Small sample size - M of Sz A03
7 studies in Glowacki et al (2016) and 3 in M&S is a small evidence base.
One issue with such a small number of studies is the file drawer problem - a bias towards publishing positive findings.
E: McMonagle & Sultana (2009) reviewed evidence for token economy & found 3 studies which used random allocation to assign patients to condition. Of these, 1 found improvement in symptoms & none found evidence of behaviour change