What is a group formation algorithm?
A group formation algorithm is a computational process used to organize individuals into groups based on predefined criteria or parameters. In educational contexts, particularly in schools or universities, group formation algorithms are employed by teachers or administrators to create balanced and effective groups of students for collaborative projects or activities.
What considerations are taken in to account?
In FeedbackFruits' Group Formation tool, the following parameters are considered within the algorithm, all of which are determined by the teacher during the survey configuration stage:
Number of students in the group: The teacher sets the desired size for each group, specifying the number of students to be included in each collaborative unit.
Question-level grouping preference: Teachers decide whether they want students to be grouped based on similarities or dissimilarities in their responses to survey questions at the individual question level. This preference influences how the algorithm evaluates and forms groups.
Handling multiple answers: In scenarios where multiple answers are allowed for a question, each answer contributes to the similarity or dissimilarity score. The teacher's configuration choice guides the algorithm in assessing the compatibility of student responses and forming groups accordingly.
How it works
At the core of the algorithm lies the normalized score, a crucial metric calculated based on the number of similar answers within a single question. This score essentially measures how similar their answers are to those of their peers.
In practical terms, when students respond to a question that allows multiple answers, the algorithm examines how many answers two students share. Subsequently, it employs normalization to factor in the total possible answers for that question.
Example: if one student selects 3 answers, another chooses 2 of the same, and there are 5 possible answers in total, the normalized similarity score would be 2 out of 5. This meticulous process is iteratively applied across all questions in the group formation algorithm, delivering a comprehensive evaluation of similarity or dissimilarity in student responses.
By normalizing the similarity score, the algorithm ensures fair comparisons across questions with different numbers of possible answers, making it a more versatile and robust measure of similarity between student responses.
Following the computation of the normalized similarity scores, the algorithm proceeds to organize students by sorting them based on the degree of similarity in their responses. This sorting process enables the algorithm to identify students with comparable answers.
Subsequently, the grouping phase commences, following to the parameters set by the teacher, particularly the specified “Number of students per group” field. The algorithm strategically forms groups by considering the calculated similarity scores, aiming to create balanced and cohesive groups that align with the teacher's preferences. This step ensures a tailored and optimized grouping outcome based on the established criteria.
When the preference is to group students based on dissimilar answers, the algorithm follows a similar process. After computing the normalized dissimilarity scores, the algorithm organizes students by sorting them based on the degree of dissimilarity in their responses. This sorting process enables the algorithm to identify students with contrasting answers. By normalizing the dissimilarity score, the algorithm ensures fair comparisons across questions with different numbers of possible answers, making it a versatile and robust measure of dissimilarity between student responses.