Peer-reviewed article · 2022
Organizational Identity and Positionality in Randomized Control Trials: Considerations and Advice for Collaborative Research Teams
Nicholas Haas, Katherine Haenschen, Tanu Kumar, Costas Panagopoulos, Kyle Peyton, Nico Ravanilla, and Michael Sierra-Arévalo · PS: Political Science & Politics 55(4): 749-753
In brief
How an organizational partner's funding, political allegiances, mandate, and reputation shape the design and ethics of a randomized trial, with practical advice for research teams working with governments, NGOs, and firms.
Research annotationQuestion, variables, design, findings, and mechanism at a glance▾
- Research question
- How does an organizational partner's position in the social and political context it operates in shape the design, implementation, and ethics of a field experiment, and what should collaborative research teams do about it?
- Main findings
- The piece reports no estimates of its own. Its substantive claims are that a partner's funding sources, political allegiances, scope of operations and mandate, legal status, identity, and reputation shape its goals and incentives; that divergence between those incentives and the researchers' can constrain feasible questions, force design changes, and create ethical problems that Institutional Review Board review does not catch; and that a partner's power over the study population can undermine genuinely uncoerced consent. It notes that between 2000 and 2017 roughly 62 percent of field experiments published in the discipline's three main journals involved some kind of external partnership.
- Mechanism
- Practical and ethical failures arise when the partner's goals and incentives diverge from those of the researchers or, more seriously, from those of the study population. Publication incentives push researchers toward maximizing measured effects, while partners answer to longer-term missions and to funders, and study participants may have no voice at all.
- Why it matters
- Extends the positionality literature, which has centered on individual researcher identity in qualitative fieldwork, to the organization as a whole in randomized trials. Its recommendations are that experiments be designed to be worth running even if the estimate is indistinguishable from zero, that immersive fieldwork and pilots precede the trial, that researchers explain the logic of random assignment rather than assume partner resistance to it, that waitlisting can resolve control-group objections, that locally embedded partners be preferred, and that a pre-analysis plan or memorandum of understanding drafted jointly with the partner ties both parties' hands.
- Speaks to
- research ethics in field experimentspositionality and reflexivityresearcher-practitioner partnershipsexperimental methods in political science
Cite
@article{haas2022positionality,
title = {Organizational Identity and Positionality in Randomized Control Trials: Considerations and Advice for Collaborative Research Teams},
author = {Nicholas Haas and Katherine Haenschen and Tanu Kumar and Costas Panagopoulos and Kyle Peyton and Nico Ravanilla and Michael Sierra-Arévalo},
journal = {PS: Political Science & Politics},
year = {2022},
doi = {10.1017/S1049096522000026},
}