Biostatistics for medical research explained

This website breaks down biostatistical methods and concepts for medical practice using simple explanations, intuitive visuals, and clear math.

Aim: We make biostatistics clear, practical, and relevant for medical professionals and students—through straightforward explanations, intuitive visuals, and essential math. Our goal is to bridge the gap between statistics and clinical research.

Research can roughly be divided into the following types:

Descriptive Analysis: To summarize the data set without interpretation.

Exploratory Analysis: To find patterns, trends, or relationships to generate hypotheses.

Inferential Analysis: To estimate how findings in a sample hold for a larger population.

Predictive Analysis: To predict outcomes for individuals based on features.

Causal Analysis: To estimate what happens on average when one variable changes another (e.g. average treatment effect).

Mechanistic Analysis: Typically to establish how one variable deterministically influences another.

We will focus on the following on two aspects:

1.  Interpretation and use of statistical evaluation measures

• Statistical evaluation measures

• p-values

• Brier score and scaled Brier score

• Pseudo R2 measures

• Calibration

2. Interpretation of model coefficients for regression models in

• the 3 archetypical statistical models

• Logistic regression

• Linear regression

• Cox regression

• Extensions:

  • Extensions logistic regression

    • Ordinal logistic regression with proportional odds assumption

  • Extensions linear regression

  • Extensions Cox regression

  • 
    

References:

Leek, J., & Peng, R. D. (2015, February 26). What is the question? Science. https://doi.org/10.1126/science.aaa6146

Kiefer, J. C. (1987). Introduction to statistical inference (G. Lorden, Ed.). Springer.