Cramming works. That is worth saying plainly, because pretending otherwise is why students stop believing study advice.
A concentrated night of work will get you through an exam the next morning. What it will not do is leave anything behind a week later, and in a programme where the material returns, that is the problem.
| Cramming | Spaced repetition | |
|---|---|---|
| Time to set up | None | Weeks |
| Works for an exam tomorrow | Yes | Only if you started earlier |
| Still there in a month | Mostly not | Yes |
| Cost per fact over a term | High, paid repeatedly | Low, paid once |
| Feels like it is working | Strongly | Barely |
Massed repetition builds a memory that decays quickly, and it does so while producing a strong feeling of competence. You have just seen the material dozens of times in four hours, so it is fluent, available and easy. That fluency is short-term availability rather than durable storage, and it is largely gone within days.
Spacing works the opposite way. Each retrieval happens when the memory has partly faded, which makes it effortful, which is what strengthens it. Intervals stretch: a day, then three, then a week, then a month. It feels less impressive at every step and it is still there in June.
Cramming for a midterm means learning the same material again for the final, and again for a licensing exam. You pay full price each time.
Spaced repetition costs more up front and then almost nothing. A card that has stretched to a two-month interval takes a few seconds a year to maintain.
For a one-off course you will never revisit, cramming may genuinely be the rational choice. For anything cumulative, it is the expensive option that looks cheap.
Most students end up doing both, and the sensible division is:
The failure is using cramming as the default and discovering in third year that the first two years are gone.
Cram, and do it properly: retrieval rather than rereading, highest-weighted material first, and sleep, because sleep is when consolidation happens and a night without it costs you more than the extra hours gain.
Then start the spaced system afterwards, for the next one.
No. It is a legitimate short-term tool with a short shelf life. The mistake is using it for material that comes back.
Partly. There is no way to compress weeks of spacing into three days, but starting late still beats not starting.
Long enough for each card to be answered correctly three or four times, which in practice means a few weeks rather than a few days.