Before You Read: About This Site
Last updated
Feedback & Forgetting is for people who look up a learning method or a psychology effect by name and want two answers: how much evidence stands behind it, and under what conditions. Each article is anchored to something you can search for, such as spaced repetition, active recall, the incubation effect or the foot-in-the-door technique. You should leave knowing what was tested, on whom, with what material and after how long a delay.
The site was built with self-taught learners and students in mind, along with teachers and writers who need a source they can cite. The tagline sums up the promise: How learning actually works — sourced, replicated, recomputed.
What the five sections cover
Learning & Memory reads the experiments behind study methods and memory: the forgetting curve, spaced repetition, active recall and retrieval practice, learning styles, the Feynman technique, and a scorecard of popular learning myths.
Language Minds covers the psychology of learning another language: whether there is a critical period, what the research says about benefits of being bilingual, and what a vocabulary size test is actually counting.
Creative Minds looks at experiments on coming up with ideas: brainstorming in groups versus alone, the incubation effect, and functional fixedness as seen in the candle problem.
Influence & Persuasion takes classic persuasion and obedience studies apart, including the foot-in-the-door and door-in-the-face techniques and Milgram's obedience experiments, and asks which results held up when they were run again. These articles explain how the techniques work so you can recognize them. They do not teach anyone to use them on other people.
Systems & Loops traces feedback loops in ordinary life, such as traffic jams and thermostats, and checks famous cautionary stories like Goodhart's law and the cobra effect against the earliest record that can be found.
The forgetting covered here
The forgetting on this site is the everyday kind studied in memory experiments: how much of a word list, a passage or a skill people keep after an hour, a day or a month. The site does not assess memory complaints, attention difficulties or any other health condition, and it gives no medical or clinical advice. If your memory or attention worries you, that is a question for a doctor.
The site also does not review study apps or tools, and it does not comment on education policy. Methods are discussed through the research on them, not through products.
What this site stands for
The way of studying that feels easy and the way that works are often not the same. When the goal is to remember something days or weeks later, this site favors methods with evidence of delayed retention, even when they feel slower or clumsier at the time. That is a stance this site takes, not a law of nature, and it applies to learning that has to last.
A few positions follow from it and come up again across the articles. When the goal is retention over days or longer, this site leans toward retrieval practice and spacing rather than rereading and highlighting. Rereading is not worthless, and retrieval practice has conditions of its own: it helps most when you can actually produce the answer or get feedback, and when the practice matches the kind of test you will face.
The idea that matching lessons to a learning style (visual, auditory and so on) improves learning has not been supported by studies designed to test it, so this site does not recommend planning lessons around it. People do have preferred ways of taking in material; the claim under question is the matching.
On the systems side, some of the cautionary cases repeated in management books rest on thin sourcing. A causal loop diagram is a sketch of a hypothesis, not evidence, and each article says which loops have been tested and which have not.
Every one of these positions is stated only as firmly as the sources that were read allow. Where only a paper's abstract could be read, the article says so and keeps its wording at that level.
How claims are checked
Each research claim is traced back to the original paper and its DOI. The claim then carries one of five replication tags (replicated, mixed, not supported, single study or retracted), and each tag states who was tested, with what material, after what delay, and on what date the tag was last checked. Every source in an article is marked by how much of it was read: the full text, the abstract only, or the citation only. Where the numbers behind a chart can be recomputed, they come with a CSV file you can download and check.
An example from the first article: the famous forgetting curve that Hermann Ebbinghaus published in 1885 tracks savings, meaning how much less time he needed to relearn a list, not how many items he could still recall. He was his own only subject. A replication published in 2015, also with a single subject, found a curve of similar shape. The article tags each of those claims separately, because a matching shape and a matching set of numbers are different findings.
The full standard, including how each tag is decided and what happens when the evidence is only an abstract, is on From paper to tag.
Three terms used on this site
The forgetting dividend is this site's label for an observation from spacing experiments: a successful recall after a gap often left better retention on a delayed test than an immediate repeat did. It is a name, not a mechanism, and the word is a label rather than a financial metaphor. It does not mean that reviewing on the edge of forgetting is best, since the useful gap depends on how long you need to remember. The term is defined on the spaced repetition evidence page.
The lecture mirage names a pattern in which the study method that feels smoother while you use it earns more confidence but does worse on a delayed test. Confidence ratings and recall scores are measured on different scales, so the site compares only their order and never subtracts one from the other. The term is defined on the active recall evidence page.
Loop literacy is the ability to spot a feedback loop unaided in an everyday scene such as a traffic jam, a thermostat or a filling bathtub. It is the teaching goal of the Systems & Loops section, not a research construct. The term is defined on the everyday feedback loops page.
Where the colors come from
The palette is taken from a classroom and a study desk: chalkboard green for the brand color and links, the gray-blue of slate and board erasers, the cream of chalk, cedar orange and graphite from a pencil, and the brick red of correcting chalk. Dark mode is the same slate after the lights go out, with the green and the chalk lifted by a desk lamp.
Found a mistake?
Email contact@feedbackandforgetting.com with the page, the sentence and what is wrong; a DOI or page number helps. There is no form, and no reply time is promised. Corrections are dated at the foot of the article they change and listed on Erased and rewritten. A replication tag that changes because new evidence arrived is handled the same way as any other correction, with the old tag and its date kept on record. More on what to send is on Raise a hand, and reuse terms for diagrams and data are on Handouts.