Open Methodology — Peer Review Welcome

Decision Load Index: The Research

Peer-reviewable methodology for measuring cognitive decision overhead. Open validation gates, live data, and explicit failure conditions.

Read the methodology → Or take the assessment yourself

All methodology is published pre-registration. Success and failure criteria defined before data collection.

The construct

What the Decision Load Index measures

The DLI operationalizes decision overhead as a measurable cognitive construct. It captures four theoretically distinct dimensions of unresolved decision weight.

Dimension 1
Open loops
Count of active decisions in working memory — tasks initiated but not yet resolved. Correlated with intrusive thought frequency (Zeigarnik, 1927).
Dimension 2
Decision velocity
Rate of new decisions entering the system per unit time. High velocity combined with low resolution rate predicts load accumulation.
Dimension 3
Overdue backlog
Decisions past their implicit or explicit deadline — the most cognitively taxing category due to compounding meta-decisions (escalate? defer? abandon?).
Dimension 4
Cognitive domain spread
Whether decision load is concentrated in one life domain or distributed. Cross-domain load is hypothesized to impair restoration more than single-domain load.
Instrument design

How the instrument was constructed

The DLI uses self-report items calibrated against a 0–100 score. Each item targets one or more of the four dimensions above.

Item generation from existing literature
Items were developed from cognitive load theory (Sweller, 1988), open-loop research (Zeigarnik effect), and GTD task management literature. Initial pool: 24 candidate items across 6 domains.
Item reduction to 5-item short form
Items were reduced based on theoretical coverage and completion rate. The 5-item form was chosen to balance signal fidelity with response burden — targeting <5 minutes to complete.
Weighted composite scoring
Items are weighted by hypothesized cognitive impact: open loops (30%), decision velocity (15%), overdue items (25%), active projects (20%), domain spread (10%). Weights are explicit and published — not derived from factor analysis yet.
Pre-registered validation criteria
Success and failure thresholds were defined before data collection began. The instrument is considered validated only if all three primary validation gates pass. Failure criteria are published alongside success criteria.
Open science commitment

All methodology, data collection protocols, and validation criteria are published prior to data analysis. We commit to publishing negative results. Peer collaboration welcome — contact via research@cognitivethoughtengine.com.

Validation

Pre-registered validation gates

The DLI is validated only if all primary gates pass. Explicit failure conditions prevent post-hoc rationalization.

Gate Criterion Status
G1: Completion rate >40% of users who start the DLI complete it within one session Measuring (n=864 signups, quick-check + DLI tracking active)
G2: Score distribution DLI scores show normal or bimodal distribution (not ceiling/floor effects) Pending sufficient n (>50 completions required)
G3: Retest reliability Test-retest correlation >0.60 within a 7-day window Pending (requires longitudinal data)
G4: Convergent validity DLI score correlates >0.45 with established burnout/overwhelm measures Pending (qualitative interviews scheduled)
G5: Discriminant validity DLI score does NOT correlate >0.80 with general anxiety measures Pending (instrument boundary test)

Full validation protocol: evidence.html →

Live data (updated continuously)

What the data shows so far

Early-stage. Report what we observe, not what we hope.

47
Mean DLI score (0–100 scale)
Based on completions to date
864
Total signups
Quick-check + DLI completion tracking active
4:32
Mean completion time
Within target (<5 min response burden)

Acknowledged limitation: 0.14% completion rate means the current dataset is not representative. All validation gates requiring >50 completions are pending. We will not claim validated status until criteria are met.

Contribute to the research.

The simplest contribution is completing the DLI yourself. Every completion improves our dataset. Research collaboration and peer review welcome.

Read the full methodology → Take the DLI (contributes to dataset)

Takes 5 minutes  ·  Results private  ·  Data policy →