Article
Is your data inclusive?
Optimizing results by eliminating the hidden costs of research participation
By Amy Shipow & Anisha Singh. Research assistance: Jennifer Adhiambo
There are many references that clearly show the difference between equality, equity and justice. They exist because many times one can be mistaken for the other — and well intentioned interventions may address a particular problem but frustrate the larger goal. Take, for example, how equality insists that we all start at the same place. Thinking biological sex doesn’t matter had medical research defaulting the “normative” human body to that of a male — to the detriment of women around the world. In order to properly address a problem we need to aim for justice by recognizing and addressing how underlying, sometimes invisible factors — across gender, race, sexual orientation, nationality (in addition to one’s intersecting identities) — contribute to different starting points and thus lead to different outcomes.
When field and lab based research fails to consider the structural factors that influence participation in a study, vulnerable populations (e.g. children, sex workers, people living with HIV, displaced people, people who are unhoused or incarcerated, members of the LGBTQI community and people who are differently abled) are frequently excluded from research as they are difficult to reach or face substantial risks to being identified. When these people are excluded, researchers perpetuate a vicious cycle whereby findings preclude — or may not apply to — the people with both the greatest needs and the most apt insights.
Childcare is a significant pain point for women — not only in finding an available caretaker on short notice, but also paying them and building in time to drop children off before participation.
What is our responsibility as behavioral economists, social psychologists, and public health and development practitioners to account for and improve gender responsiveness in our research? Similar to Western, Educated, Industrialized, Rich, and Democratic (WEIRD) labs, Busara provides participants with a show up fee to attend our lab sessions as well as an incentive for participation. But is this enough? How can we better mitigate the issue? And how deep does it run?
To better understand the difference in cost for lab attendance and our ethical responsibility to address such differences, we undertook a qualitative exercise to interview 15 current participants (10 women, 5 men) about their experience with Busara’s lab in Nairobi, Kenya. Participants were asked about the arrangements they made to come to the current session including considerations for transport, childcare, and outside obligations. Furthermore, they were asked about when they were notified to arrive, their time preference for when they want to show up, and how well we took this into account. Lastly, both men and women were asked to brainstorm factors that might affect female participants’ attendance as well as additional costs incurred.
What did we find out?
Childcare is a significant pain point for women — not only in finding an available caretaker on short notice, but also paying them and building in time to drop children off before participation. All of the men interviewed and four of the ten women explicitly stated that having a child would impact women’s ability to participate in Busara studies since they would have difficulty obtaining childcare. “Had I not found somewhere to leave the child, I would obviously not have come because I definitely wouldn’t want my child to suffer.” Another challenge that women faced with childcare was the variable length of time of lab experiments. This uncertainty proved to be a heavy burden. Women may have anticipated returning home at a certain time, but often underestimated the time requirement: “I left the youngest with my neighbour. I knew I would take just an hour and not 4 hours like I did today. I had to make a call to my neighbour to tell her that I will take longer than the hour I had promised.”
Moreover, nearly all of the women woke up early to complete household chores, including washing the dishes, fetching water, washing children’s clothes, cleaning the house, and preparing children for school: “I woke up early in the morning, did house chores, then I decided to come here. I woke up earlier than usual because I have a two year-old baby.” Men also recognized that women are responsible for more domestic responsibilities and remarked that these chores may make women late. Our interviews illustrate how women spend an average of three to six hours per day on unpaid labor. On the other hand, men spend 30 minutes to two hours on similar tasks.
Despite the impediments to fulfilling household responsibilities presented by the time of transport, most women were willing to bear the additional burden because they considered their research participation a job. One woman commented, “I am living far and I didn’t know that I would take a lot of time here and by the time I will be going back, the amount I pay for fare will be more. On a normal day, if I didn’t have work to do, I would still come because this is still work.”
Both women and men also commented that most women need to garner their husband’s permission in order to participate, detailing: “Some husbands are just stubborn and would not allow their wives to attend sessions.” Others described that later sessions are additionally challenging to navigate with their partners, as they felt they needed to explain where they were: “If one has a strict husband who questions every move, a wife is able to leave the house at 10:00 am because most husbands have left the house by that time and their wives back from the sessions before the husbands are back.” Hence, monetary costs are supplemented by the psychological cost of worrying about partner approval.
What can we do about it?
A next step for us would be to explore these findings further and attempt to quantify some of the additional costs women face (both monetary and non-monetary) and use these to inform service design improvements that might reduce these costs. In addition, biases in the built environment, which were elucidated with our qualitative research, commonly span across workplaces and public spaces.
Our recommendations thus spotlight the importance of social infrastructure: public services that propel the functioning of society, similar to the ways in which physical structures do so, as outlined here on how data is biased. By enhancing child-care considerations, emphasizing punctuality of concluding sessions, and providing more gender sensitive restrooms, Busara can help remove barriers to women’s full participation. Similarly, other organizations should reflect on other, and sometimes less visible, impediments to full female participation — whether that be at a policy, organizational, or normative level. In short, in order to make data inclusive, the structural factors that influence participation need to be addressed; the overarching goal is to optimize results from research studies to help inform and create a just system that ensures invisible factors are addressed.
Beyond our own work, we hope to further engage this issue across labs and research organizations in the Global South. We are calling on all organizations to share your ideas about how you have made your lab (or even field) research more participation-friendly for women. If you have experience with this issue and are passionate about collaborative engagement, fill out this link so that we can crowdsource ideas as a global community.
