IDinsight’s approach to ethical research practices

Before, during, and after we collect data from people – whether for research and evaluation, monitoring, or program design – we must articulate, then mitigate, the risks we impose. Many factors amplify the risk of our research activities: we work in under-resourced settings and collect information about sensitive topics, often from vulnerable populations. And in some cases, people in policy research trials may not know they are part of it (if they, for example, aren’t receiving a service being tested); in these instances, the policy is implemented by our government and NGO clients, but the research is facilitated by us. Like peer researchers in health, education, agriculture, and many other fields, our work is motivated by better outcomes for all. But the means to that well-intentioned end can cause unintended harm along the way. 

What are the risks imposed by research and what must we do to eliminate, or mitigate risks, to participants and communities?

Unethical research can harm study participants

The process of data collection is delicate. In some instances, the questions we ask in our survey are embarrassing or upsetting and can evoke strong emotional reactions, even emotional (re)trauma. We may hear about or even witness past, ongoing, or potential harm that our respondents experience. Privacy, compassion, and even on-hand support services may be required.  

The data we collect can be sensitive. This data may include sensitive financial, health, or behavioral information. Or it may include opinions about topics that can put the responder at risk if their answers fall into the wrong hands. Their data must be handled securely.

Research projects can be or feel unfair. Oftentimes, some individuals in our sample are eligible to receive a service but aren’t receiving it because of what amounts to a coin flip. Sometimes others are receiving a totally different service than what they expected because a pilot program hasn’t yet been proven effective. They should not be made knowingly worse-off for the sake of learning.

Research ethics are crucial for our mission

To help clients answer key questions to drive their decisions and actions toward social impact, IDinsight conducts applied research – gathering and analyzing data and evidence to this end. We aim to identify what works, which problems need solving, or to reveal the opinions and preferences of ordinary people whose voices are often unheard. We aim to make policy design and service delivery more effective, efficient, and responsive. Much of this work involves collecting primary data and conducting experimental research.

We must never take for granted people’s willingness to interact with us and entrust us with their information, experiences, and opinions.

It is our obligation to go beyond the legal and contractual obligations of data collection and storage to deeply understand – and in turn help potential research participants and local gatekeepers understand – the ethical trade-offs of engaging in research.

It is incumbent on us to make decisions that meet the highest ethical standards and enable potential participants to make informed decisions about whether they engage. 

Applied social science research often falls through the cracks 

There are a wide variety of regulations, organizations, and norms that govern research ethics around the world. And there are wide gaps in which types of research and which researchers are covered – and applied research run by non-academics tends to fall between them. These gaps can make it difficult for an entity like IDinsight to externally receive the ethical scrutiny we want for our work.

These gaps in standards are products of how the ethical review system evolved, with a biomedical focus and with power centered in universities and health ministries. Research, when conducted by faculty and research centers at universities, must be reviewed and approved by internal Institutional Review Boards (IRBs). These are typically housed within those universities and cater solely to their own research communities. Some of the countries in which we work have government-run IRBs and mandate IRB approval for any research activities; others do not. Regardless of this mandate, many have a limited view of what counts as “research.” Often our work, producing information for our government and non-profit clients to make decisions, falls outside of this definition, earning us an ‘exempt’ status when we genuinely want further scrutiny.

Absent tight regulations or clear guidance, organizations like IDinsight must decide to what degree it makes sense to self-impose ethical standards, policies, and practices. We have chosen to impose as high a standard as possible. In our view, an individual’s time, information, mental health, and how they are affected by the programs and services they are potentially exposed to, or denied, in the pursuit of evidence, is valuable regardless of what country they’re from (having an IRB or not), who the audience of the evidence is (private or public), whether we’re looking at public health or economic livelihoods interventions, or whether or not a university faculty member is conducting the research.

How are we establishing and enforcing research ethics?

IDinsight developed its research ethics policy and an internal ethics review process with an Internal Ethics Review Committee in 2019. In 2020 and 2021 we enlisted a panel of external research ethicists to review and update our policy. As advised by our panel, we have just hired a full-time team member to manage the Ethics Review Committee. We are now ready to formalize our ethics policy and to ensure we follow it in practice. The standards we impose on ourselves are not unique. They parallel those used by institutional review boards in academic and biomedical research. The main difference is that we are applying them to all of our own work, even when others might say it’s not necessary. 

What are some of the highlights of our ethics policy?

• Mandatory review. We require review of all evaluation and data collection activities by either an IRB or IDinsight’s internal Ethics Review Committee, regardless of whether the activity is biomedical, social or behavioral, or considered “research” (i.e., intended for publication), and even regardless of whether there are obvious study participants or “human subjects” (for example, if working with administrative data, or with information not necessarily about individuals).¹
• Informed consent and freedom to stop. When primary data is being collected from study participants, we require informed consent. We invite participants to join and we give them the opportunity to decline, stop, or withdraw participation at any point in time.²
• Addition protections when indicated. We require added protections when working with “vulnerable groups” such as children or when collecting “sensitive information” – information that can be upsetting during collection or harmful if released.
• Data security. We require de-identification and encryption of personally identifiable information “at rest” (on paper, on any device, computer, or server) and “in transit” (when connected to the internet or shared in other ways).
• Certified staff. We require all IDinsight research to receive certification in doing research with human subjects through the CITI Program.
• Process for iterative improvements. We will revisit and update our policy annually to ensure our policy reflects our experience and the ever-evolving standards and norms of research ethics.

The diagram below represents the guidance we give our project teams on what kind of review each project requires.

Striking the right balance

IDinsight’s strives to improve lives through our expertise, which is with data and evidence. We believe in the transformative power of data to inform better policy decisions but are clear-eyed on the potential risks of producing and handling that same data. With this view, we have launched our new research ethics and data security policies along with our internal Ethics Review Committee. We also established a policy to default and defer to local IRBs in the countries where we work whenever possible. We believe that combined, these approaches strike the right balance between care in our interactions with underserved populations and contributing to evidence-informed policymaking meant to improve the welfare of those same populations.

We welcome your comments below.