The rise of HuggingFace
Why you should hire an information architect
Accessible data visualizations
How to invest in junior talent
Avoiding inflated A/B test results
Explore by Category
Analytics Dispatch 12/14/2015
The fashionable side of data science
This week we're looking at best practices, Google's effect on the 2016 election, climate change, p-values, and data-related stocking stuffers.
Sign up here to get the next issue in your inbox.
DATA IN THE WILD
The Field Guide to Data Science
Booz Allen just released The Second Edition of The Field Guide to Data Science, which walks you through how to use data to generate value for your organization. The guide includes practical advice, tested processes, and insights that are helpful for anyone who touches data, whether you're a senior exec, a practioner, or a newbie. - Booz Allen Hamilton
YOU MIGHT ALSO LIKE...
Fashion Goes Deep: Data Science at Lyst
Fashion moves quickly. So, too, does the data science that powers e-commerce sites. In this interview, Lyst lead data scientist Eddie Bell shares the ins and outs of their recommendation engine. Learn how his team has tackled the challenges of constantly changing merchandise and kept suggestions fresh using machine learning and image analysis. - Fast Forward Labs
The Quartz guide to bad data
This comprehensive reference is intended for journalists, but it's a worthwhile read for anyone working with data. Familiarize yourself with common issues—ambiguous field names, inconsistent date formats, biased samples—so you can catch data quality problems early. - Quartz
SEARCH ENGINES FOR BERNIE
Why Google Search Results Favor Democrats
Apparently, Democrats are better at SEO than Republicans—Republican presidential candidates average only 5.9 favorable results in the first 10 Google search results, compared to Democratic candidates, who average seven favorable results. Since roughly 75 percent of people use the internet to get their news, how will Google's rankings affect the outcomes of the 2016 election? (And if AskJeeves were still around, what would he have to say about all this?) - Slate
The Paris Climate Conference (COP21) has, unsurprisingly, spawned a bunch of press. Here are some of our favorite articles that used data to explore climate change issues.
- Climate change data visualizations around the web (Storybench)
- A Blueprint for a Carbon-Free World (National Geographic)
- Global CO2 emissions are set to stall in 2015 (The Economist)
- Period of Soaring Emissions May Be Ending, New Data Suggest (The New York Times)
- Big Data Predicts Centuries Of Harm If Climate Warming Goes Unchecked (NPR)
Lately, the p-value has come under fire for propping up dubious study results. But the reason this so-called “misleading statistic” causes so much trouble might be due to a fundamental problem: it's incredibly hard to explain what, exactly, a p-value means.
- Not Even Scientists Can Easily Explain P-values (FiveThirtyEight)
- Blinded by Statistical Significance (KelloggInsight Blog)
- Dawkins on Saying "statistically, ... " (DataOrigami)
We analyzed millions of queries authored by Mode users to see which databases are easiest for analysis day-in and day-out.
Have a wonderful week!