Livestock is extremely important to the economy of northern Somalia. Sheep and other animals’ meat, milk, and wool account for almost 80% of the region’s annual exports. However, the grazing grounds of the region have been destroyed by years of drought. By focusing on only a few settlements, an international team is trying to figure out if those unusual accomplishments can be used as a model for restoring rangelands elsewhere.
It is necessary to rethink standard data processing in order to answer this question. According to Basma Albanna, a development expert at the University of Manchester in England, success stories like those Somali villages with sustainable grazing are statistical outliers. “The usual procedure is to remove outliers from the data.”
Yet Albanna and other proponents of the “positive deviance” approach argue that even the most extreme outliers can reveal important information. In the midst of what many consider to be noise, they comb through data in search of signals. Using “deviants,” or outliers in large datasets, researchers hope to learn why some people or societies flourish while others in similar situations fail. This information is then used to design ways to help those who are languishing achieve positive outcomes.
Megan Higgs, a statistician and independent consultant in Bozeman, Montana, believes that positive deviance has the ability to address a persistent issue. An editor of the International Statistical Institute’s blog argues, “In research generally, we have an overemphasis on quantifying averages. There may be fewer persons who fit the average in a study’s sample, she says. Averages can hide important information.
To avoid missing “a hugely important part of the picture,” Higgs says, “I just worry that we are missing a hugely important approach.”
The term “positive deviance” was coined in the 1970s, but it was not until nearly two decades later that the approach began to acquire popularity. Monique Sternin and her late husband Jerry Sternin, who worked for Save the Children as humanitarian relief workers in 1990, tested a positive deviance experiment in Vietnam in an effort to combat the country’s alarmingly high rates of childhood malnutrition. To avoid using typical but unsustainable charity methods, Vietnamese government officials encouraged the couple to assist local populations in a more sustainable manner.
They instead tried to identify youngsters in underprivileged areas who were able to maintain their nutrition in spite of the odds. The Sternins instructed locals to weigh children under the age of three in four villages in Thanh Hoa Province. These villages contained a total of 2,000 children under the age of 3. Almost seventy percent of the youngsters were found to be malnourished, with almost half of them being in imminent danger of death.
The couple then asked the villagers to identify children from the poorest homes who had a better body mass index (BMI) than the others. There were a select few families in each town who met the criteria. Monique Sternin, a positive deviance consultant in Boston, Mass., says, “We went to talk to those people.”
The Sternins noticed that families who fed their children little shrimp and crabs that lived in rice paddies and potato greens found on the roadside were more likely to have children who weighed a healthy weight. These foods were considered “taboo” or hazardous by the people of the village, according to Sternin. They also fed their kids three to four meals a day instead of the typical two meals.
On the surface, getting more families to eat these forbidden items looked like a straightforward answer. However, putting this plan into action proved to be extremely difficult. As Sternin puts it, “the positive deviants are outliers, rebels.” Sternin says. They could not ethically “out” families who were defying social conventions and traditions, because that would be unethical of them.
Instead, they offered free grains to the peasants. Hamlet women and relief workers sponsored cooking classes for the children of the locals, who were taught by the ladies of the village. For a total of 12 days, the villagers received an additional dinner as a result of these meetings. A requirement of participating was that each family in the community brought a dish to pass, which included wild greens and tiny shrimp and crab. Families and caregivers were able to see for themselves that the foods had a positive effect on their children’s health over the course of 12 days.
As a result of their efforts, the pair went on to create similar initiatives across the country. As a result, similar nutrition programs have been implemented over the world.
The work of the Sternins was inspiring, but it necessitated a tailored approach to data collection. If big data, such as satellite photos and social media material, paired with qualitative research, could get the job done at a reduced start-up cost, Albanna was curious.
Albanna lists numerous advantages of big data. The approach is less time-consuming than visiting door-to-door because the datasets are already in place. Outliers at the village or neighborhood level rather than at the individual level can be identified, which lessens privacy concerns.
Albanna also points out that as the dataset grows, it will be easier to spot positive deviations. There are very few if any ‘positive deviants’ out there. “We are talking 2% to 10% of the sample you are looking at,” she explains.
The Data Powered Positive Deviance program was co-founded by Albanna and a number of international partners in 2020. The collaborative’s pilot initiatives include locating Mexico City’s safest public spaces for women and locating the Niger communities that grow the most millet. In September, a study published in the International Journal of Environmental Research and Public Health identified the districts in Germany that were most effective at halting the spread of COVID-19. As part of the collaborative, the Somalia healthy rangeland project is also being carried out.
The team’s initial task in Somalia was to locate prosperous communities. During the 2016-2017 drought, Albanna adds, “we started, hoping that we could identify communities that are able to sustain and maintain the numbers of their livestock.” This drought was extremely severe, causing food shortages in more than half of the country.
Afterwards, the researchers focused on 314 communities in the rugged West Golis region of northern Somalia and analyzed three sets of information. Using data on rainfall and land cover, the scientists were able to group together villages that were comparable. From 2016 to 2020, Earth-observing satellite data was used to estimate the density of vegetation. Thirteen communities were identified as possible positive deviants, despite the drought, thanks to this procedure.
Distinctive conservation methods highlighted in the outliers’ detailed satellite pictures were crucial in protecting neighboring rangeland. According to a study published in Development Engineering on December 24, some villages erected shrub barriers around settlements to prevent erosion or carved half-moon-shaped basins into landscapes to hold water.
The team recruited a local pastoralism and rangeland consultant, Mohamed Jama Hussein, to investigate what motivated the locals to follow those beneficial techniques. Ten of the positive deviant villages were compared against two average-density villages and eight low-density villages — the negative deviants. Private residents have been forcibly prevented from enclosing community land for their own use by the leaders of positive deviant communities. “Squatting” on public ground was nevertheless popular in the other settlements, notwithstanding this fact.
Furthermore, Hussein saw that farmers in prosperous communities had begun to branch out beyond traditional livestock rearing. Some villages began producing their own food, including cattle feed and food for their personal consumption. Albanna points out that many women had started keeping bees, which had an unexpected benefit. People were discouraged from cutting down rangeland shrubs and trees for fuel because of the bees’ presence.
Additionally, the positive deviance method can be used to reinforce existing initiatives, adds behavioral science and public policy specialist Kai Ruggeri of Columbia University in New York City.
According to Ruggeri, who posted a commentary in Perspectives in Psychological Science in November 2021 urging researchers to consider adopting the positive deviance strategy, “It is such an easy adaptation that could potentially have major impact.”
Researchers at Rutgers University in Newark, New Jersey, Tomas Folke and Roberto Ruggeri wrote about the popular nudges intervention and how it uses simple tools to help people change their behavior and improve their lives. Using reminders to encourage patients to show up for their doctor’s visits has been studied by researchers. For example, a 2011 study published in the Journal of Telemedicine and Telecare indicated that automated recordings or text messages increased attendance by 29% across 29 studies.
Researcher Ruggeri was part of another study that looked at nearly 64,000 low-income urban patients, who can skip up to 45 percent of their visits, and found that an automated phone call and subsequent text message failed to enhance attendance for these patients. BMC Health Services Research released these findings in April 2020.
As Ruggeri points out, nudges are generally aimed at the average person. For example, “If you look at the way that nudging is mainly implemented, it speaks to an upper-middle-class population.” Ruggeri, on the other hand, argues that researchers may apply a positive deviance approach to help the most vulnerable members of society.
Finding out which participants in the preceding scenarios really visited their doctors as a result of receiving an automated reminder is what we would be looking for here. There may be nudges and other behavioral interventions that can be used to target probable no-shows in similar scenarios if the paths of those people are mapped. As a result, according to Ruggeri, policymakers would be able to reach the group most in need of preventative care.
Thus the hope in Somalia, where Hussein’s fieldwork was recently completed. With their newfound knowledge of successful outliers, the team is now looking into ways to use that knowledge to the development of behavioral and policy interventions.
Sternin argues that these interventions have an additional benefit: they allow communities to use their own expertise. She says that applying the community’s suggestions can be “transformative.”
