Fact-checked by Grok 2 weeks ago

Programmed learning

Programmed learning, also known as programmed instruction, is an educational methodology that presents instructional content in small, sequential steps—often called "frames"—designed to elicit active responses from learners, followed by immediate feedback to confirm correct answers and correct errors, thereby promoting individualized mastery of material based on principles of operant conditioning.^[1]^[2] Originating in the early 20th century, the concept traces its roots to behaviorist psychology, with early precursors in the 1920s through Sidney Pressey's mechanical testing devices that automated multiple-choice quizzing and provided rudimentary self-instruction.^[2] The approach gained prominence in the 1950s when B.F. Skinner, building on his operant conditioning research, developed "teaching machines" to deliver precisely sequenced lessons, emphasizing reinforcement to shape learning behaviors without reliance on traditional teacher-led recitation.^[1]^[2] By the 1960s, programmed learning peaked in popularity, influencing widespread experiments in schools—such as those in Denver and on Long Island—and extending to printed booklets, filmstrips, and early computer-assisted instruction, though it faced resistance due to the need for extensive content restructuring and mixed empirical results on long-term efficacy.^[2]^[3] At its core, programmed learning adheres to several key principles: content is divided into minimal units to minimize errors (ideally 95-100% success rate per step), learners actively respond via constructed answers or selections rather than passive reading, and feedback is instantaneous to provide reinforcement.^[1] Two primary styles emerged—linear programming, pioneered by Skinner, which follows a fixed sequence assuming uniform learner paths, and branching programming, introduced by Norman Crowder, which allows adaptive paths based on learner errors to address individual differences.^[2] Notable contributors include Robert Mager, who formalized the writing of precise behavioral objectives to guide program design, and Fred Keller, who integrated programmed elements into personalized systems of instruction.^[1]^[2] Despite its decline as a standalone method by the 1970s amid critiques of overemphasizing rote memorization and underemphasizing creativity, programmed learning profoundly shaped modern instructional design, laying groundwork for adaptive e-learning technologies, competency-based education, and formative assessment practices that prioritize learner agency and data-driven feedback. Recent studies as of 2023 continue to affirm its effectiveness in promoting behavior change and engagement in science education.^[1]^[3]^[4]

History

Origins and Early Influences

The origins of programmed learning can be traced to early 20th-century psychological theories emphasizing observable behavior and stimulus-response associations, which laid the groundwork for structured, incremental instruction. Ivan Pavlov's experiments on classical conditioning in the early 1900s demonstrated how neutral stimuli could elicit reflexive responses through repeated pairing with unconditioned stimuli, establishing a foundational model for associative learning that influenced later educational applications.^[5] John B. Watson, in his 1913 manifesto, advanced behaviorism by rejecting introspection and advocating for psychology as the study of stimulus-response (S-R) connections, arguing that all behavior, including learning, could be shaped through environmental stimuli and observable responses.^[6] Building on these ideas, Edward Thorndike's connectionism, introduced in his 1911 work Animal Intelligence, posited that learning occurs through the formation of bonds between stimuli and responses, strengthened by trial-and-error experiences. Thorndike outlined three key laws of learning: the law of readiness, which states that learning is more effective when the learner is motivated and prepared; the law of exercise, emphasizing that repeated S-R connections grow stronger through practice; and the law of effect, where responses followed by satisfaction are reinforced and more likely to recur.^[7] These principles provided a theoretical basis for breaking down complex learning into manageable, sequential steps with immediate feedback, core to programmed learning. In the 1920s, Sidney Pressey, a psychology professor at Ohio State University, translated these behavioral theories into practical devices for self-testing and instruction. Pressey developed early mechanical aids, such as key-press machines for multiple-choice quizzes, which automated testing and provided immediate scoring to reinforce correct responses.^[8] He first demonstrated prototypes at Ohio State in 1924 during introductory psychology courses and at professional meetings, including the American Psychological Association gatherings in 1924 and 1925, showcasing their potential to individualize drill and practice while freeing instructors from routine grading.^[9] In 1926, Pressey described his "automatic teacher" in detail—a compact device resembling a typewriter that presented questions via inserted cards, allowed students to select answers by pressing keys, and dispensed scored results or corrective prompts instantly.^[10] Despite their innovations, Pressey's devices highlighted early limitations in programmed approaches, particularly the absence of adaptive feedback mechanisms that could branch instruction based on learner errors, restricting them to linear multiple-choice formats without personalized remediation.^[11] These foundational efforts influenced later developments, such as B.F. Skinner's mid-century refinements that addressed such constraints through operant conditioning principles.^[11]

Emergence in the Mid-20th Century

The emergence of programmed learning in the mid-20th century was marked by the pioneering work of psychologist B.F. Skinner, who developed the concept of the "teaching machine" as a practical application of his operant conditioning principles. In 1953, while observing his daughter's fourth-grade classroom, Skinner recognized the limitations of traditional group instruction and conceived a mechanical device to deliver individualized learning through small, sequential steps with immediate feedback.^[12] By 1954, he had built the first prototype of this teaching machine at Harvard University, designed to present material in manageable units, require active student responses, and provide positive reinforcement for correct answers to shape behavior effectively.^[13] This approach drew on operant conditioning, where behaviors are strengthened through reinforcements, emphasizing gradual progression to build complex skills without frustration or boredom.^[11] Skinner's innovations gained visibility through public demonstrations and scholarly publication. In 1957, he showcased prototypes at Harvard, highlighting their potential to automate routine teaching tasks and allow self-paced learning.^[14] This culminated in his influential 1958 paper, "Teaching Machines," published in Science, where he advocated for devices that enable students to advance at their own speed, receiving confirmation of mastery before proceeding, thus optimizing the learning process through personalized contingencies of reinforcement.^[15] The paper built on earlier inspirations, such as Sidney Pressey's 1920s mechanical testing devices, which demonstrated automated feedback but lacked comprehensive instructional sequencing.^[16] Parallel to Skinner's linear model, Norman A. Crowder introduced an alternative branching programming approach in 1960 through his work on "automatic tutoring." In his book and related publications, Crowder described a system where learners select from multiple-choice responses, with incorrect answers directing them to remedial branches for targeted review before returning to the main path, thereby accommodating diagnostic feedback and learner variability.^[17] This intrinsic method contrasted with Skinner's emphasis on uniform steps, promoting adaptive instruction based on error analysis. These developments aligned with broader U.S. educational reforms, including the 1958 National Defense Education Act, which allocated significant federal funding—over $1 billion in the early 1960s—for innovative teaching technologies to bolster science and math education amid Cold War pressures.^[18] Central to Skinner's framework were reinforcement schedules, such as fixed-ratio (consistent rewards after set responses) and variable-ratio (unpredictable rewards to sustain engagement), which facilitated behavioral shaping in teaching machines by mimicking effective operant conditioning techniques from his laboratory research.^[11]

Core Principles

Definition and Key Characteristics

Programmed learning is a self-paced instructional method designed to facilitate effective learning by dividing educational content into small, sequential units known as frames, each followed by immediate feedback to confirm or correct the learner's response and reinforce correct behaviors. This approach ensures that learners actively engage with the material, progressing only after demonstrating comprehension of each unit.^[17]^[19] Central to programmed learning are several key characteristics that distinguish it as a structured, learner-centered technique. Learners must provide active responses, either through constructed answers (such as writing or verbalizing) or multiple-choice selections, which promotes deeper processing and retention. The method emphasizes errorless learning by presenting information in minimal steps that build gradually from simple to complex concepts, minimizing frustration and errors. Progression requires mastery of each frame, typically aiming for success rates of 95-100% per unit to build confidence and ensure solid foundational knowledge. Additionally, individualized pacing allows learners to advance at their own speed, accommodating diverse abilities without the pressure of group synchronization.^[4]^[17]^[20] The theoretical foundation of programmed learning is firmly rooted in behaviorist principles, particularly the stimulus-response-reinforcement paradigm, where learning occurs through the shaping of observable behaviors via consistent positive reinforcement. This draws from operant conditioning, positing that immediate feedback acts as a reinforcer to strengthen correct responses and extinguish errors, leading to efficient skill acquisition. In contrast to traditional lecturing, which often involves passive absorption of information with delayed or no feedback, programmed learning shifts control to the learner through iterative feedback loops that enable self-correction and personalized mastery. It primarily operates through two systems—linear programming, which follows a fixed sequence, and branching programming, which adapts paths based on responses—though these are elaborated separately.^[19]^[21]^[17]

Linear vs. Branching Programming

Linear programming, also known as Skinnerian programming, follows a fixed sequence of instructional frames that all learners progress through in the same order, emphasizing small, incremental steps to build knowledge without deviation.^[11] Developed by B.F. Skinner in the 1950s, this approach draws from operant conditioning principles, where each frame presents a brief stimulus—typically 5-10 words or a short phrase—followed by a constructed response that the learner must actively produce, such as writing or selecting a completion.^[11] The design assumes that with properly sized steps, errors can be minimized to near 100% accuracy, as immediate feedback reinforces correct responses and prevents the strengthening of incorrect behaviors; for instance, in arithmetic drills, a frame might prompt "7 + 5 = __" with the answer revealed only after the learner's input.^[11] This method promotes steady progression and habit formation by ensuring consistent reinforcement across all users. In contrast, branching programming, or Crowderian programming, introduces multiple pathways through the material, allowing the program to adapt based on the learner's responses to errors.^[22] Pioneered by Norman A. Crowder in the late 1950s, it structures content into larger frames—often 30-70 words or a full paragraph—followed by multiple-choice questions with at least three options, where incorrect selections direct the learner to remedial branches for targeted explanation, while correct ones advance to more challenging material.^[22] This intrinsic approach accommodates variability in learner prior knowledge, handling error rates of 20-30% through feedback loops and sub-sequences, as seen in diagnostic quizzes on topics like physics concepts, where a wrong answer might branch to a review of foundational principles before retrying.^[22]^[23] The primary distinction between linear and branching programming lies in their rigidity versus adaptability: linear systems enforce a uniform path to foster reliable mastery in rote or procedural skills, whereas branching enables personalized remediation but can introduce complexity in program design and navigation.^[23] Linear programming excels in scenarios requiring high accuracy and repetition, such as skill drills, by aiming for error-free advancement that builds automaticity, though it may frustrate advanced learners unable to skip ahead.^[11] Branching, however, suits diagnostic or conceptual learning by addressing individual differences, offering enrichment for correct responses and correction for errors, which supports diverse ability levels but risks overwhelming authors with extensive branching logic.^[22] Overall, linear promotes efficient, habit-based learning in controlled environments, while branching enhances engagement through tailored feedback, though both stem from mid-20th-century behaviorist influences adapted to instructional technology.^[23]

Methodologies

Development of Instructional Materials

The development of instructional materials for programmed learning begins with a thorough analysis of learning objectives, ensuring that the content aligns precisely with desired educational outcomes. This initial step involves identifying the target knowledge or skills, assessing learners' prior knowledge, and outlining the scope of the program to maintain focus and relevance.^[24] Following objective analysis, the content is sequenced into small, manageable units known as frames to facilitate gradual mastery. Each frame presents a stimulus—such as a question, statement, or prompt—followed by a required response from the learner, promoting active engagement and immediate reinforcement. Linear programming uses a fixed sequence of frames, while branching allows alternative paths based on responses, serving as foundational frameworks for this design process.^[25]^[26] Validation occurs through iterative tryouts with representative learners to refine pacing, clarity, and engagement.^[27] Key challenges include avoiding frame overload, where excessive information per step overwhelms learners and hinders retention, and ensuring feedback remains non-judgmental to foster positive reinforcement without discouraging errors. Developers must balance conciseness with completeness, often revising frames multiple times based on trial data. Standards emphasize aiming for a success rate of 95% or higher per step in final programs to minimize incorrect responses while promoting confidence, as advocated in Skinner's approach.^[28]^[29]

Implementation in Educational Settings

Programmed learning was integrated into classrooms primarily through the use of specialized teaching machines, printed booklets, and later rudimentary software, allowing students to engage with instructional materials in a structured, self-paced manner. In the 1960s, B.F. Skinner's linear teaching machines, such as those tested in Harvard's self-instruction rooms with booths for individual use, presented material in small frames on disks or cards, advancing only after correct responses. Teachers served as facilitators and monitors rather than primary lecturers, assigning tasks to machines to handle repetitive drills in subjects like spelling and arithmetic, thereby freeing time for more complex interactions. This setup enabled handling larger class sizes, with reports of classes expanding from 20-25 to 60 students while maintaining individualized progress.^[17]^[16] Self-study formats became central to programmed learning, utilizing standalone workbooks or dedicated lab stations where learners could work independently without constant supervision. By the late 1950s, materials shifted from mechanical devices to booklets like English 2600, which contained frames with questions and immediate answer reveals on the same or facing page, supporting solitary progression through topics. Lab stations equipped with branching machines, such as Norman Crowder's AutoTutor systems introduced in the early 1960s, allowed users to select responses leading to remedial or advanced paths, ideal for self-directed study in school libraries or after-hours settings. These formats emphasized the core feedback mechanisms of programmed learning, providing instant reinforcement to build mastery at the learner's rhythm.^[17]^[30] While programmed learning was designed for individual pacing, it proved scalable for group settings through rotations and monitoring tools like progress charts. In large classrooms, students rotated through machine stations or workbook areas in shifts, ensuring all could access materials without disrupting collective activities, as seen in early 1960s implementations in Denver and Long Island schools. Teachers tracked advancement via charts plotting frame completions, intervening only for persistent errors or motivation issues, which balanced individual autonomy with group management. This approach contrasted with traditional whole-class lecturing, accommodating diverse readiness levels within a single room.^[30]^[17] Effective implementation required specific prerequisites, including learner orientation to the method's demands and access to necessary hardware. Students needed initial guidance on active responding and avoiding answer peeking to foster self-discipline, often provided through brief demonstrations emphasizing the small-step progression. In 1960s setups, hardware like teaching machines was essential, with schools investing in devices for subjects such as mathematics, though by the mid-1960s, cheaper booklets reduced these barriers. Without proper orientation, adoption faced resistance, as noted in trials where unfamiliarity led to underutilization.^[17]^[30]

Evolution and Adaptations

Post-1960s Developments

In the late 1960s, programmed learning began shifting from its strict behaviorist foundations toward incorporating cognitive elements, as educators recognized the need for more structured hierarchies in skill acquisition. Robert Gagné's work on hierarchical learning, outlined in his 1965 book The Conditions of Learning, emphasized sequencing instructional content based on prerequisite knowledge and learner capabilities, influencing the design of programmed materials to better support complex cognitive processes beyond simple stimulus-response associations.^[31] Similarly, Robert F. Mager's 1962 publication Preparing Instructional Objectives promoted criterion-referenced instruction, focusing on measurable performance outcomes rather than normative comparisons, which refined programmed learning by ensuring frames targeted specific, observable behaviors.^[1] By the 1970s, programmed learning faced significant criticism for its perceived rigidity and overemphasis on rote memorization, leading to a decline in widespread adoption and publication of new materials, with annual outputs dropping sharply from earlier peaks. This period saw a resurgence in the 1980s, driven by the advent of microcomputers that extended systems like PLATO—originally developed in the 1960s at the University of Illinois—which allowed for more interactive and adaptive programming while building on linear and branching formats.^[32] Key publications advanced frame design quality during this era; for instance, Susan Meyer Markle's 1969 book Good Frames and Bad: A Grammar of Frame Writing provided guidelines for constructing effective instructional frames, stressing clarity and error avoidance to enhance learner engagement.^[33] Patrick Suppes' trials in computer-assisted instruction at Stanford, spanning the 1960s and 1970s, demonstrated practical applications in subjects like mathematics and logic, achieving high completion rates among elementary students through individualized pacing.^[34] Internationally, UNESCO promoted programmed instruction in the 1970s as an efficient tool for accelerating education in developing countries, supporting projects in regions like Central Africa to address teacher shortages and scale basic skills training.^[35] In the United States, military training programs adopted these methods extensively during the same period, with the Army integrating self-paced, programmed modules into technical and leadership courses from the mid-1970s onward to improve efficiency in large-scale soldier preparation.^[36]

Integration with Technology

Programmed learning has seen significant evolution through computer-based adaptations, particularly from the 1990s onward, building on early precursors like teaching machines to enable scalable, interactive instruction. One seminal example is the PLATO system, originally developed in the 1960s at the University of Illinois, which extended operations into the early 1990s as a networked, multi-user platform delivering individualized tutorials and simulations across thousands of terminals.^[37]^[38] This system pioneered features such as immediate feedback and branching paths, influencing later digital tools by demonstrating the feasibility of computer-assisted instruction for diverse learners. In modern contexts, learning management systems (LMS) like Moodle incorporate adaptive quizzes that emulate programmed learning principles, adjusting question difficulty based on user responses to optimize mastery.^[39] These platforms facilitate the creation of linear or branching modules, allowing educators to embed programmed sequences within broader curricula for self-paced progression. Advancements in artificial intelligence have further enhanced programmed learning by introducing machine learning-driven personalization, particularly in the 2010s. Tools like Duolingo employ adaptive algorithms that analyze user performance in real-time, dynamically selecting exercises akin to branching programming to target weaknesses and reinforce strengths, thereby mimicking a responsive tutor.^[40] This approach uses logistic regression and neural networks to predict optimal content paths, resulting in tailored language lessons that adapt to individual proficiency levels without rigid scripting. Such AI integrations extend the core tenets of programmed instruction—small steps, active response, and immediate reinforcement—into scalable, data-informed experiences accessible to millions. From 2023 onward, generative AI has advanced programmed learning by enabling dynamic generation of instructional content and hyper-personalized learning paths. Systems now use large language models to create customized frames, explanations, and assessments in real-time, adapting to learner needs while incorporating ethical guidelines for educational use. As of 2025, UNESCO's guidance emphasizes integrating generative AI to support equitable access and teacher augmentation in adaptive instruction.^[41] The proliferation of mobile and online technologies has broadened programmed learning's reach, with apps like Anki (launched in 2006) leveraging spaced repetition algorithms to schedule reviews based on user recall success, effectively implementing adaptive reinforcement schedules for long-term retention.^[42] In the 2020s, virtual reality (VR) simulations have emerged as a powerful extension for skill-based training, incorporating adaptive elements where scenarios adjust in complexity according to learner performance, such as in engineering or procedural tasks.^[43] These immersive environments provide immediate feedback through haptic and visual cues, enabling safe repetition of programmed sequences in simulated real-world contexts. Integration with technology has yielded notable benefits, including enhanced scalability for large user bases and advanced data analytics to track and refine learner paths. Studies on computer-based programmed instruction indicate efficiency gains, with meta-analyses showing reductions in instructional time by approximately 25-35% compared to traditional methods while maintaining or improving achievement outcomes.^[44] This allows for broader accessibility and personalized pacing, though implementation requires careful design to ensure equitable access across diverse settings.

Applications and Examples

Classroom and Training Uses

In K-12 education, programmed learning is commonly applied through drill-and-practice exercises to reinforce foundational skills in subjects such as mathematics and vocabulary, allowing students to progress at their own pace with immediate feedback on responses.^[45] For instance, computer-based programmed instruction has been effectively used in high school settings to teach health-related knowledge, such as AIDS prevention, by presenting sequential frames that build comprehension step-by-step.^[46] This approach supports self-paced modules in flipped classroom environments, where students complete structured instructional sequences independently before class discussions.^[2] In higher education, programmed learning facilitates remedial courses and skill-building in disciplines like languages and sciences, enabling learners to master complex concepts through incremental, interactive modules tailored to individual needs.^[47] A meta-analysis of studies confirms its effectiveness in improving academic achievement across various higher education contexts, particularly for aptitude-treatment interactions where students benefit from adaptive pacing.^[47] These applications allow instructors to focus on higher-order guidance while students engage in self-directed practice. Corporate training leverages programmed learning for compliance modules and sales simulations, providing scalable, repeatable instruction that ensures consistent skill acquisition among employees.^[48] Brief programmed instruction modules have been shown to enhance new hire training outcomes by offering immediate feedback and flexible scheduling, reducing the need for extensive instructor involvement.^[49] In military contexts, it supports procedure mastery through structured sequences that simulate operational tasks, as utilized in NATO training programs to standardize learning across diverse personnel.^[50] Key advantages of programmed learning in these settings include cost-effectiveness for large-scale rollouts, as it minimizes resource demands while maintaining instructional quality, and flexibility for diverse learners, accommodating varying paces and schedules without disrupting group dynamics.^[48]^[49] Modern integrations with technology, such as online platforms, further extend these benefits by enabling digital delivery of programmed sequences.^[3]

Real-World Case Studies

One notable historical example of programmed learning is the Roanoke Experiment conducted in 1959-1960, where B.F. Skinner's teaching machines were used to deliver linear programmed instruction in algebra to eighth-grade students at a level equivalent to ninth-grade curriculum. The project, supported by Harvard researchers, demonstrated that students could master the material through small, sequential steps with immediate feedback, paving the way for broader adoption of self-paced educational tools.^[51] In the 1960s, the U.S. Navy implemented linear programmed instruction for basic electronics training, as evaluated in a 1965 study at the Naval Air Technical Training Center in Memphis, Tennessee. Trainees using programmed materials on topics such as direct current circuits and electrical calculations completed the 26-hour conventional module in 19 hours, achieving a 27% reduction in instruction time while maintaining equivalent performance levels to traditionally instructed groups on criterion tests. This approach, emphasizing constructed responses and high success rates per frame, was part of broader efforts to individualize technical training for over 200 recruits.^[52] A modern adaptation appears in Khan Academy's mastery-based modules, launched in 2008 and refined throughout the 2010s, which incorporate branching elements—allowing adaptive paths based on performance—alongside instructional videos to reinforce concepts in subjects like mathematics. Research on implementations, such as a year-long intervention in U.S. schools, showed students using the platform improved end-of-year math scores by 0.12 to 0.22 standard deviations compared to controls, highlighting its role in addressing learning gaps through repeated practice until proficiency.^[53] IBM's adaptive e-learning programs in the 2010s, targeted at IT skills development for consultants, utilized personalized simulations that adjusted difficulty and content in real-time, blending linear progression with branching feedback to simulate real-world scenarios. This initiative upskilled over 100,000 learners, resulting in more than three times the participation rates of traditional methods and projected savings of over $13 million in training time by the end of year three, while reducing time to competence.^[54] Duolingo's language learning paths, structured as gamified programmed sequences since 2011, employ linear steps with branching remediation for errors, promoting incremental mastery of vocabulary and grammar. An independent 2012 effectiveness study found that 84.1% of participants improved their proficiency scores after an average of 34 hours of use, equivalent to a full college semester's progress, underscoring high success rates in self-directed paths.^[55]

Impact and Debates

Long-Term Effects on Education

Programmed learning significantly influenced the development of personalized learning approaches in education, laying the groundwork for competency-based education reforms in the 1970s by emphasizing individualized pacing and mastery of content before progression.^[56] This approach directly inspired Benjamin Bloom's mastery learning model outlined in his 1968 paper, which advocated for structured units of instruction with feedback and corrective procedures to ensure all students achieve proficiency, building on the principles of immediate reinforcement and sequential skill-building from programmed instruction.^[57] By prioritizing learner autonomy and tailored feedback, programmed learning shifted pedagogical focus from uniform group instruction to adaptive methods that accommodate diverse learning rates.^[58] Research on the long-term outcomes of programmed learning reveals moderate positive effects on student retention and achievement, as evidenced by meta-analyses of related mastery learning programs. For instance, a comprehensive review of 108 studies found an average effect size of 0.52 for achievement gains, indicating that students in these programs outperformed those in traditional settings by about half a standard deviation, with similar benefits observed for retention over time.^[59] However, critics have noted that programmed learning often overemphasizes rote memorization and mechanical responses, potentially limiting deeper conceptual understanding and critical thinking skills, as students focus on correct answers rather than broader application.^[60] On the policy front, the principles of programmed learning contributed to broader educational reforms emphasizing evidence-based strategies for individualized instruction. Globally, UNESCO promoted the adoption of programmed learning techniques in the mid-20th century to enhance instructional efficiency in developing regions, influencing international programs aimed at scalable, self-paced education for underserved populations.^[61] In contemporary education, programmed learning serves as a foundational concept for AI-driven tutors and intelligent tutoring systems, which automate adaptive feedback and personalization to improve access and equity. A 2016 meta-analysis reported a median effect size of 0.66 for these systems overall.^[62] Recent 2020s studies demonstrate benefits for disadvantaged students, such as a 0.3 standard deviation gain in a 2025 pilot in Nigeria and 4 percentage points higher topic mastery for students from underserved communities in a 2024 U.S. study.^[63] This evolution underscores programmed learning's enduring role in fostering inclusive pedagogies that bridge equity gaps in digital learning environments.^[63]

Learning vs. Training Distinctions

In programmed learning, the distinction between learning and training hinges on their respective emphases: learning prioritizes conceptual understanding and higher-order thinking skills, while training focuses on procedural mastery and rote behaviors. Learning-oriented applications utilize branching programs, which allow learners to explore multiple pathways based on their responses, fostering problem-solving and deeper comprehension by presenting scenarios that require analysis and adaptation. For instance, Norman Crowder's branching model (1954) enables intrinsic feedback tailored to errors, promoting exploratory engagement rather than strict sequencing.^[64] In contrast, training employs linear programs, derived from B.F. Skinner's behaviorist framework, which deliver content in fixed, incremental steps with immediate reinforcement to build automatic responses, such as compliance drills or motor skill repetition. This approach suits skill acquisition in vocational settings, like assembly line procedures, by minimizing errors through repetitive practice.^[65] The behaviorist foundations of programmed learning have drawn significant debate, with critics highlighting its inherent bias toward training over holistic learning. Rooted in operant conditioning, early programmed instruction emphasized observable behaviors and reinforcement, often sidelining cognitive processes like creativity or knowledge construction, which aligns more closely with training objectives than broader educational goals.^[66] Constructivists, such as Seymour Papert in his 1980 book Mindstorms, argued that such directive methods stifle discovery, advocating instead for child-centered environments where learners actively build understanding through tools like Logo programming, rather than passive reception of pre-structured content. Papert critiqued behaviorist "teaching machines" for treating learners as passive recipients, contrasting this with constructionist principles that encourage experimentation and personal meaning-making in the 1980s and beyond.^[67] Contemporary hybrids in programmed learning integrate training's structure with learning's exploratory elements, often through gamification, to balance procedural efficiency with conceptual depth. These blends incorporate game mechanics like points, badges, and adaptive challenges into digital platforms, allowing linear drills to evolve into interactive simulations that reward both skill mastery and creative problem-solving. For example, gamified programming education has demonstrated enhanced engagement by combining behaviorist feedback loops with constructivist exploration, leading to improved motivation and application.^[68] Evidence from instructional studies indicates that training-focused programmed methods, with their immediate reinforcement, excel in short-term retention—such as through overt responding techniques that boost delayed recall—compared to purely exploratory approaches, though long-term conceptual gains benefit from hybrid designs.^[69]

References

[1]
[PDF] A History of Instructional Media, Instructional Design, and Theories
Mar 9, 2020 · The theory and practice of adaptive teaching systems. In R. Glaser (Ed.),. Teaching machines and programmed learning II: Data and directions.
[2]
Programmed Instruction - EdTech Books
Programmed instruction (PI) was devised to make the teaching-learning process more humane by making it more effective and customized to individual differences.
[3]
The Synthesis of Programmed Instruction and Online Education - PMC
Jul 27, 2020 · Skinner (1968) laid out a science of teaching derived from operant conditioning principles and provided methods for adopting programmed ...
[4]
Classical Conditioning - StatPearls - NCBI Bookshelf - NIH
In Pavlov's experiment, the ringing of the bell became the conditioned stimulus, and salivation was the conditioned response. Essentially, the neutral stimulus ...
[5]
Psychology as the Behaviorist Views it. John B. Watson (1913).
Psychology as the behaviorist views it is a purely objective experimental branch of natural science. Its theoretical goal is the prediction and control of ...
[6]
Thorndike (1911) Chapter 5 - Classics in the History of Psychology
The Law of Exercise is that: Any response to a situation will, other things being equal, be more strongly connected with the situation in proportion to the ...
[7]
Sidney Pressey and the Automation of Education, 1924-1934
Aug 6, 2025 · Produced in the 1920s, the Automatic Teacher automated clerical and diagnostic skills and redefined productivity in terms of individualization.Missing: 1926 Ohio 1924-1926
[8]
The Automatic Teacher - Hack Education
Feb 4, 2015 · Ohio State University professor Sidney Pressey first displayed the prototype of his “automatic intelligence testing machine” at the 1924 ...
[9]
The Automatic Teacher - b2o: boundary 2 online
Here's how Pressey described the machine, which he branded as the Automatic Teacher in his 1926 School and Society article: The apparatus is about the size of ...Missing: demonstrations | Show results with:demonstrations
[10]
[PDF] Teaching Machines - B. F. Skinner Foundation
But they also had limitations which probably con- tributed to their failure. Pressey was working against a background of psycho- logical theory which had not ...
[11]
Education Technology and Skinner's Box
Feb 10, 2015 · Skinner describes how he came upon the idea of a teaching machine in 1953: Visiting his daughter's fourth grade classroom, he was struck by the ...
[12]
B.F. Skinner Demonstrates His "Teaching Machine," the 1950s ...
Jul 15, 2021 · Skinner created what he called the Teaching Machine in 1954 while he taught psychology at Harvard. He was hardly the first to design such a device.
[13]
1957: Skinner Teaching Machine. - Marketplace.org
May 14, 2014 · The below video shows Skinner making his case for why educators need the machine.Missing: Harvard | Show results with:Harvard
[14]
Teaching Machines - Science
Teaching Machines: From the experimental study of learning come devices which arrange optimal conditions for self-instruction. B. F. SkinnerAuthors Info & ...
[15]
The Far-Reaching Legacy of the First 'Teaching Machines' | NEA
Oct 9, 2021 · Before Skinner, there was Sidney Pressey, psychology professor at Ohio State University and standardized testing pioneer, who created a ...Missing: 1924 1926<|control11|><|separator|>
[16]
Programmed Instruction - EdTech Books
Pressey, S.L. (1926). A simple apparatus which gives tests and scores – and teaches. School and Society, 23 (586), 373-376. Pressey, S.L. (1927). A machine ...<|control11|><|separator|>
[17]
The Unlearned Lessons of Programmed Learning
Feb 20, 2020 · In the US, two billion dollars were spent on educational technology by the government in the decade following the passing of the National ...Missing: funding | Show results with:funding
[18]
https://adaptivelearninginelt.wordpress.com/2020/02/20/the-unlearned-lessons-of-programmed-learning/
[19]
A systematic review of programmed learning approach in science ...
Programmed instruction is learning according to a pre–developed program, which provides for the actions of both students and teacher (or a learning machine that ...Missing: key behaviorist
[20]
Mastery Learning: An Effective Teaching Strategy - Nyu
In the early 1960s the idea of mastery learning was revived in the form of programmed instruction. ... His prediction was that 95% of the students taught by the ...
[21]
[PDF] Programmed Learning.
Because the re-discovery of programmed learning principles may by one of the most significant contributions to human learning in this century. For the first ...
[22]
[PDF] A GUIDE TO PREPARING INTRINSICALLY PROGRAMMED ... - DTIC
Norman Crowder, Principal Investigator for this contract, are the authors. ... The two major approaches are called intrinsic programming and linear programming.
[23]
[PDF] linear and branch programming in programmed learning
The credit of linear programming style goes to B. F. Skinner. Linear programming style is related to “operant conditioning”.
[24]
Programmed Instruction: A Step-by-Step Approach to Learning
Dec 27, 2023 · Branching programming is more personalized than linear programming. It allows for differentiation based on the learner's performance. Learners ...
[25]
Programmed Instruction Educational Model - eLearning Industry
Apr 28, 2025 · Summary: 1954 - B. F. Skinner and J. G. Holland implemented Programmed Instruction Educational Model at Harvard.Missing: citation | Show results with:citation
[26]
[PDF] Introduction to Programed Instruction - Scholarly Commons
The material to be learned is broken down so that one small unit of information is presented in each frame. The frames develop the material step by step in ...
[27]
ED289495 - How To Design Programmed Learning Materials ... - ERIC
A sequel to the booklet "How to Produce Printed and Duplicated Materials," this booklet begins by providing an overview of programmed learning.Missing: steps | Show results with:steps
[28]
Gagné's Nine Events of Instruction - Northern Illinois University
1. Gain attention of the students · 2. Inform students of the objectives · 3. Stimulate recall of prior learning · 4. Present the content · 5. Provide learning ...
[29]
THE ROLE OF ERRORS IN LEARNING WITH FEEDBACK
in the literature on programmed instruction. Skinner (1 958) has supported a low error rate of 10 per cent. or less per frame, arguing that error tendencies.
[30]
Programmed learning - ScienceDirect.com
The linear style of programming was developed by B. F. Skinner. Linear Skinner program consists of small steps which every learner is able to pass without ...Missing: definition key aspects
[31]
(PDF) The Programmed Instruction Era: When Effectiveness Mattered
Aug 9, 2025 · Programmed instruction (PI) was devised to make the teaching-learning process more humane by making it more effective and customized to individual differences.
[32]
6.2: Robert Gagné And The Systematic Design Of Instruction
Jun 13, 2021 · Gagné posited that not all learning is equal and each distinct learning domain should be presented and assessed differently. Therefore, as an ...
[33]
PLATO | The Grainger College of Engineering | Illinois
PLATO was a distributed computer-based learning system that was created at the University of Illinois. It was the first generalized computer assisted ...Missing: revival | Show results with:revival
[34]
good frames and bad--a grammar of frame writing. - ERIC
GOOD FRAMES AND BAD--A GRAMMAR OF FRAME WRITING. MARKLE, SUSAN MEYER. THE SUBJECT OF PROGRAMING IS TREATED BY MEANS OF A PROGRAMED TEXT. ORIGINALLY WRITTEN ...Missing: 1969 | Show results with:1969
[35]
[PDF] Four Programs in Computer-assisted Instruction1 - Suppes Corpus
Four of the major programs in computer-assisted instruction developed in the Institute for Mathematical Studies in the Social Sciences at Stanford.
[36]
Programmed instruction - UNESCO Digital Library
... education in developing countries. Being an accelerated training tech- nique, it could speed the process of improving education in these countries. It ...
[37]
PLATO | Computer-Based Learning & Education System - Britannica
PLATO, computer-based education system created in 1960 by Donald L. Bitzer at the University of Illinois at Urbana-Champaign (UIUC).Missing: revival | Show results with:revival
[38]
PLATO and the History of Education Technology (That Wasn't)
Jan 25, 2018 · PLATO is remembered (in education technology circles at least), it is as an early example of computer-assisted instruction – and often, it's denigrated as such.Missing: revival 1980s
[39]
Adaptive Quiz: CAT (Computer-Adaptive Testing) implementation for ...
The Adaptive Quiz activity enables a teacher to create tests that efficiently measure the takers' abilities. Adaptive tests are comprised of questions ...
[40]
How Duolingo's AI Learns What You Need to Learn - IEEE Spectrum
Feb 5, 2023 · Duolingo uses machine learning and other cutting-edge technologies to mimic these three qualities of a good tutor. First, to ensure ...Missing: branching | Show results with:branching
[41]
Background - Anki Manual
Anki is a program that makes remembering things easy using active recall testing and spaced repetition, and is more efficient than traditional methods.
[42]
Features of adaptive training algorithms for improved complex skill ...
This study investigates the effects of responsiveness, personalization, and subtask independence on the efficacy of automated training algorithms in VR.<|control11|><|separator|>
[43]
[PDF] Effectiveness of Computer-Based College Teaching
Computer-based instruction made small but significant contributions to course achievement, had small positive effects on attitudes, and reduced instruction ...Missing: efficiency | Show results with:efficiency
[44]
[PDF] Effect of Programmed Instruction on Students' Attitude Towards ...
programmed learning for use in teaching in schools (Gagne, 1975). The terms programmed learning (PL) and programmed instruction (PI), are used synonymously ...
[45]
An Evaluation of Computer-Based Programmed Instruction for ... - NIH
An Evaluation of Computer-Based Programmed Instruction for Promoting Teachers' Greetings of Parents by Name ... The mastery criterion for all training levels was ...
[46]
Effectiveness of Programmed Instruction in Higher Education: A ...
Effectiveness of Programmed Instruction in Higher Education: A Meta-analysis of Findings. ... Descriptors: Academic Achievement, Aptitude Treatment Interaction, ...
[47]
Programmed Instruction in Industrial Training
... programmed instruction in industrial training. Programmed learning has many advantages like learning effectiveness, economy in time, economy in cost. Higher ...<|control11|><|separator|>
[48]
[PDF] Using Programmed Instruction Modules to Improve Training Outcomes
Mastery criteria of the PIMs was set at simply earning a score of 100 ... programmed instruction may help determine how long programmed instruction modules should.Missing: rate | Show results with:rate
[49]
Programmed Instruction in Military Training in the NATO Nations.
Programmed Instruction in Military Training in the NATO Nations. Nagay, John A. The state of programed instruction utilization in military training within ...
[50]
Why history matters: A review of Watters's Teaching Machines, the ...
Jun 9, 2024 · The Roanoke Experiment commenced as a pilot project that involved teaching ninth-grade algebra to eighth-grade students using teaching machines ...Missing: 9th arithmetic<|control11|><|separator|>
[51]
[PDF] COMPARISON OF CONVENTIONAL AND PROGRAMED ... - DTIC
Reduction in Training Time. It was found that basic electronics students can learn a fairly large block of material by means of pro- gramed instruction to about ...
[52]
Multiple Studies Show Khan Academy Drives Learning Gains
Nov 16, 2023 · Students who used Khan Academy as part of a year-long mastery learning intervention improved their end-of-year math scores by .12 to .22 ...Missing: quantitative | Show results with:quantitative
[53]
IBM upskill 100k learners with adaptive simulations - ETU
Being effective and efficient with learners' time was critical, so IBM chose adaptive learning simulations to maximize both efficiency and effectiveness.Missing: e- 2010s
[54]
[PDF] Duolingo Effectiveness Study - The Owl App
The research study of Duolingo effectiveness was independently conducted in September-н‐. November of 2012. The study lasted for approximately eight weeks.
[55]
What the History of Personalized Learning Teaches Us About It Today
Jan 23, 2024 · Personalized learning is an educational approach that tailors instruction to meet the unique needs and progress of each individual learner.
[56]
[PDF] Formative classroom assessment and Benjamin S. Bloom - ERIC
(Bloom, 1968, pp. 7-8). Bloom also emphasized the need for instruction and assessments in mastery learning classrooms to focus on higher level learning goals, ...
[57]
Sage Reference - Mastery Learning
Skinner's emphasis on individualized learning and the importance of feedback and reinforcement, a revival of mastery learning began with programmed instruction, ...
[58]
Effectiveness of Mastery Learning Programs: A Meta-Analysis
Mastery programs have positive effects on student attitudes toward course content and instruction but may increase student time on instructional tasks. In ...
[59]
[PDF] An Introduction to Programmed Learning - Eastern Illinois University
May 15, 2020 · The extreme difficulty of programming a branched program and the questionable sequence of steps are primary reasons for the lack of enthusiasm.
[60]
H.R.1 - 107th Congress (2001-2002): No Child Left Behind Act of 2001
No Child Left Behind Act of 2001 - Amends the Elementary and Secondary Education Act of 1965 (ESEA) to revise, reauthorize, and consolidate various programs.Text · Cosponsors (84) · Actions (172) · Committees (2)
[61]
Programmed learning: its development and structure
Meeting paper on programmed instruction as an effective means of improving parts of the learning process, reviews educational research.Missing: programs | Show results with:programs
[62]
https://journals.sagepub.com/doi/10.3102/0034654315581420
[63]
AI tutoring: Bridging the educational disadvantage gap - My College
Jun 9, 2025 · This article examines how AI tutoring could address the educational attainment gap for socio-economically disadvantaged students.Missing: 2020s | Show results with:2020s
[64]
[PDF] Types of Programmed Instruction or Programmed Learning
The size of the frames is quite large in branching programming when compare to linear programming and instructional material is divided into 'units' of ...
[65]
Learning Theories in Focus: Behaviourism and programmed learning
Oct 23, 2019 · Behaviourism assumes the learner begins their experience as a blank slate (tabula rasa) and forms their learning through positive or negative ...
[66]
Behaviorism In Instructional Design For eLearning
May 12, 2021 · As the name implies, a behavioristic approach focuses on guiding learners to reach pre-established learning outcomes. Learning is considered to ...Missing: sources | Show results with:sources
[67]
[PDF] Mindstorms: children, computers, and powerful ideas
Papert, Seymour. Mindstorms: children, computers, and powerful ideas. Includes bibliographical references and index. 1. Education. 2. Psychology.
[68]
The effectiveness of gamification in programming education
This paper aimed at constructing a systematic framework and examining the effect of gamification in programming education through a meta-analysis.
[69]
The effect of overt versus covert responding to programed instruction ...
The control group scored significantly lower on both tests. Thus, overt responding appears to increase delayed retention.