Introduction to power analysis in A/B testing: Deciding how long to run tests
Details
Jonny Brooks-Bartlett (https://www.linkedin.com/in/jonathan-brooks-bartlett-2b5bb2b1/) will be speaking on power analysis in A/B testing.
Jonny is a data scientist at Deliveroo currently working on developing algorithms to improve the consumer experience on the Deliveroo app.
__________________________________________________________
What to Expect:
Being part of an agile company means that new features are typically released iteratively. However, these features often need to be experimented with to work out if it's actually a beneficial feature to completely roll out. These experiments take time to run and we need analysts to determine how long the experiments need to run. In this presentation, we will learn about the reasons why these experiments can take a long time to run and how we work out the length of time it should take.
For those of us who would like a more hands on introduction, we will practice on some open data and perform our own A/B test. All that's required is a laptop with Python installed. We also recommend that you install Jupyter Notebooks as well but this is not a requirement.
_________________________________________________________
Agenda:
18:00 Welcome, drinks and networking
18:45 Talk and open data A/B testing practice
19:45 Announcements and Open Floor
__________________________________________________________
This event is organised in collaboration with Central Working, the UK’s premier homegrown shared workspace provider, transforming the way people work and companies grow.
Website: https://www.centralworking.com/
Twitter: @centralworking, #centralworking