The mean percentage of inconsistent reporters was 7.4%. On average, their exclusion reduced bullying victimization prevalence by 13.5% (M∆ = -13.5%; Mh = 0.107). In about two-thirds of cases, adjusted estimates fell outside corresponding unadjusted confidence intervals. The present study confirms that inconsistent reporting can substantially bias bullying victimization prevalence estimates. It should therefore be systematically addressed whenever possible.
Romain Brisson (Mon,) studied this question.