It is prone to publication bias, as the researcher may not select all the relevant studies by choosing to leave out some of the studies if they do not provide the data or results that they desire. Due to this the conclusions can be biassed as they only represent some of the data.