Calculator Inputs
Enter training and performance signals. The form uses a responsive grid: three columns on large screens, two on smaller screens, and one on mobile.
Example Data Table
Use this sample to understand typical ranges. Replace with your actual training metrics.
| Scenario | Hours | Completion | Pre | Post | Phish | Labs | Practice | Days Since | Role Risk |
|---|---|---|---|---|---|---|---|---|---|
| New hire baseline | 12 | 85% | 42% | 74% | 68% | 6 | 45 | 14 | 3 |
| High-risk admin | 18 | 92% | 55% | 82% | 77% | 10 | 60 | 21 | 5 |
| Refresher overdue | 8 | 80% | 48% | 70% | 60% | 4 | 20 | 90 | 3 |
Formula Used
The predictor uses normalized inputs, a weighted score, and a probability mapping:
- Normalize each signal to a 0–1 range (e.g., Post-test/100, Labs/12 capped).
- Weighted Score: Score = Σ(wᵢ × xᵢ) over key indicators.
- Penalties: subtract bounded adjustments for role risk, incidents, and time since training.
- Probability: P = 1 / (1 + e−k(Score−c)) with k=7 and center c=0.55.
- Retention: Retention ≈ P×100×2−Days/H, where H increases with practice, labs, and support.
How to Use This Calculator
- Gather completion, attendance, and assessment results.
- Enter phishing simulation and lab completion metrics.
- Add reinforcement signals: practice minutes and manager support.
- Click Predict Outcome to view results above the form.
- Use download buttons to export outputs for reporting.
What the Predictor Measures
This calculator converts training signals into an outcome probability, a retention estimate, and an expected risk reduction. It blends knowledge checks, behavior tests, and reinforcement indicators to summarize whether learning will transfer into safer daily actions. The adjusted score applies penalties for higher role exposure, prior incident history, and time since training, so similar test scores can still produce different outcomes across teams.
How Inputs Influence Probability
Post-test performance and learning gain drive the mastery signal, while completion and attendance reflect exposure to required content. Labs and weekly practice represent applied repetition that improves transfer to real workflows. Phishing simulation results act as a behavioral proxy, indicating how learners respond to realistic social engineering. Inputs are normalized to a 0–1 scale and capped to reduce outlier distortion and keep comparisons stable. Use consistent assessment formats and scoring rubrics to reduce noise between cohorts.
Interpreting Retention and Half‑Life
Retention decays as time passes after training, so the model uses a half-life concept to estimate how quickly skills fade. Practice minutes, labs, and manager support extend half-life, slowing decay. When days since training exceeds about 45–60 without reinforcement, short micro‑modules and scenario drills can restore retention faster than repeating the entire course. Use the retention estimate to schedule refreshers before high‑risk periods.
Using Results for Risk Planning
Combine the outcome tier with role risk level to prioritize follow‑ups. Moderate probability may be acceptable for low-exposure roles, but it can be insufficient for privileged users or teams handling sensitive records. Prior incident count highlights groups needing coaching, procedures, or oversight. Pair the prediction with controls such as MFA enforcement, least‑privilege reviews, and reporting playbooks. Re-test after interventions to confirm improvement and refine the plan.
Operationalizing Continuous Improvement
Export results to document decisions for audits and budgeting. Track probability, retention, and phishing scores quarterly, then compare changes after new modules, policy updates, or tooling rollouts. Segment by department and role risk to spot where reinforcement yields gains. Adjust weights using historical outcomes, and validate with periodic reassessments so the predictor remains calibrated, fair, and aligned with business objectives.
FAQs
What does the outcome probability represent?
It estimates the chance a learner will meet expected post‑training performance and safe‑behavior targets, given the entered signals. Use it to prioritize reinforcement, not to guarantee individual results.
How should I set the role risk level?
Choose higher levels for privileged access, sensitive data handling, payment systems, or admin tooling. Use lower levels for limited exposure roles. Align the scale with your internal risk register for consistency.
Why does time since training reduce retention?
Skills decay without practice. As days increase, the model applies a decay curve using an estimated half‑life. Refreshers, drills, and on‑the‑job practice extend half‑life and slow the decline.
Can I calibrate the calculator to my organization?
Yes. Replace weights, penalty factors, and tier thresholds using your historical assessments and incident outcomes. Validate changes with pilot groups, then monitor error rates and drift over time.
What should I export for audits and reporting?
Export the inputs, prediction tier, probability, retention, recommended actions, and the generated timestamp. Store results at an aggregated level when possible, and document any follow‑up actions taken.
How often should we reassess learners?
Reassess quarterly for most roles, and more frequently for high‑risk roles or recent incidents. Use shorter cycles after major policy changes, new threats, or tool rollouts.