Descriptive Statistics Course with Python
Descriptive Statistics is a beginner-level course of 5 lessons (about 2 h 26 min), with 15 exercises and 15 quiz questions. No prerequisites. Exercises run in the browser, with nothing to install.
- Statistics
- Beginner
- 5 lessons
- 2 h 26 min
- 15 exercises
- Exercises run in the browser
The course in brief
This course lays the statistical foundations you need for any work with data. You will learn to identify variable types, summarize a series with measures of central tendency and spread, spot outliers and describe the relationship between two variables. Every concept is practiced in Python with the standard library and pandas.
What you will be able to do
- Identify a variable's type and choose the right summaries
- Compute and interpret the mean, median, mode and weighted mean
- Measure spread with variance, standard deviation, IQR and the coefficient of variation, and master the ddof parameter
- Detect and handle outliers with good judgment
- Analyze the relationship between two variables with a cross-tabulation or a correlation coefficient
- Explain why correlation does not prove causation
Syllabus: 5 lessons
- Distinguish qualitative variables (nominal, ordinal) from quantitative ones (discrete, continuous)
- Build a count and frequency table in Python
- 02
Central Tendency: Mean, Median and Mode
- Compute the mean, median and mode in Python
- Choose the measure of central tendency that fits the shape of the data
- 03
Spread: Range, Variance and Standard Deviation
- Measure the spread of a series with the range, variance and standard deviation
- Distinguish population variance from sample variance (ddof)
- 04
Quantiles, Box Plots and Outliers
- Compute quartiles, deciles and the interquartile range
- Spot outliers with the 1.5 x IQR rule and the z-score
- 05
Relationship Between Two Variables: Cross-Tabulations and Correlation
- Summarize the relationship between two qualitative variables with a cross-tabulation
- Compute and interpret a covariance and a Pearson correlation coefficient
- Distinguish correlation from causation
Final project
A Statistical Portrait of a Shop's Sales
Generate in Python a dataset of at least 200 orders (date, channel, category, number of items, amount), deliberately including a few extreme amounts. Then produce a complete statistical portrait: variable types, frequency distributions, measures of central tendency and spread by channel, outlier detection, and a study of the relationship between number of items and amount.
Deliverables
- A commented Python script or notebook that creates the data and computes all the measures
- A summary table (mean, median, standard deviation, Q1, Q3, IQR) by channel
- The list of detected outliers, with the decision made for each one
- A boxplot of amounts by channel and an items x amount scatter plot
- A written summary of at most 10 lines with three quantified findings
Assessment criteria
- Each variable is correctly typed and summarized with suitable measures
- The choice between mean and median is justified by the shape of the distribution
- The ddof convention used is explicit and consistent
- Outliers are examined and not blindly removed
- The summary distinguishes correlation from causation
Frequently asked questions
Do I need strong math skills to take this course?
No. The course is designed for beginners: formulas are explained step by step in plain text and always come with a worked example in Python.
What is the difference between the mean and the median?
The mean uses all the values and is sensitive to extreme values; the median is the middle value of the sorted series and stays stable when extremes appear.
Why do numpy and pandas give different standard deviations?
numpy divides by n by default (ddof=0) while pandas divides by n - 1 (ddof=1), the convention for a sample. The course explains when to use each.
Do I need to install Python?
No, the exercises run directly in your browser. You can also reproduce the examples on your own computer with Python and pandas.
This course, told for your job
Say what you want to be able to do: “Descriptive Statistics” takes its place in your path, with examples from your field. The concepts stay those of the course.