Scoring Algorithms to Optimize Viewing Time Measures of Sexual Interest in Children

Area of Science:

• Psychology
• Forensic Psychology
• Psychometrics

Background:

• Viewing times (VTs) are indirect measures of sexual interests, based on differential scrutiny of stimuli.
• Current methods for assessing sexual interest in children and adults involve comparing decision latencies.
• The reliability and validity of VT measures can be influenced by the chosen scoring algorithm.

Purpose of the Study:

• To explore and compare various scoring algorithms for the VT task to improve assessment of sexual interests.
• To evaluate the impact of different scoring methods on the reliability, convergent validity, and classification of sexual interest in children.
• To determine the optimal scoring approach for VT tasks based on research or diagnostic goals.

Main Methods:

• A secondary analysis of a large German male community sample (N=8554) was conducted.
• Several scoring algorithms were tested, including standardized differences and exclusion of extreme viewing times.
• The study compared algorithms based on reliability, convergent validity, and influence of confounding variables.

Main Results:

• The validity of the VT task was robust across different scoring methods.
• Reliability and classification proportions varied significantly between scoring algorithms.
• The mean difference with exclusion of the longest viewing time in each category demonstrated the best performance.

Conclusions:

• The choice of scoring algorithm significantly influences reliability and classification rates in VT tasks.
• The optimal scoring approach depends on the specific research or diagnostic objective, balancing sensitivity and specificity.
• Measurement and scoring methods critically impact prevalence estimates of sexual interest in children.