Skip to content

Flipped Learning

Principle · All principles
Evidence · 33 claims (15 for, 17 mixed, 1 against) · 36 studies (12 causal, 10 review, 8 quant-synthesis, 3 associational, 2 theoretical, 1 qualitative), q1–q4 · 10 of 36 report an effect size · 26 claims rest on one study

Conditional relationship

This page owns one scheduling decision: where the learner's first exposure to new material happens. For learners who can take in a topic's first explanation on their own (a short video, a reading, a worked example) and whose difficulty lies in applying it, moving that first exposure before the live session and spending the freed session on application with feedback (problems, discussion, explaining, coached practice) is expected to give better course outcomes (course-assessment, general-achievement, conceptual-understanding) than spending the same live time on the same exposition. The expected gain comes from what the live time is then used for, not from the relocation: a flip that moves the lecture to video and fills class with more lecture, or with review of the video, is not expected to help.

The relationship is conditional on four things. Preparation actually happens: the design assumes learners arrive having met the material, and many learners do not do assigned pre-class reading. The pre-class material can be understood without a teacher: a novice facing material with many interacting parts may need the explanation live, with someone to ask. The live time is used for application with feedback, at a dose that changes what learners do. Learners can reach the material (device, connection, time and a place to study outside class). The horizon recorded in the evidence is the end of a course; durable retention is not established here.

The relationship is tested, in one setting and at low certainty. A Campbell systematic review of undergraduate health-professions education pooled 44 studies (n = 7813) comparing flipped classes with traditional lecture-based classes on examination scores or grades, and found the flipped classes ahead (SMD = 0.57, 95% CI 0.25 to 0.90), with very high heterogeneity (I² = 98%) (Flipped classroom learning improves academic performance of undergraduate health professional students… [+M]). The advantage held when eleven studies with imputed data were removed (SMD = 0.54, 33 studies) (The academic performance advantage of flipped classrooms persists when studies with imputed data are excluded… [+M]). Three limits travel with it: only 11 of the 45 included studies were randomised and the review judged the risk of bias in the non-randomised ones high (Only a minority of flipped-classroom studies in health professional education are RCTs… [~M]); the heterogeneity means flipped courses varied widely in how much they gained, which is what the conditions above are meant to explain, though the review as recorded does not test them; and the claim pages do not say whether home time, content or assessment were held equal, so the average may include extra hours (Time-on-task is frequently unequal… [~M]). It is one synthesis, in one field, at the end of a course. Beside it sit two flipped variants compared with each other (one calculus study), blended courses compared with face-to-face ones (a meta-analysis whose blended arms usually also had extra time and resources), and active learning compared with lecture (meta-analyses in undergraduate STEM), each read below for what it does and does not show. The steps that follow are a default design: the review supports the direction of the flip, not these doses, which remain proposals.

How the neighbouring converted pages sit beside this one: Active Learning owns the relationship between what learners do in class and what they learn; this page uses it for the live session and does not repeat it. The Blended Learning pattern owns the wider decision of dividing a course between online and in-person work; flipping is one configuration inside it (exposition online, application in person). The Team-Based Learning pattern is a complete flipped cycle with a readiness test and team application; use it when a stable team structure is wanted. Guided Practice owns how the coached practice in the live session is run.

Default design

The earlier page's guidance, made concrete. That a flipped course can beat a lecture course is tested (above); how to run one is not, so each step is a design proposal with its evidence status stated beside it. Quantities are labelled proposals where no claim gives them.

  1. Decide what moves out and what stays in. Move out only the first explanation of material learners can follow alone: definitions, a demonstration of a procedure, a worked example, a short reading. Keep in the live session anything that needs a teacher's observation, a partner, equipment or discussion, and any explanation of material learners repeatedly fail to follow alone. Untested; from the earlier page ("moving lecture out of class is not enough by itself").
  2. Keep the pre-class part short and active. Proposal: 10–15 minutes of video or 3–6 pages of reading per live session, in segments of a few minutes, with two or three questions embedded in or after it that ask the learner to predict, explain a step or answer an item before seeing the solution. Basis: a review reports that learners prefer short videos, with no length given (Guo et al. found engagement depends on video production style [+W], engagement on one platform, not learning); answering a question before studying the material improved retention of it in laboratory studies (Pretesting enhances learning [+M], carried from lab passages to pre-class work); prompting learners to explain improved immediate understanding (Prompting learners to self-explain… [+S], mostly lab and immediate tests). Designed pre-class material can beat the textbook: in one introductory physics study, students who prepared with multimedia learning modules scored about 13% higher than those who read the textbook on a retention test two weeks later (adjusted effect size 0.7, 45 completers) (The multimedia learning module advantage over the textbook persists on a retention test… [+M], one study, q2, the comparison is two forms of preparation, not flipped against unflipped). The minutes are an untested proposal.
  3. Make the preparation visibly matter, and check it at the start. Open each live session with a short individual check on the pre-class material (proposal: 3–5 items, 5 minutes, low stakes, answers shown at once), and start the first task from it. Never re-teach the whole video because some did not watch it; that teaches learners the preparation is optional. Basis: many students report rarely or never reading assigned material (Students report rarely reading their physics textbooks… [~M], one university physics course, self-report); in-class quizzing improved exam performance in school and college classes (Classroom quizzing…improves student performance on classroom exams [+M], reported second-hand in a review chapter). Learners track their own preparation better against a visible checklist and the check's results (Self-monitoring improves self-regulation… [~M], theoretical syntheses, no flipped study).
  4. Clarify only what the check shows. Proposal: at most 5–10 minutes of targeted explanation on the one or two items most learners missed, then move on. Untested.
  5. Spend most of the live session on application with feedback. Proposal: at least two thirds of the session (for a 50-minute class, about 30–35 minutes) on problems, cases, discussion or practice in pairs or small groups, with the teacher circulating and responding to what each group produces: a cue first, then a hint, then a demonstration of the step. Basis: active learning against lecture in undergraduate STEM (Active learning improves exam performance [+S], the only claim here with the class-time dose varied, and only high-intensity active learning narrowed gaps); contingent help (Contingent scaffolding improves learning more than fixed or absent support. [~M], one-to-one tutoring, untested at class scale); and one flipped calculus study in which adding peer teaching to the class time gave higher scores than a conventional flipped class (Peer-teaching flipped classroom students show higher calculus conceptual understanding… [+M]). What is done in the session matters, not only that the exposition moved.
  6. Close with an individual exit item. Proposal: one 3–5 minute unaided item on the session's target, collected from everyone, which sets the next pre-class assignment (more worked material if many fail). Untested.
  7. Count the home time and keep it within budget. Record the minutes the pre-class work takes (ask learners) and keep total home time near what the course asked before the flip; do not add the video on top of the old homework. Basis: in a review of college science comparisons, active-learning conditions often required more home time than their comparisons (Time-on-task is frequently unequal across compared conditions… [~M]), which is both a cost to learners and a confound in any comparison.
  8. Plan an arc across sessions, not one lesson. Proposal for a term: weeks 1–2, short teacher-made videos with a note-taking guide and embedded questions, a 5-minute recap at the start of class for those still learning the routine, and the check results discussed openly; middle weeks, no recap, the check only, and pre-class work with fewer prompts (one explain-why question instead of a guide); late weeks, pre-class sources learners must select or summarise themselves (a chapter section, a paper), and live time on whole tasks of the kind the course assessment asks for. A learner moves from guided to unguided preparation when they pass the opening check on two consecutive sessions; a learner who fails two in a row gets a short in-class or office-hours route to the material rather than more video. The arc and its thresholds are untested proposals; how to withdraw support is modelled in Scaffolding and Fading.

When a condition is not met: if most learners cannot or will not prepare (no access, shift work, no time), give the first 10–15 minutes of the session to the pre-class material in class ("in-class flip") and keep the application; if the material cannot be followed alone, teach it live and flip a later, more familiar topic instead.

Fitting the design to a situation

The three facts that most change the decision:

  • Whether learners will and can do the pre-class work. Access, time outside class, and whether earlier assigned reading was done decide whether the session can assume preparation. If unstated, ask: what share of these learners did last term's assigned preparation, and what device, connection and time do they have between sessions?
  • Whether the first explanation can be understood alone. A learner with prior knowledge of the topic can learn from a video; a novice facing complex material may stall with nobody to ask. If unstated, ask: when learners met similar material without a teacher before, could they answer a simple question on it afterwards?
  • Who gives feedback in the live session, and to how many. The gain is expected from coached application; one teacher with 200 learners, a volunteer tutor with ten minutes, and a small seminar call for different session designs. If unstated, ask: during the session, who will look at what each learner or group produces, and how soon?
If the brief says… Then change… Basis
Learners are novice on the topic and the material has many interacting parts Keep the first explanation live, or use a worked-example video with a pause-and-try item after each step; flip only the later, more practised topics; give the opening check before any independent application untested; Active learning improves exam performance [~S] is a course-level average that does not separate novices; the guidance limit for novices is on Active Learning. In one calculus study, prior achievement and the major taken in senior high school both predicted conceptual understanding in both flipped variants, and the peer-teaching advantage did not depend on either (Prior achievement level affects calculus conceptual understanding, and PTFC outperforms CFC across achievement levels… [~M]; Senior high school majors' background affects calculus conceptual understanding and self-efficacy… [~M]): weaker starters ended lower under either flip, which no unflipped arm tests
Learners are undergraduates in a health-professions programme (medicine, nursing, pharmacy and the like) Flip with the most confidence here, since this is the setting the one direct test covers; still check preparation, keep most live time for application, and record home time, because the pooled studies varied widely and most were not randomised Flipped classroom learning improves academic performance of undergraduate health professional students… [+M] (44 studies, low certainty, I² = 98%); Only a minority of flipped-classroom studies in health professional education are RCTs… [~M]; Both flipped and AI-enhanced flipped classrooms achieve higher in-class question accuracy than the traditional classroom [+W] (three pharmacology cohorts, 68 international students, in-class accuracy, no effect size)
Learners are advanced on the topic, or professionals updating practice Make pre-class material a short brief or a reading, and use the session for cases, disagreement and their own examples; drop the note-taking guide and the recap untested proposal
Learners are child (primary age) Do not set pre-class video as homework; run the exposure in class at a station or as a short whole-class video, then move to guided practice with the teacher circulating ("faux-flipped") untested; Contingent scaffolding… [~M] (fourth and fifth graders, one-to-one tutoring) supports the circulating teacher, not the flip
Learners are adult in a work setting (workplace-clinical), preparing between shifts Keep pre-class work to 10 minutes or less on a phone, tied to a task they will do in the session; use the session for practice on their own work tasks; do not penalise missed preparation, run a 5-minute catch-up at the start instead untested proposal
Learners have low confidence or low literacy, or study in a second language Prefer video with captions and a one-page glossary to reading; add one easy item early in the opening check; let learners show understanding by doing before explaining; make the check private untested proposal
Learners have unreliable access, or a phone only Keep each pre-class item under 5 minutes, downloadable and readable on a phone; give the same material as a printed page; schedule in-class time for those who could not access it untested; the access constraint is from the earlier page
Mixed modes: some learners join remotely while others are in the room Give remote learners the same opening check in the same tool; put remote learners in their own breakout group with a named facilitator rather than mixing one remote learner into a room group; route the teacher's circulation through the breakouts on a fixed timetable untested proposal
The live session is online (online-instructor-led) Run the check as a poll, application in small breakout rooms with shared documents the teacher can see, and the teacher visiting rooms in turn; keep total screen time per session to what was there before untested proposal; see the Blended Learning pattern for the online part's self-regulation
One teacher with a large group (100+) Use the opening check as a poll, pairs or small groups for application, and whole-class feedback on the commonest error, then go to groups still stuck; trained peer teachers or assistants can extend the circulation Active learning improves exam performance [+S] (held across class sizes, largest in small classes); Peer-teaching flipped classroom students show higher calculus conceptual understanding… [+M] (two intact classes; the procedure is not described on the claim page)
Scarce facilitator time: a volunteer, a busy shift, a tutor who checks in briefly Put the feedback into the materials: answers and worked solutions after each pre-class question and each in-class item; let the facilitator spend their minutes only on learners who failed the opening check untested proposal
Only one session, with no time before it Do not flip; give 10 minutes of exposition in the session, then application with feedback and an exit item untested proposal
A fixed short time budget (for example a 60-minute workshop with 15 minutes of preparation) Fit to it: 15 minutes pre-class (one short video, two questions); in session 5 minutes check, 5 clarification, 40 application, 10 exit and next step; cut topics before cutting application time untested proposal
The brief names its own routine (its own pre-class quiz, worksheet, readiness test or discussion protocol) Keep it; change only what the evidence speaks to: make its first item a check on the pre-class material, make sure most of the live time is application, and keep total home time within budget untested; Time-on-task is frequently unequal… [~M] for the time budget
The brief gives its own categories, criteria or rubric Use them for the opening check, the in-class tasks and the exit item; do not substitute this page's items untested proposal
Pre-class work includes private or sensitive material (reflections on practice, personal experience) The learner controls who sees it; the opening check asks only about the shared content; in class, learners choose what to bring untested proposal
The course ends in a whole-task performance or a product for a real audience (a presentation, a client case, a performance) Keep the task whole in the live sessions: pre-class work supplies only the knowledge the task needs, and each session advances the whole task with feedback, rather than splitting it into paper items untested proposal
Stakes are high (a licensing exam, a gateway course) Keep the opening check low stakes; judge readiness only on unaided exit items; record who did the preparation and who did not, and offer the catch-up route before the exam untested; Active learning improves exam performance [+S] (gateway STEM courses, lower failure under active learning)
The goal is information acquisition alone (a factual update, a policy change) A flip adds little; give the information directly (in writing, or a short talk) and use any live time for questions untested; see Active Learning on lecture and information acquisition
The brief fixes its own session structure (two short sessions a week, a retrieval check before class, a simulation in one session) Keep that structure: put the pre-class check where the brief puts it and give the application or simulation the session the brief names; change only the pre-class dose, what the opening minutes check, and how peers support each other when facilitator cover is thin untested proposal
Learners have little or no home internet, or lessons are short Make the pre-class material printable or downloadable once on school wifi (one page or a 5–8 minute clip), or do the first exposure in the first 10 minutes of class (an in-class flip); keep both practical application and a structured discussion in the live time, each with its own task untested proposal

Six of the nineteen rows cite a claim; three (health-professions undergraduates, one teacher with a large group, high stakes) rest their change on one, and the other three cite a claim for its limit. Only the health-professions row rests on a test of a flipped course against a lecture course; the others carry active-learning meta-analyses, one study comparing two flipped variants, contingent tutoring of children and a review of home time to this setting, and say so. The other thirteen rows are untested proposals, and the review's result is not carried to school-age learners or to settings outside health-professions education here.

Observation, state and explanation

What can be observed: whether each learner opened and finished the pre-class material (platform data, a short self-report), their answers on the embedded and opening-check items, what they produce during application (work, talk, questions asked), and their unaided exit item. "Prepared", "engaged", "self-regulating" and "understands" are inferences. A completed video view does not show that the learner followed it; a correct check item does not show they can apply it.

Response Plausible explanations Observation that could change the interpretation
Many learners arrive without having done the pre-class work No time or access; previous sessions did not use the preparation; the material is too long or too hard; they expect the teacher to re-teach it Check completion by learner against access and schedules; ask, privately, why; see whether completion rises after a session visibly built on the preparation.
Learners watched the video but fail the opening check The video did not teach enough for learners to follow alone; the check asks for more than the video showed; learners watched without attending Ask one learner to explain the first step aloud; compare check scores of those who answered the embedded questions with those who skipped them.
Pass the opening check, but stall on the application task The gap is in applying, not knowing (the flip is doing its job and practice is needed); the task jumps too far beyond the material; group work hides who can do it Give a closer intermediate item with a hint; record who produced each part of the group's work.
Exit items are fine, but course exam scores do not move The exit items are closer to the session than the exam; the exam measures something the sessions do not practise; too little application time Compare exit items with exam items for alignment; record the minutes of application in each session.
Learners say they prefer lectures, or rate the flipped course lower More work at home; discomfort with being exposed in class; a real goal the course does not serve (a clear account from an expert); a feeling of fluency from lectures that is not matched by performance Compare self-report with the exit items and exam; ask what they want the live time for; record home time.
Some learners fall steadily behind on preparation Competing work or family time; no place to study; low confidence; a missing prerequisite Look at when they open the material and what they skip; ask how they plan their week; offer the in-class route and see whether performance recovers.

These rows are proposals, untested with learners as diagnostics. A check is itself a learning event and changes what it measures; record what remains unresolved.

Evidence and qualifications

One claim tests the relationship directly; the rest bear on parts of it.

Learners, settings and outcomes differ across these claims (health-professions undergraduates, undergraduate STEM courses, one calculus class pair, one physics preparation study, laboratory passages, children with a tutor), so do not rank them by effect label. None measures delayed retention after a flipped course; the one delayed outcome here (two weeks) compares two forms of preparation.

The evidence by question

Further evidence, not yet read against this model

Claims from the calculus study behind a core claim, read on 2026-10-15 and left here because none bears on where the first exposure happens: each is about self-efficacy or a correlate within two flipped variants, with no unflipped arm.

Objective and learner-valued goal

The designer's objective is usually performance on the course's assessment, or on a representative task, at the end of the course, with live time spent where the teacher adds most. Ask separately what learners want from the live time and from the course: an expert's clear explanation, a chance to practise with help, fewer hours at home, flexibility to study around work, or simply the credential. Agreement and divergence are observations. A learner who values the teacher's explanation may experience a flip as being told to teach themselves; one who works shifts may value replayable video and dread fixed preparation deadlines. On average learners have not disliked it: in eight health-professions studies, satisfaction with flipped classes was higher than with traditional ones (SMD = 0.48, low certainty) (Undergraduate health professional students report more positive satisfaction with flipped class learning… [+M]), but an average says nothing about the shift worker in your course, so ask.

For example, a part-time adult in an evening accounting course may want to pass the certificate exam with as little home time as possible, while the instructor's objective is that learners can prepare a set of accounts from messy source documents. Pre-class videos that add an hour a week serve the instructor and cost the learner; a 10-minute video with two questions, an in-class check and most of the evening spent on a real set of documents may serve both. Record both aims, and do not read low preparation as low motivation before asking about time.

What would revise this model?

The expectation should weaken if comparisons of flipped and unflipped versions of the same course, with the same total time, materials and assessments, show no advantage on aligned outcomes; if gains in flipped courses disappear once home time is equated; or if courses that use the freed time for application do no better than those that use it for review of the video. If learners who do not prepare do as well as those who do, the preparation step is not doing the work credited to it. If novices consistently fail to learn the first explanation alone, the relationship should be restricted to learners with some prior knowledge of the topic. Do not protect the model by calling every failure "a bad flip" after the fact.

An end-of-course exam advantage, delayed retention, self-efficacy, preparation rates and learner satisfaction are separate claims. The present evidence establishes no best video length, no share of live time and no preparation rate needed. It records one pooled advantage of flipped over lecture-based courses in health-professions education, at low certainty, from mostly non-randomised studies that likely changed several features at once; it does not isolate an effect of relocating exposition itself. Randomised comparisons with home time equated, or a moderator analysis of the heterogeneous studies, would be the next test.

Claims

  • Scaffolding and Fading — how support in the pre-class work and the live session is withdrawn across a term
  • Digital Learning — often supplies the delivery infrastructure for pre-class exposure
  • Guided Practice — explains why the repurposed live session can be instructionally superior
  • Active Learning — flipped learning is often a scheduling strategy for creating more active class time

Examples

Illustrative

Flipped Learning — Learners complete concise pre-class exposure, then spend class time in coached application, discussion, and feedback.

Short video plus in-class problem solving — Students watch a brief explanation or walkthrough before class and then use live time to solve problems in pairs while the instructor circulates and coaches.

Pre-reading plus seminar discussion — Learners encounter foundational ideas before class so that live time can focus on interpretation, debate, and clarification rather than first exposure.

Key Sources

  • Bishop, J. L., & Verleger, M. A. (2013). The flipped classroom. In ASEE National Conference Proceedings. doi:10.18260/1-2--22585
  • Bergmann, J., & Sams, A. (2012). Flip your classroom: Reach every student in every class every day. International Society for Technology in Education.
  • Stelzer, Gladding, Mestre, and Brookes. (2008). Comparing the efficacy of multimedia modules with traditional textbooks for learning introductory physics content. https://arxiv.org/abs/0806.0405. doi:10.48550/arxiv.0806.0405
  • Bishop, J. L., & Verleger, M. A. (2013). The flipped classroom: A survey of the research. In 2013 ASEE Annual Conference & Exposition Proceedings. https://doi.org/10.18260/1-2--22585
  • Lo, C. K., & Hew, K. F. (2017). A critical review of flipped classroom challenges in K-12 education: Possible solutions and recommendations for future research. Research and Practice in Technology Enhanced Learning, 12(1), 4. https://doi.org/10.1186/s41039-016-0044-2
  • Lage, M. J., Platt, G. J., & Treglia, M. (2000). Inverting the classroom: A gateway to creating an inclusive learning environment. The Journal of Economic Education, 31(1), 30-43. An earlier "inverted classroom" paper credited with popularizing the model in higher education, using pre-class preparation and in-class economics experiments — a direct precursor to the model Bergmann and Sams later branded "flipped.". doi:10.1080/00220480009596759
  • Walvoord, B. E., & Anderson, V. J. (1998). Effective grading: A tool for learning and assessment. Jossey-Bass.
  • Arduini-Van Hoose, N. (2020). Flipped classroom. In Educational psychology. Retrieved from https://edpsych.pressbooks.sunycreate.cloud. CC BY-NC-SA 4.0.
Page metadata
Field Value
type principle
id flipped-learning
aliases multimedia-modules-for-pre-lecture-preparation
title Flipped Learning
description For learners who can follow a topic's first explanation on their own, moving that first exposure before the live session and spending the freed time on application with feedback is expected to give better course outcomes than spending live time on the same exposition, provided learners prepare and the home time is not simply added; one meta-analysis of undergraduate health-professions courses found flipped classes ahead of lecture-based ones on exam performance (SMD 0.57), at low certainty with very high heterogeneity, and no claim here tests it with home time held equal.
status review
moved_from pattern/flipped-classroom
canonical True
generated by claude/unspecified on 2026-10-15
trust unverified — no one has confirmed this page's content
sources 3 — stelzer-2008, bishop-2013, lo-2017

Frontmatter is the page's machine-readable record — see the Schema & Guide.