Communication for Development (C4D) Impact Evaluation
Last registered on October 01, 2014


Trial Information
General Information
Communication for Development (C4D) Impact Evaluation
Initial registration date
October 01, 2014
Last updated
October 01, 2014 2:06 PM EDT
Primary Investigator
Northwestern University
Other Primary Investigator(s)
PI Affiliation
ISSER, University of Ghana
PI Affiliation
Harvard School of Public Health
PI Affiliation
Yale University
Additional Trial Information
Start date
End date
Secondary IDs
The Communication for Development (C4D) program is an intervention package implemented by UNICEF and Ghana Health Services (GHS) which aims to increase awareness of common health problems in Ghana and induce behavior change. The program aims to reduce both morbidity and mortality, particularly for children under five. The key behaviors that C4D attempts to influence are: (i) Exclusive breast feeding for the first six months and complementary feeding up to age 2; (ii) Hand washing with soap at key times (before and after cleaning a baby, before preparing food, and after defecating); (iii)Malaria prevention by sleeping under an insecticide treated net; (iv) Diarrhea treatment, specifically the importance of using ORS; and (v) Birth delivery by a skilled attendant.

The C4D program includes three complementary interventions: counseling and community outreach by Community Based Agents (CBAs), community dramas, and community radio broadcasts. In addition, a Mobile for Development (M4D) campaign is currently being implemented, to further support and complement the basic C4D interventions. The M4D intervention delivers key health messages on the same five behaviors via pre-recorded voice messages. Beneficiaries receives two messages per week for 24 weeks.

Innovations for Poverty Action (IPA) is evaluating the impact of the C4D program. In 2012, IPA conducted a baseline survey of 4320 respondents (mothers with children under the age of five) in 216 communities. For the C4D program, 108 communities were randomly selected. Within these 108 communities, 20 households per community were randomly selected to be surveyed. A midline survey was conducted from July-September 2014, and an endline survey will be conducted at the end of the program.

In addition to the community-based interventions, approximately 2000 mothers were chosen for an individual level mobile phone intervention (M4D). The eligibility requirements for the M4D intervention included: (i) phone ownership or access to a phone through a family member living in the same household, and (ii) mothers with children under the age of five.Eligible mothers were randomly selected to either receive messages from the same speaker, or to receive messages from different speakers from their own area. Mobile voice messages were developed to cover the same five behaviors outlined above.

External Link(s)
Registration Citation
Fink, Gunther et al. 2014. "Communication for Development (C4D) Impact Evaluation." AEA RCT Registry. October 01.
Experimental Details
Communications for Development (C4D) is a social and behavior change communication intervention being implemented in twelve districts of the four poorest regions of Ghana : Central, Upper East, Upper West and Northern Region (IPA is evaluating in the three Northern regions). Over the next two years and with funding from UNICEF, Ghana Health Services (GHS) will continue to implement a package of C4D activities designed to encourage families to adopt and consistently practice five key behaviors to promote child and maternal health:

1. Sleeping under an insecticide-treated mosquito net (ITN)
2. Home treatment of diarrhea using Oral Rehydration Salt (ORS)
3. Hand washing with soap at critical times
4. Exclusive breastfeeding and appropriate complementary feeding
5. Delivery with a skilled birth attendant

To promote persistent and sustained behavior change, C4D utilizes three key communication activities:
1. Home visits and counseling by Community Based Agents (CBAs) affiliated with GHS
2. Ghana Community Radio Network broadcasts of focus group discussions and targeted jingles
3. Theatre dramas supported by the Center for National Culture

Currently, a mobile messaging intervention (M4D) is being implemented to complement the ongoing C4D activities. M4D targets mothers of children under five years of age who have access to mobile phones to encourage the practice of the five key behaviors outlined above. IPA has partnered with VOTO mobile to send voice messages in local languages to fifty percent of randomly selected individuals (1591 mothers) from the full eligible C4D sample. The treatment group receives two messages a week for 24 weeks. There are a total of 15 unique messages on the five key behaviors.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Knowledge and behavior change associated with five key behaviors:

1. Use of ITNs
2. Handwashing with soap
3. Use of skilled birth attendant
4. Use of ORS to treat diarrhea
5. Exclusive breast feeding and complementary feeding

The survey uses three methods to ask questions about behavior:
1. Self reported
2. Through objective proxy measures
3. List randomization

Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The sample for the C4D program in Ghana consists of 4320 households in 216 communities. 108 communities were randomly selected to receive the intervention(s) while the remaining 108 communities served as the comparison group. Twenty households were randomly selected from each community to be surveyed as a part of the RCT. In all treatment communities, GHS provided additional training to Community-Based Agents (CBAs) in C4D communications strategies, and the CBAs conducted home visits and counseling in their community. Within the treatment group, communities were further randomized into one of four groups:

1) Live community dramas and live radio discussions
2) Live community dramas and no live radio discussions (only exposure to broadcasts)
3) Recorded community dramas and live radio discussions
4) Recorded community dramas and no live radio discussions (only exposure to broadcasts)

Additionally, for the entire C4D sample (treatment and control), eligible respondents were randomly assigned to receive voice messages on the five key health behaviors (M4D). M4D followed a two-step randomization process. The first randomization assigned respondents to three groups, as follows:

Treatment 1: Receives messages from multiple voices (798 respondents)
Treatment 2: Receives messages from one voice (794 respondents)
Control: No messages (788 respondents)

The first randomization was stratified on four variables:
1. C4D treatment group
2. Level of phone access (personal, spouse, or other household member)
3. Health behavior (low, high): We used C4D baseline data to ensure that an equal proportion of respondents that indicated low health behavior were assigned to each M4D treatment and control group.
4. Health knowledge (low, high): We used C4D baseline data to ensure that an equal proportion of respondents that indicated low health knowledge were assigned to each M4D treatment and control group.

A second level randomization was conducted which assigned individuals in Treatment 2 to one of three voices. The same stratification and balance variables used in the first randomization were used in the second one.
Experimental Design Details
Randomization Method
Randomization was done both in the office using computer programs and in the field using randomization sheets
Randomization Unit
C4D: communities, then within communities, households
M4D: individual
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
216 communities
Sample size: planned number of observations
4320 households
Sample size (or number of clusters) by treatment arms
108 control communities
108 treatment communities:
TI: 27 communities will receive Live Drama
T2: 27 communities will receive Recorded Drama
T3: 27 communities will receive Live Radio + Live Drama
T4: 27 communities will receive Live Radio + Recorded Drama

Treatment 1: Receives messages from multiple voices (798 respondents)
Treatment 2: Receives messages from one voice (794 respondents)
Control: No messages (788 respondents)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB Name
Innovations for Poverty Action IRB - USA
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers