cura librorum


Google Public Data: Ethnic Breakdown of Oakland USD Elementary Schools
September 17, 2011, 20:49
Filed under: data analysis, google

Google is an extremely data-driven company – that’s no secret (I face it every day!). I stumbled upon Google Public Data while looking through the Google for Educators site, as I was thinking more about how Google works in the public sector. They’re in a really interesting and complicated situation, come to think of it. Google has so much data, the data that so many crave access to, and the computing and brain power to analyze and visualize it. Yet being a private corporation committed to dissemination & organization of information rather than decision-driving is quite the ideal postmodern position as a company… but that’s besides the point.

I present some of the data from the ethnic makeup of each of the elementary schools the Oakland USD, from 2005 – 2009. This data visualization pinpoints the location of the school when you select an “ethnicity” variable, and visualizes a bubble of a size relative to the largest proportion of data available. Provides some stark contrasts. For example, below the graph for proportion of white students clearly shows a racial divide as whites are concentrated in the Lakeshore/Piedmont areas. I drove around there today after driving through the Fruitvale/International area and indeed, all suspicions are affirmed.

Other examples by ethnicity
Hispanic/Latino
White
Asian
African-American

I think that this has just naturally occurred, but it again, is an interesting case for something like a geographical or architectural study of respective areas, as well as of disparity, social flux over time as neighborhoods grow and change, analyzing where Oakland’s revenues are coming from and if truly, the ethnic makeup of elementary schools can determine incomes/affluence of the area…

Note that you can also see this as a linear graph over time, compare schools by proportion and by time, pie/scatter plot with different x/y variables, and billions of different data visualization combinations. Interesting experiment from a ethnicity time vs. proportion graph: When I did a quick linear visualization of data by percentage of whites in the public schools in Alameda and Castro Valley Unified School districts, I saw that proportionally to the rest of the ethnicities, whites had decreased drastically (over 10%, usually) in the dataset while other ethnicities had remained mostly stagnant if not grown. Wondering about this, I remembered that this is Google Public Data – a sign that perhaps the great shift is the white migration over to private schools in the districts as opposed to public?

Questions, questions.