Table 2.

Risk of bias criteria

	Risk of bias criteria
1	Participant eligibility criteria clearly stated and adequately described, and appropriate to the aims of the research.
2	Power calculation reported, and the study was adequately powered to detect mediation or moderation.
3	Valid measure of physical activity used. Objective measures of PA, self-report measures where the validity and reliability of the measures is cited, or experimental studies whereby participants are allocated to a PA intervention relative to a control condition are all considered a valid assessment of PA.
4	Valid measure used for the mental health outcome variable, or psychometric characteristics of the outcome variable reported and within accepted ranges (e.g., Cronbach’s alpha and test–retest reliability > .60)?
5	Valid measure used to assess the mediator or moderator variable, or psychometric characteristics reported and within accepted ranges (e.g., Cronbach’s alpha and test–retest reliability > .60)?
6	Statistically appropriate/acceptable methods of data analysis used for moderation or mediation.
7	Confounding factors/covariates adjusted for in analyses (e.g., sex, age, weight status).