 Working in a communication context often involves conducting some form of research. This is obvious in the case of researchers working in universities. But communication research is also conducted, for example, in market and opinion research firms, by marketing departments, strategy and corporate communication departments in organizations, and also by data journalists. Before starting a research project, it is important to consider ethical issues, especially concerning unintended and undesired side effects. An example of an ethical principle that may apply is that of minimizing harm. This is obvious in medical research, where administering a new drug could lead to health damage. But, on the other hand, it is important that such research takes place. Otherwise, we would either not develop any new medicines, or we would give them untested to large parts of the population. Obviously, an undesirable cause of action. Therefore, the question arises as to how far research can go. This question is not a new one. Think of the Stanford Prison Experiment, an experiment from 1971 that some of you may be familiar with from the film The Experiment. In order to understand the psychology of imprisonment, researchers simulated the prison environment. The setup was simple. Put participants in a prison, let half of them be guards and half of them prisoners, and watch what happens. The researchers had to stop the experiment because the guards started to torture the prisoners. Let us now consider several possible causes of reasoning about the ethics of this experiment. Someone might argue that the suffering of a couple of participants for a limited number of days is justifiable as it is outweighed by a larger societal goal, the creation of knowledge. This would be a clearly utilitarian reasoning. But even a utilitarian might arrive at the conclusion that the experiment is unethical. For example, because the insight gained is so minimal that it is not outweighed by the suffering. A deontologist such as Kant, however, would treat such arguments as irrelevant, as in his framework, a human being and as such their suffering can never be only a means to an end. And a virtue ethics perspective might focus instead on the character of the experimenter, which does not play a role in either the utilitarian or the deontological reasoning. For instance, a virtuous character might prescribe that the researcher should avoid any unpleasant situation for participants. Because research ethics is a very broad field, we want to focus on some specific dilemmas that are specifically relevant in the communication context. In experiments, a researcher might use some stimulus materials that have an undesired long-term effect. Imagine a researcher that studies the role of emotions in politics. She uses non-mutual stimuli to induce emotions such as anger or fear to find out how this influences someone's attitudes towards immigration. How do you deal with misleading the participants? Or, to discover the effects of violent pornography, researchers might need to deliberately expose people to such content. But how do you deal with any unwanted effects of such exposure? In so-called field experiments, this becomes even more problematic. In 2010, researchers exposed 61 million Facebook users to different versions of advertisements to encourage them to vote during the US elections. This gave some unique insights in what persuades people to go to the voting booth. But even beyond the question of whether the users really consented to being participants in such experiments, it can be potentially very problematic if such studies do in fact influence voting outcomes. You might think that ethics mainly play a role in experiments because experiments by design somehow manipulate people. But also in other types of research, such as survey research, unintended and ethically problematic side effects can arise. In survey research, research ethics involve the formulation and selection of questions. You need to ensure that those questions are asked that provide the best measurement, not those that lead to the most favorable outcome. A special case of surveys is the case of polling. That is to say, the measurement of voting intentions and the publication of estimates of how many seats a party would get if we were to hold elections now. We are interested in estimating the voting intentions of the population, but by publishing the results, we might in fact change the voting intention of the public. People might decide to strategically change their voting intention based on what the expected effect of their vote is, given the poor results. While some argue that reports on how politicians or parties perform in polls help citizens in their voting intentions, others argue that they distort the voting process. Therefore, in some countries, publishing polling results close to election day is even forbidden. However one stands on this issue, it illustrates that formulation of the questions, timing and publishing decisions can have major consequences, which, as in the last videos, again opens up a lot of room for discussion about utilitarian versus deontological ethics. Some additional issues arise in research and communication contexts involving the use of so-called big data. We will discuss these in a later clip.