Reactivity in the Research Process

Uljana Feest , Leibniz Universität Hannover, Germany:

Experimental Artifacts and the Reactivity of Psychological Phenomena

Reactivity in psychological research has long been treated as potential source of artifacts: human beings react to experimental design, to experimenter, etc. in certain kinds of ways. There has been a long-standing worry within psychology that this kind of reactivity might potentially lead to distorted data and biased results. This suggests that worries about reactivity rely on a particular assumption about human psychology, namely that there is a certain way in which properties or phenomena are, independently of the investigation, and that it’s the aim of empirical research to get at those “pure” states in an undistorted way.

However, on the flipside it is also a fundamental fact about human beings that we react to social and physical stimuli, a fact that is itself worthy of empirical investigation.  For example, we know from Gestalt Psychology that the way in which a given stimulus is perceived is highly dependent on the context in which it is presented. Likewise, research on implicit bias has found that test performance is influenced by specifics of the experimental set-up (e.g., the race or gender of the experimenter). In turn, both of these facts may well reflect mechanisms that are operative outside the lab.

I will argue that if we aim for ecologically valid and socially relevant psychological research, we need to take reactivity phenomena seriously and incorporate them into our systematic explorations of the subject matter. My account will be embedded in a systematic discussion of the notion of the relationship between facts and artifacts in psychological research.

David Teira, UNED, Spain (with Niccolò Tempini, University of Exeter):

When patients revolt against medical experiments

Pharmaceutical authorities rely on a particular type of experiment (the randomized clinical trial) to test the safety and efficacy of new medical treatments. These experiments should be free from any bias introduced by the participants. For instance, patients often enrol in the trial seeking access to drugs not yet available on the market. Making sure that all these preferences do not shape the outcome of the test is crucial for its credibility. For that reason, treatments are often masked –“blinded” – so that neither physicians nor patients in the trial know who is receiving what treatments. For fifty years, blinding has worked reasonably well, even if patients disliked it.

But this has changed thanks to the internet, boosting patients' reactivity: if they dislike the experimental protocl, they can now coordinate to revolt against it. Trial participants were so far strangers often scattered all over the world, without any connection with each other. But now, thanks to digital platforms, trial participants meet online and compare treatments and outcomes, potentially unblinding the test. Can we stil rely on randomized clinical trials if patients revolt against the protocols? Can we design alternative tests, where their preferences are incorporated? What's the future of drug testing in the age of social media?