Kim 2019.

Study characteristics
Methods	Study design: quasi‐randomised controlled trial Unit of allocation: school Unit of analysis: participant Adjustment for clustering: unclear
Participants	Location: USA, specific location not reported Setting: not reported Sample size calculation: not reported Sample size: 161 elementary school teachers (Kim 2019, p 730); intervention group n and control group n not reported Mean age (SD): not reported by group; age range = 18 to 55+ years; largest age category = 35 to 44 years (35.8%) (Kim 2019, p 730) Gender: not reported by group; total sample = 91.9% women Race/ethnicity: not reported by group; total sample = 97.5% Caucasian (understood to be white) Years of experience: M = 15.4 years (SD = 7.4 years); range = 1 to 30+ years Previous child protection training: not reported Previous experience with child maltreatment reporting: (i) intervention group = 56.6% yes, (ii) control group = 65.8% yes (Kim 2019, p 730) Baseline equivalence: not reported
Interventions	Name: Second Step Child Protection Unit (Committee for Children 2021) Contents: (i) recognise indicators of child sexual abuse, (ii) respond in a supportive way, (iii) report abuse, (iv) "Recognize, Respond, and Report Abuse", (v) addressing discomfort with the topic, and (vi) how to teach student lessons and engage families. Part of a "comprehensive approach" involving "(a) school policies and procedures, (b) staff training, (c) student lessons, and (d) family education" (Kim 2019, p 728) Processes and teaching methods: methods used in online modules were not reported Delivery mode: online Trainers and qualifications: not reported Duration: 75 to 90 minutes Intensity: self‐paced Intervention integrity: not reported; online delivery offers the possibility of uniformity Comparison condition: waitlist control
Outcomes	Eligible measures (outcome domain) Knowledge: Awareness of Signs and Symptoms of Child Abuse and Knowledge subscales from the Educators and Child Abuse Questionnaire (Kenny 2001; Kenny 2004) (secondary outcome: knowledge of core concepts in child abuse and neglect), comprising 8 items with response options on a 5‐point Likert‐type scale Attitude towards reporting: Teacher Reporting Attitude Scale‐Child Sexual Abuse (Walsh 2012b) (secondary outcome: attitudes towards the duty to report child abuse and neglect), comprising 14 items with response options on a 5‐point Likert‐type scale Ineligible measures (reason) Teacher‐student relations: The Delaware School Climate Survey (Teacher/Staff) (Bear 2014) (items "assess teachers' perceptions of the extent to which teachers and other school staff members are responsive, caring and provide support" (Kim 2019, p 732); not prespecified in the protocol for this review), comprising 4 items with response options on a 4‐point Likert‐type scale Acceptability: Abbreviated Acceptability Rating Profile (Tarnowski 1992) (items assess intervention acceptability; not prespecified in the protocol for this review), comprising 8 items with response options on a 6‐point Likert‐type scale Timing of outcome assessment: pre‐test (before online training), post‐test (after online training) (Kim 2019, p 730)
Notes	Funding: Committee for Children, Seattle (WA) Author contact: yes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: adequate description of the generation of the randomised sequence Quote: "randomisation was conducted at the school (Imai, King, & Nall, 2009). For the randomisation, schools were first matched based on school characteristics such as grade levels (K‐5, pre‐K‐2, 3–5) and school size, and then randomly assigned to the intervention or wait‐list control using a computer‐generated random number list (Kim & Shin, 2014)" (Kim 2019, p 730)
Allocation concealment (selection bias)	Unclear risk	Comment: method of concealment not reported by study authors
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: performance bias due to lack of blinding, and therefore likely knowledge of the allocated intervention by participants and personnel during the study, which may have influenced subjective study outcomes (i.e. self‐report measures)
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: detection bias due to knowledge of the allocated intervention by outcome assessors, and outcome measurement is likely to be influenced by lack of blinding (pre‐post self‐report measures tied closely to intervention purpose)
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: loss of 5 participants between allocation and baseline assessments in the waitlist control group, but no losses in the experimental group. A total of 3 participants were "lost to follow‐up" (1/83 in experimental; 2/76 control condition) (p 731), and a further 3 participants in the experimental group "discontinued intervention" (p 731). Reasons for losses were not explicitly reported. Overall rate of attrition is low, and randomised sample size was used for the experimental group (n = 83), but not for the control group (n = 78) (Kim 2019, p 731).
Selective reporting (reporting bias)	High risk	Comment: study protocol not available. All outcomes in the methods section are reported in the study, but not all outcomes were reported in a sufficiently complete manner to permit effect size calculation or inclusion in meta‐analyses, or both.
Other bias	Unclear risk	Comment: additional potential sources of bias related to the specific study design have been identified
Reliability of outcome measures (measurement bias)	Low risk	Comment: pre‐existing scales were used to measure outcomes. Authors generated reliability data at pre‐ and post‐test for the Educators and Child Abuse Questionnaire (α = 0.62 and α = 0.70, respectively), the Teacher Reporting Attitude Scale‐Child Sexual Abuse (α = 0.82 and α = 0.84, respectively), and the Delaware School Climate Survey (α = 0.77 and α = 0.78, respectively) (p 732). Reliability data for the Abbreviated Acceptability Rating Profile were generated at post‐test (α = 0.91) (Kim 2019, p 733).
Group comparability (selection bias)	Unclear risk	Comment: group comparability at baseline was not reported
Contamination (contamination bias)	Low risk	Comment: measures taken to prevent or minimise the possibility that participants in the control group might receive part or all of the intervention were not described sufficiently to enable a precise assessment of contamination between groups. However, the experimental and control group participants were in different schools, thus reducing the likelihood of contamination.