Internal Structure. Results from confirmatory factor analyses (CFA) provided empirical evidence to support the theoretical structure of the CAARS 2 scales.

• Content Scales. For Self-Report and Observer, after testing various models, a 5-factor model performed best (1. Inattention/Executive Dysfunction, 2. Hyperactivity, 3. Impulsivity, 4. Emotional Dysregulation, 5. Negative Self-Concept). This 5-factor model had strong fit statistics (across both rater forms, CFI and TLI ≥ .943; SRMR ≤ .042; RMSEA ≤ .047), and all factor loadings were positive, statistically significant, and above a minimum threshold of .40. The five identified factors were all distinct from each other.
  

Relation to Conceptually Related Constructs. Evidence to support the convergent validity of the CAARS 2 was supported by moderate to strong correlations between the CAARS 2 and established tests of related constructs.

• CAARS 2 and CAARS (median r): Self-Report = .76, Observer = .83
• CAARS 2 and CEFI Adult (median |r|): Self-Report = .76, Observer = .74
• CAARS 2 and WFIRS (median r): Self-Report = .48
  

Relation to Criterion Variables. The CAARS 2 demonstrated a high degree of criterion-related validity as various clinical groups had distinctly different profiles of scores.

• Meaningful differences between average scores were found between clinical groups, such that ratings of individuals with ADHD yielded higher scores than ratings of individuals from the General Population (median Cohen’s d for significant differences: Self-Report = 2.24 and Observer = 1.29) as well as higher scores than ratings of individuals with Depression and/or Anxiety (median Cohen’s d for significant differences: Self-Report = 1.26 and Observer = 1.06).
  
  
• Scores on the CAARS 2 differentiated between ADHD Inattentive and ADHD Combined presentations (e.g., scores on scales related to hyperactivity and impulsivity tended to be higher in the ADHD Combined groups than in the ADHD Inattentive group; Self-Report median Cohen’s d = 0.93 and Observer median Cohen’s d = 0.47).
  

Classification Accuracy. Scores from the CAARS 2 were able to correctly classify individuals from the General Population and those from clinical samples into their respective groups.

• Moderate to high levels of classification accuracy were demonstrated when distinguishing between individuals from the General Population and those diagnosed with ADHD (all presentations). Across all analyses, overall correct classification rates ranged from 89.7% to 92.5% for Self-Report and 84.7% to 91.7% for Observer.