Skip to main content
Open BetaWe’re learning fast - your sessions and feedback directly shape AI CogniFit.

Help Center

How to Read Your Summary & Composite Score

Quick reference for understanding your performance metrics at a glance.

The Summary Tile Explained

Your Summary tile shows three key metrics that together tell you whether your AI-assisted work is calibrated (accurate self-assessment), sustainable (manageable workload), and improving over time.

Composite Score Formula

How we calculate your overall performance

Composite = (100 - |Δ|) × Quality Weight + TLX Penalty

The Composite score combines your calibration accuracy (Δ) with quality metrics and applies a workload penalty when TLX exceeds sustainable thresholds.

Score Components

Overestimation Δ

0-100 scale

Gap between your self-rating and reviewer score. Lower is better.

< 5
Expert calibration
5-15
Acceptable drift
> 15
Needs recalibration

micro-TLX

0-100 scale

Mental demand + frustration captured post-task. Lower is better.

< 40
Sustainable load
40-60
Moderate strain
> 60
Risk of burnout

Quality Score

0-100%

Rubric-based evaluation of your deliverable. Higher is better.

> 80%
High quality
60-80%
Acceptable
< 60%
Needs revision

Visual Reference

Reading your Summary tile at a glance

Δ
8
Drift
TLX
45
Load
Quality
82%
Score
Composite Score
74

Interpreting Your Score

High Composite, Low Δ, Low TLX

Optimal state—you're calibrated, producing quality work sustainably.

Continue current approach, try harder tasks.

Advanced packs

High Composite, High Δ

Good output but poor self-assessment. You're underselling or overselling.

Review calibration drills to align perception with reality.

Prediction practice

Low Composite, Low Δ, High TLX

Accurate self-assessment but struggling with workload.

Reduce task complexity or take breaks.

TLX guide

Low Composite, High Δ, High TLX

Warning state—both calibration and workload are off.

Stop and reset. Review Fair Trial methodology.

Fair Trial

Quick Tips

  • Track your Δ trend over 5+ sessions to spot drift patterns.
  • Compare morning vs. afternoon TLX to find your optimal work windows.
  • Review misses in the 'Review & Learn' section after each run.
  • Use the Composite score for quick comparisons; dive into components for diagnosis.

Practice Now

Apply what you've learned

Next Steps

Ready to measure your AI impact? Start with a quick demo to see your Overestimation Δ and cognitive load metrics.

PrivacyEthicsStatusOpen Beta Terms
Share feedback