Guest post: PNAS, facebook and the ethics of online experimentation

This is a guest blog post by Gerry Markopoulos. I'm posting it because I think it is an important topic that deserves wider discussion.

Recently, an article was published in the prestigious journal ‘Proceedings of the National Academy of Sciences’ (PNAS), titled ‘Experimental evidence of massive-scale emotional contagion though social networks’. The article was published online on the 2^nd of June, 2014, and it is available here.

I would like to argue that the article needs to be retracted on the basis of violating fundamental ethical principles, it should not have been considered for publication in the first place (on the basis of the journal’s stated principles), and that it could damage the reputation of psychology on an international level.  The scientific community’s disapproval needs to be made explicit in order to safeguard the public’s trust in its work and procedures especially when involving human participants.

I am very happy to report that the BPS has responded to the publication in a very timely and unambiguous fashion via a letter to The Guardian. The letter makes clear what ethical principles were violated, and how. It would have been perhaps more practical to list the principles that were not violated. (It might have been a far shorter list!) Understandably, I presume the BPS cannot go any further than merely condemn the article, considering the authors are based in the USA, and the PNAS is a US journal. I will not pretend to know anything about the legal aspect of the situation, but legality is largely irrelevant. When psychologists and other scientistis propose projects to ethics committees, they are not looking for legal loopholes. They are looking to protect their participants from any harm whether or not there is a legal provision for it. That is partly the role of ethics committees, to anticipate and try to predict how research could violate wide ethical principles such as ‘maximising benefits and minimising harm’, especially where innovative research is concerned.

The only mention of ethical issues in the PNAS article is the following:

“…it was consistent with Facebook’s Data Use Policy, to which all users agree prior to creating an account on Facebook, constituting informed consent for this research.”

One can easily see that this constitutes consent, but certainly not informed consent. It is my understanding that participants were not aware they were taking part in a psychological study, they had not been informed of the nature of the study, they had not been informed of their right to withdraw their data, and they were not debriefed. Furthermore, no steps were taken to ensure the continuing wellbeing of the participants considering the sensitive nature of the experimental manipulation – according to the article, the manipulation led to a successful induction of negative emotions. There were no exclusion criteria protecting vulnerable populations (such as depressive or emotionally unstable participants). Such issues would be raised by any informed ethics committee. Why weren’t these issues raised? One can only assume that no ethics committee approved the project. This is the only logical explanation available. I personally contacted the first author on the 29^th of June requesting clarifications, but – perhaps not surprisingly - I never received a response. At this point, I should concede that fully informed consent could compromise the outcome of the study, but such a (perhaps) necessary omission ought to be counteracted with an extensive and carefully-worded debrief minimising the risk of potential harm to participants. Having said that, in the quote above, the authors claim that accepting the terms and conditions constitutes informed consent, which it certainly does not.

The issue of the ethics committee approval (or lack thereof) leads me to what we can do as individuals to protect psychology and the reputation of the scientific community. According to the PNAS website when research with human participants is involved:

“Authors must include in the Methods section a brief statement identifying the institutional and/or licensing committee approving the experiments. For experiments involving human participants, authors must also include a statement confirming that informed consent was obtained from all participants.”

In this case, PNAS appears to have ignored its own rules. On this basis, I contacted PNAS through their contact page firmly requesting the retraction of the article.

A few days after my email, I received a response from PNAS directing me to an editorial piece on this issue. The editorial confirmed earlier suspicions that no committee had scrutinized the research proposal. There had never been a research proposal to begin with. When the article was submitted for publication, the authors stated:

“Because this experiment was conducted by Facebook, Inc. for internal purposes, the Cornell University IRB [Institutional Review Board] determined that the project did not fall under Cornell’s Human Research Protection Program”.

To summarise the editorial response, it says:

“We were aware that no ethics committee had approved the project or the data collection method, we were aware that participants were not given the opportunity to opt out, but the company that collected the data is not obligated to adhere to such rules. Therefore, we published the data. However, we are concerned”.

At this stage, I would like to reiterate this is not an issue of legality. It is an issue of ethics where loopholes have no place. Now more than ever it is obvious that the article needs to be retracted.

Unfortunately, we cannot undo the harm that potentially has been caused by this research. Considering the sample size (over 600,000) and the reported significant effect of the experimental manipulation, it is possible that vulnerable participants were harmed. What we can do is demonstrate to the public that this type of research is not representative of what we do, and that we are as indignant as they are. Can we stop private companies from conducting research in secret? I would think this is unlikely. Secret research cannot be overseen by definition. However, scientists and scientific journals should actively stay away from data collected under questionable circumstances. Publication means condoning the research process from the design stage to the write-up stage. The condoning of unethical data collection methods (through publication) only encourages such practices. This is where a difference can be made, and that is why retraction of the specific article is essential.