Kodokon kodokon.com

Getting custom quizzes and exercises made for you

Have the AI generate multiple-choice quizzes, open questions, and exercises of increasing difficulty with commented corrections.

8 min · 3 questions

Open this lesson in Kodokon

Testing yourself is the most effective way to consolidate knowledge - far more than rereading a lesson. The problem: ready-made quizzes never match exactly what you've just studied. The AI solves this: it can build, in thirty seconds, a quiz calibrated to your chapter, your level and your weak spots. You just have to ask for it with the right constraints.

PROMPT
You are a programming assessment creator.

Build me a 8-question multiple-choice quiz on loops in Python (for, while, break, continue, range).

Constraints:
1. Increasing difficulty: questions 1-3 easy, 4-6 intermediate, 7-8 hard.
2. 4 options per question, only one correct.
3. The wrong answers must be PLAUSIBLE: build them from the classic beginner mistakes (off-by-one index, inverted condition, break/continue confusion).
4. Ask ALL the questions first, without the answers.
5. Wait until I give you my 8 answers, then provide a commented correction: for each question, why the right answer is right AND why my answer was wrong if applicable.
The quiz generator: constraints 3 and 4 make all the difference.

Two constraints in this prompt deserve your attention. Constraint 3 demands plausible distractors: a quiz where the wrong answers are absurd tests nothing; a quiz where they reproduce classic mistakes reveals your real confusions. Constraint 4 forces you to separate questions from answers: if the AI shows the correction at the same time, your eye reads it before your brain has worked, and the test loses all its value.

Multiple-choice quizzes check recognition; open questions and coding exercises check production, which is more demanding and closer to professional reality. Alternate the two formats: quizzes to sweep through a chapter quickly, exercises to check that you really know how to write the code.

PROMPT
You are a teacher preparing lab exercises.

Create 3 coding exercises for me on array manipulation in JavaScript (map, filter, reduce), of increasing difficulty:
- Exercise 1: direct application of a single method.
- Exercise 2: combining two methods.
- Exercise 3: a small real-world case (for example, processing a list of orders or users).

For each exercise, give: the prompt, the input data, and the exact expected result - but NOT the solution.

I'll submit my answers to you one by one. For each attempt, reply only with: what's correct, what isn't, and a hint to make progress. You'll show a commented solution only after my SECOND unsuccessful attempt, and explaining each line.
The exercise generator: the correction only comes after real effort.

The wrong usage, so you can recognize it clearly: asking for the exercises WITH their answer key, reading everything, and telling yourself "yes, I'd have found that." It's the most widespread illusion of competence: understanding a solution has never proven that you knew how to produce it. The right usage imposes the opposite order: prompt, effort, attempt, and only then the commented correction.

Knowledge check

Make sure you remember the key points of this lesson.

  1. Why require "plausible" wrong answers in a generated quiz?
    • To make the quiz longer
    • Because distractors based on classic mistakes reveal your real confusions
    • So the AI answers faster
  2. Why ask the AI to pose the questions WITHOUT the answers?
    • To save space on screen
    • Because reading the correction before searching cancels out the effect of the test
    • Because the AI is often wrong in its corrections
    • To be able to print the quiz
  3. What is the "illusion of competence" mentioned in the lesson?
    • Believing you can produce a solution because you understood it while reading it
    • Thinking the AI is more capable than it is
    • Overestimating the difficulty of an exercise
    • Passing a quiz by luck