.
8 Conclusions
Children play outside the

health post in the village of
Wolargi, in Gemechis, a wore-
da (district) of Oromia Region,
Ethiopia.
I
The wealth of n addition to secondary analysis of data
information included in collated for the purpose of this review,
this systematic review there is also room for further improve-
© UNI CE F/NYH Q2 014 -3 635 /N ESBIT T
has opened up the ment through additional studies. Figure

20 outlines potential avenues for future re-
opportunity to conduct
search and additional studies.
further and more specific
analysis of information In many ways, this review contains much de-
already collected. tailed information on the current status of
the design, implementation and evaluation
of interventions utilizing C4D approaches to
address VAC. It further attempts to synthe-
CONCLUSIONS
size the key trends and findings to highlight ventionists and practitioners, as well as their
some imperatives that intervention planners research counterparts on the ground, can uti-
and evaluators need to keep in mind in order lize to plan, implement and evaluate interven-
to move the field forward. As such, it serves as tions using C4D approaches to address VAC
the first step. The next steps are clearly to de- that are grounded in the local realities while
vise a series of practical guidelines that inter- being informed by global best practice.
FIGURE 20: Potential directions for future research and study
Future research
• Conduct an in-depth examination of information raised through issue-specific

recommendations.
• Study the existing meta-analyses and systematic reviews to draw broad-based

solutions to local problems.
• Focus on specific C4D approaches.
• Compare the differential impact of specific communication approaches.
• Link intervention design with implementation strategies to find commonalities in

different types of implementation.
• Compare results associated with specific evaluation designs.
Additional studies
• Examine grey literature for information on innovative programming and evaluation

methods.
• Examine grey literature to contextualize the review within the context of the outcomes
of and contributors to VAC (e.g. alcohol abuse, the effects of urbanization, and other
environmental factors such as migration and displacement).
• Update the bibliography on an ongoing and periodic basis in order to have easy access
to current information.
REFERENCES
REFERENCES
Adler, Corey, Jane Chung-Do and Ophelia Ongalibang, Reflections from practice and evidence of impact’, paper
‘Safe School Task Force: University-community partner- presented at expert group meeting on Elimination of All
ship to promote student development and a safer school Forms of Discrimination and Violence Against the Girl
environment’, Progress in Community Health Partnership: Child of UNICEF, Florence, Italy, September 2006.
Research, Education, and Action, vol. 2, no. 4, pp. 301-306.
Barker, G., et al, ‘Questioning gender norms with men to
Akeo, Nani P., et al, ‘Hui malama o ke kai: Mobilizing to improve health outcomes: Evidence of impact’, Global
prevent youth violence and substance use with passion, Public Health, vol. 5, no. 5, 2010, pp. 539-553.
common goals, and culture’, American Journal of Preven-
tive Medicine, vol. 34, no. 3S, 2008, pp. S67-S71. Barron, Ian G., and Keith J. Topping, ‘School-based abuse
prevention: Effect on disclosures’, Journal of Family Vio-
Alda, Eric, Mayra Buvinic and Jorge Lamas, ‘Neighbour- lence, vol. 25, 2010, pp. 651-659.
hood peacekeeping: The Inter-American Development
Bank’s violence reduction programs in Colombia and Uru- Barron, Ian G., and Keith J. Topping, ‘Sexual abuse pre-
guay’, Civil Wars, vol. 8, no. 2, 2006, pp. 197-214. vention programme fidelity: Video analysis of interac-
tions. Child Abuse Review, vol. 20, 2011, pp. 134-151.
Alsaker, Françoise D., and Stefan Valkanover, ‘The Ber-
nese program against victimization in kindergarten and Barsoum, Ghada, et al, Toward FGM-Free Villages in
elementary school’, New Directions for Youth Develop- Egypt: A Mid-Term Evaluation and Documentation of the
ment, vol. 133, 2012, pp. 15-28. FGM-Free Village Project, Population Council, Inc., Cairo,
Egypt 2009.
Amin, Sajeda, and Luciana Suran, ‘Program Efforts to
Delay Marriage Through Improved Opportunities: Some Beier, K. M., et. al,. ‘Encouraging self-identified paedo-
Evidence from Rural Bangladesh’, report, Population philes and hebephiles to seek professional help: First
Council, New York, 2005. results of the Prevention Project Dunkelfeld (PPD)’, Child
Abuse & Neglect, vol. 33, no. 8, 2009, pp. 545-549.
Anderson, Lorin W., et. al, eds., A Taxonomy for Learning,
Teaching, and Assessing: A Revision of Bloom’s Taxono- Beran, Tanya N., Leslie Tutty and Greg Steinrath, ‘An eval-
my of Educational Objectives, Allyn & Bacon and Pearson uation of a bullying prevention program for elementary
Education Group, Boston, 2001. schools’, Canadian Journal of Psychology, vol. 19, no. 1,
2004, pp. 99-116.
Auemaneekul, Naruemon, et al, ‘Participative model of
child protection in Northern Thailand’, Thai Journal of Berg, Rigmor C., and Eva Denison, ‘Interventions to Re-
Nursing Research, vol. 133, 2009, pp. 216-226. duce the Prevalence of Female Genital Mutilation/Cutting
in African countries’, Campbell Systematic Reviews, vol.
Babalola, Stella, et al, ‘Impact of a communication pro- 9, Oslo, Norway, 2012.
gramme on female genital cutting in eastern Nigeria’,
Tropical Medicine and International Health’, vol. 11, no. Bess, Kimberly D., Paul Speer, and Douglas Perkins, ‘Eco-
10, 2006, pp. 1594-1603. logical contexts in the development of coalitions for youth
violence prevention: An organizational network analysis’,
Bailey, A., and O Ngwenyama, ‘Community mediation and Health Education & Behavior, vol. 39, no. 5, 2012, pp. 526-
violence prevention through telecentre usage: ICTs me- 537.
diating the ‘Border Line’, Paper presented at SIG Global
Development, Third Annual Workshop, Saint Louis, Mo., Bigelow, Kathryn M., Judith J. Carta, and Jennifer Burke
12 December 2010. Lefever, ‘Txt u ltr: Using cellular phone technology to en-
hance a parenting intervention for families at risk for ne-
Baker, Charlene K., et al, ‘Increasing knowledge of sex- glect’, Child Maltreatment, vol. 13, no. 4, 2008, pp. 362-367.
ual abuse: A study with elementary school children in
Hawai’i’, Research on Social Work Practice, vol. 23, no. Boothby, Neil, and Lindsay Stark, ‘Data surveillance in
2, 2013, pp. 167-178. child protection systems development: An Indonesian
case study’, Child Abuse & Neglect, vol. 35, no. 12, 2011,
Ball, Barbara, ‘Expect Respect support groups: Prelimi- pp. 993-1001.
nary evaluation of a dating violence prevention program
for at-risk youth’ Violence Against Women, vol. 18, no. 7, Brady, Martha, et al, Providing New Opportunities to Ado-
2012, pp. 746-762. lescent Girls in Socially Conservative Settings: The Ishraq
Program in Rural Upper Egypt, Population Council, New
Barker, Gary, ‘Engaging boys and men to empower girls: York, 2007.
REFERENCES
Bugental, Daphne B., and Alex Schwartz, ‘A cognitive ap- Evans, Catrin, and Pankaja Bhattarai, ‘A Comparative
proach to child mistreatment prevention among medically Analysis of Anti-Trafficking Intervention Approaches in
at-risk infants’, Developmental Psychology, vol. 45, no.1, Nepal’, Asia Foundation, Population Council, 2000.
2009, pp. 284-288.
Evans, Doug W., ‘Efficacy of child abuse and neglect pre-
Casey, Erin A., et al, ‘Context, challenges, and tensions in vention messages in the Florida Winds of Change cam-
global efforts to engage men in the prevention of violence paign’, Journal of Health Communication, vol. 17, no. 4,
against women: An ecological analysis’, Men and Mas- 2012, pp. 413-431.
culinities, 2013, pp. 1-24.
Evans, Steven W., et al, ‘Feasability of the MindMatters
Charmaz, Kathy, Constructing Grounded Theory: A Prac- school mental health promotion program in American
tical Guide Through Qualitative Analysis, Sage Publica- schools’, Journal of Youth and Adolescence, vol. 34, no.
tions, Thousand Oaks, CA, 2006. 1, 2005, pp. 51-58.
Chesterton, Paul, Evaluation of the Meena Communica- Flannery, Daniel J., et al, ‘Initial behaviour outcomes for
tion Initiative: UNICEF regional office for South Asia Kath- the PeaceBuilders universal school-based violence pre-
mandu, UNICEF, Kathmandu, Nepal, 2004. vention program’, Developmental Psychology, vol. 39, no.
2, 2009, pp. 292-308.
Constantino, R., et al, ‘Exploring the feasibility of email-
mediated interaction in survivors of abuse’, Journal of Frederick, John, ‘Sexual Abuse and Exploitation of Boys
Psychiatric and Mental Health Nursing, vol. 14, no. 3, in South Asia: A review of research findings, legislation,
2007, pp. 291-301. policy and programme responses’, working paper, no.
2010-02, UNICEF, Innocenti Research Centre, Florence,
Cornell, Dewey G., ‘A developmental perspective on the Italy, 2010.
Virginia Student Threat Assessment Guidelines’, New Di-
rections for Youth Development, vol. 129, 2009, pp. 43-59. Gage, Anastacia J., ‘Coverage and Effects of Child Marriage
Prevention Activities in Amhara Region, Ethiopia: Findings
Daigneault, Isabelle, et al, ‘Evaluation of a sexual abuse From a 2007 Study’, USAID, Washington, D.C., 2009.
prevention workshop in a multicultural, impoverished ur-
ban area’, Journal of Child Sexual Abuse, vol. 21, no. 5, Governance and Social Development Research Centre,
2011, pp. 521-542. ‘Helpdesk Research Report: Evidence-Based Strategies
for Preventing Child/Early Marriage’, Governance and
Ditlopo, P., et al, Testing the Effectiveness of the Men as Social Development Research Centre, Birmingham, Ala-
Partners Program (MAP) in Soweto, South Africa, Fron- bama, 2011.
tiers in Reproductive Health Program Population Council,
Hope Worldwide, EngenderHealth, and USAID, 2007. Graft, Auralice, Nicole Haberland, and Rachel Goldberg,
‘Married adolescents: A review of programmes’, Paper
Doek, Jaap E., ‘The CRC 20 years: An overview of some presented at the WHO/UNFPA/Population Council Techni-
of the major achievements and remaining challenges’, cal Consultation on Married Adolescents, Geneva, Swit-
Child Abuse & Neglect, vol. 33, no. 11, 2009, pp. 771-782. zerland, 9-12 December 2003.
Domino, Meg, ‘Measuring the impact of an alternative ap- Greenwood, Peter W., The California Wellness Founda-
proach to school bullying’, Journal of School Health, vol. tion’s Violence Prevention Initiative: Findings from an
83, no. 6, 2013, pp. 430-437. Evaluation of the First Five Years, RAND Corporation.
Duerr Berrick, Jill, Carole Shauffer, and Jennifer Rodri- Santa Monica, California, 2001.
guez, ‘Recruiting for excellence in foster care: Marrying Griffith, Derek M., et al, ‘Organizational empowerment
child welfare research with brand marketing strategies’, in community mobilization to address youth violence’,
Journal of Public Child Welfare, vol. 5, 2011, pp. 271-281. American Journal of Preventive Medicine, vol. 34, no. 3S,
Erulkar, Annabel S., and Eunice Muthengi. ‘Evaluation of 2008, pp. S89-S99.
Berhane Hewan: A program to delay child marriage in Gupta, Sreela D., et al, ‘Knot Ready: Lessons From India
rural Ethiopia’, International Perspectives on Sexual and on Delaying Marriage for Girls’, International Centre for
Reproductive Health, vol. 35, no. 1, 2009, pp. 6-14. Research on Women, New Delhi, India, 2008.
Erulkar, Annabel, and Francis Ayuka, ‘Addressing Early Hahlweg, Kurt, et al, ‘Long-term outcome of a randomized
Marriage in Areas of High HIV Prevalence: A Program to controlled universal prevention trial through a positive pre-
Delay Marriage and Support Married Girls in Rural Nyan- vention program: Is it worth the effort?’ Child & Adolescent
za, Kenya,’ USAID, Population Council, 2007. Psychiatry & Mental Health, vol. 4, no. 14, 2008, pp. 1-14.
REFERENCES
Haines, Jess, et al, ‘V.I.K. (Very Important Kids): A school- Keleher, Helen, and Lucinda Franklin, ‘Changing gendered
based program designed to reduce teasing and unhealthy norms about women and girls at the level of household
weight-control behaviours’, Health Education Research, and community: A review of the evidence’, Global Public
vol. 21, no. 6, 2006, pp. 884-895. Health: An International Journal for Research, Policy, and
Practice, vol. 3, no. S1, 2008, pp. 42-57.
Haner, Dilys, et al, ‘The role of arts-based curricula in bully-
ing prevention: Elijah’s Kite - A children’s opera’, Canadian Kenny, Maureen C., Sandy K., Wurtele and Laura Alonso,
Journal of School Psychology, vol. 25, no. 1, 2010, pp. 55-69. ‘Evaluation of a personal safety program with Latino pre-
schoolers’ Journal of Child Sexual Abuse, vol. 21, no. 4,
Hervey, Lenore, and Rena Kornblum, ‘An evaluation of 2012, pp. 368-385.
Kornblum’s body-based violence prevention curriculum
for children’, The Arts in Psychotherapy, vol. 33, 2006, pp. Kolko, David J., Anne-Marie R. Iselin and Kevin J. Gully,
113-129. ‘Evaluation of the sustainability and clinical outcome of
alternatives for families: A cognitive-behavioural therapy
International Centre for Research on Women, ‘How to (AF-CBT) in a child protection centre’, Child Abuse & Ne-
End Child Marriage: Action Strategies for Prevention and glect, vol. 35, no. 2, 2011, pp. 105-116.
Protection’, International Centre for Research on Women,
Washington, D.C., 2007. Kraft, Ellen M., and Jinchang Wang, ‘Effectiveness of cy-
ber bullying prevention strategies: A study on students’
International Centre for Research on Women, ‘Delay- perspectives’, International Journal of Cyber Criminology,
ing Marriage for Girls in India: A Formative Research to vol. 3, no. 2, 2009, pp. 513-535.
Design Interventions for Changing Norms’, UNICEF, New
Delhi, India, 2011. Krathwohl, David R., ‘A Revision of Bloom’s Taxonomy: An
Overview’, Theory into Practice, vol. 41, no. 4, 2002, pp.
Jabaley, Julie J., et al, ‘Using iPhones to enhance and re- 212-218.
duce face-to-face home safety sessions within SafeCare:
An evidence-based child maltreatment prevention program’, Kroovaud Hipple, Natalie, et al, ‘Project Safe Neighbour-
Journal of Family Violence, vol. 26, no. 5, 2011, pp. 377-385. hoods Case Study Report: Middle District of North Caro-
lina’, Unpublished case study report, U.S. Department of
James, Deborah, et al, ‘A friend in deed? Can adolescent Justice, 2007.
girls be taught to understand relational bullying?’, Child
Abuse Review, vol. 20, no. 6, 2011, pp. 439-454. Landis, J. Richard, and Gary G. Koch, ‘The measurement
of observer agreement for categorical data’, Biomet-
Jejeebhoy, Shireen J., ‘Broadening Girls’ Horizons: Effects of rics vol. 33, no. 1, 1977, pp. 159–174.
a Life Skills Education Programme in Rural Uttar Pradesh’,
The Population Council, Inc., New Delhi, India, 2009. Law, B. M. F., and Shek, D. T., ‘Process evaluation of a posi-
tive youth development program in Hong Kong based on
John, Jessy, ‘The impact of the Life Skills Development different cohorts’, Scientific World Journal, 2012, pp. 1-9.
Programme on the behavioural aspects of children in
need of care and protection’, Loyola Journal of Social Leach, Fiona, Eric Slade, and Máiréad Dunne, Desk Re-
Sciences, vol. 23, no. 2, 2009, pp. 177-194. view for Concern: Promising Practice in School Related
Gender Based Violence (SRGBV) Prevention and Re-
Johnston, B. J., et al, ‘Individual and group-based parent- sponse Programming Globally, University of Sussex Cen-
ing programmes for the treatment of physical child abuse tre for International Education, East Sussex, UK, 2012.
and neglect (review)’, The Cochrane Library, vol. 4, 2008,
pp. 1-23. Leathers, Sonya, et al, ‘Use of a parent management
training intervention with urban foster parents: A pilot
Karna, Antti, et al, ‘A large-scale evaluation of the KiVa study’, Child Youth Services Review, vol. 33, no. 7, 2011,
antibullying program: Grades 4-6’, Child Development, vol. pp. 1270-1279.
82, no. 1, 2011, pp. 311-330.
Leech, M., ‘Pathways to Prevention Project’, in Strategies
Karna, Antti, et al, ‘Going to scale: A nonrandomized na- and Best Practices in Crime Prevention in Particular in
tionwide trial of the KiVa antibullying program for grades Relation to Urban Areas and Youth at Risk: Proceedings
1-9’, Journal of Consulting and Clinical Psychology, vol. of the Workshop held at the 11th UN Congress on Crime
79, no. 6, 2011, pp. 796-805. Prevention and Criminal Justice, (Bangkok, Thailand,
Kaufman, Michelle R., and Mary Crawford, ‘Sex traffick- 18-25 April 2005), edited by Margaret Shaw Katherine
ing in Nepal: A review of intervention and prevention pro- Travers, International Centre for the Prevention of Crime,
grams’, Violence Against Women, Research and Activism Montreal, Canada, 2007, pp. 118-125.
Review, vol. 17, no. 5, 2011, pp. 651-665.
REFERENCES
Leff, Stephen S., et al, ‘Using community-based participa- Pepler, Debra, and Wendy Craig, ‘Promoting relationships
tory research to develop the PARTNERS youth violence and eliminating violence in Canada’, International Journal
prevention program’, Progress in Community Health Part- of Behavioral Development, vol. 35, no. 5, 2011, pp. 389-397.
nerships: Research, Education, and Action, vol. 4, no. 3,
2010, pp. 207-216. Perkins, H. Wesley, David Craig, and Jessica M. Perkins, ‘Us-
ing social norms to reduce bullying: A research intervention
Love, Susan M., et al, ‘Enhancing accessibility and en- among adolescents in five middle schools’, Group Process-
gagement in evidence-based parenting programs to re- es & Intergroup Relations, vol. 14, no. 5, 2011, pp. 703-722.
duce maltreatment: Conversations with vulnerable par-
ents’, Journal of Public Child Welfare, vol. 7, no. 20, 2013, Peterson, Dana, and Finn-Aage Esbensen, ‘The outlook is
pp. 20-38. G.R.E.A.T.: What educators say about school-based pre-
vention and the Gang Resistance Education and Training
Malti, Tina, Denis Ribeaud, and Manuel Eisner, ‘The ef- (G.R.E.A.T.) Program’, Evaluation Review, vol. 28, no. 3,
fectiveness of two universal preventive interventions in 2004, pp. 218-245.
reducing children’s externalizing behaviour: A cluster
randomized trial’, Journal of Clinical Child & Adolescent Mer, Carol A., ‘Prevention of child sexual abuse: A survey
Psychology, vol. 40, no. 5, 2011, pp. 677-692. of 87 programs’, Violence and Victims, vol. 16, no. 5, 2009,
pp. 575-588.
McGarrell, Edmund F., et al, ‘Attempting to reduce fire-
arms violence through a Comprehensive Anti-Gang Initia- Portwood, Shannon G., ‘What we know--and don’t know-
tive (CAGI): An evaluation of process and impact’, Journal -about preventing child maltreatment’, Journal of Aggres-
of Criminal Justice, vol. 41, no. 1, 2013, pp. 33-43. sion, Maltreatment & Trauma, vol. 12, no. 3/4, 2006.
Meyer, Aleta L., et al, ‘Developing a comprehensive ap- Quinby, Rose K., et al, ‘Installing the Communities That
proach to youth violence prevention in a small city’, Care prevention system: Implementation progress and
American Journal of Preventive Medicine, vol. 34, no. 3S, fidelity in a randomized controlled trial’, Journal of Com-
2008, pp. S13-S20. munity Psychology, vol. 36, no. 3, 2008, pp. 313-332.
Mikton, Christopher and Alexander Butchart, ‘Child mal- Quinn, G. P., et al, ‘Adolescent perceptions of violence:
treatment prevention: A systematic review of reviews’, Bul- Formative research findings from a social marketing cam-
letin World Health Organization, vol. 87, 2009, pp. 353-361. paign to reduce violence among middle school youth’,
Public Health, vol. 121, 2007, pp. 357-366.
Mitchell, Claudia, ‘Where are the Youth? Creating Youth
Action Spaces for Countering Gender Inequality in Edu- Reich, Stephanie M., et al, ‘Using baby books to change
cation’, Paper commissioned for Education for All Global new mothers’ attitudes about corporal punishment’, Child
Monitoring Report, UNESCO, 2003-2004. Abuse & Neglect, vol. 36, no. 2, 2006, pp. 108-117.
Morrison, Gary, Steven Ross, Jerrold E., Designing effec- Rheingold, Alyssa A., et al, ‘Prevention of child sexual
tive instruction, 5th ed., John Wiley & Sons, New York, 2007. abuse: Evaluation of a community media campaign’, Child
Maltreatment, vol. 12, no. 4, 2007, pp. 352-363.
Morrongiello, Barbara A., et al, ‘Video messaging: What
works to persuade mothers to supervise young children Rijsdijk, Liesbeth E., et al, ‘The World Starts With Me: A
more closely in order to reduce injury risk?’, Social Sci- multilevel evaluation of a comprehensive sex education
ence & Medicine, vol. 68, no. 6, 2009, pp. 1030-1037. programme targeting adolescents in Uganda’, BMC Pub-
lic Health, vol. 11, 2011, pp. 334-345.
Ortega-Ruiz, Rosario, Rosario Del Rey, and José A. Casas,
‘Knowing, building and living together on Internet and Russell, Beth S., Jeremiah Trudeau, and Preston A. Brit-
Social Networks: The ConRed cyberbullying prevention ner, ‘Intervention type matters in primary prevention of
program’, International Journal of Conflict and Violence, abusive head injury: Event history analysis results’, Child
vol. 6, no. 2, 2012, pp. 303-312. Abuse & Neglect, vol. 32, no. 10, 2008, pp. 949-957.
Osakue Stevenson, Osakue, and Kayode Rasak Awosola, Saxena, Ruchi, Nalini Srivastava, and Abha Ahuja, ‘Edu-
‘Child abuse and media: An assessment of Oredo Local cating adolescent girls and young women on family life
Government Area of Edo State Nigeria’, Pakistan Journal education issues with the use of communication aids in a
of Social Sciences, vol. 5, no. 1, 2008, pp. 128-133. village of Uttrakhand’, Journal of Social Sciences, vol. 21,
no. 1, 2009, 73-78.
Payne, Pedro R., and Kirk Williams, ‘Building social capi-
tal through neighbourhood mobilization: Challenges and Schmid, Jeanette, Theresa Wilson, and Rayna Taback,
lessons learned’, American Journal of Preventive Medi- ‘Soul Buddyz Clubs: A social development innovation’,
cine, vol. 34, no. 3S, 2009, pp. S42-S47. International Social Work, vol. 52, no. 2, 2011, pp. 272-286.
END NOTES
Schober, Daniel J., Stephen B. Fawcett and Jetta Bernier, Taylor, L. R., ‘Dangerous trade-offs: The behavioural ecol-
‘The Enough Abuse Campaign: Building the movement to ogy of child labour and prostitution in rural North Thailand’,
prevent child sexual abuse in Massachusetts’, Journal of Current Anthropology, vol., 46, no. 3, 2005, pp. 411-432.
Child Sexual Abuse, vol. 21, 2012, pp. 456-469.
Tucci, Joe, Janise Mitchell and Chris Goddard, More
Scott, Kathreena L, and Claire V. Crooks, ‘Preliminary action-Less talk! Community responses to child abuse
evaluation of an intervention program for maltreating fa- prevention, Australian Childhood Foundation Melbourne,
thers’, Brief Treatment and Crisis Intervention, vol. 7, no. Australia, 2001.
3, 2007, pp. 224-238.
United Nations Children’s Fund ‘C4D Global Programme
Self-Brown, Shannon, et al, ‘A media campaign preven- Guidance’ [in print] UNICEF, New York, 2019.
tion program for child sexual abuse: Community mem-
bers’ perspectives’, Journal of Interpersonal Violence, Vuijk, Patricia, et al, ‘Testing sex-specific pathways from
vol. 23, no. 6, 2008, pp. 728-743. peer victimization to anxiety and depression in early ado-
lescents through a randomized control trial’, Journal of
Simon, Thomas R., et al, ‘The multisite violence preven- Affective Disorders, vol. 100, 2006, pp. 221-226.
tion project: Impact of a universal school-based violence
prevention program on social-cognitive outcomes’, Pre- Walker, Judith-Ann, ‘Early marriage in Africa: Trends,
vention Science, vol. 9, no. 4, 2008, pp. 231-244. harmful effects and interventions’, African Journal of Re-
productive Health, vol. 16, no. 2, 2012, pp. 231-240.
Sidey, M., and T. Lynch, ‘Creating new choices: A violence
prevention project for schools in Australia’, Innodata Wessells, Mike, ‘What Are We Learning About Community-
Monographs-9, UNESCO, Geneva, 2001. Based Child Protection Mechanisms?: An inter-agency re-
view of the evidence from humanitarian and development
Kraemer Smothers, Melissa, and D. Brian Smothers, ‘A settings’, inter-agency report, UNICEF, New York, 2009.
sexual assault primary prevention model with diverse ur-
ban youth’, Journal of Child Sexual Abuse, vol. 20, no. 6, Wessler, Stephen L., and Lelia L. De Andrade, ‘Slurs, ste-
2011, pp. 708-727. reotypes, and student interventions: Examining the dy-
namics, impact, and prevention of harassment in middle
Soleiman Ekhaiari, Yalda, et al, ‘The effect of an interven- and high school’, Journal of Social Issues, vol. 62, no. 3,
tion based on the PRECEDE-PROCEED Model on preven- 2006, pp. 511-532.
tative behaviours of domestic violence among Iranian
high school girls’, Iranian Red Crescent Medical Journal, Wilson, Shandra J., and Mark W. Lipsey, ‘School-based
vol. 15, no. 1, 2013, pp. 21-28. interventions for aggressive and disruptive behaviour:
Update of a meta-analysis’, American Journal of Preven-
tive Medicine, vol. 33, no. 2S, 2007, pp. S130-S143.
END NOTES
1. UNICEF defines communication for development lence’ is used in the broad sense to include all
as ‘an evidence-based and participatory process forms of violence, abuse, and exploitation.
that facilitates the engagement of children, fami-
4. Available at: https://www.ohchr.org/en/profes-
lies, communities, the public and decision makers
sionalinterest/pages/crc.aspx
for positive social and behavioural change in both
development and humanitarian contexts through 5. As famously noted by Richard Horton, the editor
a mix of available communication platforms and in chief of The Lancet: “We portray peer review
tools’. In this study, the term communication for to the public as a quasi-sacred process that helps
development (C4D) is used to cover a wide variety to make science our most objective truth teller.
of communication approaches and strategies, as But we know that the system of peer review is
detailed in Table 1. Key terms of the systematic biased, unjust, unaccountable, incomplete, easily
review. fixed, often insulting, usually ignorant, occasion-
ally foolish, and frequently wrong” [Horton, R.
2. The overarching research question of the study
(2000). ‘Genetically modified food: consternation,
was: ‘What are the effects of communication for
confusion and crack up.’, Medical Journal of Aus-
development (C4D) approaches to address vio-
tralia, 2000, 172 (4): 148-9.]
lence against children?’.
6. Some pertinent examples include: a recent re-
3. For the purpose of this assignment, the term ‘vio-
view of efficacious nutrition interventions (dif-
END NOTES
ferent approaches) in addressing child mortality stages of change: Precontemplation is the stage at
(single outcome), for which the authors catego- which individuals are unaware of their problem and
rize 10 evidence based nutrition interventions that have no intention to change behavior in the fore-
impact child mortality [Bhutta, Z.A. et al. (2013). seeable future. Contemplation is the stage in which
Evidence based interventions for improvement people are aware of the problem and are seri-
of maternal and child nutrition: what can be done ously thinking about overcoming it but have not yet
and at what cost. The Lancet, 382 (9890). 452- made a commitment. Preparation is a stage when
477]. Similarly a 2010 study by Wakefield, Loken individuals are intending to take action. Action
and Hornik, examined specifically the use of mass is the stage in which individuals modify their be-
media campaigns (single approach) to change havior, experiences, or environment to overcome
health behavior (multiple outcomes) [Wakefield, their problems. Maintenance is the stage in which
M. Loken, B. and Hornik, R. (2010). Use of mass people work to prevent relapse and consolidate the
media campaigns to change health behavior. The gains attained during action.
Lancet, 376. 1261-1271].
13. Bronfenbrenner’s (1994) social ecological model
7. Data on the type of publication indicates that over postulates that the entire ecological system in
75 per cent of the manuscripts reviewed belonged which growth occurs needs to be taken into ac-
to the ‘peer-reviewed category’ comprised pri- count in order to understand human development.
marily of peer-reviewed journal articles, and to This system is composed of five socially orga-
some extent book chapters and literature re- nized and interrelated subsystems: microsys-
views. Only 25 per cent of the systematic review tem, mesosystem, exosystem and macrosystem.
consisted of project reports. [Bronfenbrenner, U (1994). Ecological Models of
Human Development. In International Encyclope-
8. The 10/90 gap popularized by the Global Forum for
dia of Education, Vol. 3 2nd Ed. Oxford: Elsevier.]
Health Research posits that health research ap-
plied to the needs of low and middle income coun- 14. As with any systematic review, some subjectivity
tries is grossly under resourced with less than is to be expected when coding data. For example,
10 per cent of the world’s resources for health when entering data on the implicit conceptual
research being applied to the health problems of frameworks/ models the coders may have found
low and middle income countries where over 90 the implicit reference to the social ecological
per cent of the world’s preventable deaths occur. model as reflecting their personal commitment to
the social ecological approach, a big focus of the
9. Available at: http://childinfo.org/fgmc_progress.
MPH programme at Drexel.
html
15. SMART is a mnemonic for criteria used to guide
10. See National Cancer Institute (2005). Theory at
the setting of objectives. The letters broadly con-
a Glance: A Guide for Health Promotion Practice
form to the words: Specific – Measurable – Attain-
(second edition) for a summary of conceptual
able - Relevant and Time-bound (Poister, 2003).The
frameworks utilized in this review.
SPICED approach is a useful tool for thinking about
11. Social Learning Theory: Albert Bandura’s social how project objectives can be set in a participa-
learning theory states that human behaviour is tory and inclusive way with local communities. The
learned by observing the actions of others while letters broadly conform to the words: Subjective
being influenced by the environment and person- - Participatory - Interpreted and communicable -
al qualities of the person. Another key concept Cross-checked and compared - Empowering - Di-
within social learning theory is modeling, which verse and disaggregated (Roche, 1999).
involves four steps: paying attention, retaining
16. Some manuscripts mentioned various levels of
details about a behavior, ability to reproduce the
influence, therefore the totals do not add up to
behaviour for example through practice and fi-
a 100 per cent.
nally motivation to engage in a behaviour. A final
contribution of social learning theory is the idea 17. Despite the fact that some behaviours are crimi-
of self-efficacy or an individual’s confidence that nalized, a research and treatment project from
they are capable of performing a behaviour. Germany, encouraged self-identified paedophiles
and hebephiles to seek professional help. The
12. The term ‘stages of change’ is a central construct
first results from this Prevention Project Dun-
in the transtheoretical model. The stages examine
kelfeld (PPD) published in 2009 by Beier et al.
an individual’s readiness to act on a new healthier
found that a notable portion of men who admit
behavior, and provides strategies, or processes
to being attracted to minors could be success-
of change to guide the individual through the five
END NOTES
fully reached via a media campaign. The authors 22. Ideation refers to new ways of thinking and the
concluded that many could be encouraged to diffusion of those ways of thinking by means of
seek clinical diagnosis and a majority had already social interaction in local, culturally homoge-
sought professional help. Another intervention neous communities. The ideation approach has
from Canada specifically targeted maltreating been used to assess the behavioural impact of
fathers by providing counselling and training. various issues including female genital cutting
The objectives were to change abusive parent- (Babalola et al., 2006). The theory of ideation sug-
ing strategies, attitudes and beliefs that support gests that ‘ideation’ variables determine the likeli-
unhealthy parenting, and increase the men’s ap- hood of a person adopting a particular behaviour.
preciation of the impact of violence on children The more favourable the ideation variables re-
among men who have maltreated their children lated to a particular behaviour, the more likely
and/ or who have exposed their children to abuse a person is to adopt and practice the behaviour.
of their mother. The intervention achieved a high Ideation variables include cognitive (knowledge,
number of referrals indicating a high level of felt belief, values, etc.), emotional (emotional re-
need for such interventions. The evaluation of this sponse, self-efficacy) and social (social influence
intervention found that fathers’ level of hostil- and personal advocacy) factors.
ity, denigration, and rejection of their child and
23. For more information on the PRECEDE-PROCEED
their level of angry arousal towards the child and
Model see: http://ctb.ku.edu/en/tablecontents/
family situations decreased significantly over the
sub_section_main_1008.aspx
course of the intervention. Men’s level of stress
was also reported to have decreased although not 24. For more information on global child marriage in-
significantly (Scott and Crooks, 2007). terventions see: Amin, 2005; Erulkar, 2009; Erulkar
and Ayuka, 2007; Gage, 2009; Governance and
18. Formative evaluation is used to design an evidence-
Social Development Research Centre, 2011; Graft,
based intervention. Process evaluation is more
Haberland and Goldberg, 2003; Gupta, Mukherjee,
commonly referred to as monitoring and commonly
Singh, Pande and Basu 2008; and International
used to determine if an intervention is being deliv-
Centre for Research on Women (ICRW), 2001.
ered according to plan and to provide a means to
make mid-course corrections. Impact evaluation is 25. For example, Evans and Bhattarai (2000) provide
used to determine effectiveness of a intervention. a comparative analysis of anti-trafficking inter-
vention approaches in Nepal and Kaufman (2011)
19. See for example: Law and Shek, 2012; Vuijk, van
reviews research and activism addressing sex
Lier, Crijnen and Huizink, 2006; Simon et al., 2008;
trafficking in Nepal.
Malti, Ribeaud and Eisner, 2011.
26. See Brady (2007) for an examination of an inter-
20. For a summary of FGM/ C programmes, see Berg
vention in rural Egypt and Jeejeebhoy, Acharya,
and Denison, 2013.
Kalyanwala, Nathani and Bala (2009) for the eval-
21. A Type 1 error is when an effect is detected when uation findings of a life skills education interven-
in fact none exists. Conversely, a Type 2 error re- tion in rural Uttar Pradesh, India.
fers to the failure to detect an effect when one is
present. A Type 3 error occurs when an interven-
tion is not implemented according to plan.
Boys at Bhagyanagar Chil-
dren’s Home in India: two near-
by children’s homes separately
house girls and boys mostly
between 6-14 years who are
orphans, children of migrant
labourers and potential child
labourers.
Communication for Development Section
Programme Division
UNICEF New York
Email: c4dhq@unicef.org
Website: www.unicef.org/cbsc

C4D VAC Systematic Review Report

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

C4D VAC Systematic Review Report

Загружено:

Авторское право:

Доступные форматы

C4D WORKING

Communication for Development

tle: 3: C4D and Violence Against Children - working paper

Communication for Development Approaches

Authors: Suruchi Sood and Carmen Cronin

© United Nations Children’s Fund (UNICEF),

Suggested citation: Sood, S. and C. Cronin,

The designations in this publication do not imply

Cover photo: © UNICEF/UNI159383/Pirozzi

For further information, contact:

Communication for Development and Child

ISBN Number 978-0-9911961-6-6

In the State of Palestine, five-

RAFAEL OBREGÓN STEPHEN BLIGHT

5 Limitations of this systematic review ..................................................... 20

SECTION 1: Sampling information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

SECTION 2: Programme design elements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

SECTION 3: Programme implementation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

SECTION 4: Overall programme evaluation processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

SECTION 5: Thematic analysis of key results by type of research. . . . . . . . . . . . . . . . . . . . . . 44

TABLE 2: Inclusion and exclusion criteria for the systematic review..............................................17

TABLE 3: Search engines utilized for the systematic review..........................................................................17

TABLE 4: Location of interventions for manuscripts in the

TABLE 5: Conceptual frameworks in the manuscripts included in the

TABLE 6: Implementing agencies and partners for manuscripts in the

TABLE 7: Intended audiences for manuscripts in the systematic review.............................. 35

TABLE 8: Evaluation design for manuscripts in the systematic review...................................... 38

TABLE 9: Research methods for manuscripts in the systematic review................................. 39

TABLE 10: Analysis methods for manuscripts in systematic review................................................ 39

FIGURE 2: Structure of this report’s findings section .........................................................................................................15

FIGURE 3: Summary of manuscripts in the systematic review......................................................................18

FIGURE 5: Year of publication for manuscripts included in the

FIGURE 6: Summary of sampling findings....................................................................................................................................................... 26

FIGURE 7: Example of conceptual model from KiVA intervention.......................................................... 28

FIGURE 8: Levels of influence included in the systematic review database.................. 30

FIGURE 9: Results on levels of influence from the manuscripts included in the

FIGURE 10: Examples of C4D interventions influencing policy .................................................................... 32

FIGURE 11: Summary of programme design findings............................................................................................................ 34

FIGURE 12: Results on the strategic communication

FIGURE 13: Results on the communication channels used in the

FIGURE 14: Summary of programme implementation findings....................................................................... 37

FIGURE 16: Summary of overall evaluation processes findings ................................................................. 43

FIGURE 17: Summary of findings from formative research results .................................................... 48

FIGURE 18: Summary of findings from process evaluation results .................................................... 52

FIGURE 19: Summary of findings from impact evaluation results ...........................................................61

FIGURE 20: Potential directions for future research and study.................................................................... 68

CO-PRINCIPAL INVESTIGATOR: Günay Salazar

CO-PRINCIPAL INVESTIGATOR: Ami Sengupta

INPUTS AND CONTRIBUTIONS:

EDITING AND PROOFREADING

DESIGN AND LAYOUT

Children and their teacher

Communication for Development (C4D)1

VAC is ubiquitous and manifests itself in differ-

Overlapping results for different searches re-

by university-based researchers) tended to use

Many of the interventions focus on interven-

The paucity of robust and rigorous evaluations in

The review shows that prevention efforts can

A group of children play ëis

by the Programme Division at effectiveness of C4D approaches in

Children hold books and other

The following chapters summarize the methodology

The central section presents the key findings from the