massed practice looks efficient, decays fastest; spacing and interleaving
Cramming fails the week after because the memory was encoded during a brief window of stability and never re-stabilized. Retrieval after partial forgetting is what turns a short-term trace into a durable one, and massed practice skips that step entirely.
When you first encounter new material, your brain forms a temporary association between the new information and what you already know. That association is fragile. Hermann Ebbinghaus documented this in the 1880s by testing himself on nonsense syllables at increasing delays, and the curve he drew still shapes memory research today: retention drops sharply within the first day, then levels off. The drop is not a failure of learning. It is the signal your brain uses to decide what is worth keeping.
Each time you retrieve something that has begun to fade, the act of pulling it back up triggers a process called reconsolidation. The memory is rebuilt, and the rebuilt version is woven more tightly into existing knowledge. Cramming short-circuits this because nothing has had time to fade. You feel as though you learned more, but you only confirmed that you could recognize the material when it was still fresh.
If you have roughly eight hours total and three weeks before the exam, a workable split looks like this:
That uses about six hours, not eight, and reaches the exam with material that has survived four separate forgetting intervals. The same eight hours spent the night before leaves the material at its peak freshness and its lowest durability.
Students often abandon spaced practice because the early sessions feel awful. You sit down on Day 3 and discover you can barely recall what you "knew" on Day 1. That discomfort is informative, not a sign of failure. It tells you the memory has begun to decay, which means the next retrieval will be the kind that actually strengthens it. Smooth rereading feels productive because recognition is effortless, but effortless processing leaves a shallow trace. Recognition is not recall, and a page you have read four times will still feel unfamiliar the moment you close the book.
The spacing principle is robust, but it assumes the material will matter again. For a one-off presentation tomorrow, a phone number you will dial once, or material so unfamiliar that you cannot yet generate useful retrieval cues, some massed exposure is reasonable. Cramming is also the correct response to a deadline that has already arrived. The mistake is treating massed practice as a default study strategy rather than an emergency tool, and then acting surprised when the material expires the moment the pressure does.