Reexamining ADHD-Related Self-Reporting Problems Using Polynomial Regression.

Area of Science:

• Psychiatry
• Developmental Psychology
• Biostatistics

Background:

• Individuals with attention deficit hyperactivity disorder (ADHD) frequently underreport their symptoms compared to external informants and objective assessments.
• Understanding the factors contributing to this underreporting is crucial for accurate diagnosis and effective treatment of ADHD in adolescents.

Purpose of the Study:

• To investigate the factors contributing to symptom underreporting in adolescents diagnosed with ADHD.
• To compare the efficacy of polynomial regression with traditional difference score models in analyzing ADHD self-reporting.

Main Methods:

• Employed polynomial regression, an enhanced statistical methodology, to analyze self-reported ADHD symptoms in 107 adolescents (ages 11-15).
• Nested traditional difference score models within polynomial regression models to evaluate the influence of modeling strategy on results.
• Assessed discrepancies between adolescent and parent symptom reports and their association with hypothesized predictors.

Main Results:

• Sixty-six percent of adolescents with ADHD substantially underreported symptoms compared to parental reports; 23.6% denied all symptoms.
• Polynomial regression models found no meaningful linear associations between the discrepancy in symptom reports and hypothesized predictors.
• Difference score models demonstrated poor model fit and an increased risk of Type I errors in examining ADHD underreporting.

Conclusions:

• The study highlights limitations of traditional difference score methods for analyzing ADHD symptom underreporting in adolescents.
• Polynomial regression offers a more robust statistical approach, suggesting previous findings using difference scores may require re-evaluation.
• Accurate assessment of ADHD symptoms in adolescents necessitates advanced statistical techniques to avoid artifactual results.