Skip to content

Selection Algorithms

Octav Popescu edited this page Nov 10, 2026 · 11 revisions

TutorShop has a number of built-in problem selection algorithms an author could choose from. Besides that it also allows authors to implement their own custom selection algorithms and link it to the system in one of several ways.

Mastery Learning means choosing the problem that could advance the student furthest toward completion. If the problem set's skills of interest is the set S and the student has mastered no skills in S, the algo should select first the problem having the most opportunities to demonstrate all skills in S.

Mastery Learning - Easier First means choose the problem closest to the student's current skills. If the problem set's skills of interest is the set S and the student has mastered no skills in S, the algo should select first the problem having opportunities to demonstrate the fewest skills from S, and, of those, the skill(s) the student is already closest to mastering.

Basically the system computes a score for each problem, which is a weighted average of the missing skill probabilities. Where a missing probability is the difference between the mastery level (.95) and the current skill p-known, and the weights are given by the number of opportunities for each skill. And then the first algorithm takes the problem with the highest score, while the second one takes the problem with the lowest score. If there are more problems with the same highest/lowest score, they will choose one randomly. So in your example it all depends on what the p-knowns for the two skills are after finishing the first two problems.

Mastery is hard coded to 0.95 probability. The formula to update skills is found in some of Albert's papers that I can never find a reference to, you might have learned about it in school. Basically it all depends on what the skill parameters are set to. With the default parameters you need about 3 sequential correct answers to get to mastery. But if you set p-learn close to 1 and p-guess and p-slip close to 0 you should be able to get to mastery after 1 correct answer. These parameters can be adjusted for each skill if you go to the skills' Settings pages in Tutorshop, you might have seen that.

We do have a way for authors to customize the problem selection algorithm, by replacing the standard ones with their own. This could be implemented either as a script running on the same server, or an internet service running on any other server. Tutorshop would send to this service the student model and some data about the problems to choose from in JSON format, and the service would return the next problem choice using whatever strategy it wants.

For details about the TutorShop interface to custom problem selectors see Interface to Custom Problem Selectors

Clone this wiki locally