"'Personalization' for All": Artificial Intelligence in Education as a Means for Personalized Learning on a Mass Scale, and How it Ensures Inclusive Education and Sustainability
Overview
This OER goes about to relate Personalized-Learning, Artificial Intelligence in Education, Inclusive Education, and Massive Open Online Courses, as well as talk about sustainabality.
Introduction
In this lesson, you will be introduced to mind-opening topics which interlink in a way to shape possible digital futures in learning. First topic to be discussed is inclusive education (IE), it will be defined and explored in relation to its position against personalized-learning (PL). Next, PL is presented as
"Tailoring of learning to each students' strengths, needs, and interests-including giving students voice and choice in what, how, when, and where they learn - to provide flexibility and support, to ensure mastery of the highest standards possible" (Aurora Institute, 2016)
and demonstrated in context of Artificial Intelligence in Education (AIEd). After that, this lesson goes on to showcase how massive open online courses' (MOOCs) provide PL experiences for students, and how these experiences are viewed with regards to sustainability.
By the end of this lesson, you will have engaged with the topics mentioned above enough to have an idea and position concerning:
- Where do IE and PL meet and/or differ?
- How is PL demonstrated in AIEd? And What are the criticism and challenges AIEd and PL face?
- What makes PL and MOOCs sustainable or unsustainable?
(Title image: "Computer AI - Why I keep losing" by Si-MOCs is licensed under CC BY-NC-SA 2.0)
Inclusive Education (IE)
For this section, you will watch the video of Prof. Lani Florian giving a conference on Inclusive Pedagogy. As you watch the lecture, keep in mind what Florian has to say about the students' individual differences, unique educational needs, and social inclusion where marginalization is concerned.
Section Activity:
Answer the question found in the link below (recommended to open in a new tab):
Inclusive Education Section Activity
Optional Additional Resource on Inclusive Education: (Note: if interested in the following video seminar, come back to it after the end of the OER session)
Personalized-Learning (PL)
In 2004, Pollard & James claimed that personalized-learning was the "Big Idea" back then. Now, what is PL exactly, and how did it come to be?
PL was said to be originated from Gardner's multliple intelligences theory (Guldberg, 2004; Johnson, 2004). Meaning that students' learning experiences must be tailored based on their individual abilities, style, interests and needs, for the best learning outcomes (Good & Brophy, 1990). Moreover, in the ‘A National Conversation about Personalised Learning’ document, PL is explained as:
"... the drive to tailor education to individual need, interest and aptitude so as to fulfil every young person’s potential." (DfES, 2004a: 4)
Also worth noting, another goal for PL is to promote "equity and social justice" (DfES, 2004a), and cancel out the distinctions that may arise from gender bias, social economic status or ethnicity. Although the whole idea of PL still seems ambiguous, remember that was all mostly in 2004, and recent studies have lead to its evolution (which we will be discussing in the sections).
Attached is an optional reading (Free Access) on PL
Personalized-Learning (PL) and Artificial Intelligence in Education (AIEd)
"Artificial Intelligence & AI & Machine Learning" by mikemacmarketing is licensed under CC BY 2.0
In recent advancements in the field of education, Artificial Intelligence (AI) is starting to make an appearance, and it is believed that AIEd technology are presenting new scientific precision in analyzing educational activities. Which, in turn, can provide 'deeper, and more fine-grained understandings of how learning actually happens' and 'an intelligent, personal tutor for every learner' (Luckin et al., 2016). Moreover, PL has stepped up a level with the novel AI involvement in education. With the addition of 'intelligent' and data-driven technology, tailored educational contents and automatic teacher feedback are generated faster, leading to an enhanced and fast-forward learning experience, and to easier access. Furthermore, by adapting to each student, an ideal AI tutor could become an ultimate pedagogue (Knox et al., 2019).
Watch the videos below in order:
Introducing IBM Watson Education
Example of AI tutor Watson at work
Additional fun yet informative TEDx talk by Ashok Goel
Based on Luckin & Cukurova (2019) statements, evidence of well-designed AI working in Education are increasing. Some of these AI's are OLI learning courses (Lovett, Meyer, & Thille, 2008), SQL-Tutor (Mitrovic & Ohlsson, 1999), ALEKS (Craig et al., 2013), Cognitive Tutor (Pane, Griffin, McCaffrey, & Karam, 2014) and ASSISTments (Koedinger, McLaughlin, & Heffernan, 2010), they have all shown statistically significant positive results towards students learning.
On another note, AIEd and intelligent tutoring systems (ITS) are both aiming to produce intelligent tutoring software for every student. They are encouraging effective skill development and promoting more engagement with learning content by including 'interactive software, adaptive learning delivers paced, customized instructions with real-time feedback that allow faster student progression' (Six benefits of adaptive learning 2013). Furthermore, technology-enhanced learning (TEL) is making the job for teachers easier and allowing them to be more effective in their teaching and mentoring, as it removes the administration and bureaucracy workloads by 'marking, planning their lessons' and many other similar functions (Ferster 2014). Also, because the quality and range of the adaptive customization differs throughout varied TEL environments, intelligent learning environments (ILEs) were introduced. ILEs are unique TEL systems that work to create 'interactive and adaptive learning experiences' that are suitable for a student using a variety of AI techniques (Brusilovsky 1994).
Criticism and Challenges involving PL and AIEd
To begin with, it is beneficial to understand the discourses between IE and PL through AIEd. As previously mentioned, AIEd supports the idea of 'an automated, and personalized, one-to-one tutor for every learner', whereas IE tends to focus on methods to include 'marginalized and excluded individuals' and creating an inclusive educational environment (Knox, Wang, & Gallagher, 2019). That is to say, IE argues against a one-to-one tutor/student approach and for a ‘common ground’, as it views that PL marginalizes students further instead of bringing them all together and foster acceptance and growth as a community.
Also, Luckin & Cukurova (2019) have noted that AI technologies in Education are lacking enough evidence at scale (Baker, 2016) and are less effective in providing more complex instructions (eg. Collaboration or self-explanation) (Koedinger et al., 2012). Additionally, AI may still create forms of injustice, as mostly rich students and organizations could get access to it and make use of its better digital skills (Wood, 2019), possibly widening the already present gap.
Section Activity:
Watch video below and take notes of the challenges Cynthia Breazeal presents where AI and PL are involved.
Massive Open Online Courses (MOOCs)
MOOCs Logos - by Programmableweb.com
Watch Video and take notes of what makes MOOCs ideal and to who and in what occasions:
Introduction to MOOCs
Section Activity:
After watching the video, answer the question provided in the link below (recommended to open in a new tab):
Education for Sustainable Development (ESD) and MOOCs
Read the discussion:
In 2005, the Education for Sustainable Development (ESD) campaign, promoted by UNESCO, claimed education as a significant contributor to sustainable development. Moreover, assuming sustainable development is achievable, the required knowledge and skills can be described particularly through technical innovation, efficiency, and different consumer habits (cf. Kopnina 2014). As a result, the future becomes measurable and predictable (Holfelder, 2019). With the rise of new technology, including AI, massive open online courses (MOOCS), machine learning and data mining, further improvements in teaching and learning are occurring. Noticeably, the fast increase in popularity of the MOOCS did raise concerns in ESD. In comparison to MOOCS, the semantic analysis branch of AI supports both flexibility in information management and learning behavior mining better (Ling Wang, 2018). Besides that, MOOCS allows access to affordable personalized education, which can support sustainable development and decrease poverty (Ling Wang, 2018). Unfortunately, learner attrition is limiting MOOCS development, and there is also not enough evidence its capacity of sustainable development (Ling Wang, 2018). Nonetheless, higher education institutions have recently been researching through different teaching modes, as some types of education have not been backing sustainable development appropriately. With traditional teaching methods not taking into consideration the diversity of abilities available, this slows the development of traditional education since it inhibits diverse students to use their own abilities for their own advantage. Additionally, the widespread of new technologies is altering living habits, ways of thinking and values, and with it greatly changing the traditional way of teaching in higher education. MOOCs is enabling access to new knowledge and courses of interests and/or need for students anywhere at any time. Unluckily, studies show that students drop out of these courses after two weeks, and the main reasons are 'lack of time, lack of motivation, feelings of isolation, lack of interactivity, and insufficient background or skills" (Ling Wang, 2018). There is no question that the traditional education mode offers its learners learning platforms for their development, and positively impacts their social and scientific growth. But with their increasing need for free development, their demand for personalized learning experiences also increases. Furthermore, there are several indications that there is a need to implement new technology within the traditional education model to adjust to the sustainable development in education. And so, the surfacing of MOOCs does not only give a selection of course choices to the students, but also it gives them a sense of autonomy in learning and fosters their individual learning plans depending on their interests and needs. The only thing left, in this case, is to improve MOOCs applicability and serviceability (Ling Wang, 2018).
"Grumpy MOOC cat" by ryan2point0 is licensed under CC BY-NC-ND 2.0
References - Readings List
Ainscow, M., Slee, R., & Best, M. (2019). Editorial: the Salamanca Statement: 25 years on. International journal of inclusive education, 23(7-8), 671-676. doi:10.1080/13603116.2019.1622800
Courcier, I. (2007). Teachers' Perceptions of Personalised Learning. Evaluation & Research in Education, 20(2), 59-80. doi:10.2167/eri405.0
Holfelder, A.-K. (2019). Towards a sustainable future with education? Sustainability Science, 14(4), 943-952. doi:10.1007/s11625-019-00682-z
Knox, J., Wang, Y., & Gallagher, M. (2019). Introduction: AI, Inclusion, and ‘Everyone Learning Everything’. In J. Knox, Y. Wang, & M. Gallagher (Eds.), Artificial Intelligence and Inclusive Education: Speculative Futures and Emerging Practices (pp. 1-13). Singapore: Springer Singapore.
Ling Wang, G. H. a. T. Z. (2018). Semantic Analysis of Learners’ Emotional Tendencies on Online MOOC Education. Sustainability.
Luckin, R. (2018). AI is coming: use it or lose to it. The Times Educational Supplement(5306). Retrieved from https://search.proquest.com/docview/2064293840?accountid=10673
Luckin, R., & Cukurova, M. (2019). Designing educational technologies in the age of AI: A learning sciences-driven approach. British Journal of Educational Technology, 50(6), 2824-2838. doi:10.1111/bjet.12861
Nganji, J. T. (2018). Towards learner-constructed e-learning environments for effective personal learning experiences. Behaviour & Information Technology, 37(7), 647-657. doi:10.1080/0144929X.2018.1470673