Docs / Why it works this way
Why it works this way.
Some of what Ignia does is counter-intuitive on purpose. It asks you questions instead of explaining again. It waits before bringing something back, then brings it back just as you were about to lose it. It refuses to tell you that you have mastered anything.
None of that is a style choice. Each one is a response to a specific obstacle, and each page in this section is built the same way: here is what you want, here is the obvious way to get it, here is why the obvious way fails, and then the mechanism we use instead.
-
Why you are asked, not told.
Rereading a lesson makes it feel familiar without making it recallable — and familiarity is the signal you use to judge that you know something. Why answering is the app’s default verb, and what that does not buy.
Read
-
Why the app makes you wait.
Reviewing something often, soon after learning it, is the obvious move and it is the wrong one twice over. How the schedule decides, what it is actually estimating, and where the estimate is weakest.
Read
-
Why it is harder than it needs to be.
A right answer in an easy format is worth less to your schedule, and the support goes away one question after it arrives. Why smoothing practice out damages the schedule as well as the learning — and which frictions are not for your benefit at all.
Read
-
Why it mixes topics up.
Finishing one topic before starting the next makes practice go better and learning go worse. How the mixing is gated on what you have already met, and the two conditions under which the effect is known to reverse.
Read
How to read this section.
- Every principle comes with its limit. A page here does not end on the mechanism. It ends on what the mechanism does not buy you, because a claim with no stated limit is a claim you have no way to check.
- When a number is a convention, we say so. Not every choice in the app is forced by evidence. Some are calibration, some are inherited, some are just a reasonable line drawn somewhere. Those are marked as such rather than dressed up as findings.
- Claims about the research point at the paper. Every page ends on a list of its sources, with a line saying what each one is doing there — including the ones that cut against us, and the one whose famous number we deliberately do not use. What you will not find is an effect size quoted on the page: an average across studies does not predict your result, and a number without its conditions reads as more precise than it is. They are in the papers, where the conditions are too.
- Claims about the app point at the code. Where a page describes something the engine does, it names the function or constant so you can go and read it. The technical section documents the same mechanisms at source level, including the parts that are not finished.