Participant observation (whether overt or
covert) is not the most reliable
research method. Such studies are, by their very nature, impossible to
repeat and the data they
produce is, when all'ssaid and done, simply the opinion of one observer.
In addition, the reliability of overt participant observation can be further questioned in terms of the extent to which the presence of the observer actually changes the behaviour of those being studied.
However, while such studies may lack reliability it is evident that the validity of the data gained can be impressive.