§1Why this page exists
If you came here looking for the flipped-classroom literature, this is the right page. If you're trying to figure out whether to flip your own classroom, the pillar guide is where I argue the practice case. This page is the receipts.
What follows is a chronological tour of the formal flipped-classroom math studies I've authored or co-authored, from 2014 to the present, plus the community-college research I'm running this academic year. There are four published peer-reviewed papers and one in-progress study. None of them is a meta-analysis. Together they're a practitioner's research arc — papers written by a math teacher who started flipping in 2012 and has been studying what happens ever since.
A note on what's a finding and what isn't. The four peer-reviewed papers are findings. The 2025–2026 community-college extension I'm collecting data on right now is in progress; I won't claim findings the data doesn't yet support. And the personal teaching outcome from my AP Statistics classroom — the one-to-ten-students-passing number you may have seen on the pillar or in interviews — is a story, not research. I explain why in section seven.
§22014 — Where the formal research started
I started flipping my math classroom in 2012, but the first time I wrote it up for an academic audience was 2014. That paper, with my colleague Criselda Toto, asked a question I'd been getting in faculty rooms for two years: even if test scores move, what about the students who already hated math going in? Does flipping make them like it more, or hate it more?
The paper — Does the Implementation of a Flipped Classroom Pedagogy change the Students' Attitude towards Mathematics? — was published in the Proceedings of the World Conference on E-Learning (AACE), pages 552 to 557, and presented at the conference in New Orleans. The full text is in the AACE digital library at editlib.org/p/148784.
What we did. Two cohorts of high-school math students, attitudes measured at the start and end of a year-long flipped intervention, compared to historical baselines. The finding was a measurable shift on the attitude instrument — not the headline result of my arc, but an early signal that flipping wasn't only moving test scores, it was moving how students felt about being in a math classroom.
What the paper is and isn't. It's an early descriptive study from a working teacher, with the strengths and weaknesses that implies. It's not the RCT — that came two years later. It's the paper I point grad students to when they ask where to start with the question, because the design is straightforward and replicable in any school willing to run a flipped unit. If you're a researcher reading this and looking for the rigorous version, the next section is where to go.
§32016 — The flagship RCT (n=91, p<.05)
If a colleague is going to read just one of my papers, this is the one I want them to read.
Flipped Classroom Model: Effects on Performance, Attitudes and Perceptions in High School Algebra, co-authored with Khristin Fabian and Criselda Toto, was published in 2016 in Springer's Lecture Notes in Computer Science, Volume 9891 — the proceedings of the European Conference on Technology Enhanced Learning (EC-TEL) for that year. The DOI is 10.1007/978-3-319-45153-4_7.
The study was a year-long randomized controlled trial. Ninety-one high-school algebra students, split between a flipped condition and a traditional-lecture condition, with the same teacher (me) running both sections. We measured performance on common assessments and student attitudes on standardized instruments, pre- and post-intervention.
The treatment effect on the performance measure was statistically significant in favor of the flipped condition, with a p-value below .05. Attitudes also moved in the flipped condition's favor, on multiple sub-scales of the instrument. The full numbers, instruments, and methods are in the chapter itself; I'd rather you read the paper than have me paraphrase the table here.
Across a year-long intervention with ninety-one high-school algebra students, the flipped-classroom condition produced statistically significant gains in mathematics performance compared to the traditional-lecture condition (p < .05), along with measurable improvements in student attitudes toward mathematics.
Esperanza, P., Fabian, K., & Toto, C. (2016). Springer LNCS Vol. 9891.(Springer)
A few honest qualifications. n = 91 is a real sample for a single-classroom study, but it's not a meta-analysis. One school, one teacher running both conditions, one academic year. The result is consistent with the broader meta-analytic literature on the flipped classroom that has accumulated since 2016 — studies in math, chemistry, undergraduate physics, nursing — but the citation a researcher should actually pull is the original Springer chapter, not someone else's gloss of it.
§42016 — A companion paper on AP perceptions
The same year as the Springer RCT, my co-author Khristin Fabian and I published a smaller paper — Student Perceptions on the Use of the Flipped Classroom Model for Advanced Placement Mathematics — in the proceedings of the International Congress on Mathematical Education (ICME).
Where the Springer chapter measured outcomes, this one measured how students experienced the model. The population was AP-level high-school math students — calculus and statistics — and the question was whether the flipped structure made the AP workload more manageable or less, in the students' own assessment.
The reason I mention this paper alongside the Springer one isn't that it's the more rigorous study — it isn't. The reason is that AP teachers and college faculty asking about flipping for advanced math tend to want this paper specifically, because the population is closer to theirs than a general algebra class. It's the paper I direct people to when the question is "does this work for AP?"
The AP Calculus FRQ Reviewer book series I later put on Apple Books — six volumes covering the standard FRQ topic areas — was partly informed by the perception data here. Specifically, students reported that the rewatch-and-pause behavior was worth more to them on FRQ practice than on procedural drills. That observation is what made me decide the FRQ books should be multi-touch with embedded video, rather than text-only. More on the books on the Books spoke and the dedicated AP Calc FRQ page.
§52021 — Extending the question across a wider sample
Five years after the Springer paper, I co-authored a follow-up with a research team based in Cebu — The Utility of a Flipped Classroom in Secondary Mathematics Education, with C. Himang, M. Bongo, E. Selerio Jr., and L. Ocampo — published in the International Journal of Mathematical Education in Science and Technology (Taylor & Francis).
The point of the 2021 paper, in a sentence, is to test whether the effect from the 2016 RCT holds up in a wider sample, across multiple teachers and contexts. The 2016 paper was one school, one teacher; that's the right way to start a research line, and the wrong way to end one. If a model only works when the original investigator runs it, the field has a problem.
The 2021 paper is the most recent peer-reviewed work I've co-authored on flipped classrooms in mathematics. It moves the conversation past "does the effect appear?" and toward "is the effect stable across teachers and student populations?" That's the question I'd want a research team in 2026 to be designing studies around, including the one I'm running myself.
A note on adjacent work. During the COVID period, I co-authored a separate study on Open Education Resources in Filipino-language mathematics video instruction — significant gain-score differences in favor of Filipino-language video for ninth-graders (t = 4.146, p < .001) and a smaller but still significant effect for tenth-graders (t = 2.170, p < .031). That paper sits more naturally on the Distance Learning spoke, where I treat it in detail. I mention it here so the citation chain is complete.
§62025–2026 — What I'm collecting right now
The piece of research I most want to write up, and can't yet, is the community-college extension of the 2016 RCT.
I'm currently teaching at Barstow Community College, with adjunct loads at Victor Valley College and West Coast University. Through the 2025–2026 academic year, I'm running an action-research extension of the K–12 flipped-classroom study into the community-college population. The design borrows directly from the 2016 RCT — same instruments where they translate, same year-long structure, same focus on performance and attitude as outcomes.
Three questions I'm asking, specifically. First, does the treatment effect from the K–12 RCT replicate in a population of adult learners? The constraints are different — jobs, kids, transportation, financial-aid timing — and any of those could moderate the effect. Second, does the in-class redesign that I rely on — the warm-up, the circulating, the one-on-one — survive the move to a fully online or hybrid course? About half my current load is online or hybrid, which is a different thing from face-to-face in ways that matter. Third, does the asynchronous video produce the same rewatch-and-recover behavior in college students that I saw in high schoolers? The early signal is that it does, and is in some ways more pronounced — but "early signal" is not a result.
What I'm not claiming. The data isn't all in, and I'm not pre-publishing on this page. The qualitative signal so far is consistent with the K–12 work — students show up to the in-person or synchronous-online sessions more prepared, and those sessions are doing more of the work that used to happen at the kitchen table. The formal claim waits for the data, and the paper goes through peer review like the others.
There's also an earlier, smaller two-semester study I ran at the community-college level on additional online-based materials in an online Elementary Algebra class — listed on my LinkedIn profile and the seed of the 2025–2026 extension. The Higher-Ed spoke goes deeper on the teaching practice; this section is the research frame. When the paper lands, it'll be added to the references below.
§7The personal data point that isn't research
There's a number I cite in talks and that has appeared in press coverage of me, including the Asian Journal piece in 2018. I want to address it here, on the research page, so it's clear what it is.
When I started teaching AP Statistics in 2009, exactly one of my students passed the AP exam that year. After I started flipping my classroom in 2012 and gaining confidence in the model, that number grew to about ten students passing each year — with scores of 3, 4, and 5.
That is not a research finding. It's a teaching outcome from a single classroom. There's no control condition; I changed many things about my AP Stats teaching between 2009 and the post-2012 cohorts; my own skill as a teacher was different by year five than it was by year one; the students who walked into my room each year were different cohorts with different backgrounds. The variables I changed weren't isolated to flipping. The number isn't bad — it just isn't what a study would let you conclude.
What it is: the data that convinced me, before I had any peer review behind it, that the model was doing something the lecture wasn't. It's the data point a teacher has on their own students. I cite it in talks because it's mine and it's true. I don't cite it as research because that's not what it is.
I'm flagging this distinction explicitly because when practitioners write up classroom outcomes as research, the field gets noisier and harder to draw on. The 2016 Springer paper is the citation. The AP Stats number is a story. Both are true. They are different kinds of true.
§8FAQ
Is one RCT with n=91 enough evidence to change my practice?
On its own, no. The 2016 paper is one study in one school with one teacher. What it does is sit alongside a broader meta-analytic literature on the flipped classroom that has accumulated since the mid-2010s — studies in math, chemistry, undergraduate physics, nursing, and beyond, with mixed but generally positive effect sizes. The short version: act on the practice signal where you can pilot it cheaply (one unit, per the pillar's section three), and watch whether the attendance, attitude, and assessment signals move in your room. The literature is informative; your room is decisive.
Where can I get the 2016 Springer paper?
The DOI in the references list resolves to the publisher's page on Springer Link. Most university libraries have institutional access. If you're an educator without institutional access, send me an email through the contact page and I'll point you to a legal copy.
Can I cite the in-progress community-college work on a grant or proposal?
Not as a finding — the data isn't published. You can cite the 2016 Springer paper as the design basis and reference the community-college extension as "in progress at Barstow Community College, expected publication 2026–2027" if a forward-looking note is appropriate.
How do I find the rest of your research?
The references list below has the four core flipped-classroom papers. My Google Scholar profile aggregates citations across all of those plus a few smaller pieces — adjacent papers on student perceptions, the COVID-era Open Education Resources work, and conference proceedings.