The U.S. Department of Education has increasingly encouraged and funded states to collect and analyze information about students: grades, state test scores, attendance, behavior, lateness, graduation rates and school climate measures like surveys of student engagement.
The argument in favor of all this is that the more we know about how students are doing, the better we can target instruction and other interventions. And sharing that information with parents and the community at large is crucial. It can motivate big changes.
what might be lost when schools focus too much on data. Here are five arguments against the excesses of data-driven instruction.
The National Education Policy Center releases annual reports on commercialization and marketing in public schools. In its most recent report in May, researchers there raised concerns about targeted marketing to students using computers for schoolwork and homework. Companies like Google pledge not to track the content of schoolwork for the purposes of advertising. But in reality these boundaries can be a lot more porous. For example, a high school student profiled in the NEPC report often consulted commercial programs like dictionary.com and Sparknotes: “Once when she had been looking at shoes, she mentioned, an ad for shoes appeared in the middle of a Sparknotes chapter summary.”
4) Missing What Data Can’t Capture
Computer systems are most comfortable recording and analyzing quantifiable, structured data. The number of absences in a semester, say; or a three-digit score on a multiple-choice test that can be graded by machine, where every question has just one right answer.
5) Exposing Students’ “Permanent Records”
In the past few years several states have passed laws banning employers from looking at the credit reports of job applicants. Employers want people who are reliable and responsible. But privacy advocates argue that a past medical issue or even a bankruptcy shouldn’t unfairly dun a person who needs a fresh start.
There are two types of Lexile measures: a person’s reading ability and the text’s difficulty. Students who are tested against state standards receive a Lexile reader measure from the Kansas Reading Assessment. Books and other texts receive a Lexile text measure from a MetaMetrics software tool called the Lexile Analyzer, which describes the book’s reading demand or complexity. When used together, the two measures are intended to help match a reader with reading material that is at an appropriate difficulty or will at least help give an idea of how well a reader should comprehend text. The reader should encounter some level of difficulty with the text, but not enough to get frustrated. The Lexile reader measure is used to monitor reader progress.
My note: is this another way / attempt to replace humans as educators? Or it is a supplemental approach to improve students’ reading abilities.
At a Ford Foundation conference dubbed Fairness by Design, officials, academics and advocates discussed how to address the problem of encoding human bias in algorithmic analysis. The White House recently issued a report on the topic to accelerate research into the issue.
U.S. CTO Megan Smith said the government has been “creating a seat for these techies,” but that training future generations of data scientists to tackle these issues depends on what we do today. “It’s how did we teach our children?” she said. “Why don’t we teach math and science the way we teach P.E. and art and music and make it fun?”
“Ethics is not just an elective, but some portion of the main core curriculum.”
The U.S. Bureau of Labor Statistics backs that up, predicting that employment of statisticians will grow 34 percent between 2014 and 2024. Not surprisingly, the bureau notes, that is “much faster than the average for all occupations.”
ACRL e-Learning webcast series: Learning Analytics – Strategies for Optimizing Student Data on Your Campus
This three-part webinar series, co-sponsored by the ACRL Value of Academic Libraries Committee, the Student Learning and Information Committee, and the ACRL Instruction Section, will explore the advantages and opportunities of learning analytics as a tool which uses student data to demonstrate library impact and to identify learning weaknesses. How can librarians initiate learning analytics initiatives on their campuses and contribute to existing collaborations? The first webinar will provide an introduction to learning analytics and an overview of important issues. The second will focus on privacy issues and other ethical considerations as well as responsible practice, and the third will include a panel of librarians who are successfully using learning analytics on their campuses.
Webcast One: Learning Analytics and the Academic Library: The State of the Art and the Art of Connecting the Library with Campus Initiatives
March 29, 2016
Learning analytics are used nationwide to augment student success initiatives as well as bolster other institutional priorities. As a key aspect of educational reform and institutional improvement, learning analytics are essential to defining the value of higher education, and academic librarians can be both of great service to and well served by institutional learning analytics teams. In addition, librarians who seek to demonstrate, articulate, and grow the value of academic libraries should become more aware of how they can dovetail their efforts with institutional learning analytics projects. However, all too often, academic librarians are not asked to be part of initial learning analytics teams on their campuses, despite the benefits of library inclusion in these efforts. Librarians can counteract this trend by being conversant in learning analytics goals, advantages/disadvantages, and challenges as well as aware of existing examples of library successes in learning analytics projects.
Learn about the state of the art in learning analytics in higher education with an emphasis on 1) current models, 2) best practices, 3) ethics, privacy, and other difficult issues. The webcast will also focus on current academic library projects and successes in gaining access to and inclusion in learning analytics initiatives on their campus. Benefit from the inclusion of a “short list” of must-read resources as well as a clearly defined list of ways in which librarians can leverage their skills to be both contributing members of learning analytics teams, suitable for use in advocating on their campuses.
student’s opinion of this process
benefits: self-assessment, personal learning, empwerment
analytics and data privacy – students are OK with harvesting the data (only 6% not happy)
8 in 10 are interested in personal dashboard, which will help them perform
Big Mother vs Big Brother: creepy vs helpful. tracking classes, helpful, out of class (where on campus, social media etc) is creepy. 87% see that having access to their data is positive
recognize metrics, assessment, analytics, data. visualization, data literacy, data science, interpretation
INSTRUCTION DEPARTMENT – N.B.
determine who is the key leader: director of institutional research, president, CIO
who does analyics services: institutional research, information technology, dedicated center
analytic maturity: data drivin, decision making culture; senior leadership commitment,; policy supporting (data ollection, accsess, use): data efficacy; investment and resourcefs; staffing; technical infrastrcture; information technology interaction
student success maturity: senior leader commited; fudning of student success efforts; mechanism for making student success decisions; interdepart collaboration; undrestanding of students success goals; advising and student support ability; policies; information systems
developing learning analytics strategy
understand institutional challenges; identify stakeholders; identify inhibitors/challenges; consider tools; scan the environment and see what other done; develop a plan; communicate the plan to stakeholders; start small and build
ways librarians can help
idenfify institu partners; be the partners; hone relevant learning analytics; participate in institutional analytics; identify questions and problems; access and work to improve institu culture; volunteer to be early adopters;
questions to ask: environmental scanning
do we have a learning analytics system? does our culture support? leaders present? stakeholders need to know?
questions to ask: Data
questions to ask: Library role
learning analytics & the academic library: the state of the art of connecting the library with campus initiatives
7 Things You Should Know About First-Generation Learning Analytics. Published:
causation versus correlation studies. speakers claims that it is difficult to establish causation argument. institutions try to predict as accurately as possible via correlation, versus “if you do that it will happen what.”
Instruction and Liaison Librarian, University of Northern Iowa
games and gamification. the semantics are important. using the right terms can be crucial in the next several years.
gamification for the enthusiasm. credit course with buffet. the pper-to-peer is very important
affordability; east to use; speed to create.
assessment. if you want heavy duty, SPSS kind of assessment, use polldaddy or polleverywhere.
Kahoot only Youtube, does not allow to upload own video or use Kaltura AKA Medispace, text versus multimedia
Kahoot is replacing Voicethread at K12, use the wave
Kahoot allows to share the quizzes and surveys
Kahoot is not about assessment, it is not about drilling knowledge, it is about conversation starter. why do we read an article? there is no shame in wrong answer.
the carrot: when they reach the 1000 points, they can leave the class
Kahoot music can be turned off, how short, the answers are limited like in Twitter
screenshot their final score and reach 80%
gravity is hard, scatter start with. auditory output
1st day is Kahoot, second day is Team challange and test
embed across the curriculum
gaming toolkit for campus
what to take home: have students facing students from differnt library
In the age of Big Data, there is an abundance of free or cheap data sources available to libraries about their users’ behavior across the many components that make up their web presence. Data from vendors, data from Google Analytics or other third-party tracking software, and data from user testing are all things libraries have access to at little or no cost. However, just like many students can become overloaded when they do not know how to navigate the many information sources available to them, many libraries can become overloaded by the continuous stream of data pouring in from these sources. This session will aim to help librarians understand 1) what sorts of data their library already has (or easily could have) access to about how their users use their various web tools, 2) what that data can and cannot tell them, and 3) how to use the datasets they are collecting in a holistic manner to help them make design decisions. The presentation will feature examples from the presenters’ own experience of incorporating user data in decisions related to design the Bethel University Libraries’ web presence.
data tools: user testing, google analytics, click trakcer vendor data
user testing, free, no visualization, cross-domain, easy to use, requires scripts
qualitative q/s : why people do what they do and how will users think about your content
3 versions: variables: options on book search and order/wording of the sections in the articles tab
Findings: big difference between tabs versus single-page. Lil difference btw single-page options. Take-aways it won’t tell how to fix the problem, how to be empathetic how the user is using the page
Like to do in the future: FAQ and Chat. Problem: low use. Question how to make it be used (see PPT details)
Crazy Egg – Click Trackers. not a free tool, lowest tier, less $10/m.
see PPT for details>
interaction with the pates, clicks and scrollings
not easy to use, steep learning curve
“blob” GAnalytics recognize the three different domains that r clicked through as one.
vendor data: springshare
chat and FAQ
is there a dashboard tool that can combine all these tools?
optimal workshop: reframe, but it is more about qualitative data.
how long does it take to build this? about two years in general, but in the last 6 months focused.
the study of digital tools linked to these new literacies is absolutely necessary, particularly because Web 2.0 allow users to interact and cooperate together as content creators in a virtual community. Although this concept may suggest a new version of the World Wide Web (WWW), it really does not refer to an update of the technical features, but rather to the changes concerning the use and interaction through the Web.
IBM Predictive Analytics Solutions for Education can help you improve outcomes
Your data is a record of what’s already happened. But did you know that the same data—combined with the right analytical tools—can give you a forward-looking view of your situation, along with recommendations for decision making?
Read this white paperto learn how predictive analytics can help your institution address a range of challenges, from increasing graduation rates student by student to optimizing recruitment, fundraising and the performance measures that matter most.
real-time impact on curriculum structure, instruction delivery and student learning, permitting change and improvement. It can also provide insight into important trends that affect present and future resource needs.
Big Data: Traditionally described as high-volume, high-velocity and high-variety information.
Learning or Data Analytics: The measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs.
Educational Data Mining: The techniques, tools and research designed for automatically extracting meaning from large repositories of data generated by or related to people’s learning activities in educational settings.
Predictive Analytics: Algorithms that help analysts predict behavior or events based on data.
Predictive Modeling: The process of creating, testing and validating a model to best predict the probability of an outcome.
Data analytics, or the measurement, collection, analysis and reporting of data, is driving decisionmaking in many institutions. However, because of the unique nature of each district’s or college’s data needs, many are building their own solutions.
For example, in 2014 the nonprofit company inBloom, Inc., backed by $100 million from the Gates Foundation and the Carnegie Foundation for the Advancement of Teaching, closed its doors amid controversy regarding its plan to store, clean and aggregate a range of student information for states and districts and then make the data available to district-approved third parties to develop tools and dashboards so the data could be used by classroom educators.22
Tips for Student Data Privacy
Know the Laws and Regulations
There are many regulations on the books intended to protect student privacy and safety: the Family Educational Rights and Privacy Act (FERPA), the Protection of Pupil Rights Amendment (PPRA), the Children’s Internet Protection Act (CIPA), the Children’s Online Privacy Protection Act (COPPA) and the Health Insurance Portability and Accountability Act (HIPAA)
— as well as state, district and community laws. Because technology changes so rapidly, it is unlikely laws and regulations will keep pace with new data protection needs. Establish a committee to ascertain your institution’s level of understanding of and compliance with these laws, along with additional safeguard measures.
Make a Checklist Your institution’s privacy policies should cover security, user safety, communications, social media, access, identification rules, and intrusion detection and prevention.
Communicate, Communicate, Communicate
Students, staff, faculty and parents all need to know their rights and responsibilities regarding data privacy. Convey your technology plans, policies and requirements and then assess and re-communicate those throughout each year.
“Anything-as-a-Service” or “X-as-a-Service” solutions can help K-12 and higher education institutions cope with big data by offering storage, analytics capabilities and more. These include:
• Infrastructure-as-a-Service (IaaS): Providers offer cloud-based storage, similar to a campus storage area network (SAN)
• Platform-as-a-Service (PaaS): Opens up application platforms — as opposed to the applications themselves — so others can build their own applications
using underlying operating systems, data models and databases; pre-built application components and interfaces
• Software-as-a-Service (SaaS): The hosting of applications in the cloud
• Big-Data-as-a-Service (BDaaS): Mix all the above together, upscale the amount of data involved by an enormous amount and you’ve got BDaaS
Use accurate data correctly
Define goals and develop metrics
Eliminate silos, integrate data
Remember, intelligence is the goal
Maintain a robust, supportive enterprise infrastructure.
Prioritize student privacy
Develop bullet-proof data governance guidelines
Create a culture of collaboration and sharing, not compliance.
Tim Brugger (Big Data): In part because the world around us is becoming “connected” through a growing number of IoT sensors, mobile devices, and the world’s affinity for the Internet, the sheer volume of information available is already staggering.
Daniel B. Kline (endless payment): While subscriptions have always been a factor on the enterprise side of the software business, they’re now moving into the consumer end of things. The leader has been Microsoft (NASDAQ:MSFT), which has managed to move a large part of its Office customer base into a subscription model.