joseph.by

This diagram that I put together is my first attempt to organize the analytics, growth, testing and optimization tool landscape. This slide is also available as a PDF. Discussion on Hacker News.

The Mississauga (Half) Marathon is Sunday. Here’s my Race Plan:

rowingsoundsfun:

Being “up to date” beyond a certain point - that is, being aware of all of the latest tweets, news stories, stock quotes, etcetera - can actually hurt one’s understanding of the universe. This is partially explained by the lingering impact of misinformation.

In trying to get a story first, publishers - whether old-line media or your brother-in-law on twitter - will inevitably get some facts wrong. Journalism is, after all, the first draft of history - and citizen journalism even more so. These mistakes will (hopefully) eventually be corrected in the public record. However, the damage that the initial, incorrect facts do to public understanding seems to be permanent. That’s because the human brain does not process corrections as well as it processed the original story.

This is beautifully spelled out in an article in Psychological Science in the Public Interest from December, 2012. I have no idea how the article came to my attention, but I’m glad it did. Worth a quick read.

Now I don’t feel quite as bad for thinking that Tamerlan and Dzhokhar Tsarnaev were not the guys in brown dri-fit and a white baseball cap.

I’ll try to remember this the next time I’m tempted to retweet some half-baked “fact.”

Another of my clients is about embark on another round of A/B testing, so I thought I’d post some of my favorite A/B testing resources, summarizing each as a single rule:

1. Why you should always set your sample size (n) in advance:
   In How Not to Run an A/B Test @evmill makes the case for always fixing your sample size in advance. I generally agree with him — this is good research design.
2. How to set your sample size:
   A/B Testing Tech Note: determining sample size — This post from @noahlorang at 37 Signals explains how to use a power law function to calculate the required sample size in order to measure an affect in some downstream outcome.
   Remember - the “significance” measures generated by a testing tool are only as accurate as your “conversion” measure. In other words, if Optimizely is looking at “Sales,” but you want to look for changes in “Revenue”, then the significance measure that a tool spits out will be irrelevant.
3. Balancing Speed vs. Certainty in A/B testing - a good post summarizing the inevitable tradeoffs from @jfarmer

I’ve been doing a lot of data analysis for a couple clients lately; I’m using R pretty heavily, and thought that I would create this post to keep track of the R resources I use most.

This post will be updated frequently over the coming days as I fine-tune my R installation.

Things I use or recommend:

• Stats Mini-Courses from the UCLA Statistics Center, including a lot of valuable information on the basics of loading and manipulating data in R

Getting Started (on Mac):

• Install R
• Install R Studio Desktop (an IDE)
• Install R Commander (a GUI for various statistical analysis functions, that also runs as a plug-in for R studio)

Handy resources that I’ve used:

• R for Mac OS FAQ

Things I am meaning to look at:

• One Page R

(via fuckyeahtoronto)

Festa San Giuseppe on Flickr.

ccindecision: