There are two types of observations: Participant (where researcher takes part in actions of the group) and non-participant (observer is detached from the group). There are also different ways these observations can be conducted: overtly (group has knowledge that they are being observed) and covertly (group lacks knowledge of observation)
Observations allow the researcher to witness behaviour first-hand, as well as it happening in a natural environment which increases the validity of the findings
Observations are a good method of reaching 'hard-to-access' groups like deviant subcultures
Observations have practical issues:
Gaining access into some groups- researcher will need to dress and act differently in some situations and potentially change their identity
Recording behaviour- the way they record the behaviour may put the researcher at risk
Can be costly- use of costly equipment and specialist techniques
Observations have ethical issues:
Deception in covert operations
Informed consent- have the people being observed consented
Protection from harm- researchers undercover may come into contact with dangerous groups
Illegal or immoral activities- witnessing these may cause an ethical dilemma for the researcher
Observations have theoretical issues:
Reliability- observations are difficult to replicate
Lack objectivity- researcher can become attached to a group
Validity- overt observations may suffer from the Hawthorne effect
Less representative- usually small samples
An example of an observation was Zimbardo'sStanford Prison Experiment. It was an overt, participant observation where Zimbardo was the Warden and he wanted to investigate the influence of social roles on human behaviour