Jacobsen 1993.

Study characteristics
Methods	Study design: controlled before‐and‐after study Unit of allocation: district Unit of analysis: participant Adjustment for clustering: no (participants received the intervention in a group. Participants were from schools in 1 school district)
Participants	Location: rural western school district, USA Setting: school district (Jacobsen 1993, p 10) Sample size calculation: not reported Sample size: 40 kindergarten through 6th‐grade regular and special education teachers (Jacobsen 1993, p 23); intervention group n = 20, control group n = 20 (Jacobsen 1993, p 23) Mean age (SD): (i) intervention group = 40 years (SD not reported), (ii) control group = 37.9 years (SD not reported) (p 23) Gender: (i) intervention group = 75% women, (ii) control group = 85% women (Jacobsen 1993, p 23) Race/ethnicity: (i) intervention group = 70% white, (ii) control group = 75% white (Jacobsen 1993, p 22) Years of experience: (i) intervention group = 12.7 years, (ii) control group = 9.7 years (Jacobsen 1993, p 23) Previous child protection training: (i) intervention group = 75% received at least 1 hour of prior education about child sexual abuse, (ii) control group = 70% received at least 1 hour of prior education about child sexual abuse (Jacobsen 1993, p 23) Previous experience with child maltreatment reporting: (i) intervention group = 60% no, (ii) control group = 60% no (Jacobsen 1993, p 23) Baseline equivalence: not reported
Interventions	Name: 3‐hour inservice training on child sexual abuse (adapted from Kleemeier 1988) Contents: (i) prevalence, laws, and reporting, (ii) definitions, myths and facts about child sexual abuse, (iii) indicators of child sexual abuse, (iv) long‐term effects of child sexual abuse, (v) identifying, reporting, and handling disclosure, and (vi) child sexual abuse prevention programmes (Jacobsen 1993, pp 18‐20) (adapted from Kleemeier 1988) Processes and teaching methods: (i) specification of workshop goals, (i) didactic presentation, (iii) practical application of concepts, (iv) video presentation, and (v) group discussion Delivery mode: face‐to‐face workshop Trainers and qualifications: 2 x facilitators (school psychology graduate interns) Duration: 3 hours Intensity: not reported Intervention integrity: not reported Comparison condition: not reported
Outcomes	Eligible measures (outcome domain) Teacher vignettes measure (primary outcome: number of reported cases of child abuse and neglect, measured subjectively by participant responses to vignettes), comprising 8 vignettes with open‐ended responses to questions about behavioural indicators, most appropriate course of action, and initiating a conversation with the child Teacher knowledge scale (secondary outcome: knowledge of core concepts in child abuse and neglect), comprising a 30‐item scale with true/false response options Ineligible measures (reason): teacher opinion scale (items assess attitudes towards child sexual abuse rather than attitudes towards the reporting duty; not prespecified in the protocol for this review), comprising a 25‐item scale with response options on 4‐point Likert‐type scale Timing of outcome assessment: pre‐test (details not reported), post‐test (details not reported)
Notes	Funding: not reported Author contact: no
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Comment: selection bias due to non‐randomised allocation of participants Quote: "the treatment group was not randomly assigned but took part in the study based on interest and the degree to which site administrator deemed the information important to the start ... . the control group consisted of 20 randomly selected elementary school teachers ..." (Jacobsen 1993, p 10)
Allocation concealment (selection bias)	High risk	Comment: selection bias (biased allocation to interventions) due to inadequate concealment of allocations prior to assignment attributable to research design (non‐randomised study). Participants were assigned based on rotation (participants and investigators could foresee assignment).
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: performance bias due to lack of blinding and almost certain knowledge of the allocated intervention by participants and personnel during the study, which may have influenced subjectively measured study outcomes (i.e. self‐report measures)
Blinding of outcome assessment (detection bias) All outcomes	High risk	Comment: detection bias due to likely 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)
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: reported sample is 40; however, the author did not report on attrition over time (i.e. at recruitment, intervention, outcome assessment) Quote: "Control Group (n = 20)"; "Treatment Group (n = 20)" (Jacobsen 1993, p 23)
Selective reporting (reporting bias)	Low risk	Comment: although the study protocol was not available, all outcomes described in the methods were fully reported in the study. Appropriate data and comparisons were offered (Jacobsen 1993, p 23; p 26).
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: authors reported coefficient alphas for the Teacher Knowledge Scale (α = 0.84), Teacher Opinion Scale (α = 0.78), and Teacher Vignettes Measure (α = 0.78) (Jacobsen 1993, pp 10‐3)
Group comparability (selection bias)	Unclear risk	Comment: authors reported differences between groups on demographic characteristics likely to influence results, including: mean years teaching experience, prior experience with child sexual abuse, and previous child protection education. Groups appear comparable on all characteristics apart from experience (control group 9.7 years; treatment group 12.7 years). Authors reported group comparability data for each of the study outcomes, but did not assess whether baseline equivalence was achieved (e.g. via statistical testing) (Jacobsen 1993, pp 23, 25).
Contamination (contamination bias)	Unclear risk	Comment: study author does not report the ways in which contamination may have been possible, or what may have been done to prevent or minimise this. It is unclear whether experimental and control participants had contact with each other in their workplaces.