Skip to main content

Experimental Design in Python

Learn about experimental design, and how to explore your data to ask and answer meaningful questions.

Start Course for Free
4 Hours16 Videos53 Exercises4,850 Learners4400 XP

Create Your Free Account

GoogleLinkedInFacebook

or

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA. You confirm you are at least 16 years old (13 if you are an authorized Classrooms user).

Loved by learners at thousands of companies

Course Description

Data is all around us and can help us to understand many things. Making a pretty graph is great, but how can we tell the difference between a few outliers on a graph and a real, reliable effect? Is a trend that we see on a graph a reliable result or just random chance playing tricks? In this course, you will learn how to interrogate datasets in a rigorous way, giving clear answers to your questions. You will learn a range of statistical tests, how to apply them, how to understand their results, and how to deal with their shortcomings. Along the way, you will explore Olympic athlete data and the differences between populations of continents.

  1. 1

    The Basics of Statistical Hypothesis Testing

    Free

    In this chapter, you will learn how to explore your data and ask meaningful questions. Then, you will discover how to answer these question by using your first statistical hypothesis tests: the t-test, the Chi-Square test, the Fisher exact test, and the Pearson correlation test.

    Play Chapter Now
    Welcome to the course!
    50 xp
    Getting started with plotnine
    100 xp
    Boxplots
    100 xp
    Density plots
    100 xp
    Student's t-test
    50 xp
    Your first t-test
    100 xp
    One-sample t-test
    100 xp
    Two-sample t-test
    100 xp
    Testing proportion and correlation
    50 xp
    Chi-square test
    100 xp
    Fisher's exact test
    100 xp
    Pearson correlation
    100 xp
  2. 2

    Design Considerations in Experimental Design

    In this chapter, you will learn how to examine and multiple factors at once, controlling for the effect of confounding variables and examining interactions between variables. You will learn how to use randomization and blocking to build robust tests and how to use the powerful ANOVA method.

    Play Chapter Now
    Confounding variables
    50 xp
    Exploring confounding variables
    100 xp
    Finding confounding variables
    100 xp
    Blocking and randomization
    50 xp
    Random sampling
    100 xp
    Blocking
    100 xp
    Paired t-test
    100 xp
    ANOVA
    50 xp
    One-way ANOVA
    100 xp
    Two-way ANOVA
    100 xp
    Interactive effects
    50 xp
    Two-way ANOVA with interactive effects
    100 xp
    Choosing an appropriate test
    100 xp
  3. 3

    Sample size, Power analysis, and Effect size

    In this chapter, you will focus on ways to avoid drawing false conclusions, whether false positives (type I errors) or false negatives (type II errors). Central to avoiding false negatives is understanding the interplay between sample size, power analysis, and effect size.

    Play Chapter Now
    Type I errors
    50 xp
    Bonferroni correction
    100 xp
    Šídák correction
    100 xp
    Sample size
    50 xp
    Exploring sample size
    100 xp
    Sample size for a t-test
    100 xp
    Effect size
    50 xp
    Effect size for a t-test
    100 xp
    Computing Cohen's d
    100 xp
    Effect size for a Fisher exact test
    100 xp
    Effect sizes for Pearson correlation
    100 xp
    Power
    50 xp
    Power analysis for a t-test
    100 xp
    Interpreting power analyses
    50 xp
  4. 4

    Testing Normality: Parametric and Non-parametric Tests

    In this final chapter, you will examine the assumptions underlying statistical tests and learn about how that influences your experimental design. This will include learning whether a variable follows a normal distribution and when you should use non-parametric statistical tests like the Wilcoxon rank-sum test and the Spearman correlation test.

    Play Chapter Now
    Assumptions and normal distributions
    50 xp
    Exploring distributions with summary stats
    100 xp
    Q-Q plot
    100 xp
    Testing for normality
    50 xp
    Shapiro-Wilk test
    100 xp
    Choosing tests and normality
    50 xp
    Non-parametric tests: Wilcoxon rank-sum test
    50 xp
    Wilcoxon rank-sum test
    100 xp
    Wilcoxon signed-rank test
    100 xp
    Parametric vs non-parametric tests
    100 xp
    More non-parametric tests: Spearman correlation
    50 xp
    Spearman correlation
    100 xp
    Choosing the correct correlation test
    100 xp
    Summary
    50 xp

Datasets

Olympic datasetUN dataset

Collaborators

chesterChester Ismayamy-4121b590-cc52-442a-9779-03eb58089e08Amy Peterson

Prerequisites

Foundations of Probability in Python
Luke Hayden Headshot

Luke Hayden

Postdoctoral Researcher

Luke is a Senior Data Scientist at Conjura, where he applies statistics, machine learning and data visualisation to commercial problems. While working as a researcher, his research projects included using machine learning methods and gene expression data to model biological age and measure tissue rejuvenation and developing statistical methods to measure developmental variability. He has worked in the Averof Lab and Pantalacci/Sémon labs at the ENS de Lyon and holds a PhD from the National University of Ireland, Galway. His personal ambition is to find a really good excuse to use a polar area graph.
See More

What do other learners have to say?

I've used other sites—Coursera, Udacity, things like that—but DataCamp's been the one that I've stuck with.

Devon Edwards Joseph
Lloyds Banking Group

DataCamp is the top resource I recommend for learning data science.

Louis Maiden
Harvard Business School

DataCamp is by far my favorite website to learn from.

Ronald Bowers
Decision Science Analytics, USAA