Predictive Analytics from Childhood to Higher Ed
Low cognitive skills at age 3 were found to predict the likelihood that a person would commit a crime or rely on social welfare services later in life. A team of researchers studied a cohort of 1000 children over a 35 year period to determine who might struggle later in life and identify potential interventions. In response to concerns that this data could be used to predict criminal behavior, Professor Avshalom Caspi from King's College London, one of the lead researchers said, “I hope what our study does does not feed into prejudice. I hope that our research will create the public compassion and political will to intervene with children and more importantly offer services to families of children so they can get a better start in life.”
EdTech magazine listed predictive analytics as the #1 trend in higher education. Colleges are using predictive analytics to determine which students are most likely to perform well and graduate. As reported by Business Insider in October, some universities rank students on a 0-10 scale using demographic factors that also predict whether the student will choose to enroll if accepted. A report published by New America finds that a growing number of colleges are using predictive analytics for targeted student advising, enrollment management, and adaptive learning. The authors of the report caution, however, that predictive systems may increase the likelihood of discrimination and labeling and encourage colleges to make the decision processes transparent. MIT professors advocate for an academic ‘Moneyball,’ the use of predictive analytics for hiring and personnel decisions, although others contest the use of algorithmic decision-making to measure scholarly potential.
Eric Westervelt of NPR examined personalized interventions made possible by student data analysis. Personalized emails from professors were found to improve retention levels. Westervelt compared “autopsy data”–static spreadsheets that may be circulated to limited administrative personnel–with data tools that provide more timely feedback that aggregates a student’s demographic, past performance, and current performance data. This is a shift from how colleges previously used data – retroactively analyzing data to develop reports or using it for alumni fundraising, marketing, and reports for the board of directors.
Lingering Digital Divide
In an article for MIT Technical Review, David Talbot explores the reason why the digital divide still lingers. The article calls attention to how data showing that most have access to wifi fails to tell the full story. Using the example of data on households with incomes under $20,000 in Cleveland, of which 58% have neither home broadband or mobile Internet access, the article reveals the way that the gap in access continues to disadvantage children from poor families. The article describes a student who was assigned online math problems for homework but did not have access to the Internet at home. Although there are attempts by government municipalities and service providers to increase access for these families, Talbot argues that these attempted solutions are inadequate.
Findings around the social and cognitive contexts of the digital divide complement Talbot’s work. Ellen Helsper of London School of Economics and Political Science finds that social exclusion, rather than digital exclusion, should be taken as a starting point when addressing digital inequality (starts on p.175). Rebecca Eynon and Anne Geniets of the University of Oxford similarly find that limited support networks, lack of experience, and broader social networks contribute to the digital divide.Happenings. A text messaging campaign seeks to help disadvantaged students meet college application deadlines and improve their decision-making options. A study by the Regional Educational Laboratory (REL) Northwest of over a million students enrolled in Washington State high schools finds that improving content mastery for Spanish-speaking students in earlier grades can have positive impacts on their enrollment in advanced courses in high school and college. Researchers found that while Spanish-speaking students earned lower grades and enrolled in fewer advanced courses than English-only speakers, when compared using GPAs from the previous year, differences in course enrollment disappear. A new study used smartphone GPS data, survey responses, and administrative data to examine racial disparity in exposure to areas prone to violence in urban youth. The National Center for Education Statistics finds that in 2014-2015, 20% of U.S. students surveyed were bullied, of which 11% were bullied online. This article debates how funding and money spent on schools impacts school performance, citing a lack of information about how and what money is spent on in studies showing an association between increased funding and performance.
The New York Times reports that violence against disabled persons usually goes unreported. Data about corporal punishment has spurred a Houston Congressman to file a federal bill to ban physical punishment in schools. Currently, corporal punishment is explicitly illegal in only 28 states (children of color and children with disabilities are disproportionately abused). The Houston Chronicle investigated Houston Independent School District and found that low enrollment numbers of special education students are due to deliberate discouragement and delay of evaluations to pursue goals that have kept services away from thousands of children with disabilities. A study released last month from UCSF and Johns Hopkins finds that the percentage of medical students with disabilities is higher than thought because many students choose not to disclose non-apparent disabilities due to fear of stigma, concerns over residency placement, and privacy. The U.S. Department of Education’s Office for Civil Rights has released three new guidance documents for protection for students with disabilities, referencing ways in which data has informed the new guidance.A Virginia legislator has filed a bill to regulate transgender people’s use of bathrooms in schools, highway rest stops, and other government buildings, requiring a provision to “provide physical privacy from members of the opposite sex.” Laurie Penny challenges the “generation snowflake” rhetoric that has been leveled at organizations like Black Lives Matter, tracing its origins from Fight Club to Tumblr: “language of oppression is not about sensitivity and victimhood...it is an expression of strength and knowledge through shared suffering, a cry of rage and outrage that contains its own demand for change.”
Understanding the Actual Rules and the Unofficial Rules for Youth on Instagram
When it comes to figuring out the rules for Instagram posts, teenagers and their parents both have a lot of learning to do. The actual rules for Instagram posts– that is, the company’s Terms of Service– are difficult to parse. According to “Growing Up Digital,” a report released earlier this month from the U.K. Children’s Commissioner’s Growing Up Digital Taskforce, “only a [person with a] postgraduate [education] could be expected to understand” Instagram’s terms and conditions. While some parents have postgraduate degrees, not all do– and kids almost never do– thus a significant number of Instagram users are not able to understand the terms of the site. When an attorney translated these terms of service into plain language, teens were startled to learn about the scope of Instagram’s powers, including the power to “send you adverts connected to your interests which we are monitoring. You cannot stop us doing this and it will not always be obvious that it is an advert.” The Taskforce has proposed several reforms for the UK to address the difficulty of understanding social media terms of service and related digital challenges, including the creation of a “compulsory ‘digital citizenship’ program for kids aged 4 to 14, which would be led by older kids, and to a lesser degree, teachers."
Teens will likely make great instructors for their younger colleagues, as they often have complex and thoughtful systems for navigating certain aspects of social media and other parts of digital life. As a U.S.-based educational consultant recently explained in the New York Times: “While parents sometimes impose rules for using social media on their kids, the most important rules are those that children create for themselves.” These rules include: “Be careful which vacation photos you share so you don’t brag. It’s O.K. to post photos from a fun event, but not too many.” While “these often unspoken rules can be dizzying,” these and other youth-created norms have considerable value for positively shaping online culture for kids and teens, as well as informing the digital tool-kits of the parents, teachers, and other trusted adults who care for and mentor them. (Further reading: an article out this week from Professor Sonia Livingstone et al. offers new insights into the role that parents can play in children’s digital lives, finding that there are “2 parental mediation strategies [for youth digital participation]. Enabling mediation is associated with increased online opportunities but also risks. This strategy incorporates safety efforts, responds to child agency, and is employed when the parent or child is relatively digitally skilled . . . Restrictive mediation is associated with fewer online risks but at the cost of opportunities . . . It is favored when parent or child digital skills are lower, potentially keeping vulnerable children safe yet undermining their digital inclusion.”)
Happenings. New York City has a new “literacy pilot program” where teachers run after-school reading clubs in homeless shelters to connect with kids there; data will be gathered to determine “whether it [the program] has any impact on students' attendance or reading performance.” An op-ed last month in Teen Vogue offers teens a blueprint for identifying and combatting the “President-elect[‘s] . . . deliberate attempt to destabilize journalism as a check on the power of government” through the dissemination of fake facts and the rejection of real ones, including “fact-checking every Trump statement you read, every headline you share or even relay to a friend over coffee. If you find factual inaccuracies in an article, send an email to the editor, and explain how things should have been clearer. Inform yourself what outlets are trustworthy and which aren’t. If you need extra help, seek out a browser extension that flags misleading sites or print out a list of fake outlets, such as the one by communications professor Melissa Zimdars, and tape it to your laptop.” (A recent thought piece in Quartz discusses this op-ed, concluding that “[t]he feminist-blog movement, and the women’s media revolution that followed, has trained the exact press corps we need for this moment in history.”) Teens are receiving media literacy instruction in some classrooms too, such as a nationally-recognized program at Concord High School in Concord, NH; according to a Dartmouth College government professor, “Trump can create a kind of informational nihilism where people throw up their hands and say, ‘I don’t know what to believe,’” a prospect that seems to make a skill-set for analyzing information veracity increasingly crucial for young people to learn. But danah boyd of Data & Society cautions that media literacy training could backfire “because [it] fail[s] to take into consideration the cultural context of information consumption that we’ve created over the last thirty years.” Last month, the U.S. Department of Education hosted an “Innovators’ Briefing” for select schools that have committed to “integrate technology across [their] curriculum.” Pepperdine University has new funding from the National Science Foundation’s Advancing Informal STEM Learning (AISL) program to “establish a network of 12 makerspace clubs in the United States, Europe and Africa.” The Berkman Klein Center is looking for summer “Berkterns” for the Youth and Media team and other projects; applications are due on February 13, 2017.
Monica Bulger: Monica Bulger is a Researcher in the Enabling Connected Learning initiative at the Data & Society Research Institute where she studies data use in education, focusing on personalized learning, privacy, equity, and digital literacy. She is a research consultant for UNICEF and serves on the International Advisory Board for Global Kids Online, a collaboration between UNICEF, LSE, and international partners to study the risks and opportunities of children’s use of the internet. Her first experience of education technology was Speak & Spell, which seemed uninspiring until E.T. demonstrated its extraordinary possibilities.
Mikaela Pitcan: Mikaela Pitcan is a Research Analyst at the Data & Society Research Institute, where she researches and writes about the impact of technology on learning, the ways in which educators and students interact with technology, and the intersection of race and gender within schooling. She is also a 4th year Counseling Psychology Doctoral Candidate. Ms. Pitcan's first experience with education technology was being mandated by her father to spend an hour every week-day playing "Where in the World is Carmen San Diego" after her family bought their first home computer. She was 7 years old but her father hoped it would give her the upper hand in the Geographic Bee competition at school. Unfortunately, she lost to her older brother every single year.
Leah Plunkett: Leah A. Plunkett is a Fellow with the Youth and Media team at the Berkman Klein Center, where she researches and writes about digital citizenship with a focus on student privacy. She is also an Associate Professor & Director of Academic Success at University of New Hampshire School of Law. Ms. Plunkett's first experience with connected ed tech was playing "Six Degrees of Kevin Bacon" online in her high school library when she was supposed to be doing research for history class. At that time, she and her classmates didn't even have email accounts, but they soon had a lot of knowledge about Kevin Bacon’s social network.
Alba Hancock: Alba Sophia Hancock is a Program Coordinator at the Berkman Klein Center, where she works on digital security and privacy law research, in addition to supporting the global Network of Internet & Society Centers, a particularly meaningful project for her as a proud third culture kid. Her background is in International Relations and Middle Eastern Studies, with a focus on using statistical methods to tease out the impact of information and communication technologies on civil resistance. Ms. Hancock’s first experience with edtech was receiving a beloved Tamagotchi when she was six years old. Having a needy pocket-sized digital pet taught her principles of responsibility and service to a greater cause that serve her to this day.