I'm conducting a study about the effects of a verbally administered placebo.
The actual setting and intervention are different from the example I give here, but similar enough that the example may serve as a basis for discussion.
In the study, a team of psychologists running a local practice offers a new experimental service. The psychologists have developed a diagnostic tool that allows them to predict with surprising accuracy if a single person will fall in love within the upcoming four weeks. To further test and refine this diagnostic tool they now offer a public service after having tested it on a small sample of students. So they invite local residents to their practice to be tested through an advertisement and offer them a free diagnosis, whether or not they will fall in love within the next two weeks. The participants are told that the diagnostic tool for this prognosis is being evaluated in its predictive power and that they should report back after four weeks and tell the researchers if they had fallen in love or not.
Of course there is no diagnostic tool. The psychologists believe that they are being perceived as experts by the laypeople, and that if they make such a prediction after some impressive psychological examination, this prediction will cause people to fall in love, if without this intervention they would not have, or remain "out of love", if they had fallen in love.
Obviously, the researchers do not want to give negative (not fall in love) prognoses, because they do not want anyone to refuse and miss out on an "offer" they would have been happy about. (They do not believe that they will cause people to fall in love with unsuitable suitors and cause grief, only that participants with a will-fall-in-love prognosis will react more benevolently and accepting to offers they receive.) And, in the absence of negative prognoses, the researchers will need a control group that receives no prediction, but still reports back after two weeks.
The psychologists do not know how to set up this experiment so that they can give positive (fall in love) diagnoses or no diagnosis to random participants, without raising suspicion. Some of the patients will obviously know each other and talk about this experimental "service", and they will notice that they know no one with a negative diagnosis. That would seem odd to them. Also the reason to not give a diagnosis to many participants must be valid without raising doubt regarding the validity of the prognoses that were given.
What would you tell your participants to allay their doubts over the absence of negative prognoses? And how would you explain that often you do not give a prognosis, without invalidating the prognoses you give?
Note
Since the ethic concerns are being pushed to the forefront of the discussion, I have edited my question to include an example that more closely resembles the real research. Yet, I want to repeat that this is an example only, so please don't get hung up on the specifics of falling in love. The real study includes psychologists and a behavior prediction, but not about falling in love. Similar research has been done in a laboratory setting, the difference is I want to do it out "in the field".
If you are worried about psychologists lying to their participants, you need to understand that cover stories are a common method in psychological research. They are in fact so very common, that it is easier to list instances where cover stories are not used: only if the knowledge of the purpose of the study cannot falsify the results are participants told in advance what is being studied and how. In all other cases the researchers lie to the participants and debrief them after the experient.
Of course, if there is any danger that this lie would have lasting or strong negative effects, such research will not be conducted.
You can safely assume that
- participants will be informed that they are taking part in an experiment and asked to give their consent (this is easily part of the cover story, as it allows to not give a prognosis to some of the participants)
- participants will be debriefed by a trained psychologist
- participants will be given contact information and told to contact a team member when they observe any kind of effect at all, positive or negative (since we want to measure these effects, this is easily part of the cover story)
- an ethics commitee of our university will evaluate the study design
Since similar "lies" have been told to participants of other experiments before, I am quite confident that this study will be approved by our ethics committee also. Here is an example of a study where the researchers lied to the participants about their sleep quality: http://www.ncbi.nlm.nih.gov/pubmed/24417326
There are also many examples of research, where researchers actually induce negative (i.e. unwanted by the participants) effects in their participants. A recent example is experimentally lowering self-esteem: http://www.sciencedirect.com/science/article/pii/S0165178113007774.
