Documentation
Everything you need to get started and use Erhebimus
Quick Start
Installation
Download the latest version of Erhebimus and install it on your server.
# Download and extract
wget https://erhebimus.de/download/latest.tar.gz
tar -xzf latest.tar.gz
cd erhebimusConfiguration
Adjust the configuration file to your needs.
# Edit config
nano config.toml
# Start server
./erhebimus --port 8080First experiment
Sign in and create your first experiment via the web UI.
Documentation categories
Handbook
Complete guide to Erhebimus from getting started to advanced usage
- Introduction
- Getting Started
- User Guide
- Data & Reports
- Administration
- Troubleshooting
Installation & setup
System requirements, installation, and initial configuration
- System requirements
- Install on Linux
- Security configuration
- Set up SSL certificates
Create experiments
Step-by-step guide to building and configuring surveys
- Create a new experiment
- Manage stimuli/items
- Configure randomization
- Configure response modes
Manage participants
Generate participant codes and manage access
- Create participant codes
- Access control
- Monitor progress
- Share participation URLs
Data export
Export data and feed it into analysis tools
- CSV export
- Excel export
- SPSS integration
- R scripts
Licensing
Manage and activate licenses
- License models
- Activation
- Upgrade options
- Renewals
API documentation
RESTful API for integration and automation
- Authentication
- Endpoints
- Sample code
- OpenAPI spec
Frequently asked questions (FAQ)
What are the system requirements?
Recommended: Linux server (Ubuntu 20.04+ or Debian 11+), 4 GB RAM (8 GB with Whisper), modern CPU. Runs as a standalone webserver; no extra infra needed.
Can I run Erhebimus without a server?
Yes, you can run it locally for tests or small studies. For production with many participants, we recommend a dedicated server.
How does speech recognition work?
Erhebimus uses the open-source Whisper model. Transcription runs locally on your server—no audio is sent to third parties.
Is Erhebimus GDPR compliant?
Yes. As a self-hosted solution, all data stays on your server. No telemetry or tracking. Final compliance depends on your infrastructure and privacy policy.
Can I run multiple experiments in parallel?
Yes—any number. Each experiment has its own settings, participant codes, and data. Check your license for limits.
Which languages does transcription support?
Whisper supports 90+ languages (great quality for German/English) and handles dialects/accents.
How do I back up my data?
Data lives in SQLite. Back up the DB and config regularly. Deployment scripts can automate backups.
Do you offer support?
Commercial license: email support included. Demo: community docs. Enterprise: optional priority support with SLAs.
Glossary
Consistent terminology used throughout Erhebimus. These terms are used in the user interface and documentation.
The top-level container for a survey. A Study contains Questions, optionally organized into Question Groups. Participants are linked to a Study via Assignments and provide Responses there. A Study can have multiple versions and produces aggregated Results.
A single stimulus or sentence stem that participants respond to. Each Question belongs to a Study and can optionally be assigned to a Question Group. The Response Type determines how the Question is answered (free text, scale, single choice, or multiple choice). For choice-based questions, a Choice Set is assigned.
A logical grouping of Questions within a Study. Question Groups enable thematic organization and can have their own introduction text or audio. The order of Question Groups can be customized, and Questions within a group can be randomized.
A reusable collection of Choice Options. Choice Sets are assigned to Questions that use the single choice or multiple choice Response Type. A Choice Set can be used across multiple Questions or even across different Studies.
A single selectable item within a Choice Set. Each Choice Option has display text and optionally a value for data analysis. The order of options can be fixed or randomized.
A participant's input to a Question. Depending on the Response Type, a Response can contain text, an audio recording, a scale value, or selected options. Audio Responses can have an automatic Transcription.
Defines how a Question is answered. Available types: Free Text (text or speech input), Scale (numeric rating), Single Choice (exactly one option from a Choice Set), and Multiple Choice (several options from a Choice Set).
Aggregated data and analytics from all Responses of a Study. Results include statistical analyses and can be exported in various formats (CSV, Excel).
A person who takes part in a Study. Participants gain access via a participant code that is linked to the Study through an Assignment. They provide Responses to the Study's Questions.
A person who creates and manages Studies. Researchers have a user account with a password and can create Studies, configure Questions, assign Participants, and analyze Results.
The automatic conversion of speech to text for audio Responses. Transcription runs server-side using the Whisper model and operates entirely locally, without sending data to third parties.
Links Participants to a Study. An Assignment contains a participant code and optionally class information. Assignments control who can participate in which Study.
Need help?
Our team is here to help. Contact us for technical support or tailored guidance.