Introducing Data360R — data to the power of R
From the post:
Last January 2017, the World Bank launched TCdata360 (tcdata360.worldbank.org/), a new open data platform that features more than 2,000 trade and competitiveness indicators from 40+ data sources inside and outside the World Bank Group. Users of the website can compare countries, download raw data, create and share data visualizations on social media, get country snapshots and thematic reports, read data stories, connect through an application programming interface (API), and more.
The response to the site has been overwhelmingly enthusiastic, and this growing user base continually inspires us to develop better tools to increase data accessibility and usability. After all, open data isn’t useful unless it’s accessed and used for actionable insights.
One such tool we recently developed is data360r, an R package that allows users to interact with the TCdata360 API and query TCdata360 data, metadata, and more using easy, single-line functions.
…
So long as you remember the World Bank has an agenda and all the data it releases serves that agenda, you should suffer no permanent harm.
Don’t take that as meaning other sources of data have less of an agenda, although you may find their agendas differ from that of the World Bank.
The recent “discovery” that machine learning algorithms can conceal social or racist bias, was long overdue.
Anyone who took survey work in social science methodology in the last half of the 20th century would report that data collection itself, much less its processing, is fraught with unavoidable bias.
It is certainly possible, in the physical sense, to give students standardized tests, but what test results mean for any given question, such as teacher competence, is far from clear.
Or to put it differently, just because something can be measured is no guarantee the measurement is meaningful. The same applied to the data that results from any measurement process.
Take advantage of data360r certainly, but keep a wary eye on data from any source.