The register

All twenty-one entries. Every entry states the effect, what it predicts, how well it replicates, and where it breaks. No effect size is printed where the published estimates disagree.

Sec.	Entry	What it predicts	Where it breaks	Evidence
01	The testing effect Retrieval	Retrieval practice beats re-reading on any delayed test of the same material.	Immediate tests, multiple-choice formats, and far transfer.	Replicates broadly
01	Retrieval difficulty Retrieval	More effortful successful recall produces more durable memory than easy recall.	When the attempt fails and no correction follows.	Robust, bounded
01	Feedback timing Retrieval	Delayed corrective feedback is remembered better than immediate feedback.	Procedural skills, and learners very early in acquisition.	Mixed
02	Distributed practice Spacing	The same total study time spread across occasions beats the same time massed.	Very short retention intervals, where massing can win.	Replicates broadly
02	Expanding intervals Spacing	Review gaps that lengthen as material is relearned outperform fixed gaps.	Comparisons with equal-interval schedules, which often tie.	Mixed
03	Mixing problem types Interleaving	Mixed practice lowers accuracy during practice and raises it on a later test.	Types with nothing to confuse, and very short practice sets.	Robust, bounded
03	Discrimination Interleaving	Interleaving works by forcing the choice of method, not its execution.	It is a mechanism claim, and it is tested indirectly.	Mixed
04	Cognitive load theory Cognitive Load	Working memory is narrow, and the design of an explanation adds to what it must hold.	Its partitions of load resist independent measurement.	Robust, bounded
04	Split attention Cognitive Load	A diagram separated from the text that explains it worsens comprehension of both.	Readers already fluent in the material.	Robust, bounded
04	The expertise reversal Cognitive Load	Support that helps a beginner measurably hinders someone who already knows the material.	Where expertise is assumed rather than measured.	Robust, bounded
05	The Ebbinghaus curve Forgetting	Retention falls steeply at first and then progressively more slowly.	As a source of exact intervals: nonsense syllables, one subject.	Shape replicates
05	Bartlett's reconstruction Forgetting	Recall of meaningful material is reconstruction shaped by expectation, not replay.	As a quantitative model; the original studies were loosely controlled.	Replicates broadly
06	Fluency Illusions	Ease of processing raises confidence in later recall without raising recall.	When judgements follow a delayed practice test rather than a reading.	Robust, bounded
06	Judgements of learning Illusions	People predict their own future recall badly, and in a predictable direction.	Delayed judgements are markedly better calibrated.	Replicates broadly
07	Near and far transfer Transfer	Gains from practice extend to genuinely novel problems.	Most tests of far transfer; near transfer is the safer claim.	Contested
07	Desirable difficulties Transfer	Conditions that slow acquisition can improve long-term retention.	Difficulty beyond what the learner can resolve unaided.	Robust, bounded
08	Replication Evidence Standards	Which of these findings survived the replication crisis, which shrank, and which did not.	A single failed replication settles less than it appears to.	Method note
08	Effect size Evidence Standards	A statistically real effect can still be too small to matter, and several here are.	Effect sizes travel badly from the laboratory to real study.	Method note
08	Lab versus classroom Evidence Standards	Effects measured over an afternoon on word lists do not automatically hold over a term.	Field studies trade control for realism and usually find less.	Method note
09	The self-experiment How We Know	One rigorous subject can found a field and still represent nobody but himself.	Any generalisation from a single participant.	Method note
09	Working memory How We Know	A limited-capacity system with separable components handles information in use.	It was never a model of long-term learning.	Robust, bounded

21 entries shown.