Education Week - April 19, 2017 - 8
DIGITAL DIRECTIONS > Tracking news and ideas in educational technology
Algorithmic Bias a Rising Concern for K-12 Ed-Tech Field
RAND study shows
tech industry trends
By Benjamin Herold
& Sarah Schwartz
From criminal sentencing to
credit scores, algorithms and artificial intelligence increasingly make
high-stakes decisions that have big
implications for people's freedom,
privacy, and access to opportunity.
Despite the almost-blind faith
people can put in such "artificial
agents," it's no secret that they are
often biased, according to a report
from the RAND Corp. that has implications for education.
More than ever, RAND researchers Osonde Osoba and Bill Welser
said in an interview, it's important
to raise awareness about the role
that algorithms play and to push
for a public accounting of their impact-particularly in areas that involve the public interest, including
"For the longest time, any time
questions of bias came up, hard-core
researchers in artificial intelligence
and algorithms dismissed them because they were not 'engineering
concerns,' " Osoba said. "That was
OK for commercial toys, but the
moment the switch was made to applying algorithms to public-policy
systems, the issue of bias no longer
became a triviality."
The new RAND report, "An Intelligence in Our Image: The Risks of Bias
and Errors in Artificial Intelligence,"
does not focus on education. Instead,
the authors lay out examples such as
the algorithmic bias in criminal sentencing and the problems with Tay, a
chatbot developed by Microsoft that
was supposed to learn the art of conversation by interacting with Twitter
users-and quickly began spewing
racist and vulgar hate speech.
Artificial agents can process the immense streams of data now running
through society in ways that humans
can't, making them a necessary tool
for modern society, the RAND researchers write. But too often, they
say, the public ascribes objectivity and
neutrality to algorithms and artificial
intelligence, even though most function as a "black box" and some have
been shown to result in different outcomes for different groups of people.
Origins of Digital Bias
Where does such bias come from?
The individual humans who program the artificial agents, who may
have biases they are not even aware
of; a pool of computer and data scientists that is far less diverse than
the populations their products eventually affect; and biases in the data
that are used to train the artificial
agents to "learn" by finding patterns, RAND concluded.
All those issues are present in the
One example: the growing field of
software programs that rely on algorithms to choose what types of
instructional content and learning
experiences students have each day
in the classroom. Algorithm-driven
Pinpoints Top Priorities
Open vs. Proprietary Resources
The "2017 K-12 IT Leadership Survey Report" also highlights shifts in
other district technology priorities.
Improving mobile learning, which has
been listed consistently as a top issue
for district officials, notched the highest spot this year. Increasing broadband and network capacity, the No. 1
priority in 2016, took the second spot
this year. Rural districts in geographically isolated areas face an especially
great challenge, said Weeks.
Despite the recent scaling-up of the
open educational resources movement, the report found a "slight" shift
in district officials' preferences for proprietary digital resources. CoSN's survey shows a decrease in the percent
of respondents planning to evenly divide their use of OER and proprietary
digital materials within the next three
years, from 46 percent to 43 percent.
At the same time, the share of respon-
8 | EDUCATION WEEK | April 19, 2017 | www.edweek.org
The Consortium for School Networking conducted a survey of 495 school and district ed-tech leaders to
identify their top priorities and challenges for this year.
Broadband and network
Budget constraints and
lack of resources
Only 13 percent of IT
leaders report staffing is
matched to needs.
IT leaders said this was their
top challenge for the third year
in a row.
Efforts to improve mobile learning, boost broadband capacity, and
tighten cybersecurity are the three
top priorities in 2017 for educational
technology leaders in K-12 school systems, according to a new survey by the
Consortium for School Networking,
which represents school district chief
technology officers around the country. For the third year in a row, leaders
identified budget constraints as the
No. 1 challenge facing their districts.
Leaders' top priorities reflect the
increasing digitization of the learning environment, said Tracy Weeks,
the executive director of the State
Educational Technology Directors Association. "We want to make sure that
students have access to the content,
to [technology] tools, both inside and
outside the school building."
CoSN has tracked districts' tech
priorities via its IT Leadership survey since 2013. This marks the first
year that cybersecurity has ranked in
the top three, according to the report.
Sixty-one percent of district technology leaders reported the concern as
more important than it was last year,
and 30 percent said it was "much
Experts point out that cybersecurity is a growing concern as criminals
increasingly target schools. A recent
study of IT infrastructure in several
industries by BitSight, a security ratings company, found that K-12 and
higher education experienced the
highest rate of ransomware, an approach in which hackers insert a virus
into a computer system and then ask
for ransom payments to get rid of the
virus. (See Education Week, Jan. 11,
2017.) The CoSN report also points to
a recent warning from the IRS about
phishing scams-which are attempts
by scammers to get personal information such as bank-account numbers-
targeting school districts.
In trying to help districts, CoSN
has developed a cybersecurity selfassessment and planning template
for IT leaders looking for a strategy to
address such issues. "Cybercriminals
are getting very sophisticated," said
Keith Krueger, the CEO of CoSN.
algorithms will never replace a
human teacher's evaluation.
"It's not to say that you can't use
artificial agents to help you identify
particular potential gaps in instruction and learning," Luftglass said.
The ed-tech product-development
process provides opportunities to
detect bias that are unavailable to
companies in other industries, said
Bridget Foster, the senior vice president for the Education Technology
Industry Network, a division of the
Software & Information Industry
Association. "In education, developers are right there in the school, in
the classroom, working with educators," she said.
Foster's organization recommends
that ed-tech companies figure out
how to mitigate bias from the early
planning and development stages
Welser said it's too early to try to
regulate the field or mandate bias
testing across the board. But, he argued, it is time for a conversation to
begin in K-12 about how to address
the potential biases in algorithms.
WHAT'S ON SCHOOL TECH ADMINISTRATORS' MINDS
By Sarah Schwartz
tools are also use by some districts to
provide career and college guidance
and to hire teachers.
What if such tools are biased
against students of color, or students
with special needs? How would educators, parents, and students even
Such questions are both realistic
and important for the field to be
asking, Osoba and Welser said.
"Educators need to not cede complete control to the computer," Welser
said. That means being aware which
products used in the classroom,
school, or district rely on algorithms
and artificial intelligence to make decisions; understanding what decisions
they are making; and paying attention to how different groups of students are experiencing the products.
Maribeth Luftglass, the assistant
superintendent for information
technology for the Fairfax County,
Va., schools, said it is the district's
responsibility to make sure digital
tools driven by algorithms remain
bias-free. When it comes to assessing students, she said, adaptive
SOURCE: Consortium for School Networking
dents who predicted use of proprietary
materials that are "only" digital increased, from 31 percent to 36 percent.
Krueger said that while there has
been some "overhype" around OER,
districts are still interested in the free
resources. Seventy-nine percent of
survey respondents, for instance, indicated that OER was a part of their
districts' digital content strategies.
OER is an important piece of curriculum, said Krueger, but it's unlikely to
replace all proprietary digital content
in the near future.
BYOD Loses Ground
were at their least popular in the
survey's history, in last place on the
priorities list. Though the proportion
of districts with fully implemented
BYOD programs increased to 24
percent this year, from 16 percent in
2016, the percent of districts with no
interest in BYOD reached a high of
"Device prices have dramatically
come down, and so you see much
lower [priced] options-Chromebooks and things like that-that
[have] changed the economics for
school districts," said Krueger. He
predicts that districts will gradually move to a dual model, in which
students are invited to use their own
devices and the district provides devices for those who are not able to
bring their own.
Henry County Schools is implementing a version of this dual
model. The 42,000-student dis-
trict in Georgia has emphasized
BYOD over the past several years,
said Brian Blanton, the assistant
superintendent for technology
services. Next year, a local option
sales tax will fund a 1-to-1 program in the district, but students
will still have the choice to bring
their own devices.
Budget constraints were identified by district officials as the top
challenge in the report, for the
third year in a row. Though a majority of leaders, 59 percent, said
their technology budgets remained
the same over the past year, more
district officials are reporting decreases in their technology budgets. Since 2015, the percent of districts reporting budget decreases
has gone up each year.