A Meta-Analysis of Viewing Time Measures of Sexual Interest in Children

Due to unobtrusiveness and ease of implementation, viewing time (VT) measures of sexual interest in children have sparked increasing research interest in forensic contexts over the last two decades. The current study presents two meta-analyses of VT measures adapted to assess pedophilic interest to determine their discrimination between sexual offenders against children (SOC) and non-SOC groups as well as convergent validity (associations with other measures of sexual interest in children). On average, VT measures showed moderate discrimination between criterion groups (fixed-effect d = 0.60, 95 % CI [0.51, 0.68], N = 2705, k = 14) and significant convergent validity with self-reports, penile plethysmography, Implicit Association Tests, and offence behavioral measures ranging from r = .18 to r = .38. VT measures, however, provided better discrimination for adults (fixed-effect d = 0.78, 95 % CI [0.64, 0.92]) than adolescent samples (fixed-effect d = 0.50, 95 % CI [0.40, 0.61]), Q _between = 9.37, p = .002. Moreover, compared to absolute scores, using pedophilic difference scores within adult samples substantially increased VT measures' validity (fixed-effect d = 1.03, 95 % CI [0.82, 1.25], N = 414, k = 7). Results are discussed in terms of their theoretical and applied implications for forensic contexts.