Can anyone advise me on which type of statistical analysis I should use? I want to look at the association between receiving a particular service (yes/no) and subsequent self-harm (yes/no) and whether the effect is mediated by social support (continuous), controlling for baseline SS. Thanks in advance!