Skip to content
SciBLIND
Part of the ScAIentist® ecosystem

Research. Simplified.
AI-enhanced. Scientifically rigorous.

Design experiments, surveys and vignette studies in minutes – no coding required. AI helps you set up, and science keeps it rigorous.

10 free tokens daily · No credit card required

9 languages · consent, instructions and UI – all translated, zero extra work

Seven paradigms. One platform.

⚖️

Blind Comparisons

ELO-ranked pairwise or group comparisons. No order bias, no position effects.

📋

Surveys & Scales

Likert scales, multiple choice and open text with randomized ordering.

📖

Vignette Studies

Scenario-based surveys with pinned stimulus text. Latin square counterbalancing built in.

⏱️

Reaction Time

Millisecond-precise timing for Stroop, priming, go/no-go and visual search tasks.

🔀

A/B Preference

Two-alternative forced choice between variants with built-in significance testing. Between-subjects random assignment ships in v2.7.

🎯

Categorical Perception

Match-to-sample paradigm for within/cross-category discrimination with accuracy and RT.

📏

Method of Adjustment

Drag a stimulus until it matches a reference. Continuous-error DV for Müller-Lyer, Ebbinghaus and Ponzo illusions.

AI-Enhanced

AI that assists, never decides

AI handles setup so you can focus on research. Every AI suggestion is reviewed by you before it reaches participants.

💬

AI study configurator

Describe your research question in plain language. AI recommends the right paradigm and pre-fills your study.

📝

Scenario generation

Generate vignette scenarios from a theme description. Edit, reorder and approve before publishing.

📊

Item analysis

Post-pilot Cronbach’s α, item-total correlations and revision suggestions – computed automatically.

🔊

Read-aloud narration

Browser-native text-to-speech reads scenarios and questions aloud for accessibility.

Scientific rigor

Built for publication

SciBLIND is designed to produce data that meets peer-review standards.

Counterbalanced designs (Latin square, blocked random)
Adaptive staircase methods for threshold estimation
Position bias correction and order randomization
Configuration change log with participant counts
Multilevel-ready exports for HLM, lme4 and SPSS MIXED
Bradley-Terry and ELO ranking with confidence intervals
Reproducibility

Every study ships its own audit trail

SciBLIND produces the three artifacts reviewers ask for – automatically, on every publish.

📄

Methodology PDF

One-click pre-registration-ready report. Paradigm, stimuli, counterbalancing scheme, randomization seed, inclusion criteria – formatted for journal appendices and ethics boards.

📦

Reproducibility bundle

JSON export with study definition, stimulus hashes, RNG seed and platform version. Reimportable into any SciBLIND instance. The .psyexp equivalent – without the code.

⏱️

Timing audit

Per-participant jitter histogram grounded in Reimers and Stewart 2015. Plain-English grade – excellent, good, fair, poor – so you can filter trials on empirical fidelity, not assumptions.

Shipped April 2026. Live on every published study.

Why SciBLIND?

Sci · Bias-Limited INdependent Decision

Every decision deserves clean data. Whether you’re a researcher running a psychophysics study or a product team testing two designs – SciBLIND removes the noise so you can trust the signal.

Start collecting clean data

Free to start. No credit card. No coding.

Create a study