About

Purpose

This blog is dedicated to sharing techniques for identifying and analyzing data that can help shed light on social and political ills, as well as their potential remedies.  Thus, my primary focus is to demonstrate best practices for interpreting and communicating results from data analysis in order to inform and potentially persuade a diverse audience.  

Why democratic Data Science?  
Early on, data analysis was exclusively conducted by those with expensive computers and software. With improvements in personal computing and open source software (R and python in particular), as well as the recent addition of cloud computing, individuals and small organizations can conduct meaningful analyses.  Small "d" democratic means rule by the people and I believe data science can and should be "ruled by the people."

Tools

     Data

Most of the data that I use is publicly available and can be accessed via an API, or application programmers interface, which greatly decreases the time and effort required to prepare data for final analysis.  The cliche is that 90% of data analysis is in the data preparation stage so accessibility (of data) should not be underestimated in any research plan.  Aside from easily accessible data behind APIs, I will also demonstrate how to find other, more 'wild' data sources that require additional effort harder to prepare for final analysis. 

     Examples of Data Sources

• St. Louis Federal Reserve (FRED)
• World Bank
• Bureau of Labor Statistics

     Software

Outside of my normal 9-5 job, I primarily use a free, open source statistical programming language called R to do the vast majority of data preparation, analysis, and output.  I also use Excel, Word, and an open source language called LaTex primarily for display purposes (though only for text/table output).  
I will explain how R can be used to clean and analyze data using a problem-based approach.  

The comprehensive R archive network (CRAN) has the most recent versions of R to download.

Example Problems

• Investigate trends relating to infant mortality rates in Sub-Saharan Africa since 1980
• Quantify the relationship between local unemployment rates and community college enrollment

I will explain (a) how one should go about obtaining and preparing data; (b) what sort of analyses should be considered; and (c) some of the best ways to present findings.

Other sources

• r-bloggers.com

Summary

Stay tuned for more updates -- my goal will be to make at least a short post every week, if not more.