On subjective data, why datasets should expire, & data sabotage.

Expert says technology deployed is based on outdated science and therefore is unreliable

More companies are using automated job screening systems to vet candidates, forcing jobseekers to learn new and absurd tricks to have their résumés seen by a human.

Guest post by Ken Mathis, Simmons University Library No library has a perfect system of organization and occasional checks of the stacks are necessary. It's important to check on how the collection has grown and understand when it needs to...

On over- and underestimating how many of a minority group exist in a larger mixed maj/min group.

Financial companies often use data on borrowers' higher education to determine access to credit and the price of consumer financial products. And those education data can lead to redlining, a form of discrimination against borrowers who attended community colleges, historically black colleges and universities, or Hispanic-serving institutions. That was the key finding of a new report from the Student Borrower Protection Center, a nonprofit group led by Seth Frotman, the former student loan ombudsman for the federal Consumer Financial Protection Bureau. For example, the group said, borrowers who take out student loans may pay a penalty for attending a community college. "Wells Fargo charges a hypothetical community college borrower an additional $1,134 on a $10,000 loan when compared to a similarly situated borrower enrolled at a four-year college," according to the report. Likewise, borrowers who refinance student loans with a company that uses education data may pay a penalty for having attended an HBCU. "When refinancing with Upstart, a hypothetical Howard University graduate is charged nearly $3,499 more over the life of a five-year loan than a similarly situated [New York University] graduate," the group said. Upstart responded with a statement, saying the company conducts rigorous, regular fair-lending tests across millions of borrowers, in collaboration with the CFPB. "According to the CFPB, Upstart’s AI model increases access to credit across all tested race, ethnicity and gender segments by 23-29 percent while also decreasing average rates by 15-17 percent," Upstart said. "Too many people are unfairly disadvantaged by traditional lending models primarily based on credit score and income, and solving that requires considering more data." The report called on federal and state regulators to enhance oversight of the financial sector's use of education data. "The use of education data in underwriting could charge borrowers more for a loan simply for choosing the most accessible path for pursuing the American Dream," the center said. "Is this what is meant by a mission of ‘innovation’? Access to credit should not simply mean ‘more people getting more loans.’ It is imperative to examine the variance in the cost of those loans. Otherwise, expanded access to credit will not expand equity."

Hosted on the Open Science Framework

During a trip to Mexico to visit family, writer Myriam Gurba took “American Dirt,” a novel about immigration and cartel violence that was being touted as one of the biggest U.S. releases of 2020. The writer was of mostly white descent, and Gurba felt the book didn't ring true.

The clumsy high-profile rollout of the polarizing novel points to a larger issue concerning how new books are promoted.

In a growing list of criminal cases, attorneys for Facebook, Twitter and other social...

Addressing inequality requires more than consciousness raising.

A deletion battle over a black female chemist's biography has put Wikipedia's gender gap in the spotlight

If a hacker knows your interests and demographic information, it becomes much easier to target you with scams across the board. And that's just what happened recently when hackers pilfered data from a specific audience of internet users. Unlike previous campaigns, this one targets a vulnerable population with ruthless efficiency. Is this a new low for data theft? If not, it's definitely a close call.

Discovery and Web Services Librarian, University of Alberta

Understanding how race is historically and structurally built into the workplace

“The larger point is that it’s unaccountable and opaque, and Apple doesn’t really care.”

OA ebook!

It won’t end employment discrimination. In fact, it could make it worse.