stats4science 1.0.0

Statistical tools for scientists

StatisticsForScientists

StatisticsForScientists is a small Python project plus companion paper "From Estimands to Inference: A Practical Tutorial for Robust Statistical Analysis in Scientific Papers" for teaching practical statistical inference with an estimand-first perspective.

The repository includes:

• inferential_stats.py: the main lightweight analysis module
• examples.py: reproducible worked examples used in the paper
• tests/: unit tests for the statistical helpers and reporting functions
• latex/estimands_to_inference_tutorial.tex: the source of the tutorial paper
• latex/estimands_to_inference_tutorial.pdf: the compiled PDF version of the paper

Disclaimer: It's a Work In Progress!

It's a new project I've been working on in my spare time; there are still many concepts that need to be presented. For example, paired/repeated-measures analyses, regression, contingency tables, ANOVA / mixed models, etc.

Paper

The tutorial manuscript is available here:

• PDF: From Estimands to Inference: A Practical Tutorial for Robust Statistical Analysis in Scientific Papers

Installation

Create or activate your environment, then install the required dependencies:

pip install -r requirements.txt

Run The Examples

The paper's worked examples are implemented in examples.py. Please refer to this file for guidalines on how to use module inferential_stats.py.

python3 examples.py

Useful options:

• python examples.py --show-data prints the raw datasets used in the paper
• python examples.py --json emits machine-readable results for all worked examples

Run The Tests

python3 -m unittest discover -s tests

Rebuild The Paper

From the latex/ directory:

pdflatex -interaction=nonstopmode -halt-on-error estimands_to_inference_tutorial.tex
bibtex estimands_to_inference_tutorial
pdflatex -interaction=nonstopmode -halt-on-error estimands_to_inference_tutorial.tex
pdflatex -interaction=nonstopmode -halt-on-error estimands_to_inference_tutorial.tex

This produces latex/estimands_to_inference_tutorial.pdf.

Acknowledgments

My warmest thanks to Josh Starmer, the author of StatQuest YouTube channel and Tang et al., the authors of the paper Misuse, Misreporting, Misinterpretation of Statistical Methods in Usable Privacy and Security Papers.

Your journey continues with these two resources.