With all the buzz around data science, machine learning, Python and R, learning old-fashioned statistics in a vanilla Excel workbook may seem near-obsolete.
In fact, learning data analysis in this stripped-down environment is perfect: it cuts down on the hype, and focuses on the essentials of what it means to explore sample variables, and make inferences about their population.
In the below half-day workshop outline, students learn how to make the leap from descriptive to inferential statistics. They learn how to size up a dataset, develop hypotheses and test them.
Finally, students will take what they’ve learned about hypothesis testing and apply it to a foundational practice of digital analytics: A/B testing.
Again, don’t let the hype fool you! A/B testing is an application of classical statistics, and the analyst needs to be able to design an experiment and test a hypothesis, using statistics as a mode of inquiry.
Lesson 1: Descriptive statistics
Objective: Student can classify, visualize and explore variables in a dataset using descriptive statistics.
Description:
- What is a variable and how do you measure it?
- Visualizing a variable’s distribution
- Describing a variable’s central tendency
Exercises: Summarizing a dataset with descriptive statistics
Assets needed: Home prices dataset
Time: 50 minutes
Lesson 2: Inferential statistics
Objective: Student can convert business questions into testable hypotheses
Description:
- Central limit theorem
- Law of large numbers
- Hypothesis testing
- Framing an independent samples t-test
Exercises: Crafting hypotheses from data and checking assumptions
Assets needed: Home prices dataset
Time: 50 minutes
Lesson 3: Hypothesis testing and experiment design
Objective: Student can conduct statistical research for meaningful business outcomes
Description:
- Conducting and independent samples t-test
- Evaluating for substantive and statistical significance
- Analyzing results for informed business decisions
- Presenting results for business impact
Exercises: Evaluating a business experiment
Assets needed: Home prices dataset
Time: 50 minutes
Lesson 4: Capstone — A/B testing
Objective: Student acquires framework for designing, implementing and analyzing A/B tests
Description:
- A/B testing and lean business methods
- Designing an A/B test
- Evaluating an A/B test
- Rolling out an A/B test
Exercises: End-to-end A/B test case study
Assets needed: E-commerce dataset
Time: 50 minutes
Leave a Reply