Educational Technology

Educational technology is the deliberate design and use of tools, media, and systems to improve learning.

10 frameworks, from 1900 to today · Education · Social Sciences

still a working view ended · each leads to its entry below

Overview

Educational technology is the deliberate design and use of tools, media, and systems to improve learning1. The phrase can suggest only computers, but the field is older and broader than that: it begins with charts, models, and film, passes through teaching machines and programmed textbooks, and only later becomes the digital enterprise most people picture today. What is at stake throughout is a recurring question — whether a new medium can make teaching more effective, more widely available, or more individualized, and what has to change in the design of instruction for that promise to be kept. Its history is a long record of enthusiasm, disappointment, and slow accumulation of design knowledge.

Visual Instruction and the Museum of Things

Around 1900, schooling in Europe and North America was still overwhelmingly verbal: children learned from recitation and textbooks. Reformers argued that understanding came through the senses, and that a lesson on volcanoes or cotton manufacture should be seen, not merely read. The result was the visual instruction movement, which built collections of lantern slides, photographs, models, and specimens for classroom use. School museums, such as the St. Louis Educational Museum founded in 1905 under the direction of Carl Rathmann, lent such materials to teachers across a district. The idea spread through the early twentieth century, and the phrase "visual education" became a recognized specialty with its own journals and training.

The movement's limitation was that pictures could illustrate but not explain sequence or process. A slide of a machine showed what it looked like, not how it worked. That gap pushed the field toward motion and sound.

Audiovisual Instruction and the Wartime Surge

From the 1920s, film, radio, and later sound recording entered the classroom. The 1930s and 1940s saw the rise of audiovisual instruction, with film libraries, projection equipment, and research on whether films taught as well as teachers. The decisive test came in the Second World War, when the U.S. military needed to train millions of recruits quickly. The Army produced training films such as Frank Capra's Why We Fight series, and studies of military training films by researchers including Carl Hovland and his colleagues examined what made instructional film persuasive and memorable. The war demonstrated that media could carry instruction at scale, and it trained a generation of educators in their use.

The postwar decades brought overhead projectors, language laboratories with tape recorders, educational television, and the belief — often overstated — that each new medium would transform schooling. A recurring finding of the research, later summarized as "no significant difference," was that media rarely taught better than a competent teacher by themselves; what mattered was how the instruction was designed. That conclusion quietly redirected the field.

Programmed Instruction and the Teaching Machine

The pivotal figure was B. F. Skinner, the psychologist of operant conditioning. In a 1954 article, "The Science of Learning and the Art of Teaching," Skinner argued that classroom teaching violated nearly every principle of learning he had established in the laboratory: pupils waited too long for feedback, moved at a pace set by the class rather than themselves, and were rarely reinforced for each correct step. His remedy was programmed instruction: content broken into small frames, each presenting a bit of information and requiring an active response, with immediate confirmation of the answer, so that the learner advanced only when correct.

Skinner built a mechanical teaching machine to present such frames, and the 1950s and 1960s saw a wave of programmed textbooks, some running to hundreds of frames, in subjects from arithmetic to psychology. The approach had real strengths — clear objectives, small steps, active responding, immediate feedback — and real weaknesses: programs were laborious to write, often rigid, and better suited to factual recall than to complex reasoning. Its lasting contribution was conceptual. It made educators think of instruction as something that could be designed in advance, tested, and revised.

Instructional Design as a Discipline

That way of thinking crystallized into instructional design, the systematic planning of learning experiences2. Its intellectual foundations were laid in the 1950s and 1960s by figures such as Robert Gagné, whose Conditions of Learning (1965) argued that different kinds of learning outcomes — verbal information, intellectual skills, attitudes, motor skills — required different instructional conditions, and by Benjamin Bloom's taxonomy of educational objectives (1956), which classified learning goals from recall to evaluation. In the 1970s, models such as the ADDIE framework — analysis, design, development, implementation, evaluation — gave practitioners a repeatable process, and the field professionalized through graduate programs and journals.

Instructional design was the field's answer to the "no significant difference" problem: if media did not teach by themselves, then the design of the instruction around them was what mattered. This insight carried the field through every subsequent technology.

Computer-Assisted Instruction and Its Limits

Computers entered instruction in the 1960s. At Stanford, Patrick Suppes developed drill-and-practice programs in elementary mathematics; at the University of Illinois, the PLATO system, begun in 1960, grew into a large networked teaching system with terminals, courseware, and its own authoring language. Computer-assisted instruction (CAI) promised patience, individual pacing, and record-keeping no teacher could match. Early evaluations were mixed: drill programs could improve basic skills, but the systems were expensive, and the software often reproduced the worst of programmed instruction — rigid, text-bound, and shallow.

The 1980s brought the microcomputer into schools and with it a debate about what computers were for. Some saw them as tutors; others, following Seymour Papert, saw them as materials for construction.

Constructionism and the Learner as Builder

Papert, a mathematician who had worked with Jean Piaget, argued in Mindstorms (1980) that children learn most powerfully when they build things that matter to them and that others can see and share. His constructionism extended Piaget's constructivism — the view that knowledge is actively built by the learner rather than transmitted — by insisting that building a public artifact, especially a computer program, makes that construction visible and sharable3. The vehicle was Logo, a programming language designed in the late 1960s with a "turtle" that children could command to draw shapes, teaching geometry and procedural thinking through play.

Constructionism's influence outlasted Logo itself. It shaped later movements toward project-based learning, makerspaces, and the use of programmable toys, and it stood as a standing critique of drill-and-practice CAI: the computer as a tool for expression, not only for delivery.

Online and Blended Learning

The internet changed the field's scale. From the 1990s, learning management systems such as WebCT and Blackboard allowed courses to be delivered partly or wholly online, and blended learning — combining face-to-face and online instruction — became widespread in higher education and, later, schools. The 2000s and 2010s brought massive open online courses (MOOCs), which enrolled hundreds of thousands in single courses, and the COVID-19 pandemic of 2020 forced a sudden global shift to remote instruction, exposing both the reach and the inequities of online learning.

The design questions remained the old ones in new clothes: how to keep learners active, how to give feedback at scale, how to build community at a distance. The field's accumulated design knowledge proved more valuable than any single platform.

Connectivism and the Networked Claim

In the 2000s, George Siemens and Stephen Downes proposed connectivism, arguing that in a networked, rapidly changing world, learning consists less in acquiring a body of knowledge than in forming and maintaining connections — to people, resources, and information sources — and in the capacity to know where to find knowledge when needed4. MOOCs, in their original "cMOOC" form, were built on connectivist principles of learner-driven networks5. Connectivism was influential as a description of learning in a networked age, though critics questioned whether it was a learning theory at all or a pedagogy for a particular context.

Learning Analytics and the Data Turn

As online learning generated streams of data, a new specialty emerged in the 2010s: learning analytics, the measurement, collection, and analysis of learner data to understand and improve learning. Its practitioners mine log data from learning management systems to predict which students are at risk of failing, to visualize engagement, and to test which instructional interventions work. Learning analytics descends directly from the field's oldest ambitions — individualization, feedback, and evidence — but it raises new questions about privacy, the ethics of prediction, and whether what is easily measured is what matters.

The Present

Educational technology today is less a single method than a set of overlapping practices. Instructional designers work in schools, universities, corporations, and online course providers, using systematic models to plan instruction. Constructionist and project-based approaches flourish in makerspaces and coding curricula. Online and blended learning are ordinary parts of institutional life. Learning analytics and, more recently, machine learning and generative AI tools are pushing the field toward automated feedback and adaptive systems, reviving the old dream of instruction fitted to each learner.

The questions that remain open are the ones its history keeps returning to: whether new technologies change learning or merely deliver it differently; how to design for deep understanding rather than surface recall; and who benefits when instruction is scaled, automated, and measured. The field's most durable lesson is that the medium is never enough — the design is what teaches.

References

  1. Educational technology (introduction, Definition, Related terms), Wikipedia
  2. Instructional design, Wikipedia
  3. Constructionism (learning theory), Wikipedia
  4. Connectivism (History), Wikipedia
  5. Massive open online course, Wikipedia

The frameworks of Educational Technology

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1900-1930

    Visual Instruction

    Should lessons be seen rather than only read?

    Fly to its tower in Nousfarer

  2. 1925-1970

    Audiovisual education

    Can film and sound teach as well as a teacher?

    Fly to its tower in Nousfarer

  3. 1954-1975

    Programmed learning

    Can instruction be broken into small reinforced steps?

    Fly to its tower in Nousfarer

  4. 1960-1995

    Computer-Assisted Instruction

    Can a machine tutor each learner individually?

    Fly to its tower in Nousfarer

  5. 1965-Present

    Instructional Design

    How should instruction be planned and tested in advance?

    Fly to its tower in Nousfarer

  6. 1980-Present

    Constructionism

    Do children learn best by building shareable things?

    Fly to its tower in Nousfarer

  7. 1995-Present

    Blended Learning

    How should face-to-face and online teaching be combined?

    Fly to its tower in Nousfarer

  8. 1995-Present

    Online and Blended Learning

    Can courses be taught well at a distance?

    Fly to its tower in Nousfarer

  9. 2005-Present

    Connectivism

    Is learning mainly forming connections in a network?

    Fly to its tower in Nousfarer

  10. 2010-Present

    Learning Analytics

    What can learner data reveal about improving learning?

    About Learning Analytics
    Learning analytics emerged in the early 2010s as educators and institutions faced an unprecedented flood of data from digital learning environments. The core question driving the framework is simple but transformative: what can we learn from the traces learners leave behind as they interact with educational systems, and how can those insights improve learning outcomes, institutional effectiveness, and learner agency? Unlike earlier educational technology frameworks that focused on designing instruction or supporting learner cognition, learning analytics treats the data generated by learning itself as a primary resource for understanding and improving education. It is not a theory of how learning happens—it draws on existing theories for that—but a practical, data-intensive approach to making learning more visible, more responsive, and more equitable.
    At its heart, learning analytics follows a cycle: the learning analytics cycle begins by capturing data from learning environments (such as learning management systems, online forums, or digital textbooks), then processing that data to produce metrics, visualizing patterns, and finally using those insights to intervene in and improve the learning process. The cycle reflects a commitment to data-driven decision making as the foundation of educational improvement. Within this cycle, several clusters of concepts form the core curriculum.
    First, the foundational cluster: data-driven decision making asks educators and administrators to base choices on evidence rather than intuition alone. Student success analytics applies this principle specifically to predicting and supporting positive outcomes such as retention, graduation, and course completion. Educational data mining provides the computational methods—clustering, classification, pattern detection—that turn raw log files into interpretable patterns. All three rest on the premise that learning produces measurable signals worth interpreting.
    Second, the predictive and intervention cluster: predictive modeling uses historical data to forecast individual learner outcomes, such as the likelihood of dropping a course or earning a low grade. These models feed into early warning systems, which flag at-risk students before they fail. The goal is to generate actionable insights—not just interesting reports, but findings that directly inform intervention strategies, such as targeted tutoring, course redesign, or academic advising. Course recommendation systems extend this logic by suggesting optimal learning paths based on a learner's profile and past behavior.
    Third, the visualization and tools cluster: learning dashboards display analytics results to instructors, students, or advisors in real time, often using visualization of learning data—charts, networks, progress bars—to make patterns immediately understandable. Learning analytics tools and platforms (such as learning record stores, analytics engines, and dashboard interfaces) are the technical infrastructure that makes the cycle operational. Assessment analytics applies the same data-driven lens to evaluating both learner performance and the quality of assessments themselves.
    Fourth, the ethical and profiling cluster: learner profiling aggregates data to build composite pictures of student behavior, preferences, and risk levels. This capability raises ethical considerations in learning analytics, a crucial area of the curriculum that addresses privacy, consent, algorithmic bias, transparency, and the risk of reducing learners to data points. The field's practitioners must constantly balance the promise of personalized insight against the dangers of surveillance and stereotyping.
    The intellectual history of learning analytics in educational technology begins not with a single founder but with a convergence of pressures and capabilities. By the late 2000s, the growth of learning management systems and digital courseware had generated vast data sets—what came to be called big data—that simply did not exist in earlier eras. Concurrently, techniques from machine learning, data mining, and predictive analytics matured enough to be applied to educational questions. The field was not born in a laboratory but at conferences and in policy documents. In 2011, the first International Conference on Learning Analytics and Knowledge (LAK) was held, and the Society for Learning Analytics Research (SoLAR) was founded soon after, giving the field an institutional identity. Seminal early works included the 2011 LAK proceedings and a 2012 special issue of the Journal of Educational Technology & Society, which together articulated the learning analytics cycle as the field's organizing logic.
    Key figures in the framework's formation include George Siemens (a key codifier of connectivism as well), who helped convene the early analytics community; Phil Long, who co-authored early pieces linking analytics to educational transformation; and researchers such as Ryan Baker, who developed rigorous educational data mining methods, and Abelardo Pardo, who advanced the practical application of analytics dashboards. The framework drew inspiration from business intelligence but adapted it for pedagogical goals rather than corporate efficiency. It also built on earlier work in user modeling and adaptive learning systems, though those traditions had focused more on real-time adaptation than on institutional-level pattern recognition.
    Learning analytics coexists with, rather than replaces, earlier frameworks such as connectivism (which emphasizes networked learning) and constructivist learning environments (which stress learner agency). What distinguishes learning analytics is its insistence on empirical, data-backed feedback as a driver of improvement, whether for individual learners, instructors, or entire institutions. By the 2020s, the framework had become institutionalized: learning analytics courses and programs appear in graduate curricula, and the language of "actionable insights" and "early warning systems" is now standard in educational technology discourse. However, the field remains acutely aware of its risks. Critics within the community have warned against learner profiling that reinforces inequity, and the ethical considerations in learning analytics are increasingly treated not as an afterthought but as a foundational concern that must shape the design of tools and interventions from the start.
    The 13 concepts its tower teaches
    • Actionable insights
    • Course recommendation systems
    • Data-driven decision making
    • Early warning systems
    • Educational data mining
    • Ethical considerations in learning analytics
    • Intervention strategies
    • Learning analytics cycle
    • Learning analytics tools and platforms
    • Learning dashboards
    • Predictive modeling
    • Student success analytics
    • Visualization of learning data

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Visual InstructionAudiovisual educationProgrammed learningComputer-Assisted InstructionInstructional DesignConstructionismBlended LearningOnline and Blended LearningConnectivismLearning Analytics

Beyond the field

Where Educational Technology's ideas came from in other fields, and what they led to there: the strongest 12 of 56 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Educational Technology is a planet in Nousfarer

Its city has a tower for each of these 10 frameworks. Land, walk in, and climb one concept at a time.

Land on Educational Technology

Related fields

Fields its ideas are linked to

More in Education

See where Educational Technology sits on the Map of Human Knowledge