Shuffling and selecting students or team members for individual or group-based activities
Fair Play is a simple Android app, intended for use by teachers who need to call on individual students in randomized order or organize students randomly into groups for collaborative activities.
Teachers who frequently have students solve problems at the whiteboard in front of the class, but want to be unbiased in their selections.
As a math teacher with an interactive teaching style, I want to select individual students in a random order for solving problems at the whiteboard, to lessen the likelihood that some students will feel picked on by being called on more frequently than they perceive to be fair.
Teachers or facilitators who conduct group-oriented activities, but don’t want this to encourage the formation of long-term cliques within the class.
As a programming teacher who frequently organizes students into groups for brainstorming/problem-solving activities, I need to be able to assign students to groups randomly, but with weighting of previous assignments biasing the process against repeating the assignment of any given student subset too often, in order to keep the groupings dynamic and encourage cross-pollination of ideas and problem-solving approaches.
The fundamental objective, in developing this app, is to provide a tool for use by teachers, facilitators, etc., to address—in a straightforward, easy-to-understand and use manner—the concerns below.
Selection & shuffling
Many classroom activities require random selection and ordering of students, so that they be called on in a random order. However, as the developers of music player software (for example) have learned, human beings often perceive patterns even in purely random sequences, so it’s sometimes helpful to bias this process. One approach is a sort of weighted perpetual shuffle: Rather than pure random sampling-with-replacement (in which even the most recently selected student is just as likely to be selected next as any other given student), or pure sampling-without-replacement (in which no student that has been called on in the current sequence will be called on again until the previous shuffle is completely exhausted), this is a sampling-with-replacement scheme in which the probability of selection is not uniform across all students, but is weighted so that more recently selected students are less likely to be selected again—but still may be selected.
The hope is that this approach will lessen (even if it never wholly eliminates) the likelihood that a student will perceives themselves to be selected more often than their peers, while still keeping the order from being entirely predictable.
Assignment to groups
Just as an unbiased sampling-with-replacement scheme will occasionally result in seemingly long runs (where the same item is selected multiple times in a row) or near-runs (where the same item appears much more frequently in a subsequence than would typically be expected), random groupings will sometimes result in some combination of items being grouped together much more frequently than expected. Part of this a perception problem: human beings are highly susceptible to confirmation bias, and they often see patterns when there is nothing more than coincidence. But beyond attempting to combat such perceptions, there is value in biasing successive partitioning of members of a set into subsets, to reduce the likelihood of repeated assignments—occurring closely together—of any given combination of 2 or more elements together in a subset. In social or educational contexts, such an approach might be employed to minimize the formation of long-lived cliques, or to promote cross-pollination of ideas.
A recent application of this concept was seen in Friday Networking Lunch (no longer in operation), which used a biased rotation scheme to group members together for weekly lunches at different restaurants, with the promise to the members that (among other things) they would have opportunities to network, face-to-face, with a frequently changing group of other members.
Import, creation, and editing of student rosters.
Flexible configuration of seating diagrams, for visual highlighting of student selection and group assignments.
Configurable weighting of past selections and assignments in new selection and assignments, to control the bias (or lack of same) against repeated selections and assignment groupings.
Manual override of generated shuffles and group assignments.
Student rosters
Seating layouts
Past selections/shuffles.
Past group assignments.
Weighting schemes.
Reading external files for importing student rosters.
Export of student shuffles and group assignments.
Import of rosters from Android contacts.
Export of selections and group assignments to Android messaging or e-mail.
Export (to SVG files) of seating diagrams with group assignments highlighted using different colors.