Chapter 5: Adaptive Learning Systems in Modern Classrooms

Simo Zarkovic

Introduction

In today’s modern educational systems, it is very common to have two to three dozen students in a classroom but only one teacher. In post-secondary institutions, this can easily turn into 200 to 300 students for every one professor. Class size is compounded in today’s classrooms, which are filled with diverse learners with a multitude of needs along with unique strengths and weaknesses. The complexity and range of student needs can result in an overworked teacher who gives teaching their best effort and still finds that some students fail to meet an acceptable level of concept mastery. Simply put, there are not enough teachers to meet the multitude of students’ needs so that every single student can succeed in every single classroom. In recent years, artificial intelligence (AI) has begun to take ever-stronger roots in education, giving rise to Adaptive Learning Systems (ALS). ALSs, created through the fusion of AI software with handheld, network-connected computers, aim to deliver personalized learning. These systems can “help teachers reallocate 20 to 30 percent of their time so they can focus more on student-centric activities such as building deeper one-on-one relationships, refining individual lesson plans, or providing real-time personalized feedback to students” (Microsoft [New Tab], 2018, p.5). These ALS educational technologies are synonymous with intelligent tutoring systems, student-centred learning, intelligent instructional design, and personalized learning applications. Marr’s [New Tab] (2018) article defines such systems as “digital platforms that use AI to provide learning, testing and feedback to students from pre-K to college level that gives them the challenges they are ready for, identifies gaps in knowledge and redirects to new topics when appropriate” (para. 3). In the current era of heightened calls for greater personalization and better results in education, “personalized learning applications are currently among the most heavily marketed, exciting and controversial applications of edtech” (Regan & Jesse [New Tab], 2018, p. 168).

Educational Ethical Issue

This chapter will discuss the ethical considerations of Adaptive Learning Systems (ALS) and their impact on students, teachers, and the merchants that provide them. The following question will be examined though consequentialist, deontological, and virtue ethics perspectives (Farrow, 2016), as well as my own perspective: should Adaptive Learning Systems (ALS) be implemented in modern classrooms?

Elevated Calls for Personalized Learning

Personalized learning — where students control pace, content, and assessment — was one of the two emerging findings in Microsoft’s (2018) report The Class of 2030 and Life-Ready Learning: The Technology Imperative [New Tab]. The report found that “nearly 70 percent of [2,000 surveyed] teachers cited time constraints as their biggest hurdle to providing more personalized content to their students” (Holzapfel [New Tab], 2018b, para. 14). Automated grading and personalized feedback are common features of ALS, so this two-pronged approach promises to provide more instructional time to teachers to aid students in places where their exceptionally smart digital tutors cannot. The calls for personalization do not stop there; tech companies or ‘philanthrocapitalists’ such as the Bill and Melinda Gates Foundation, Chan Zuckerberg Initiative, Dell, Hewlett, and Google.org emphasize the differences in ways students learn and the importance of flexible learning opportunities. These five companies use captivating statements such as “a truly transformative, personalized learning experience (Chan Zuckerberg), and real-time assessments for gauging student learning (Gates)” (Regan & Steeves [New Tab], 2019, para. 33), to describe ALSs.

Table 5.1 Examples of real-world applications of artificial intelligence (AI) in educational settings
Adaptive Learning System
(Parent Company)
Brief Marketed Description of ALS
iReady
(Curriculum Associates)
Delivers online lessons that provide tailored instruction and practice for each student to accelerate growth, while supporting teachers with in-the-moment resources for remediation and re-teaching (Curriculum Associates [New Tab], n.d., para. 4).
MATHia
(Carnegie Learning)
Using sophisticated AI technology to adapt at a very detailed, skill-by-skill level, MATHia personalizes the learning and keeps students engaged with customized just-in-time feedback and contextual hints, while providing you with all the real-time feedback and assessments you need to understand where your students are at and where they’re headed (Carnegie Learning [New Tab], n.d.a, para. 3).
Exact Path
(Edmentum)
Combines adaptive diagnostic assessments with individualized learning pathways to promote growth for K-12 grade students in math, reading, and language arts, as students receive a unique testing experience that precisely pinpoints their instructional level, strengths, and needs (Edmentum [New Tab], 2020, para. 2).
“Jill” Watson
(IBM)
A graduate-level teaching assistant who can hold office hours 24/7/365, where “she” spends her days helping students in the online M.S. in Computer Science program’s Knowledge-Based Artificial Intelligence course (Georgia Institute of Technology College of Computing [New Tab], n.d., para. 1).
Cognitive Immersive Room (IBM) An immersive classroom environment, where students feel as though they are in a restaurant in China, a garden, or a Tai Chi class, where they can practice speaking Mandarin with an AI chat agent through immersive technologies (IBM [New Tab], 2019, para. 1).

AI Policy in Education

No educational district wants to be left behind; leaders are on a constant pursuit to bring in new literature and reports, which likely shape their board’s learning directive for the next few years. Alberta Education’s (2018) Leadership Quality Standard expects K-12 leaders to embody visionary leadership and to lead a learning community. Similarly, boards of governors at both the University of Calgary and the University of Alberta support forward-thinking programs and collaborations to meet AI learning needs in society (Pascoe [New Tab], 2019). Additionally, in 2017, Canada’s federal government created a Pan-Canadian Artificial Intelligence Strategy [New Tab] — the world’s first national AI strategy. This led to conferences where, thus far, more than 150 researchers, thought leaders, and policy makers examined the social, economic, ethical, and legal implications of AI (Barron et al. [New Tab], 2019). One crucial observation from these gatherings is that many policymakers lack awareness of current AI capabilities and applications, and their associated policy implications. Despite the policy shortcomings with respect to AI, the participants proposed a general framework to guide policy development for public education and responsible innovation, including to:

  • Promote awareness of data protection rights and regulations among the general public;
  • increase the digital literacy of the public, particularly among traditionally marginalized and vulnerable populations;
  • provide government funding to incentivize companies to incorporate transparency into the design of their applications; and
  • encourage open source algorithms to mitigate inequality (Villeneuve et al. [New Tab], 2019, pg. 7).

Consequentialist Perspective

Easily interpreted, visual data is a part of every ALS, and these data sets are designed to help teachers recognize learning gaps. This newfound awareness is supposed to lead to greater efficiency when it comes to the time spent during an interaction between a teacher and their student. ALS can also be connected to personal mobile devices that students carry; AI can augment the physical world, overlaying the real environment with virtual information (Luckin et al. [New Tab], 2016). This augmented reality is designed to engage students as it moves lessons from hypothetical scenarios to real-life, and diverges from one-size-fits-all content delivery to a tailored and dynamic learning experience.

Deontological Perspective

Experienced teachers can quickly identify the skills and curricular knowledge that their students both possess and lack. From an equity perspective, teachers want to narrow the gaps as much as possible before moving to the next sequential outcome, and ALS can certainly help with that. Furthermore, “teachers will be able to record their observations of students — and benefit from the observations of other teachers” (CoSN [New Tab], 2018, p. 34) in their work to ensure the most efficient use of resources, both human and artificial.

Virtue Ethics Perspective

The shared narrative between people, schooling institutions, and government agencies seems to be that people “who are unfamiliar with the use of AI-driven technology, will not receive the same benefits as those who have adopted these tools” (Villeneuve et al. [New Tab], 2019, p. 9). It makes sense to employ ALS, which can automate grading and thus free teachers to forge deeper socio-emotional bonds with their students. Instructional flexibility offered by ALS can provide access for students to progress at their own pace; not only to catch up, but also to accelerate learning. An industry-sponsored white paper called The Equity Equation [New Tab] by McGraw-Hill showcases “institutions like Columbus State, Arizona State University and Triton Community College in Illinois, among others, which are improving educational equity by applying new learning methods and tools that adapt to individual student needs” (Neelakantan [New Tab], 2019, para. 4).

My Own Perspective

As teachers, students, and parents become aware of ALS, the pressure to purchase will likely increase, and the delay of implementation may result in frustration. I am surprised that the marketing teams of various ALS have not advertised as heavily in Canada, where they could leave the impressionable public feeling disadvantaged if their school boards do not acquire ‘the latest and greatest’ for their pupils. This will be reminiscent of computer purchases that occurred just before the turn of the millennium, with schools hurriedly purchasing both PCs and Macs to accommodate their student populations. The leadership of modern school districts must demonstrate that they are aware of powerful and influential corporations’ potential hidden intentions, and honour the expectation to disclose and justify their choice of one ALS vendor over others. Over the last two decades that I have spent in classrooms as a student and as a teacher, I have noticed significant amounts of technology brought in, with limited rationale provided about the purchasing decisions. I have encountered bulky desktops, then slim laptops, and, most recently, Smart Boards in classrooms, with little time allocated to learning about the rationale for the purchase, insufficient resources invested in preparing users, and inadequate discussion of concerns behind the proposed enhancements. As ALS arrive in classrooms, many teachers will need to be trained how to effectively incorporate them in teaching routines in order to gain “new ways of understanding and interacting with their students” (Microsoft [New Tab], 2018, p. 5). Otherwise, these costly systems will remain in boxes at the back of classrooms, as untrained teachers continue to teach, undisrupted, and without using the latest technology gadgets.

Benefits and Challenges of Implementation of Privacy Protection, Data Security, and Informed Consent in Adaptive Learning Systems

Predicting the future is challenging work, and this applies especially to the personalized predictive lessons generated in ALS immediately after students input their unique responses. The concept of accurate error diagnosis is fundamental to all successful tutoring, and ALS will only be as good as the size of a database that it is connected to (Ferster [New Tab], 2017). This “entails collection of more, and more granular, information about students, teachers, and families, as well as administrative details regarding the functioning of educational institutions” (Regan & Jesse [New Tab], 2018, p. 168). As ALS collect and analyze multiple streams of data in real time, “there is a real possibility of continuous improvement via multiple feedback loops that operate at different time scales — immediate to the student for the next problem, [and] to the teacher for the next day’s teaching” (Bienkowski et al. [New Tab], 2012, p. viii). Recent big data aggregator systems such as InBloom had “some 400 ‘optional fields’ that schools could choose to fill in and that included sensitive information such as disability status, social security numbers, family relationships, reasons for enrolment changes, and disciplinary actions” (Regan & Jesse [New Tab], 2018, p. 169). ALS databases need to strike a balance between not asking for too much data and going defunct — as in the case of InBloom — while also requesting as much data as possible in order to better personalize the next ALS-generated learning task.

Fuelled by renewable student populations, the market for student data is only projected to increase; “analysts forecast the Artificial Intelligence Market in the US Education Sector to grow at a CAGR (Compound Annual Growth Rate) of 47.77% during the period 2018-2022″ (Research and Markets [New Tab], 2018, para. 2). This equates to an incremental [four-year] growth of 253.79 million USD, with one of the key market drivers being “an increased adoption of ITS [Intelligent Tutoring Systems] . . . in the education sector” (TechNavio [New Tab], 2018, para. 2). As private student data transfers from schools to remote servers, “we need to be particularly careful about educational technologies which store and/or access information outside of Canada; these educational technologies are not always bound by Canadian law” (University of Victoria [New Tab], 2019, para. 2). This could lead to foreign corporations selling the data to other merchants or even to foreign governments nefariously targeting people by deeply digging into their personal data.

Research Data of Various Adaptive Learning Systems and How They Impact Instruction and Learning

Getting ALS raw data is a challenge. No peer-reviewed data exists, only industry-sponsored data. Additionally, “there are no established standards for describing or evaluating the extent to which a learning experience is personalized, and often the difference between responsiveness and adaptiveness is not accounted for in product descriptions” (Bulger [New Tab], 2016, p. 4). ALS companies such as iReady state that rigorous and scientific analysis is their priority, but this has been slow going, because the process requires “extensive data sharing, privacy safeguards, significant funding, and longstanding relationships with districts and schools” (McKinnon [New Tab], 2018, para. 16).

Table 5.2 Self-reported results of adaptive learning systems listed in Table 5.1
Adaptive Learning System
(Parent Company)
Excerpts from Self-Reported Data
iReady
(Curriculum Associates)
Students using iReady Personalized Instruction for an average of 45 minutes or more per subject per week for at least 18 weeks showed statistically significantly greater growth than the average student who did not receive iReady Personalized Instruction during the 2017–2018 school year (Curriculum Associates [New Tab], 2020, p. 2).
MATHia
(Carnegie Learning)
An independent study by the RAND Corporation and the U.S. Department of Education found that MATHia’s blended approach nearly doubled growth in performance on standardized tests in the second year of implementation (Carnegie Learning [New Tab], n.d.b, para. 1 )
Exact Path
(Edmentum)
Results indicated that use of Edmentum Exact Path is positively associated with student achievement outcomes in math, reading, and language arts. Statistically significant effects were found linking the amount of time spent on Exact Path and end-of-year diagnostic scores (Edmentum [New Tab], 2017, p. 2).
“Jill” Watson
(IBM)
Jill was a highly effective teaching assistant for students, answering questions with a 97% success rate. Out of 10,000 queries that require little thinking, Jill’s aim was to answer 40% of all these questions (Maderer [New Tab], 2016).
Cognitive Immersive Room (IBM) Acquiring a new language naturally, through cultural immersion, may be more effective than non-immersive practices. One of the biggest obstacles in learning a foreign language through immersion is students’ fears of being judged by native speakers (IBM [New Tab], 2019, para. 2). 


Consequentialist Perspective

According to Neelakantan [New Tab] (2019), “powered by advanced algorithms, adaptive learning technologies boost completion rates and give students confidence” (para. 1). Furthermore, mining and analyzing data gathered by an ALS “can analyze underlying patterns in order to predict student outcomes such as dropping out, needing extra help, or being capable of more demanding assignments” (West [New Tab], 2012, p. 2). Despite a teacher’s best efforts to provide formative feedback in lectures, and summative feedback on an occasional assignment or quiz, use of the ALS enables more frequent, real-time feedback to students on tasks and interactions in the system. While it would be unrealistic to expect a teacher of two to three dozen students to adjust a lesson based on each student’s needs, ALS makes this type of responsiveness possible using the data that is collected from students.

Deontological Perspective

Data, upon which the ALS relies, can be insufficient, meaning that unseen threats are pending to everyone who is overlooked. Students can also become overlooked and underserved by teachers who rely heavily on thinking that AI knows best, without using their own intelligence to think, probe, and teach. Lerman [New Tab] (2013) suggests that the “big data revolution may create new forms of inequality and subordination, and thus raise broad democracy concerns” (p. 60). Furthermore, continued differences in decisions about how best to hold ALS vendors accountable remain uncertain, especially when it comes to critical issues of data security and privacy protection. As Regan & Jesse [New Tab] (2018) point out, “one of the most problematic issues involves whether edtech companies should be able to use data generated by students’ use of their software programs to improve those programs” (p. 173). This supposed mutually beneficial practice blurs the lines on whether student learning profiles are used for a student’s own good or for corporate profit.

Virtue Ethics Perspective

Big data’s use of mathematical algorithms and artificial intelligence to make predictions about individuals based on their information and that of others raises questions about treating individuals as individuals fairly, accurately, and in ways they can understand (Citron & Pasquale [New Tab], 2014). More so, big data critics worry that “the world’s increasing ‘datafication’ ignores or even smothers the unquantifiable, immeasurable, ineffable parts of human experience.” (Lerman [New Tab], 2013, p. 56). Any teacher can attest that it takes attentive emotional intellect and relationships with students to read the cues that students put out, and that students’ learning is affected when they are in an escalated state of mind. Consequently, there has been a recent interest in supporting teachers to become better acquainted with Trauma Informed Practice (TIP) and student mental wellness.

My Own Perspective

Regan & Jesse [New Tab] (2018) explain that “a critical ethical concern raised with personalized learning is whether such programs constitute tracking and sorting of students that might be considered discriminatory” (p. 168), as in the 1950s, when children were divided by race, ethnicity, gender, and class. Some of these divisive factors might come back or can even be encouraged, especially in less socially progressive countries. Though such divisions are opposed in Western societies, the wealth gap seems to be widening (Litwin [New Tab], 2019). Data gathered by an ALS could serve as a foundation for parents or guardians of very affluent students to advocate for redirection of the school’s limited budget to fund elite, exclusionary classes for their children. ALS are often marketed as having the potential to accelerate brains, so influential parents within any particular school could demand a special settings class where “as the pace of change accelerates, learners will demand more ways to convert learning to earning” (Consortium for School Networking [New Tab], 2018, p. 8).

Benefits and Challenges of Implementation of Respect for Participant Autonomy and Independence Concerns When Using Adaptive Learning Systems

Through the use of big data and smart algorithms, ALS can be used to help teachers find their own blind spots and even reveal unorthodox thinking and teaching strategies for students. Students can link ALS profiles to their own cellular devices, which can then remind them of deadlines and offer opportunities to continue tutoring and learning at home. Students, teachers, smart machines, and software increasingly interact in new and deeper ways, and may be reshaping our brains in intended and unintended ways. KnowledgeWorks [New Tab] (2018) explains that “repeated use of Google Search has been shown to stimulate the use of short- over long-term memory in ways that may undermine critical thinking” (p. 12). Additionally, a continuous connection to smart devices can lead to mental exhaustion due to lack of downtime — we are wired and tired all the time (Brody, 2017). Both parents and school administration need to ensure that students and teachers have allocated time for themselves by respecting their right to disconnect from the non-stop digital realm. Furthermore, all parties need to find a balance when it comes to the wealth of data available to track various stages of student development; it is easy to lose sight of the big picture in daily performance metrics supplied by ALS. It is also important to note that the youngest, pre-kindergarten learners tend to learn better watching real, face-to-face events, since they have trouble transferring information from a screen to the real world (Troseth & Strouse [New Tab], 2017).

Avoiding Harm and Minimizing Risk to Educational Integrity When Using Adaptive Learning Systems

With so much attention being paid to improving the standard metrics of success in school, less attention may be paid to students’ psychological well-being and social health. However, educators should indeed keep student well-being in sharp focus, since “the strongest signal from [Microsoft’s] study was the need for teachers, schools, and school leaders to help students develop stronger social-emotional skills” (Microsoft [New Tab], 2018, p. 10). These skills help establish successful team collaborations and enable discourse plus discussion —not just online but in person. We need our students to be resilient to life’s many curveballs, to have the capacity to receive constructive criticism, and to come back stronger after a failure. Furthermore, the same report estimated that 30-40% of future jobs will require explicit social skills and emotional literacy (Microsoft [New Tab], 2018). According to Holzapfel (2018), “social-emotional skills provide students with the perspective and flexibility necessary to function at a high level even when faced with uncertainty, change, pressure, stress, and other work and life challenges” (p. 11). Enhanced socio-emotional capacity will come in handy within the tumultuous gig economy, since people will shift from having a single job to having short-term contracts with multiple employers. Social-emotional and other relational and well-being skills will not be evaluated by ALS, but they still need to be practiced and built, so that students can have happy, healthy, and well-rounded lives.

Consequentialist Perspective

Education is moving toward outcome-based and competency-based learning, meaning that students do not need to sit for hours listening to content being presented, but can instead optimize their time by focusing on mastering the cognitive skills and the socio-emotional and content knowledge they lack. When a student gets ill, connecting to their ALS account from home can enable them to engage with the curricular outcomes missed at school. More mature, part-time students can use their time to learn other skills, which are constantly fluctuating as workplace demands and employability skills change. In their report on the future of work and learning, D2L (2018) writes that the “constantly fluctuating skills market means employee skills are becoming outdated more rapidly and require ongoing training and development” (p. 11)

Deontological Perspective

Traumatic events like war conflict, poverty, family instability, homelessness, and a lack of opportunities frequently result in educational disruptions. ALS can help to mend the gaps in learning, so schools should budget funds to close those gaps for the students in need. The academic part of student well-being is quickly captured and reported, resulting in instantaneous hope and jubilation in those students. KnowledgeWorks [New Tab] (2018) explains that “as students gain the rights to own their own data, data asset advisors help students and their families manage, present and exchange data related to students’ learning, locations and device and platform usage” (p. 20).

Virtue Ethics Perspective

Personal autonomy deals with the extent of one’s freedom to make choices, and can be affected by the algorithms and intelligence that exist within ALS. When students spend a dozen years of schooling with ALS, these computers can become a long-lasting crutch for them to lean on. As students move on to the workforce, the validations that ALS provided them are gone, and they may feel lost without the devices that they grew up with. At the same time, teachers cannot rely too much on ALS, as these programs and algorithms lack empathic capacities, and therefore provoke feelings of dehumanization that endanger personal autonomy (Royakkers et al. [New Tab], 2018).

My Own Perspective

The recent COVID-19 pandemic has essentially disrupted every single school in the world for months. The first few days and weeks were spent in uncertainty, with many educational leaders drafting corrective directives to enable education from April until the end of the 2019-2020 school year. Through communication with colleagues from different countries, the discrepancies in educational delivery were incredible. Some school boards scrambled to get an exponential number of online learning platforms just as a starting point, while others had ALS programs like iReady and MATHia loaded on their students’ tablets, minimizing the educational disruption. On one end, some parents and students were stressed and anxious as many aspects of their world suddenly stopped operating, including schools. On the other end, students continued to log into their ALS accounts daily, ready for timely and targeted curation of learning content for the day. Depending on grade level, the Government of Alberta [New Tab] (2020) mandated that students spend between five to 12 hours per week learning online. With some parents already concerned about too much screen time, school boards needed to ensure that the hours students spent online were engaged through responsive, adaptive, and learner-centric pedagogy.

Conclusion

The cost-effective, scalable, yet personable and impactful learning experience that ALS can provide means the spread of these technologies in K-12 may be unstoppable. “These systems continually assess skill and confidence levels and provide precise direction to fill knowledge gaps, accelerate mastery, and adapt to each student’s individual learning styles and unique circumstances” (McGraw-Hill, 2019, p. 11). School boards generally set goals to increase achievement, while students face multidirectional pressures to bolster their own test scores. It is important not to get caught up in the perceived novelty of any new technology, such that “we’re so hopeful about upgrades that we rarely look at the practices that technology does not change, [and] those that it changes for the worse.” (Watters [New Tab], 2015, para. 9). Before district leaders and other agents of change bring ALS to schools, they need to ask questions about big data, learner profiles, and algorithmic analytics, and to evaluate the marketed information though multiple ethical considerations.

Neither teachers nor parents should agonize about teachers getting replaced by machines, as “most experts believe that teachers will remain irreplaceable, but there will be many changes to a teacher’s job and to educational best practices” (Marr [New Tab], 2019). In fact, “actual study findings indicate that personalized learning systems are best deployed as supplements to teachers, rather than their replacements” (Bulger [New Tab], 2016, p. 12). The role of teachers will continue to change, and it is likely they will be expected to incorporate computer tablets containing ALS and associated data into their daily routines sooner or later. The shift to ALS can enable teachers to respond directly to questions not answered by ALS, freeing time for greater emphasis on socio-emotional human connections that empower teachers to nurture other human intelligences in their students. Due to many complex societal problems, schools have become crucial communal hubs that offer supports for essential services and human connections. With an ever-increasing population and widening wealth gap, the need for such services in school will likely increase in the future. Though schools will continue to have standardized tests, there is no such thing as a standardized student, and this is where the hyper-personalization offered by ALS can be a helpful addition to teaching and learning supports. We cannot let fear hold back innovation in the educational technology field. We need to embrace new pedagogical models for the 21st century — to do otherwise would be unethical.

Questions to Consider

  • Which other major societal shifts — comparable to the COVID-19 pandemic — will impact the future of learning?
  • What might be the ethical and long-term health and educational implications of using neural enhancement technologies?
  • What policies, processes, and protocols will prevent recorded student learning profiles from being sold to prospective employers, financial institutions, and health insurance providers of those students?
  • In what ways will classrooms change for present learners, and to better prepare graduates for their futures?

References

Alberta Education. (2018). Leadership quality standard. Alberta Government. https://education.alberta.ca/media/3739621/standardsdoc-lqs-_fa-web-2018-01-17.pdf

Barron, B., Chowdhury, N., Davidson, K., & Kleiner, K. (2019). Annual report of the CIFAR pan-canadian AI strategy. CIFAR. https://www.amii.ca/wp-content/uploads/2019/04/ai_annualreport2019_web.pdf

Bienkowski, M., Feng, M., & Means, B. (2012). Enhancing teaching and learning through educational analytics and data mining. U.S. Department of Education. https://tech.ed.gov/wp-content/uploads/2014/03/edm-la-brief.pdf

Brody, J. E. (2017, January 9). Hooked on our smartphones. The New York Times. https://www.nytimes.com/2017/01/09/well/live/hooked-on-our-smartphones.html

Bulger, M. (2016). Personalized learning: The conversations we’re not having. Data and Society Research Institute. https://datasociety.net/pubs/ecl/PersonalizedLearning_primer_2016.pdf

Carnegie Learning. (n.d.a) MATHia. Retrieved February 27, 2020, from https://www.carnegielearning.com/products/software-platform/mathia-learning-software/

Carnegie Learning. (n.d.b). Welcome to the new Carnegie Learning Sample Center! Retrieved March 28, 2020, from https://www.carnegielearning.com/sample-center/?redirected=/sample-center/why-cl/the-cl-story#our-research

Citron, D. K., & Pasquale, F. (2014). The scored society: Due process for automated predictions. Washington Law Review, 89(1), 101–133. https://digitalcommons.law.uw.edu/cgi/viewcontent.cgi?article=4796&context=wlr

Consortium for School Networking (2018, March). The future of work and learning: 2030. https://imgsvr.eventrebels.com/ERImg/02/08/06/WorkandLearning2030ToolkitFINAL.pdf

Curriculum Associates. (2020). iReady Diagnostic: Linking study with the Massachusetts comprehensive assessment system (MCAS). https://www.curriculumassociates.com/-/media/mainsite/files/i-ready/iready-diagnostic-assessments-linking-study-overview-massachusetts-2020.pdf?la=en&hash=CAE0B4E43A145D8513CE3198005D8990

Curriculum Associates. (n.d.). iReady: Personalized instruction and teacher resources. Retrieved February 26, 2020, from https://www.curriculumassociates.com/products/i-ready/i-ready-learning

Desire2Learn. (2018, November 6). The future of work and learning in the age of the 4th industrial revolution. https://www.d2l.com/future-of-work/

Edmentum. (n.d.). Exact Path. Retrieved February 29, 2020, from https://www.edmentum.com/products/exact-path

Ferster, B. (2017, January 29). Intelligent Tutoring Systems: What happened? Bill Ferster. http://www.stagetools.com/bill/intelligent-tutoring-systems-what-happened/

Georgia Institute of Technology College of Computing. (n.d.). Jill Watson. Retrieved February 22, 2020, from https://www.cc.gatech.edu/holiday/jill-watson

Government of Alberta. (2020). Student learning during COVID-19. Retrieved April 8, 2020, from https://www.alberta.ca/student-learning-during-covid-19.aspx

Holzapfel, B. (2018, May 24). How can technology empower the class of 2030? Microsoft Education Blog. https://educationblog.microsoft.com/en-us/2018/05/technology-empower-class-of-2030/

IBM. (2018, August 23). Mandarin language learners get a boost from AI. IBM Research Blog. https://www.ibm.com/blogs/research/2018/08/mandarin-language-ai/

KnowledgeWorks. (2018, November 27). Navigating the future of learning: KnowledgeWorks Future Forecast 5.0. https://knowledgeworks.org/resources/forecast-5/

Lerman, J. (2013, September). Big data and its exclusions. Stanford Law Review Online, 66, 55–63. https://www.stanfordlawreview.org/online/privacy-and-big-data-big-data-and-its-exclusions/

Litwin, K. (2019, January 21). Obscene gap between rich and poor, says Oxfam. National Observer. https://www.nationalobserver.com/2019/01/21/news/obscene-gap-between-rich-and-poor-says-oxfam

Luckin, R., Holmes, W., Griffiths, M., & Forcier, L.B. (2016). Intelligence unleashed: An argument for AI in education. Pearson. https://www.pearson.com/content/dam/one-dot-com/one-dot-com/global/Files/about-pearson/innovation/open-ideas/Intelligence-Unleashed-v15-Web.pdf

Maderer, J. (2016, May 9). Artificial intelligence course creates AI teaching assistant. Georgia Tech News Center. https://www.news.gatech.edu/2016/05/09/artificial-intelligence-course-creates-ai-teaching-assistant

Marr, B. (2018, July 25). How is AI used in education — Real world examples of today and a peek into the future. Forbes. https://www.forbes.com/sites/bernardmarr/2018/07/25/how-is-ai-used-in-education-real-world-examples-of-today-and-a-peek-into-the-future/#d1d407d586e8

Marr, B. (2019). How is AI used in education — Real world examples of today and a peek into the future. Bernard Marr & Co. https://bernardmarr.com/default.asp?contentID=1541

McGraw-Hill. (2019). The equity equation. https://s3.amazonaws.com/ecommerce-prod.mheducation.com/unitas/highered/explore/equity/equity-equation-digital.pdf 

McKinnon, R. (2018, September 25). More teachers, less tech, say parents wary of i-Ready. Gainesville Sun. https://www.gainesville.com/news/20180921/more-teachers-less-tech-say-parents-wary-of-i-ready?template=ampart

McLeod, J. (2017). Exact path research brief: Effectiveness study. Edmentum. https://www.edmentum.com/sites/edmentum.com/files/resource/media/Exact Path Effectiveness Paper FINAL_0.pdf

Microsoft. (2018). The class of 2030 and life-ready learning: The technology imperative. https://education.minecraft.net/wp-content/uploads/13679_EDU_Thought_Leadership_Summary_revisions_5.10.18.pdf

Neelakantan, S. (2019, November 25). Colleges see equity success with adaptive learning systems. EdTech Magazine. https://edtechmagazine.com/higher/article/2019/11/colleges-see-equity-success-adaptive-learning-systems

Pascoe, J. (2019, July 25). Meeting society’s AI learning needs. University of Alberta. https://www.ualberta.ca/science/news/2019/july/reinforcement-learning-online-course

Regan, P.M., & Jesse, J. (2018). Ethical challenges of edtech, big data and personalized learning: Twenty-first century student sorting and tracking. Ethics and Information Technology, 21(3), 167-179. https://doi.org/10.1007/s10676-018-9492-2

Regan, P., & Steeves, V. (2019). Education, privacy, and big data algorithms: Taking the persons out of personalized learning. First Monday, 24(11). https://doi.org/10.5210/fm.v24i11.10094

Research and Markets. (2018, August). Artificial intelligence market in the US education sector
2018-2022. https://www.researchandmarkets.com/reports/4613290/artificial-intelligence-market-in-
the-us?utm_code=5lshzz&utm_medium=BW

Royakkers, L., Timmer, J., Kool, L., & van Est, R. (2018). Societal and ethical issues of digitization. Ethics and Information Technology, 20(2), 127-142. https://doi.org/10.1007/s10676-018-9452-x

TechNavio. (2018a, August). Artificial intelligence market in the US education sector 2018-2022. https://www.technavio.com/report/global-artificial-intelligence-market-in-education-sector-analysis-share-2018

TechNavio. (2018b, August). Global artificial intelligence market in education sector 2018-2022. https://www.technavio.com/report/global-artificial-intelligence-market-in-education-sector-analysis-share-2018

Troseth, G. L., & Strouse, G. A. (2017). Designing and using digital books for learning: The informative case of young children and video. International Journal of Child-Computer Interaction, 12, 3–7. https://doi.org/10.1016/j.ijcci.2016.12.002

University of Victoria. (n.d.). Student privacy. Retrieved March 22, 2020, from https://www.uvic.ca/library/featured/copyright/faculty/studentprivacy/index.php

Villeneuve, S., Barron, B., & Boskovic, G. (2019, May). Rebooting regulation: Exploring the future of AI Policy in Canada. CIFAR & Brookfield Institute. https://www.cifar.ca/docs/default-source/ai-futures-policy-labs/rebooting-regulation-exploring-the-future-of-ai-policy-in-canada.pdf?sfvrsn=1b627b74_14

Watters, A. (2015, August 10). Teaching machines and Turing machines: The history of the future of labor and learning. Hack Education. http://hackeducation.com/2015/08/10/digpedlab

West, D. M. (2012, September). Big data for education: Data mining, data analytics, and web dashboards. Governance Studies at Brookings. https://www.brookings.edu/wp-content/uploads/2016/06/04-education-technology-west.pdf

Appendix A

Table 5.3 Completed ethical framework for use of adaptive learning systems in modern classrooms, based on Farrow’s (2016) Uncompleted Framework
Principle Consequentialist Theory Deontological Theory Virtue Ethics Theory
Full disclosure
  • ALS deliver all sorts of new data, which provide teachers with new insights.
  • Teachers are able to use their time with students in more efficient way, to fulfil the gaps that ALS identifies.
  • Student engagement increases as ALS incorporates student’s own data into otherwise static questions and makes the visualization appears on screen.
  • Practicing equity, teachers use real-time data to narrow the gaps in student comprehension before moving to the next curricula outcome.
  • Teachers utilize the recorded ALS observations to better address student needs.
  • Ongoing information flow about student’s learning profile mean no need to reinvent the wheel and waste time asking what was already captured.
  • Students who attend schools that do not have funds to purchase ALS become disadvantaged by not having the opportunity to use AI in ALS.
  • Teacher is freed from spending time marking so more of it can be spent on other pedagogical tasks.
  • Flexibility offered by ALS can  inspire students to go at faster or slower pace, depending on each curricular outcome studied.
Privacy, data security and informed consent
  • Data can be analyzed to forewarn about risks of students dropping out.
  • Thanks to instantaneous feedback offered by ALS, students do not wait for days to get their quizzes or assignments evaluated and recorded.
  • AI within ALS uses student input to generate personalized lessons for every student in class.
  • Gaps in data can lead to students becoming overlooked, erroneously resulting in no interventions.
  • Regulations are vague or lacking when it comes to ethic expectations  and lawful  compliance with respect to student data.
  • ALS merchants could use student data to improve their own AI systems and sell more units, creating a positive feedback loop to generate more money.
  • There are many aspects of a human being that go beyond the ALS perimeters, and those unincorporated bits of information can lead to dehumanization of its users.
  • Automated ALS cannot read emotional cues presented by students, leading to futile attempts to engage actively.
  • Teachers need to utilize their training to address ALS shortcomings.
Respect for autonomy and independence while avoiding harm and minimizing risk
  • Following the competency-based model of instruction, students use their time in school to progress in their own unique pathways.
  • For times when students cannot be at school, they can remotely connect to their ALS devices to and carry on with learning.
  • Time spent learning is optimized, leaving students with time to learn other skills that can assist them in the future.
  • The spectrum of educational disruption is wide, but customized learning content, generated by ALS can close gaps in a timely manner.
  • Parents get learning reports frequently, enabling them to focus further at home on topics that students find challenging.
  • ALS distributes data to students, teachers, and parents, ensuring that all involved parties are on the same page.
  • Constant use of ALS can lead students to become dependent on the system for validation.
  • Heavy reliance on ALS for confirmation can hinder development of student’s own internal sense of checks and balances.
  • Teachers need to use their professional judgement and not depend solely on ALS to drive the direction of student’s learning.

License

Share This Book

Feedback/Errata

Comments are closed.