One of the key areas of social psychology is the study of prosocial behaviour. However, it is difficult to study how people would react in an emergency situation and meet ethical standards in psychology.  Using people’s personal stories gives us insight into how people behave, but the data is anecdotal and open to biases. Designing experiments leads to potential undue stress and often requires a certain level of deception.

In the study of bystanderism, Latané proposed the Social Impact Theory – that is, factors in one’s environment or the power of a situation determines whether an individual will help someone or be a bystander.  In his classic study, Latané asked participants to take part in a discussion of college life, focusing on worries and problems.  The participants were each given headphones and a microphone and were put alone in a cubicle supposedly to keep their answers anonymous.  The participants were told that they were speaking with one other person, two people or a group of a five people.

Deception was used in this experiment. A few minutes into the conversation, a person in the group was acting as though he was having an epileptic seizure and called out for help. The participants who thought they were alone sought help in 85% of the cases. When they thought they were in a group of three, it dropped to 64%.  And when they thought that they were in a larger group, the level dropped to 31%.  Deception was used in order set up a “realistic” situation and control for demand characteristics.  However, there is a concern that there was a high level of stress for the participants – both during and after the experiment.  Latané videoed the participants. In the larger groups, the participants displayed high levels of anxiety as they worried about the person having the seizure, but assuming that someone else would help.  After the study, they were told in the debriefing that the situation was not real.  However, this could lead to a sense of embarrassment as a result of their failure to act.  Although the study was highly controlled and was rather “unnatural” in its situation, it had a high level of ecological validity as the participants believed that the other student was having a seizure.  The question is whether it would be possible to study bystanderism while not causing this level of stress.

Drury decided to carry out a study of bystanderism by using virtual reality.  The goal was to see if social identity played a role in helping behaviour.  Participants were told that the study was to find out how they would behave in an emergency situation.  Unlike Latané’s study that used deception by commission, Drury’s study used deception by omission.  They were not told an incorrect aim, but they were not told that the aspect of their response that was being studied was whether they would help someone or not.  

Before the experiment, they were asked to read a text about a historic fire in a London metro. After being primed for the emergency, the participants were given one of two scenarios.  In the first scenario, they were told that they had just been to a English football match and in the metro on the way home from the match, someone screams fire! In the second scenario, the situation was they same, except that they had just been shopping.  They were then put into a virtual reality simulation in which the football fans were all wearing the same jersey as the participant; in the shopping scenario, there were no matching t-shirts.   In the “in-group” scenario, they found that the participants were less likely to push people out of their way and were more likely to help people in need of help. In the shopping scenario, in which the participant identifies as an individual instead of as part of a group, they were more likely to push and less likely to help.  

Drury’s study has much lower sense of “undue stress.”  Although a virtual reality experience can be very realistic – and even result in signs of stress or emotion – the participants gave informed consent in order to be part of the study.  Although there may be a stress response to being in the study, the participant is aware at all times that this is a simulation and the other “people” in the situation are only characters.  In this case, the internal validity is high in that the IV is manipulated and extraneous variables are controlled, but the study may lack ecological validity.  The lack of authentic emotion, knowing that none of these people are real, means that it may not predict what happens in a real emergency situation.  That being said, virtual reality does appear to be more “real” to people than other experimental situations.

Ethical considerations in prosocial research are meant to protect the rights of participants, but at the same time may actually lower the validity of a study. Both of the studies above manipulate an IV with the goal of determining a cause and effect relationship. It also allows the researchers to create an “emergency” situation, rather than having to wait for one to naturally occur.  Emotion is a key factor in a real emergency situation, so if the experiment lacks an emotional element, it is doubtful that this will predict what happens in a real emergency.  And yet, ethics limits the ability for researchers to create situations which would lead to this level of emotional response – making all experimental research on bystanderism suspect.   Deception is important to control for demand characteristics; if a participant knows that we are studying their willingness to help, they may engage in the expectancy effect or social desirability effect, and not act and they would naturally have reacted in such a situation. In both cases, debriefing is the best solution we can offer to address both the deception and any stress experienced – but this may lead to embarrassment.  Drury’s study is only “virtual”, so in this case, it may prevent this problem.