The role of seed product performance information on farmers’ seed choice

Last registered on March 10, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

The role of seed product performance information on farmers’ seed choice

RCT ID

AEARCTR-0015495

Initial registration date

March 10, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

March 10, 2025, 9:49 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Kenya

Region

Embu and Kirinyaga Counties

Primary Investigator

Name

Michael Ndegwa

Affiliation

International Maize and Wheat Improvement Center (CIMMYT)

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Collins Bulinda

PI Affiliation

CIMMYT

Contact Investigator

PI Name

Sarah Kariuki

PI Affiliation

CIMMYT

Contact Investigator

PI Name

Fredrick Otiato

PI Affiliation

CIMMYT

Contact Investigator

PI Name

Moti Jaleta

PI Affiliation

CIMMYT

Contact Investigator

Additional Trial Information

Status

In development

Start date

2025-03-17

End date

2025-04-04

Keywords

Agriculture

Additional Keywords

Seed performance infomration, Seed choice

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

This study will examine the role of third-party seed product performance information in influencing smallholder farmers’ seed choices at the point of sale in Kenya. Farmers often lack reliable, objective data to guide their seed selection, leading many to continue using outdated or uncertified maize seed varieties despite the availability of improved options. To address this gap, the study will employ a randomized controlled trial (RCT) in Embu and Kirinyaga counties, involving 2,250 farmers across 50 agro-dealer shops. Farmers will be randomly assigned to one of four groups: a comparison group receiving general farming information and three treatment groups receiving different types of seed performance data, including county-level average yields, top 25% yield potential, or both. The study will measure the impact of performance information on farmers’ purchasing decisions, including variety selection, switching behavior, and seed quantity purchased. The findings will inform future efforts to establish independent seed performance testing and dissemination systems, ultimately supporting more informed decision-making among farmers.

External Link(s)

Registration Citation

Citation

Bulinda, Collins et al. 2025. "The role of seed product performance information on farmers’ seed choice." AEA RCT Registry. March 10. https://doi.org/10.1257/rct.15495-1.0

Sponsors & Partners

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

Request Information

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Intervention Start Date

2025-03-17

Intervention End Date

2025-04-04

Primary Outcomes

Primary Outcomes (end points)

- Demand for listed and top performing hybrids (extensive and intensive marging)
- Actual purchases for listed and top performing hybrids (Extensive and intensive margins)
- Switching behavior - demand for and/or purchasing products that was initally intended

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

The study will be conducted in Embu and Kirinyaga counties in Kenya. About 50 agro-dealers (25 per county) will be enlisted to participate in the study. Approximately 45 farmers per agro-dealer will be invited to participate in the study, giving a total of 2250 farmers as the sample size. Farmers will be randomly assigned to four experimental groups.

- Comparison Group (T0) – Placebo Treatment
Farmers will receive general farming information unrelated to maize seed selection, such as fun facts about farming in Kenya or pest management tips.

- Treatment 1 (T1) – County Average Yield Information
Farmers will receive information on the county-level average yield performance of various maize seed products, based on data collected from farmers in their county.

- Treatment 2 (T2) – Varietal Yield Potential (Top 25%)
Farmers will receive information on the yield potential of maize seed products based on the top 25 percent of farmers who achieved the highest yields, demonstrating the varieties' full potential.

- Treatment 3 (T3) – Both County Average and Top 25 Percent Yield Information
Farmers will receive both the county-level average yield data and the top 25 percent yield performance data, providing a more comprehensive performance comparison of maize seed varieties.

Experimental Design Details

Randomization Method

We shall use computer generate random numbers to assign farmers into experimental groups

Randomization Unit

Individual level randomization

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

2250 farmers. No clusters

Sample size: planned number of observations

2250 farmers

Sample size (or number of clusters) by treatment arms

562

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We define the main outcome of interest as the probability of purchasing the top two listed varieties (as observed in a similar information study) and use distributional parameters from the POS information study implemented by our team in 2024 to compute power and sample size for the current study. The probability of the comparison group purchasing the top two listed varieties in our previous studies was 0.059 (6%) with a standard deviation of 0.236. In that study, information alone achieved a treatment effect of 0.07, meaning that information increased the likelihood of purchasing the top two varieties by approximately seven percentage points relative to the mean of the comparison. We therefore consider a more modest minimum detectable effect (MDE) of 0.04, meaning that we expect our information treatments to increase the likelihood of purchasing a promoted variety by at least four percentage points relative to the comparison group mean. To achieve sufficient power (0.8) to detect the desired MDE (0.04), we need a minimum of 551 individuals in both the control and treatment groups for pairwise comparisons. Since we aim to compare each treatment against the control, we maximize power by keeping each group the same size. This results in a total sample of 2,204 farmers, with 551 farmers in the control group and 551 in each of the three treatment arms.

Supporting Documents and Materials

IRB