An institutional review board (IRB) is an organization that reviews and approves (or disapproves) any research study involving human subjects. A human subject is any individual about whom the research team collects data through an intervention, or interaction with the individual. Institutional review boards (or IRBs) can recommend changes in all aspects of an impact evaluation, including protocols, content, and data security. The research team must seek IRB approval to ensure that they are protecting the rights and welfare of human subjects.
- Research ethics.
- Protecting human research subjects.
- DIME Analytics has compiled the DIME Research Ethics Standards to outline guidelines for ethical research.
- The research team must obtain IRB approval before starting an impact evaluation, as well as for any changes that they make during the study.
Institutional review boards (IRBs) review and monitor research studies to protect the rights of human subjects. The IRB can also request the research team for information on additional measures to protect the interests of vulnerable populations, such as pregnant women, prisoners, children, or economically or educationally disadvantaged people. In general, the research team must seek an IRB approval for all studies that interact with human subjects, and collect (or use) personally identifiable information about individuals.
The research team must prepare the applications for IRB approval in advance, since it can be a time-consuming process. Further, the research team must set aside enough time for obtaining all necessary approvals in the timeline for the entire study. The entire process of obtaining approvals also involves applying for local approvals with government agencies, and budgetary approvals from the organization funding the study.
Institutional review boards (IRBs) can take between 1 day to 4 months to approve a study. When submitting a project for review, the research team should allow enough time to ensure that the study is not interrupted. Further, the research team may need to obtain new approvals from the IRB at various stages of a study, so it is important to allocate sufficient time for this in the study timeline. The three broad stages of approval are as follows:
- Initial protocol review
- Amendment review
- Continuing review
Initial protocol review
The first review occurs at the beginning of the study, before starting with data collection. Ideally, the research team should seek IRB approval after conducting the pre-pilot, and finalizing survey protocols and instruments. Since the pre-pilot is conducted before obtaining IRB approval, it should not collect or store personally identifiable information (PII). The pre-pilot should only be used to assess the feasibility of the survey protocols, and to make sure all members of the research team have clearly understood the purpose of the study.
The research team must submit an amendment application to an IRB whenever it makes changes to the following aspects of a study:
- General study design. This covers changes to location of study, method of sampling, and changes to the primary hypothesis of the study (or theory of change).
- Method of data collection. This covers changes in the method of data collection. For example, the research team might decide to transition from a field survey to a remote survey.
- Survey instrument. This covers changes to the content, translations, and programmed version of the questionnaire.
- Training. This covers changes in the methods, and medium of conducting enumerator training. For example, the research team decide to conduct the training virtually, instead of in-person.
- Protocols. This covers changes to survey protocols. For example, if the research team decides to transition to a phone survey, then they will need to update the protocols to include audio audits.
- Informed consent procedures. This covers changes to the manner in which the research team obtains informed consent from participants of a study.
Finally, since risks to the human subjects are constantly changing over time, the research team must submit the study for a yearly review. As part of this, the research team must reports on the progress of the approved research every year, even if no aspect of the study has changed. Every IRB clearly lays out the format of the application for the yearly review, which the research team must follow.
Applications for approval from an IRB can be quite demanding, especially in cases where research is taking place outside the borders of the United States. IRBs require the research team to submit details about various aspects, including study design, instruments, training guidelines, data collection protocols, and documents for Preparing for Field Data Collection#Procurement. In addition to these documents, the research team must also submit the following information for approval of a study involving human subjects:
- Informed consent
- Protecting human subjects and certification
- Vulnerable populations and expected enrollment
Before involving a human subject in a research study, the research team must obtain informed consent from each individual. This means that the research team must clearly mention all possible risks and benefits from participating in a study, either for the survey pilot or as a respondent in the actual study. The research team must clearly lay out the procedure for obtaining informed consent, and clearly state that human subjects can take back this consent at any point during the study. DIME Analytics has created templates for obtaining oral, as well as written consent.
Protecting human subjects and certification
The application should contain a separate section that deals with protecting human subjects. This section should specifically mention that human subjects are a part of the study. It should also include an assessment of possible risks and benefits to human subjects, steps taken to protect the human subjects from these risks, and the key insights that the research team hopes to gain from conducting the study.
Finally, every member of the research team who handles personally identifiable information must obtain a human research subjects certification before the research team submits an application for approval. Two organizations that provide the human research subjects certification are:
- Protecting Human Research Participants (PHRP)
- Collaborative Institutional Training Initiative (CITI)
Vulnerable populations and expected enrollment
The proposal should also include a plan detailing how it will protect the interests of vulnerable members of the population. These may include women, children, or minority communities in the area of the study.
The proposal should also include a table which lists indicating expected enrollment figures, especially for vulnerable categories, like women, children and minorities. Enrollment rate (or take-up rate) is the number of individuals who actually participated in a study, out of the total number of individuals who were eligible to participate in a study. The National Institute of Mental Health (NIMH) has created a template that research teams can use to create such a table.
Often, experimental and quasi-experimental methods exclude certain sections of the population from the study. In such cases, the proposal must provide a justification for excluding a particular section of the population, especially if participating in the study can provide certain financial, or health benefits. For example, a study designed to provide cash transfers to farmers might propose to exclude farmers who cultivate more than 10 hectares of land. In this case, the research team must provide strong reasoning for excluding this group.
Institutional review boards (IRBs) follow the basic principles for IRB approval that are outlined in the Federal Policy for Protection of Human Subjects, or the 'Common Rule'. Research teams often submit applications to institutional review boards (IRBs) that are based in partner academic institutions, or research organizations. Some examples of these IRBs are given below:
- MIT has its own IRB, called Committee on the Use of Humans as Experimental Subjects (COUHES).
- Innovations for Poverty Action (IPA) has its own IRB, the IPA IRB.
- There are also several certified independent IRBs, like Western IRB.
Since the World Bank does not have its own IRB, World Bank projects typically seek IRB approval from one of these above IRBs. In many cases, the research team may also need multiple IRB approvals, especially when different members of the research team are associated with different institutions. Sometimes, additional approvals are needed based on the requirements in the country in which the study is being conducted.
- Additional IRB approval. In some cases, a particular country will require the research team to obtain approval from local IRBs as well.
- Local approvals. The research team may also need to obtain other permissions from relevant government agencies and ministries in the area of the study. These can include permissions to access administrative, or geospatial data. These approvals can allow the research team to work smoothly in the location of the study.
- Budgetary approval. The research team must also seek approval for the study budget from the organization funding the study. This includes obtaining approval for various expenses, including administrative costs like salaries, costs for purchasing physical equipment like tablets, as well as country-specific costs like taxes.
There are very few circumstances under which a research team can be exempted from needing an IRB approval for a study that interacts with human subjects. In practice, exemptions can only be obtained for continuing reviews, since the research team must submit an initial application to an institutional review board (IRB) for future exemption.
- Berk Özler (World Bank), Taking ethics seriously: Response #1
- Berk Özler (World Bank), Research with adolescents: Issues surrounding consent
- DIME Analytics (World Bank), Research Ethics & Data Security
- J-PAL, Ethics
- Martin Ravallion (World Bank), Taking ethical validity seriously
- National Institute of Mental Health, Human Subjects in Research: Things to Consider