COMP 2303 · Data Science: Statistics

Led by Ronald Fisher Simulacrum

Descriptive and inferential statistics for data science — from mean and variance through confidence intervals to hypothesis testing and the p-value crisis.

1. Module 1

   Descriptive Statistics: Describing What You Have

   Led by Ronald Fisher Simulacrum

   The question

   Mean, median, mode — three summaries of the same data. Variance, standard deviation — how spread out it is. Correlation — how two variables move together. Why does the sample variance formula divide by n-1, and why does it matter?

   Outcome

   The student can compute the core descriptive statistics and explain Bessel's correction.

   Sub-units

   1. ○ 1.1 Compute the Basics Open this unit → ✓ ↻
   2. ○ 1.2 Why n-1? Open this unit → ✓ ↻
   Begin the course → Print certificate of completion
2. Module 2

   Distributions and the Normal Curve

   Led by Ronald Fisher Simulacrum

   The question

   The central limit theorem guarantees that sample means approach normality regardless of the underlying distribution. This is why the normal distribution appears everywhere in statistics. What does "sampling distribution" mean — and when do you use t instead of z?

   Outcome

   The student can compute z-scores, apply the CLT, and choose between z and t distributions.

   Sub-units

   1. ○ 2.1 Z-scores and the Normal Table Open this unit → ✓ ↻
   Open this module → Print certificate of completion
3. Module 3

   Confidence Intervals

   Led by Ronald Fisher Simulacrum

   The question

   A 95% confidence interval means that 95% of intervals constructed this way contain the true parameter. It does NOT mean there is a 95% probability the parameter is in this interval. What is the difference — and why does it matter?

   Outcome

   The student can construct confidence intervals and explain precisely what 95% confidence means.

   Sub-units

   1. ○ 3.1 Construct and Interpret Open this unit → ✓ ↻
   Open this module → Print certificate of completion
4. Module 4

   Hypothesis Testing

   Led by Ronald Fisher Simulacrum

   The question

   p < 0.05: reject the null. But what does a p-value actually say — and what are the most dangerous ways people misread it?

   Outcome

   The student can conduct a hypothesis test and explain the common misinterpretations of p-values.

   Sub-units

   1. ○ 4.1 Test the Claim Open this unit → ✓ ↻
   2. ○ 4.2 The p-value Problem Open this unit → ✓ ↻
   Open this module → Print certificate of completion
5. Module 5

   Statistics in Practice: From Description to Decision

   Led by Ronald Fisher Simulacrum

   The question

   The replication crisis was partly caused by over-reliance on p-values. A study with p=0.04 and n=12 is different from p=0.04 and n=1200. What should accompany a p-value — and how do you design studies that produce trustworthy conclusions?

   Outcome

   The student can report effect sizes, check assumptions, and take a defended position on statistical best practice.

   Sub-units

   1. ○ 5.1 Final Essay: Beyond the p-value Open this unit → ✓ ↻
   Open this module → Print certificate of completion