PontisThe science
The science behind Pontis.
Six study methods that research supports, each one something the lessons really do, and two ideas behind them. For each: what it is, what the studies found, and how it shows up on screen.
i
Retrieval practice
Pulling an answer out of memory, instead of looking at it again.
What the research found
In a well known experiment, students who took a recall test on a passage remembered more a week later than students who studied it again. On a test five minutes later, the restudy group did better, so the benefit shows up over time [1]. Another study put short memory tests between the parts of an online lecture. Learners wandered off less, took more notes and did better on the final test [2]. A large review of study techniques rated practice testing among the most useful [3].
How Pontis does it
Each lesson video stops at planned moments and asks a question about what you just saw. You answer before the video continues, and you cannot skip ahead past a question. Practice sets after the lessons ask you to recall again, and unit tests draw their questions at random from a bank.
Solve the equation. What is x?
Sources: [1] [2] [3]
ii
Feedback at the right moment
Finding out at once whether you were right, and what to do about it.
What the research found
A major review of research on feedback found that its effect varies a great deal, and that it helps most when it shows the learner what to do next [4]. A broad review of classroom studies on checking understanding while teaching goes on, called formative assessment, reported learning gains when it was done well [5].
How Pontis does it
A first wrong answer on a checkpoint sends you back to rewatch the part that teaches it, starting a little before the question. Then you try again, on a fresh version of the question where the bank has one. A second miss shows a worked explanation, you confirm the right answer, and the question joins your review list.
Solve the equation. What is x?
Sources: [4] [5]
iii
Spaced review
Coming back to an idea after a gap, instead of all at once.
What the research found
Ebbinghaus charted how quickly memories fade in 1885 [6], and a 2015 replication found a similar curve: steep at first, then slower [7]. A review of studies on spreading practice over time found that spaced study generally beat the same amount of study packed together, and that the best gap depends on how long you need to remember [8]. The curve on the front page illustrates this pattern. It is not Pontis data.
How Pontis does it
Pontis’s spacing is simple, and we say so. Questions you miss come back a few items later in the same practice set, and wait on a review list that comes first in your next practice set. A few questions from earlier units are mixed into new practice. A learner can start only a set number of new lessons a day, and a unit test allows one attempt a day, so a course cannot be crammed into a weekend. Pontis does not yet schedule reviews by calendar day.
Solve the equation. What is x?
Sources: [3] [6] [7] [8]
iv
Mastery before moving on
Reaching a set standard on one step before the next one opens.
What the research found
A meta-analysis of mastery learning programs found that students in them generally scored higher on examinations than students in conventional courses. The size of the benefit varied from program to program [9].
How Pontis does it
Each lesson has a pass mark on its checkpoint questions, and each unit has a timed test with its own pass mark. For signed-in students, passing the test opens the next unit. Questions missed on a test go on your review list. The open demo lets visitors try any unit, so the gate applies to real accounts.
Sources: [9]
v
In your own words
Putting an idea into words of your own, in writing or speech.
What the research found
In one study, students who were prompted to explain a text to themselves as they read it came to understand it more deeply than students who were not prompted [10]. The large review of study techniques rated self-explanation as moderately useful [3]. Those studies looked at explaining while reading, so Pontis’s tasks are related to them but not identical.
How Pontis does it
Lessons include typed reflection questions. Each unit has a discussion post and a project with a written part and a short recorded explanation, scored on a rubric. Pasting is turned off in typed answers, so the words are yours.
Sources: [3] [10]
vi
Start where you are
Matching where a learner begins to what they already know.
What the research found
Researchers have found that teaching which helps beginners can lose its benefit, or get in the way, for people who already know the material [11].
How Pontis does it
A placement assessment suggests where to start. It only recommends, and it never locks you into a level. In sections that offer it, unit pretests let you show what you already know.
Sources: [11]
vii
Working memory is the bottleneck
What you can hold in mind at once is small, so how much arrives at once matters.
What the research found
Cognitive load theory starts from a simple picture. Working memory, where you hold and work on new information, is limited in how much it can handle at once and for how long, while long-term memory is very large [12, 13]. A well known review of studies concluded that when people are stopped from rehearsing or grouping information, the limit is about four chunks, lower than the “seven plus or minus two” often quoted from a 1956 paper [14, 16]. Researchers who study multimedia lessons proposed ways to lower the load, including breaking a lesson into learner-paced segments and cutting extra material that does not help the learner [15].
How Pontis does it
Lessons are short video segments, and the video pauses at a checkpoint with a question before it goes on. Assessments show one question at a time. A unit test passed opens the next unit, so new material builds on what you have already shown you know. For some courses there is help in Somali, which can reduce the load for a learner working in a second language; that last point is our reading of the theory, not a finding we cite.
What this does not show
The “about four” figure comes from laboratory tasks and is debated; real lessons are not tested items. Cognitive load theory is a model, and Pontis has not measured the load on its own learners.
Sources: [12], [13], [14], [15], [16]
viii
What top-performing systems have in common
Practices described in the research and curriculum documents of high scoring systems, and in Pontis.
Which systems
In the OECD’s PISA 2022 assessment, six education systems scored highest in mathematics, in this order: Singapore, Macao (China), Chinese Taipei, Hong Kong (China), Japan and Korea [17]. The IEA’s TIMSS study is the other major international comparison of mathematics achievement [18]; this page names systems only from the PISA 2022 results above. They are consistently among the highest scoring systems. Pontis is not affiliated with, endorsed by or equivalent to any of them, and does not produce their results.
What the research describes
Analyses of the earlier TIMSS study found that high scoring countries taught a coherent, carefully sequenced mathematics curriculum, and that they taught a very similar core curriculum to nearly all of their students [19]. Singapore’s official mathematics syllabus puts understanding concepts and solving problems alongside fluency with facts and procedures [20].
How Pontis does it
Units are sequenced, and the next one opens only when the previous unit test is passed. Learners in a course work toward the same units, and assessments are in English so a credit means the same everywhere. Typed explanations and project write-ups ask you to put ideas in your own words.
What this does not show
These are associations, not proof that a practice causes high scores, and the curriculum analyses used data from the 1990s. A syllabus states an aim, not a result. Practices we looked at but left out: mastery before moving on and the concrete-pictorial-abstract sequence, because we found no source that ties them to these systems in the way we would need, and frequent low-stakes checking, for the same reason. Pontis has not published results from its own learners.
Sources: [17], [18], [19], [20]
What we do not claim.
- No method here makes a promise about any learner. Studies report averages, and individual results vary.
- The studies used particular tasks, ages and settings. Pontis uses each method in its own way, so the match is close but not exact.
- Pontis has not published results from its own learners.
- Only methods the platform does today are listed. Ideas we have not built are left out.
- The two ideas on working memory and on high scoring systems are context, not claims that Pontis gives the same results.
References.
- Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. doi:10.1111/j.1467-9280.2006.01693.x
- Szpunar, K. K., Khan, N. Y., & Schacter, D. L. (2013). Interpolated memory tests reduce mind wandering and improve learning of online lectures. Proceedings of the National Academy of Sciences, 110(16), 6313–6317. doi:10.1073/pnas.1221764110
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58. doi:10.1177/1529100612453266
- Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77(1), 81–112. doi:10.3102/003465430298487
- Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy & Practice, 5(1), 7–74. doi:10.1080/0969595980050102
- Ebbinghaus, H. (1885). Über das Gedächtnis: Untersuchungen zur experimentellen Psychologie. Leipzig: Duncker & Humblot.
- Murre, J. M. J., & Dros, J. (2015). Replication and analysis of Ebbinghaus’ forgetting curve. PLOS ONE, 10(7), e0120644. doi:10.1371/journal.pone.0120644
- Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. doi:10.1037/0033-2909.132.3.354
- Kulik, C.-L. C., Kulik, J. A., & Bangert-Drowns, R. L. (1990). Effectiveness of mastery learning programs: A meta-analysis. Review of Educational Research, 60(2), 265–299. doi:10.3102/00346543060002265
- Chi, M. T. H., de Leeuw, N., Chiu, M.-H., & LaVancher, C. (1994). Eliciting self-explanations improves understanding. Cognitive Science, 18(3), 439–477. doi:10.1207/s15516709cog1803_3
- Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. doi:10.1207/s15326985ep3801_4
- Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. doi:10.1207/s15516709cog1202_4
- Sweller, J., van Merriënboer, J. J. G., & Paas, F. G. W. C. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251–296. doi:10.1023/A:1022193728205
- Cowan, N. (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity. Behavioral and Brain Sciences, 24(1), 87–114. doi:10.1017/S0140525X01003922
- Mayer, R. E., & Moreno, R. (2003). Nine ways to reduce cognitive load in multimedia learning. Educational Psychologist, 38(1), 43–52. doi:10.1207/S15326985EP3801_6
- Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97. doi:10.1037/h0043158 (cited for history only)
- OECD (2023). PISA 2022 Results (Volume I): The State of Learning and Equity in Education. Paris: OECD Publishing, 5 December 2023. doi:10.1787/53f23881-en
- IEA / TIMSS & PIRLS International Study Center, Boston College (2024). TIMSS 2023 International Results in Mathematics and Science. timss2023.org/results (no DOI checked)
- Schmidt, W. H., Houang, R., & Cogan, L. (2002). A coherent curriculum: The case of mathematics. American Educator, 26(2), Summer 2002. aft.org/ae/summer2002 (no DOI)
- Ministry of Education, Singapore. Mathematics Syllabus, Primary One to Six (2021 syllabus). moe.gov.sg (no DOI)