We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
Impact Evaluation of ISFM Technology Dissemination Program in Burkina Faso
Last registered on March 28, 2019


Trial Information
General Information
Impact Evaluation of ISFM Technology Dissemination Program in Burkina Faso
Initial registration date
March 20, 2019
Last updated
March 28, 2019 6:45 PM EDT

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
UC Berkeley
Other Primary Investigator(s)
PI Affiliation
Université de Koudougou
PI Affiliation
University of Mannheim
PI Affiliation
World Bank Group
Additional Trial Information
On going
Start date
End date
Secondary IDs
Farmers in many Sub-Saharan African countries face chronically low crop yields. A number of field studies have been conducted to further understand the sources of these low yields and the constraints that farmers face not only to increasing their yields, but also to translating this gain into higher income and overall improved well-being. Evidence has shown that the uptake of capital intensive technologies such as chemical fertilizers or improved seeds is low, especially in areas where capital is scarce relative to labor. Integrated Soil Fertility Management, or ISFM as it will henceforth be referred to, offers an alternative approach that complements capital intensive approaches with labor intensive techniques such as field preparation, innovative planting, and organic fertilizers.

The focus on this study is on cowpea farmers in Burkina Faso. The evaluation will take a closer look at ISFM techniques and answer two critical questions: 1. How do these techniques compare with traditional practices (pratique paysannes, or PP), and how do they compare with strict capital intensive approaches, and 2. If labor intensive technologies prove to be effective, are they easily adopted by all farmers? If not, what are the constraints to their adoption?

The evaluation will compare traditional (PP) practices both with labor-intensive technologies alone, and then with a combination of labor and capital intensive technologies. This comparison will help researchers answer another critical question: are labor intensive techniques sufficient on their own to generate higher yields, or are they more effective when combined with capital intensive technologies? In other words, do these approaches work as substitutes or complements?

To conduct the randomized evaluation, the implementing partner, Groupe de Recherche et d'Actions pour le Développement (GRAD), will select 99 farming organizations from the existing 262 in the Sanmentenga region of Burkina Faso. After identifying these 99 farming organizations, each will be asked to nominate a potential demonstrator, i.e. a farmer willing to perform the demonstration to the farmers within his/her network. 40 of these farming organizations and their selected demonstrator will be assigned to treatment, and the remaining 59 will be assigned to control. Those assigned to the treatment will receive assistance and expertise from GRAD in setting up the demonstrations, and those assigned to control will not. The outcomes of interest in the first stage are crop yields, which will be measured precisely using crop cut techniques. Researchers will compare the yields of farmers in the control group to those in the treatment group, and further compare the outcomes of the labor intensive techniques alone to the combination of labor and capital intensive techniques. In the second stage, researchers will investigate how each technology is adopted throughout the demonstrators’ farming organization, comprised of roughly 17 farmers each. This will be done through quantitative and qualitative surveys, which will capture information about agricultural knowledge and practices as well as social networks.
External Link(s)
Registration Citation
Bouguen, Adrien et al. 2019. "Impact Evaluation of ISFM Technology Dissemination Program in Burkina Faso ." AEA RCT Registry. March 28. https://doi.org/10.1257/rct.2312-1.0.
Former Citation
Bouguen, Adrien et al. 2019. "Impact Evaluation of ISFM Technology Dissemination Program in Burkina Faso ." AEA RCT Registry. March 28. https://www.socialscienceregistry.org/trials/2312/history/44301.
Experimental Details
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
The main questions that this randomized experiment aims to answer are: 1. How do ISFM labor intensive farming techniques (LI) compare to traditional farming practices (PP) and to capital intensive ones? 2. Do farmers who employ labor intensive practices alone fare better or worse than those who employ a combination of capital and labor intensive techniques (LI+)? 3. If ISFM practices do indeed generate greater yields for farmers, does this translate necessarily into improved wellbeing in the long run? 4. Are ISFM technologies easy to adopt among all farmers, and how does this adoption of knowledge and new practices spread through social networks?

These questions will be answered by looking at key variables such as crop yields, revenue generated by the last harvest of cowpeas, and household spending. Data will be collected through a combination of qualitative and quantitative surveys as well as secondary data from sources such as the Agricultural Census conducted by the Directorate of Agricultural Statistic and other annual surveys.

Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
From the 262 cowpea framing organizations in the Sanmatenga area of Burkina Faso, 99 will be selected. These 99 are selected on the criteria that they have not been previously exposed to the interventions of the partner organization GRAD (Groupe de Recherche et d'Actions pour le Développement), and that they are sufficiently spread out as to limit the possibility of spillover. Each of the 99 FOs will nominate one demonstrator.

After this, 40 FOs will be randomly assigned to the treatment group, in which the selected demonstrator will receive expert training and instruction from GRAD on ISFM techniques, and 59 will be assigned to the control group, in which they will receive nothing. The randomization will allow researchers to assess: 1. The comparative yields of farmers performing different practices within the treatment group, 2. The comparative yields of farmers in treatment versus control FOs, 3. How quickly the new technologies trickle down within the same FO.

About 17 farmers from each of the 99 FOs will be surveyed, for a total sample of 1,782 farmers (99 demonstrators + 17 farmers within their direct network).
Experimental Design Details
Not available
Randomization Method
Re-randomization method: we first randomized the 99 OP into 2 groups. We then regressed the treatment variable against 5 stratification variables ( total number of FO members, gender, whether the FO had support from an organization, number of FO meetings last year, name of the commune). We only kept the draws that present a F-test lower than 1 for each of the 5 stratification variable regression and that presented a balanced number of treatment and control in each commune (the treatment ratio could not be below 25 % or above 67% in each commune). We draw 9999 randomizations and took the median value of the 10 first one.
Randomization Unit
The unit of randomization is the farming organization. 99 will be selected from the existing 262 in the region.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
Sample size: planned number of observations
The study will be conducted on 1510 farmers in total: 99 demonstrators from different farming organizations plus 18 members of each of their networks.
Sample size (or number of clusters) by treatment arms
Out of 99 total farming organizations, 40 were assigned to treatment, and 59 to control.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
MDE 20.17 % sd with full compliance, and a baseline capturing 20% of the total variation. With a more reasonable compliance of 80%, MDE is reduced to 25% SD. ICC= 10 %
IRB Name
Mannheim Review Committee
IRB Approval Date
IRB Approval Number
Analysis Plan

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information