Practical disadvantages - very time-consuming, requires training, personal characteristics may restrict what groups can be studied, groups may not wish to be studied
Ethical problems - covert observation involves deception and participation in illegal/immoral activities
Representativeness - small, haphazardly selected groups, not a sound basis for generalization
Reliability - difficult to replicate due to dependence on personal skills of the researcher, qualitative data makes comparisons difficult
Bias and lack of objectivity - risk of 'going native' and presenting a one-sided view, sympathies with the underdog may lead to biased accounts
Validity - positivists argue findings are merely subjective impressions of the observer, Hawthorne Effect undermines naturalistic account
Lack of a concept of structure - focuses on micro-level interactions and meanings, ignores wider structural forces