Demand characteristics and their prevalence in psychological research has long been a contested issue. What with the increase in ethical guidelines that means the use of deception is limited it becomes increasingly harder to ensure that the results gathered in experiments are ‘true’. One of the aspects blamed for increased demand characteristics is the sampling methods used in particular volunteer sampling.
The argument goes that volunteer sampling results in an increased amount of demand characteristics as the participant knows they are in an experiment and so will try to judge what is wanted of them from the researcher and either act accordingly or do the opposite of what they think is wanted, known to some as the ‘screw you’ effect. However is this actually what happens? Just because we make participants aware of a study going on doesn’t mean they will try to change their behaviour. This is where I’m going to have to apologise in advance as I’m about to use a study that is so over used in psychology however it does illustrate my point so again sorry. This is apparent in Zimbardos prison study. In this study though participants did know they were taking part in an experiment they didn’t know what the nature of it was. They were then split into prison guards and prisoners and to simplify it acted their roles in the way they thought it should be (though I don’t think this is what they were told to do it is what happened). The behaviour exhibited by the participants was shocking even to Zimbardo himself with participants allocated the role of prisoner going on hunger strikes and prison guards using harsh behaviour against the prisoners. Due to the nature of the behaviour he actually cut the experiment short, he himself now acknowledging that it created a kind of spell even on him that made him not realise how extreme the behaviour had become. Was the behaviour anything to do with demand characteristics though? Zimbardo himself couldn’t predict the reaction in his research. So is it plausible to assume that novice participants would be act in the way they did? However I do concede that this is an extreme case and that sometimes demand characteristics, though slight, do occur. Another question though is: are they avoidable? It’s not just sampling that can effect results but also the experiment design. For example if the experiment uses repeat measures design then the participant is given the chance to repeat the task (say it’s remembering a set of words) this gives them the chance to remember the words better and effect the results in the same way as demand characteristics would.
It seems to me that demand characteristics and slight interference with results is unavoidable in psychology and that no manipulation of sampling method and experimental design will ever be able to get rid of them completely and that the best way to deal with this, without the use of deception, is to use the high measures of control that are being used I psychology.