Bullet points summary
Introduction
The landscape of educational technology is in the midst of a transformative shift, as the recent widespread adoption of public access large language models has catalyzed a renewed reflection on the historical patterns of technological integration within the education sector. As an economic historian, specialized in technological innovation and large technological systems, I believe that a deeper understanding of these historical trends is crucial in shaping our approach to the emerging technological advancements (Schram, 1997). As I wrote elsewhere, this is particularly relevant for educational institutions in low-income countries in order to tackle the global education crisis and make progress towards achieving the UN's Sustainable Development Goal four (SDG -4"access to quality education").
The article by Molenda, Subramony, Clark and Stallkamp (MSCS - 2023) provides an insightful overview of the state of the question, organizing the complex subject into broad "paradigms" - guiding principles that have inspired both research and practice in educational technology. Today, the "models vs. media" paradigm is particularly relevant, since we tend to forget that all technology are tools, or means to an end. If the pedagogical model is deeply flawed, and does not promote students' engagement and active learning, adopting new technology is like the proverbial "putting lipstick on a pig".
In the 1990s, Professor Carl Wieman, a Nobel Prize winner in physics and innovator in university science education, asked himself why PhD candidates at Stanford University were still unable to do physics. He concluded that they had been taught in the wrong manner and demonstrated that by putting the problem first, and giving learners more agency in an improved pedagogical approach, learning outcomes would improve dramatically.
The origins of giving learners more agency and autonomy can be traced back to the movement started in primary and secondary schools by Maria Montessori in Italy at the beginning of the 20th century. However, due to pressure from the fascist regime, the movement's headquarters moved to the Netherlands. The emergence of more decentralized and challenge oriented pedagogical models would have been impossible without Maria Montessori's work. Much later in the 1970s, at McMaster University in Canada, Problem-Based Learning a similar approach was developed for adults in tertiary education, initially only in the medical field.
However, in MSCS I find their conclusions regarding the recent developments during the introduction of mobile and internet technologies to be somewhat unsatisfactory, particularly in the context of what they describe as the "emerging technologies paradigm" from around 1998 to the present.
At the core of this paradigm is the tendency to readily embrace and adopt the continuous stream of new applications and technologies that emerge at an increasingly rapid pace, without sufficient critical evaluation. While the initial phases may be marked by an uncritical hype surrounding a new technology, the limitations of these innovations often become apparent in subsequent phases. It is only then that a more serious examination of use cases and value propositions takes place.
In this article, we examine the effects of the most recent wave of technological innovation on the educational sector - the advent of multimodal, large language models. By drawing upon historical models of technological diffusion and adoption, we aim to shed light on the complex interplay between emerging technologies and the transformation of teaching and learning practices.
Models of technological adoption and diffusion
Broadly, there are three main theoretical models about how new technology in general is adopted.
Core-Periphery Model
The core - periphery model is frequently used by economic historians to describe the geographical spread of technology associated with the first (steam), second (electricity), third (computers), and fourth (AI) industrial revolutions. Understanding core-periphery models provides insights into the historical patterns of technological diffusion and the challenges faced by less developed regions in gaining access to and mastering new technologies. A variety of this model is are formal gravitational models, which predict a higher adoption rate in areas of concentration of, for example, population or wealth. Uncritical application these models can lead to a degree of geographical determinism, and disregard opportunities or alternative strategies that may emerge in peripheral regions.
Product life cycle model
The product life cycle model comes from the world of marketing and applies best to technologies for which there is a potentially mass demand. Understanding these broad patterns can help companies and individuals anticipate technological changes and make more informed decisions. The spread of technology from invention to widespread adoption and eventual obsolescence can be described in the following stages:
- Invention: A new technology or innovation is created, often through research and development efforts.
- Early Adoption: The technology is first introduced to the market. Adoption is typically slow as it may be expensive, complex, or unfamiliar to most consumers.
- Early Majority Adoption: As the technology improves and prices decline, it starts gaining mainstream acceptance. More people begin adopting the technology at a faster pace.
- Late Majority Adoption: The technology becomes more affordable and accessible. Even those who are more skeptical or risk-averse start adopting it.
- Laggard Adoption: The last group to adopt the technology are the laggards, who may be resistant to change or have limited access.
- Maturity: The technology reaches widespread usage and saturation in the market. Innovation and improvements slow down.
- Decline: New, more advanced technologies emerge and start replacing the older technology. Demand for the older technology decreases, leading to its eventual obsolescence.
- Obsolescence: The technology is no longer used or supported, often phased out as newer alternatives become available and more widely adopted by consumers and businesses.
The disruptive innovation model
Disruptive innovation is a concept developed by Clayton Christensen, a Harvard Business School professor, in the early 1990s. It refers to a process through which a product or service in a competitive market initially takes root in simple applications at the bottom of a market and then gradually moves upmarket, ultimately displacing established competitors. This model highlights how new entrants can successfully challenge and overtake established firms by targeting overlooked or underserved market segments. It key characteristics are:
- Market Entry: Disruptive innovations typically begin by serving low-end or niche markets that are not appealing to established companies, often due to lower profit margins.
- Initial Inferiority: These innovations usually offer simpler, less sophisticated products or services that may not initially meet the performance expectations of mainstream consumers.
- Upward Trajectory: Over time, as the disruptor improves its offerings, it begins to attract more demanding customers from the mainstream market.
- Market Transformation: Eventually, these innovations redefine market standards, leading to significant shifts in consumer behaviour and industry dynamics.
Examples of disruptive innovation are the personal computers. Initially targeted at hobbyists and small businesses, they eventually displaced mainframe computers. Another example is Netflix and streaming video in general. Started as a DVD rental service and disrupted traditional video rental stores like Blockbuster by transitioning to streaming services. Disruptive innovation is a model with explanatory power and some predictive power.
In my view, each of these models can be valid in specific historical situations, and reveal part of the truth. Therefore, I take an eclectic approach and take elements from all three.
The productivity paradox
Robert Solow, who won the Nobel Memorial Prize in Economic Sciences in 1987 famously remarked in 1998, "You can see the computer age everywhere but in the productivity statistics." This quote encapsulates the productivity paradox, which refers to the observed phenomenon where significant investments in information technology (IT) and computing do not correspond with expected increases in productivity.
In the statistics that in the 20th century the adoption of a new technology by 80% US households took between 30 and 40 years. The only exception was the mobile phone that took 10 years or less. This is in line with Paul David's findings on the "productivity paradox". His article "The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox," published in the American Economic Review in 1989, examines the historical adoption of electric dynamos in factories and draws parallels to contemporary issues surrounding computer technology and productivity. He postulates that it takes an average of 40 years before a new technology leads to substantial increase in productivity or output per worker.
According to Paul David, it can take approximately 40 years for new technologies to manifest their effects in productivity statistics due to several intertwined factors related to the adoption and integration of these technologies into existing systems. His insights, particularly drawn from the historical example of the electric dynamo, illustrate this phenomenon.
 |
| Flat and steep adoption curves |
The reasons for this time lag, according to Paul David are:
Initial Disruption: When a new technology is introduced, it often disrupts established production processes. For instance, the introduction of electric motors allowed for a more efficient arrangement of machinery compared to centralized steam engines. However, this transition required significant reorganization of factory layouts and workflows, which initially led to a decrease in productivity as firms adjusted to the new system
Incremental Improvement: New technologies often undergo a period of incremental improvement before their full potential is realized. In David's analysis, while electric motors were available, factories did not immediately adopt them due to the complexity of integrating them into existing operations. It took time for businesses to recognize and implement the more efficient configurations that electric power enabled.
Path Dependence: The process of technological adoption is path-dependent, meaning that past decisions and existing infrastructures influence how quickly and effectively new technologies can be integrated. This historical context can slow down the realization of productivity gains as firms navigate the complexities of changing their operational frameworks.
Measurement Challenges: The initial effects of new technologies may not be easily captured by conventional productivity measures. As firms invest in new technologies but do not see immediate output increases, it creates a paradox where rapid technological advancement does not correlate with observable productivity growth.
Paul David emphasizes that while technological innovations like electricity and computing hold great promise for enhancing productivity, realizing these benefits requires time for adaptation, reorganization, and overcoming initial disruptions within industries.
The decade of the 2020s can be described as the decade where the productivity gains and new business models of the distributed internet become apparent, and reach a stage of maturity. However the release of ChatGPT on 30 November 2022 and its rapid and massive adoption created new business opportunities and accelerated this adoption process. The relationship between human and computer has fundamentally changed: it is much easier to interact through a "chat" than through a programing language. The costs of creating specialized IT tools and online platforms dramatically decreased.
Educational technology
After exploring the process of adoption of technology, innovation and transition in general, and its inherent challenges, we must now shift our perspective to the psychological and organizational challenges of technological transitions in educational institutions. The social factors are probably equally important as those directly connected to the technology.
On the surface, the day-to-day practices in schools and universities may appear to have changed little over the past 50 years. Instructors still write on whiteboards (the successor to chalkboards), creating learning materials in real-time. Their pedagogical approach is still "chalk and talk" and their persona is the "sage on the stage", and not providing support for learners as "guide on side". Similarly, students continue to handwrite essays on paper, which are then marked up by teachers with written feedback. This process can become highly inefficient when the handwriting is poor on either side.
However, the transformative effects of information and communication technology are quietly reshaping educational practices in the background. While photocopied materials remain commonplace, the digital distribution of course content on screens, for example, has become widespread. Furthermore, the use of learning management systems encourages instructors to more carefully prepare their teaching materials in advance of classes and assessments, rather than in real time.
Beyond the direct impacts of new technologies, the integration of educational technology is also influenced by psychological factors, such as individual resistance to change, as well as organizational rigidity within educational institutions. Overcoming these barriers is crucial for realizing the full potential of technological innovations in teaching and learning.
Psychological barriers
The integration of modern educational technologies into teaching and learning practices is often hindered by various psychological barriers. One of the primary obstacles is the natural human resistance to change. Many educators and students are accustomed to traditional methods, such as lecturing and handwritten assignments, and are reluctant to disrupt their established routines and adopt unfamiliar approaches, even if the new technologies offer potential benefits.
Closely related to this resistance to change is the technological anxiety and aversion experienced by some individuals. Some educators and students may feel intimidated by the perceived complexity of new technologies, lacking confidence in their ability to effectively utilize and integrate these tools into their teaching and learning. This sense of technological inadequacy can lead to a reluctance to engage with educational technologies altogether.
Furthermore, the generational divide between younger, tech-savvy students and older, less tech-fluent educators can create additional barriers to the widespread adoption of educational technologies. Older instructors, in particular, may be more skeptical of the educational value and learning outcomes associated with technology-based approaches, preferring the familiar and "proven" traditional methods.
Concerns about privacy and data security also play a role in hindering the adoption of educational technologies. Educators and institutions may be hesitant to embrace technologies that they perceive as posing risks to student data protection and privacy, further contributing to the reluctance to integrate these tools into the educational landscape.
Alongside these psychological factors, some educators may also harbor concerns that the integration of educational technologies could ultimately lead to the replacement of teachers. The fear that technology-driven instruction could diminish the role and relevance of human educators can further exacerbate the resistance to adopting these innovations. This apprehension stems from the perception that technology-based learning may undermine the personal connections, mentorship, and individualized support that teachers provide to their students.
Overcoming these psychological barriers is crucial for the successful implementation of modern educational technologies. Comprehensive training, professional development, and the cultivation of a culture that encourages the exploration and adoption of these tools can help address the concerns and build confidence among both educators and students. By addressing these psychological hurdles, educational institutions can unlock the full potential of technological innovations in enhancing teaching and learning experiences.Conclusions and recommendations
Organizational impediments
In addition to the psychological barriers that impede the adoption of educational technologies, educational institutions also face significant organizational impediments that hinder the integration of these innovations, particularly in the context of recent advancements in AI-powered tools.
One of the primary organizational challenges lies in the inherent rigidity and bureaucracy that often characterize schools and universities. These institutions are typically structured with well-established policies, procedures, and hierarchical decision-making processes that can be slow to adapt to the rapid pace of technological change. Implementing new technologies, especially those with disruptive potential, can be a complex and arduous process that requires navigating layers of administrative approval, budgetary allocations, and institutional inertia.
This organizational rigidity is further compounded by the difficulty these institutions face in providing detailed guidelines and clear policies around the use of emerging AI-powered tools. For teaching, I have been inspired by the practical and transparent approach developed by Leon Furze who has been conducting research at the University of Melbourne in Australia on the topic for years. As part of this work he has developed the AI assessment Scale (AIAS), that can be applied at all levels on any subject and he has proposed to UNESCO. For research, I have enjoyed the contributions of Dr. Mustaq Bilal, a humanities scholar at the University of Odense in Denmark. Lately, he has developed ResearchKick, an all in one AI-powered solution for academic writing and research, or in short ChatGPT for researchers.
The rapid development and evolving capabilities of AI-based technologies, such as language models, virtual assistants, and automated grading systems, present a unique set of challenges for educational institutions. Crafting comprehensive, future-proof guidelines that address the ethical, privacy, and pedagogical implications of these tools can be a daunting task for administrators and policymakers.
Moreover, the integration of AI-powered tools often requires significant investment in infrastructure, faculty training, and ongoing technical support. Educational institutions, which typically operate under constrained budgets and limited resources, may struggle to allocate the necessary funds and personnel to effectively implement and maintain these advanced technologies. The competing priorities and budget allocations within the institution can further impede the widespread adoption of AI-powered educational tools.
Additionally, the lack of clear regulatory frameworks and standardized best practices around the use of AI in education can create uncertainty and hesitation among educational leaders. Without well-defined guidelines and industry-wide standards, institutions may be reluctant to embrace these technologies, fearing potential legal or reputational risks.
To overcome these organizational impediments, educational institutions must cultivate a culture of innovation, invest in cross-functional teams to navigate technological complexities, and establish robust governance frameworks to guide the responsible and effective integration of AI-powered tools. Proactive engagement with policymakers, industry experts, and peer institutions can also help develop the necessary frameworks and guidelines to support the widespread adoption of educational technologies within these complex organizational structures.
Conclusions and recommendations
Recommendations for Teachers and Lecturers:
- Seek out opportunities for continuous learning and professional development to stay abreast of educational technology advancements.
- Experiment with new tools and approaches in a low-stakes, iterative manner to build familiarity and confidence.
- Advocate for institutional support, resources, and policies that enable the effective integration of educational technologies.
- Collaborate with peers to share best practices and lessons learned in the use of educational technologies.
Policy Recommendations for Educational Leadership:
- Invest in cross-functional teams and task forces to navigate the complexities of educational technology integration.
- Establish clear, future-proof guidelines and policies that address the ethical, privacy, and pedagogical implications of emerging technologies, particularly AI-powered tools.
- Allocate sufficient resources for infrastructure, faculty training, and ongoing technical support to enable the successful implementation of educational technologies.
- Foster a culture of innovation and continuous improvement that empowers educators to explore and adopt new technologies.
- Engage with policymakers, industry leaders, and peer institutions to develop industry-wide standards and best practices for the responsible use of educational technologies.
Recommendations for Leadership in Low-Income Countries:
- Prioritize strategic partnerships with international organizations and higher-income institutions to facilitate knowledge sharing and capacity building.
- Advocate for increased investment in fundamental digital infrastructure and connectivity to ensure equitable access to educational technologies.
- Develop contextualized training programs that address the unique challenges faced by educators in resource-constrained environments.
- Collaborate with local communities and policymakers to create adaptive, low-cost solutions that leverage educational technologies.
- Improve teaching methods first, before adopting technology, in order to avoid the "putting lipstick on a pig" fallacy.
- Establish regional networks and knowledge-sharing platforms to leverage the collective experiences and innovations of low-income countries.
Sources
Molenda, M., Subramony, D. P., & Clark-Stallkamp, R. (2023). A Brief History of Educational Technology. EdTech Books. Retrieved from https://edtechbooks.org/foundations_of_learn/history_of_lidt
Schram, A. (1997). Railways and the Formation of the Italian State in the Nineteenth Century. Cambridge University Press. Retrieved from https://www.cambridge.org/us/universitypress/subjects/history/european-history-after-1450/railways-and-formation-italian-state-nineteenth-century
Terzian, S. (2019). The History of Technology and Education. OUP Academic. http://doi.org/10.1093/oxfordhb/9780199340033.013.33
Wieman, Carl ((2015, November 12) Active Learning and new pedagogies for higher education. Youtube. Retrieved from https://www.youtube.com/watch?v=9A13RWOs6oA&list=PLiJAu6VY_CMejV78gmeiqIrKfJ8x8Xdb8&index=29&t=15s
No comments:
Post a Comment
Note: only a member of this blog may post a comment.