Here are a few of the data stories that caught my eye this week.
Another acquisition for AVOS
Last month, news broke that Yahoo had sold the popular bookmarking service Delicious to YouTube founders Chad Hurley and Steve Chen. Hurley and Chen have founded a new company AVOS, and while details are still fuzzy exactly what Delicious will look like under new ownership, that picture became a little clearer this week when AVOS announced another acquisition this week, this time purchasing the social media analytics startup Tap11.
“Our vision is to create the world’s best platform for users to save, share, and discover new content,” said Hurley in a statement on AVOS blog. “With the acquisition of Tap11, we will be able to provide consumer and enterprise users with powerful tools to publish and analyze their links’ impact in real-time.”
With the Delicious and Tap11 tools in its toolbox, AVOS could build a sophisticated system for recommending content and monitoring sentiment.
Data and mathematical intimidation
The idea of mathematical analysis being misused and misconstrued is hardly new, but as data-driven decision-making moves into new sectors, there are likely to be new controversies over the ways in which math and data are wielded. That’s certainly the case with a string of recent stories written by The LA Times in which the newspaper has constructed a mathematical model to rate teachers’ impact on their students. The model uses students’ test scores to devise a teacher’s “value-add.” The analysis compares a student’s performance on tests with their prior performance, and that difference — for better or worse or the same — is attributed to the teacher.
Teachers have balked at the method in part because The LA Times has published teachers’ names and scores. But now some mathematicians are pushing back as well. An editorial written by John Ewing, president of Math for America, traces the history of the value-added systems and looks at several of the problems with these models (and with sweeping judgments made on standardized test scores).
Value-added modeling is promoted because it has the right pedigree — because it is based on “sophisticated mathematics.” As a consequence, mathematics that ought to be used to illuminate ends up being used to intimidate. When that happens, mathematicians have a responsibility to speak out.
Data and scientific storytelling
Scientific papers are stories that persuade with data. That was the topic of a very interesting presentation by Anita de Waard, Disruptive Technologies Director at Elsevier Labs, at the recent Harvard Digital Scholarship Summit. Her talk provided a story analysis of scientific text, looking at the ways in which citations create facts and how linked data can help better support this knowledge (or story) creation process.
The research paper isn’t going away any time soon, de Waard argues, but she does point to several other ways in which the citation, review and publication processes of scientific papers can be improved. Ideas include a science data app store and executable scientific papers.
The slides from her talk are embedded below
Got data news?
Feel free to email me.