Overview

Split deadlines can provide equal hours while producing different experiences because information timing, support queues, neighboring workload, and student pacing shape the usable work window. This chapter examines the CS 124 Early/Later policy and its Spring 2023 survey, which received 141 responses from 746 students; respondents generally rated it fair and effective, but the volunteer sample could not represent the whole class. It also shows why subgroup effects and non-significant grade differences do not establish equivalence or lasting habit change. Fairness is a condition to test, not an assumption from equal duration.

Learning outcomes

Equal working time is necessary for a fair split deadline, but equal information and aligned test access are also needed to make that time usable.
A one-day split can interact with public discussion, support timing, neighboring assignments, and whether students delay work.
CS 124 respondents generally viewed the policy as fair and effective, but 141 of 746 volunteer responses cannot establish identical perceptions or universal fairness.

Keywords

fairnessinformationworkloadsupportperceptionssubgroups

Fairness Is Part of the Mechanism

A split deadline can be formally equal and still feel unequal in practice. The policy may give two groups the same number of hours, yet students may encounter different information, neighboring assignments, support queues, or emotional pressure on those timelines. This chapter examines fairness as part of deadline design in computing education: how the CS 124 case handled equal time and information, how students perceived the result, why average success can hide unequal effects, and what the evidence still cannot establish.

Equal time is a starting condition, not the whole experience

In the CS 124 policy, the Early and Later groups received the same amount of time to complete each project checkpoint. Release dates and test-suite access were staggered with the due dates, and students were randomly assigned to groups. Those choices address several obvious sources of unfairness. One group was not meant to receive extra working time, better tests, or a hand-picked deadline. The policy was also explained to students and kept the separation to one day.

Those conditions are important because a later due date can otherwise carry a hidden advantage. If the Later group sees the assignment materials earlier, students in that group may have both more calendar time and more opportunity to observe how the task is being interpreted. If the test suites are not released at the same time relative to each group's start, the groups have unequal evidence about whether their programs work. Equal hours only mean equal opportunity if the information needed to use those hours is also aligned.

Yet a formal work window is not experienced in a vacuum. Students who habitually wait until the last day may treat the Early group's nominal start as a practical loss, even when the Early group has the same duration on paper. In the survey responses, some students explicitly said that procrastination made the Early group feel as if it had less time. Others worried that the Later group could learn from friends, public questions, or a more established course forum before reaching its own deadline. On Early days, Later students could still seek help because their deadline was approaching; on Later days, Early students were beginning a new assignment and generated less comparable demand. The same one-day separation could therefore create different support environments.

The interaction with other deadlines adds another layer. Split deadlines applied to programming assignments, while daily exercises and quizzes remained shared. Some students reported that one group could face a programming assignment alongside a quiz and homework on the same day while the other group did not. A policy can be equal within the assignment it splits and unequal within the student's broader workload. Fairness must therefore be checked across the course calendar, not inferred from the assignment duration alone.

What students reported about the policy

The Spring 2023 survey collected 141 responses from a class of 746 students, an 18.9% response rate.

Published by DeepMox

FAQ

Does equal deadline duration guarantee equal opportunity?

No; equal duration supports equal opportunity only when information, test access, support, neighboring workload, and practical timing are also aligned.

How did students perceive the CS 124 split-deadline policy?

In the Spring 2023 survey, respondents generally rated it fair and effective, with no statistically significant Early/Later difference; the 141 of 746 volunteer responses limit generalization.

Why does a non-significant grade difference not prove equivalence?

A non-significant result can reflect small differences, limited power, ceiling effects, or an insensitive measure; it is compatible with no meaningful disadvantage but does not prove equivalence.

Previous11 minCS 124 Split Deadlines and Support Queues

Next9 minDo Deadlines Reduce Procrastination?

Fairness in Split Deadlines

Fairness Is Part of the Mechanism

Equal time is a starting condition, not the whole experience

What students reported about the policy

FAQ