StickForStats is an advanced, modular statistical toolkit designed to bridge the gap between complex statistical inference and researchers with minimal statistical background. Built using Python and Streamlit, the platform provides an intuitive interface for performing various statistical analyses with proper guidance on test selection, assumption verification, and result interpretation.
The project originated from a need identified during laboratory collaborations at CSIR-IGIB, where many researchers struggled with choosing appropriate statistical methods for their experiments. StickForStats makes advanced statistical approaches more accessible while maintaining mathematical rigor.
Recommends appropriate statistical tests based on data characteristics and research questions
Automatically checks test assumptions and suggests alternatives when assumptions are violated
Generates publication-ready visualizations with customizable options
Explains statistical concepts and test selection logic for better understanding
Interactive tutorial and visualization tool for understanding confidence intervals through simulations and practical applications. Helps researchers understand the nuances of statistical inference through direct manipulation of parameters.
View DemoComprehensive tools for monitoring and analyzing process quality, including control charts, process capability analysis, and measurement systems analysis. Particularly useful for biotechnology lab quality control.
Interactive PCA tool with detailed visualizations, step-by-step guides, and intuitive interpretation help. Features interactive biplots and scree plots with direct manipulation.
View DemoEducational tool for exploring statistical distributions and their properties with interactive visualizations and practical examples from biological sciences.
View DemoArchitectural refactoring from individual Streamlit modules to a cohesive, integrated Flask-based web application with comprehensive project structure, API endpoints, and authentication system.
Creation of individual statistical modules (Confidence Intervals, PCA, SQC, Probability) with standardized interfaces and educational components. Focus on intuitive design and proper statistical guidance.
Initial research on statistical education needs among researchers, technology selection, and prototyping of core concepts. Surveyed researchers at CSIR-IGIB to identify key statistical pain points.
The latest addition to StickForStats is a Retrieval Augmented Generation (RAG) system for contextual AI assistance in statistical analysis. This system provides intelligent guidance tailored to each user's specific analysis context.
Implemented efficient vector storage using SentenceTransformers for similarity searching, optimized for statistical terminology and concepts.
Comprehensive knowledge items for all statistical domains with module-component relationships for contextual suggestions.
System to monitor user activity and provide relevant assistance based on current module and actions, enabling intelligent content discovery.
Tiered access approach (Basic, Premium, Enterprise) with secure API key management for premium features.
StickForStats is an ongoing project with several exciting developments planned for the future:
Whether you're interested in using StickForStats for your research, collaborating on its development, or have questions about the platform, I'd love to hear from you.