Clinician tool

CPT score interpretation calculator

Enter raw CPT scores to get population-percentile interpretation across the four most-reported metrics: omission errors, commission errors, mean reaction time, and RT variability.

Patient age group

Raw scores

Enter values from the patient's CPT report. Leave blank for any metric you don't have.

Methodology

This calculator uses approximate published norms compiled across multiple CPT validation studies (TOVA, Conners CPT-3, QbTest, IVA-2 normative literature). The norms are conservative population estimates; they are not specific to any single test or platform.

Approximate norms used
MetricAdult M (SD)Pediatric M (SD)
Omission errors2.5 (4)4 (6)
Commission errors15 (8)18 (10)
Mean RT (ms)420 (80)480 (100)
RT variability — SD (ms)95 (30)110 (35)

These are approximate. Specific tools have their own age- and sex-stratified norms based on larger samples — use those for clinical decisions.

Why these metrics?

  • Omission errors map most directly to vigilance / inattention — the patient's ability to detect target stimuli over time.
  • Commission errors map most directly to response inhibition / impulsivity.
  • Mean reaction time reflects processing speed and task engagement; elevated values can suggest cautious responding, slowed processing, or sub-effortful engagement.
  • RT variability is one of the most ADHD-discriminative CPT metrics in the literature — intra-individual variability is consistently higher in ADHD groups than mean RT alone.

Caveats

  • CPT performance is sensitive to engagement, fatigue, time-of-day, sleep, caffeine, and medication state. A single bad performance does not establish a diagnosis.
  • CPT is one input among many. Diagnostic decisions integrate clinical interview, developmental history, rating scales, functional impairment, and rule-outs.
  • Within-subject change (test–retest with treatment) is often more meaningful than cross-sectional comparison to norms.
  • Norms used here are approximations from the published literature, not vendor-proprietary normative scoring.