Bioversity International: research for development in agricultural and tree biodiversity

ClimMob – a software for crowdsourcing climate smart-agriculture

28 Sep 2016

A new approach to participatory research will make it easier for farmers to contribute to new agricultural solutions. ClimMob, developed by Jacob van Etten and his colleagues at Bioversity International, turns the research paradigm on its head; instead of a few researchers designing complicated trials to compare several technologies in search of the best solutions, it enables many farmers to carry out reasonably simple experiments that taken together can offer even more information.

Van Etten says he was inspired to develop ClimMob by his frustration at some of the participatory research he saw firsthand in the field. "A lot of it was trying to imitate what would happen on experimental stations in farmers' fields," van Etten recalled. Asking farmers to produce statistically meaningful results often complicated matters too much.

He set out to make something so simple that farmers can understand it and feel that they are in charge. "We adapt our statistics to work with what the farmers are able to observe, rather than the other way round."

The name ClimMob reflects another driving force: climate change means farmers need to mobilize locally valuable options for adaptation, and more efficient participatory research is a good way to find those. 

To look for an improved variety, for example, ClimMob offers each farmer a package of three different varieties. The farmer has to note which of the three is best and which worst on a list of characteristics that they develop together with the researchers. The varieties are drawn from a pool of several varieties, so while one farmer receives A, B and C, another receives A, B and D and a third B, C and D. Even though no farmer compared A and D directly, statistical methods can reveal whether A or D is better. Additional variables, such as whether a farmer has access to irrigation, or the altitude of the plot, can also be examined to see whether they affect the performance of the varieties. The researchers call these citizen science experiments 'tricot' trials, for triadic comparisons of technologies.

In a preliminary trial of beans in Honduras, farmers looked at growth habit, response to pests and diseases, productivity, taste and marketability, among other factors. Van Etten and his colleagues compared the rankings of individual farmers with the 'gold standard' ranking of an expert on beans. Although each individual farmer's observations did not reliably mirror the expert's ranking, on average the farmers' rankings of the varieties were almost identical to the expert's. Other trials in Ethiopia, India and Nicaragua confirmed the value of the approach, the quality of the data it could produce and the willingness of farmers to participate.

These early trials also resulted in modifications to the process. At first the farmers were asked to rank the varieties in order, from first to third on each of the characteristics. "Farmers found that confusing," van Etten said "so we changed it so that they have to decide which is best and which is worst." That worked better, and has an added advantage that it is easier for the researcher to collect data by phone.

The development of ClimMob to date has followed twin tracks, refining the processes and statistical approaches and developing an online platform to make it easier for researchers to manage and analyze tricot trials. The online platform is up and running for anyone to use, with a full set of explanatory videos (in English and Spanish) showing how to use it, and there is a mobile app that researchers can use on their smartphones.

One key advantage of the platform is that it makes it easy for the researchers to prepare the packages of three different varieties that will go to the farmers. More importantly, van Etten says, it enables very rapid analysis of the results. These preliminary results may not be the last word, but they can be shared with the farmers soon after the end of the trial.

"There's a short time window to give feedback to the farmers," van Etten says. "It has to be quick, so they can decide if they want to grow one of the trial varieties next season. That's what keeps them engaged, and engagement is the engine of participatory research."

At the moment ClimMob is being used mostly in trials of annual crops, finding the best varieties for farmers to use. Van Etten says it can do far more, examining almost any technology. Strains of beneficial fungi, grafting methods for perennial crops and different approaches to integrated pest management are three he cites off the top of his head.

Visit the ClimMob platform

Watch the videos in English and Spanish

Download the manual in English and Spanish


This research is part of the CGIAR Research Program on Climate Change, Agriculture and Food Security and supported by CGIAR Fund Donors. It was financially supported by USAID Development Innovation Ventures.