Page 485 -

15.3 Care and handling of research participants 477

participate in research. Furthermore, Cornell University’s IRB determined that the
study was not under their jurisdiction, as it was conducted by Facebook. As a result,
participants were not informed of their participation, which the editor considered to
be “not fully consistent with informed consent” (Verma, 2014). Reports of user con-
cern quickly spread throughout the new media (Goel, 2014), as users complained that
they may have been manipulated without their knowledge or consent. Subsequent
soul-searching in the academic literature (Fiske and Hauser, 2014; Puschmann and
Bozdag, 2014; Ross, 2014) examined the implications for evolving research ethics
in the age of social media.
Many issues raised by the Facebook study are thorny questions that are not easily
resolved. Should Facebook have informed users and obtained consent? Would that
have biased results, as users might have been more sensitized to positive or negative
content in posts? Would a design involving consent with some amount of deception
(see 15.3.6.1) have been more appropriate?
The participation of Facebook as a corporate sponsor of research complicates
matters further. As acknowledged in the paper, and in the Editor’s expression of con-
cern, Facebook is a private company and therefore not subject to the requirements of
the Common Rule (Verma, 2014). This raises the interesting question of corporate
ethics and conflicts of interest—specifically, what are the obligations of corporations
that conduct human subjects research that are not subject to external regulation?
How, if at all, did the Facebook study differ from the widely used techniques of
showing different web site designs to different sets of users to determine which is
preferred (so-called “A/B testing”? (Merritt et al., 2010)).
Discussions of corporate research behavior are likely to continue and evolve for
the foreseeable future. Not long after the publication of the Facebook paper, online
dating site OkCupid published a blog post describing the many ways that they have
experimented with manipulations of content, in the hopes of understanding how par-
ticipants respond to postings describing potential dates (Rudder, 2014). In contrast with
Facebook’s effort, these experiments did not involve academic researchers as partners,
and were not published in a scientific journal. Do these differences change our percep-
tions of the ethical implications of the work, or our interpretations of the results?
These questions do not have simple answers, but they do illustrate concerns that
most researchers would be well advised to consider carefully. Corporations such as
Facebook and OkCupid may be able to weather the publicity associated with these
potentially controversial research studies, but many academic researchers—specifi-
cally, those working with public funds in public in university settings—might want
to think twice before conducting studies that might lead participants to feel as if they
had not been treated appropriately.
Perhaps an application of the golden rule to study design might be appropriate.
Before conducting a study, you might ask how you might feel if asked to participate,
or, as in the Facebook study, you later found that your actions might have been part
of the study without your knowledge. If you decide that you might not be comfort-
able, others might have the same reaction, and you might consider revising your
study design.