NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
The effect of training on knowledge about, attitude towards and adoption of organic farming practices: A randomized field experiment in Indonesia
Initial registration date
February 11, 2019
July 24, 2019 3:23 AM EDT
University of Passau
Other Primary Investigator(s)
University of Passau
Additional Trial Information
The agricultural sector in Indonesia faces several challenges, including food security and sustainability for a growing population, environmental degradation and climate change, improving income and living standards for smallholders and an aging workforce in the agricultural sector. Organic farming has the potential to address some of these challenges. It is more environmentally friendly than conventional farming and potentially it could increase the attractiveness of farming for the younger population if it is perceived as a new modern technology. However, the adoption of organic farming is still at a very low level in Indonesia. This research explores the effectiveness of interventions on knowledge about, attitude towards and adoption of organic farming using a randomized control trial. The study is conducted in two provinces in Java, Indonesia enrolling a sample of 1,200 farmers. Farmers in treatment groups are exposed to alternative treatments that either offer (1) only training on organic farming practices and the marketing of organic products, (2) the same training but augmented by a video intervention that raises awareness for potential health and environmental consequences of conventional farming or (3) again the training but this time augmented by an intervention that strengthens mutual support within farmer groups with respect to a conversion towards organic farming. Interventions are randomized at the village level and not the individual level to limit the occurrence of unwanted treatment spillovers. The study aims to provide non-government institutions and policy makers with evidence on the effectiveness of training on organic farming.
Grimm, Michael and Nathalie Luck . 2019. "The effect of training on knowledge about, attitude towards and adoption of organic farming practices: A randomized field experiment in Indonesia." AEA RCT Registry. July 24.
The study comprises three treatment groups (200 farmers each) and the control group (600 farmers). The treatment consists of a three-day training on organic farming practices, organic farming principles and marketing of organic products. The aim is to encourage the practice of organic farming methods, to increase the awareness for problematic aspects of conventional farming and to increase the knowledge on principles. AOI, national NGO, designed the training content and training schedule together with two local partners in the research regions
Treatment 1 (training): Intensive 3-day technical training incl. training on marketing
Treatment 2 (training + video & discussion): Intensive 3-day technical training incl. training on marketing plus intense awareness training on health and environmental impacts of conventional agriculture.
Treatment 3 (training + role play): Intensive 3-day technical training incl. training on marketing plus an intervention that strengthens the solidarity within farmer groups for mutual support with respect to the conversion towards organic farming.
We further visited each training village again around eight months later for a two-hour “coaching clinic” session. This coaching clinic session offered farmers the opportunity to discuss problems and questions related to our training with their trainers from the three-day training.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
a) Adoption of organic farming practices and reduction in chemical input use
b) Knowledge on organic farming (technical aspects and marketing)
c) Attitude towards organic farming
Primary Outcomes (explanation)
a) Organic farming practices refer in particular to the use and production of organic fertilizer and pesticides. This is an important practical part of the all three types of training interventions. Further, we measured chemical input use at the baseline and will explore whether organic farming training leads to a reduction in the use of chemical fertilizers and chemical pesticides.
b) The questionnaire contains knowledge questions on organic farming practices and regulations that qualify for a certification (e.g. whether it is permitted to burn land) as well questions on organic labels. The answers to these questions were discussed during the training. The questionnaire also contains question on potential negative effects of chemical input use on health and environment. These aspects were discussed in more depth in the video treatment group. c) Attitudes relate to whether the respondent perceives organic farming as a promising farming technology irrespective of adoption status. This includes aspects such as the perception of demand for organic products or classification of organic farming as modern or old-fashioned technology. It further includes price expectations for organic farming products.
Secondary Outcomes (end points)
Heterogeneous effects of the intervention
Secondary Outcomes (explanation)
We will test for heterogeneous effects of the intervention on all outcomes for the following:
• Age of the respondent • Education of the respondent
• Social capital of the respondent
• Awareness regarding agricultural environmental pollution (measured at baseline) • Risk aversion (measured at baseline)
• Land size (measured at baseline) • Location
To select 1,200 respondents, we applied a three stage random sampling design. In the first stage we sampled villages and in the second stage we sampled farmer groups from the selected villages. In each village in which farmer groups were sampled AOI held a two hours information session on organic farming to which all members of the respective farmer groups were invited. In the third stage, we sampled participants from all members who attended the information session. Hence, these information sessions served two purposes (1) allow for self-selection based on a basic interest in organic farming and the willingness to participate in farmer group events (2) collect contact details on prospective respondents.
The treatment was randomized at the village level. Half of the villages (30 villages) formed the control group while the other 30 villages received the trainings. Among the training villages we again randomized the type of training intervention. To assess the impact of the intervention we will compare farmers in treated farmer groups and farmers in control farmer groups at the follow up stage. Given the randomization of the intervention, this should yield unbiased treatment effects.
We invited the 20 interviewed farmers in each treatment village to participate in the training. In each farmer group we sampled the farmer group head whenever it was possible (as one of the 20), i.e. when the farmer group head was an active farmer. Given the importance of the farmer group, we expect that this enhances the uptake of any organic farming methods. The training was held on three consecutive days in each farmer group’s village. Organising the training in the respective villages minimized the barriers for farmers to attend. The farmers received Rp 50,000 for each day of the training (only if they attended) to cover any transport costs and to compensate them for their potential forgone earnings on that day. After the training, the farmers received a printed manual on organic farming.
We further visited each training village again around eight months later for the two-hour “coaching clinic” session.
Experimental Design Details
Stratified randomization. Randomization was done by a computer.
The training interventions were randomized at the village level (desa level). Depending on the size of the farmer groups in the respective villages, a village group included one, two or three farmer groups. However in most cases, the first farmer group drawn was sufficiently large and thus there was only one farmer group sampled in most villages.
We stratified the sample according to urban and rural status and village agricultural land area. In one of the two regions, we additionally used “travel distance to the district capital” as stratification criteria as this region is characterized by less developed infrastructure and long travel times within the district. Based on these criteria we created five cells à six villages in each region. Within each cell, we randomly allocated three villages to the control group, one village to treatment 1, one village to treatment 2 and one village to treatment 3.
Was the treatment clustered?
Sample size: planned number of clusters
60 villages, 30 villages in each region
Sample size: planned number of observations
1,200 farmers (20 per village)
Sample size (or number of clusters) by treatment arms
30 villages (15 in each region) did not receive training on organic farming
10 villages (5 in each region) received the training 10 villages (5 in each region) received the training and the video treatment 10 villages (5 in each region) received the training and the role play
In total, 600 farmers in 30 villages received invitations to one of the three training intervention types.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
A sample of 1,200 participants (600 in control group and 600 in treatment groups) would give 80% power (alpha=0.05, 2-tailed) to detect
- a difference of 16.4% in organic fertilizer use (dummy variable 1=use 0=no use; use at baseline 37.2%; intraclass correlation at baseline 0.164)
- a difference of 17.17% in organic pesticide use (dummy variable 1=use 0=no use; use at baseline 8.6%; intraclass correlation at baseline 0.185)
- a difference of 16.8% in perception of agricultural pollution as problematic (dummy variable 1=use 0=no use; perception as problematic at baseline 45.5%; intraclass correlation at baseline 0.175)
INSTITUTIONAL REVIEW BOARDS (IRBs)
Post Trial Information
Is the intervention completed?
Intervention Completion Date
January 31, 2019, 12:00 AM +00:00
Is data collection complete?
Data Collection Completion Date
April 26, 2019, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
Final Sample Size (or Number of Clusters) by Treatment Arms
30 villages control, 30 villages treatment