How students leverage assignment submission flexibility — A case study

Abstract

Recent pedagogical trends in post-secondary education focus on how providing students with greater autonomy through assignment submission flexibility offers benefits ranging from increased learning to stress reduction. Unfortunately, the relationship between submission flexibility and any specific benefit is not firmly established. One explanation for this is a potential misalignment between anticipated benefits and an understanding of how most students leverage extended opportunities for assignment completion. The goal of this study was to investigate the relationship between assignment submission flexibility and how students used the opportunity. Quantitative evidence reveals that most students routinely maximized the time taken before submitting assignments. This occurred independent of assignment type, teaching modality, or the duration of assignment availability. The results support a conclusion that most students do not capitalize on increased flexibility to meet the demands of their unique schedules. Instead, they appear to adapt their schedules to submit assignments shortly before a perceived deadline.

Highlights

• •Mose students maximize time taken before submitting assignments.
• •Scheduling constraints do not explain delayed assignment submission patterns.
• •Students do not appear to leverage course flexibility to adapt to their unique schedules.
• •Delayed assignment submission patterns result independent of teaching modality.

1. Introduction

A recent survey of 3004 college students from 128 two- and four-year colleges indicates that the majority of students feel that greater assignment submission flexibility would facilitate their academic success [1]. Consistent with this preference, a variety of studies suggest that flexible assignment deadlines may offer compelling benefits. For instance, some studies support the conclusion that flexible deadlines increase student satisfaction and learning outcomes [[2], [3], [4], [5], [6], [7], [8], [9], [10]]. Flexible assignment submissions dates have also been argued to support student mental health by reducing stress [5,[11], [12], [13]], while also enhancing equity [10,12,14,15].

Given the student-articulated desire for greater flexibility and the potential benefits outlined above, it seems obvious that instructors should reevaluate how assignment deadlines are structured in post-secondary academic settings. Yet important questions remain. For example, how much flexibility is sufficient to provide a student benefit? Is there a level of flexibility that is counter productive? Critically, how can instructors develop strategies that optimize benefit and minimize downside without first understanding how students manage the flexibility? Accordingly, do students leverage assignment submission flexibility to meet their unique scheduling needs or do students adapt their schedule around the ‘flexibility’?

Current pedagogical trends push for recognition that students are capable, self-regulated learners who would benefit in myriad ways from greater assignment submission flexibility [10,12]. Studies investigating the potential benefits of this flexibility often use some variation of a theme where students are allowed to submit assignments after a deadline without penalty and/or without specifying a reason [3,6,8,9]. Unfortunately, realizing the potential benefits of increased flexibility will likely come to rely more heavily of student time management skills if different courses begin to implement flexibility strategies more broadly. Thus, understanding how students use flexibility will be critical to ensuring alignment between instructor-intended benefit and the kind of flexibility offered.

The objective of this case study is to determine how students manage submission deadlines when provided with early assignment access and broad windows of time in which to complete those assignments. Three assignment types were evaluated where each type had a different level of impact on final grades ranging from no impact, voluntary assignments like practice questions to high-impact exams. To better understand the relationship between the flexibility provided and when each assignment type was submitted, anonymized online data logs were queried. This was done for multiple sections of a non-majors undergraduate biology course over the course of six semesters. Moreover, to investigate potential differences in how students might navigate courses with distinct teaching modalities, in-person and online sections of the course were compared.

2. Methodology

2.1. No-stakes practice questions

The first assignment type came in the form of formative practice questions. These questions were intended to boost the depth and breadth of student understanding and were provided in response to increasing student demand for additional study resources. Practice questions were offered on a voluntary basis where one practice question set was provided for each textbook chapter covered during the semester. All practice questions were available online starting the first day of class to allow each student scheduling flexibility. Moreover, question sets could be repeated without time constraint and students were allowed access to any resource of their choosing. To encourage student engagement, a small bonus incentive was offered. Students could earn one bonus point for each practice question set. However, to receive the bonus incentive, students had to 1) correctly answer at least 70 % of the questions from each question set and 2) complete the practice questions by 5 p.m. the day before the exam covering the associated content. To enable feedback in an asynchronous setting, practice question answers became available automatically following the initial question set submission.

Studies indicate that students who distribute their learning/effort over time [16] or complete assignments early [4,17] attain higher levels of achievement. Thus, students were allowed access to practice questions over a broad period of time. For example, the first exam covered 5 chapters over the course of 5 weeks. As such, students had a 5 week window of time to complete 5 practice question sets (∼40 multiple-choice questions per chapter). Over the course of the 14 week semester, 13 question sets were available to students. Thus, successful completion of all practice sets could boost student grades by ∼3 % (13 points out of 420–430 total course points). Given that practice questions were completely voluntary and there was no negative impact on final grades for those students who chose not to participate, this assignment type was classified as ‘no-stakes.’

2.2. Low-stakes reading quizzes

To encourage students to keep pace with the course, a short, ten-question, multiple-choice reading quiz was given each week that exclusively covered content from the assigned reading. The average reading load for each quiz was 25-30 pages from the textbook. With few exceptions, quizzes were available until Friday at midnight each week. Once started, students were provided 20 min to complete each quiz, although the average time needed by students was <45 % (∼9 min) of the time allotted. Due to structural differences, in-person sections were assigned 12 quizzes while online sections were assigned 13 quizzes during the course of each semester. Like practice questions, students had online access to all of the reading quizzes starting the first day of class to allow students autonomy in balancing course requirements with other obligations. For example, students could readily read ahead and take a reading quiz in advance to mitigate the impact and stress of an exceptionally busy week. Individual reading quizzes were required, but classified as ‘low-stakes’ given that each quiz was worth 10 points or ∼2 % of the course total.

2.3. High-stakes exams

The third assignment type was a summative exam designed to evaluate overall mastery of course material. There were 3 exams, approximately 1 exam every 4–5 weeks, during each semester. Each exam consisted of a combination of question types including multiple choice, fill-in-the-blank, matching, short answer, and essay. Students were provided 75 min to complete each exam although, on average, students only used ∼70 % (∼53 min) of the time allotted. To provide scheduling flexibility, exams were available online for 2 days, starting at 8 a.m. on the first day and ending at 5 p.m. on the second. With regard to in-person sections, the first day always overlapped with a normally scheduled class period. However, no meeting was held to ensure students could take the exam during that time if they chose. Each exam was worth 100 points (∼25 % of the total course points). Given that each exam had a significant impact on overall grades they were classified as ‘high-stakes’ assignments.

2.4. Data acquisition

Anonymized assignment and activity data was collected from 604 students who were enrolled in two in-person and five online sections of an introductory cell biology course between the Spring semester of 2021 until the Spring semester of 2023. The university-supported, Canvas Learning Management System (LMS) was used for routine administration of each course section. Therefore, assignment data were obtained for each section using the Canvas LMS REST API for Quiz Submissions (https://canvas.instructure.com/doc/api/quiz_submissions.html). These datasets included information related to each assignment type (practice questions, quiz, or exam) and the time of submission. Course activity data were collected from the Canvas New Analytics reports and course activity features. Pre-recorded course videos were stored on the university-supported Kaltura platform (https://www.kaltura.com) which students accessed through integration with Canvas. Video access and streaming activity data were obtained from Kaltura by downloading the ‘User Engagement’ data. Downloaded datasets were then automatically parsed to capture content type and access time.

2.5. Data availability

The anonymized datasets used for generating each figure in this study are freely available and can be downloaded from Mendeley Data (https://doi.org/10.17632/ftz87pkvfx.1; https://data.mendeley.com/datasets/ftz87pkvfx/1).

2.6. Variables and assignment submission data analysis

To relate how students manage assignment deadline flexibility between three different assignment types, each having a unique availability timeframe, the idea of an elapsed opportunity window (EOW) was developed as a quantitative method for comparison. Only three variables were considered in calculating the EOW: 1) the initial time when an assignment become available (T1), 2) the time when the opportunity window ended (T2; defined due date), and 3) the time when an assignment was submitted (T3). The EOW was calculated by first determining the time difference between when an assignment became available (T1) and when the assignment was submitted (T3). This difference was divided by the time difference between assignment availability (T1) and when the assignment became due (T2) and then multiplied by 100 [EOW = (T3-T1)/(T2-T1)∗100]. In essence, the EOW represents the percentage of available time that students allow to pass before submitting any particular assignment. EOW calculations for each assignment were plotted on a histogram using ggplot2 (https://ggplot2.tidyverse.org/), a data visualization package for the open source R programming language (https://www.r-project.org/). All statistical analysis was performed with rstatix (https://cran.r-project.org/web/packages/rstatix/index.html), a framework for basic statistical tests.

2.7. Analysis of online activity

Canvas course activity logs record the time and frequency in which course content was accessed, as well as the duration of activity. These anonymized activity logs were queried for all non-assessment-based activity as a proxy for student schedule availability and course engagement. For analysis, each activity timestamp was transformed into a percentage of a week where Friday was used to mark the end of the week or the weekly deadline. This day was chosen to be consistent with the course schedule and assignment structure. An EOW was calculated for each recorded event to better align and compare student activity with assignment submissions.

2.8. Ethics and consent

Institutional review board (IRB) staff evaluated of the proposed research leading to this study and determined that the activity was not subject to the requirements of the Department of Health and Human Services or the Federal Drug Administration for research involving human subjects.

3. Results

To investigate how assignment submission flexibility is used, students were provided specific time windows in which to complete each assignment type. This time window was defined as the ‘opportunity window’ and ranged from days to weeks, depending on the assignment type. The opportunity window was then used to calculate the percentage of time that elapsed before the assignment was submitted (the elapsed opportunity window; EOW, see Methodology). A negative EOW indicates an assignment was submitted early, an EOW between 0 and 100 indicates the submission occurred on time (e.g. within the opportunity window), and an EOW >100 indicates the assignment was submitted late. Importantly, given the broad temporal range in which assignments were available, the end of the opportunity window served as a fixed due date. As such, assignments were not accepted late without a student-initiated extension request. Moreover, while students could readily submit most assignments in advance, early exam submissions required a student-initiated request.

3.1. No-stakes practice questions

From the 604 total enrolled students (56, in-person; 548, online), 5320 (in-person, 585, ; online, 4735) practice question sets were completed. No practice question sets were submitted early in the in-person sections, however, 42 (0.9 %) question sets were submitted early in the online sections (Table 1). By comparison, 97.4 % of question set submissions occurred on time in both in-person and online sections, while 2.6 % (in-person) and 1.8 % (online) of submissions occurred late (Table 1). 82.9 % (504/604) of students successfully completed at least 1 practice question set, where the average number of submitted sets was >10 (5320/504) out of 13 available. This indicates good participation rates, although 17.1 % of students chose not to participate, despite the bonus incentive.

Table 1.

Percent of submissions for each indicted assignment from students enrolled in in-person (IP) and online (OL) sections.

	Early % (#)		On Time % (#)		Late % (#)
	IPL	ODL	IPL	ODL	IPL	ODL
Practice	0.0 (0)	0.9 (42)	97.4 (570)	97.4 (4610)	2.6 (15)	1.8 (83)
Quiz	0.0 (0)	1.8 (104)	97.8 (578)	96.4 (5634)	2.2 (13)	1.8 (104)
Exam	0.6 (1)	0.5 (8)	94.5 (155)	96.9 (1525)	4.9 (8)	2.6 (41)

To better understand how students managed question set assignment submission flexibility, an EOW was calculated for each submission within the opportunity window, a four or five week window before each associated exam. The results were grouped and plotted according to course modality (in-person vs online). The average EOW was 81.2 % (median: 93.8 %) for in-person sections (Fig. 1A) and 81.0 % (median: 93.5 %) for online sections (Fig. 1B). This indicates that final submission of practice question sets, on average, occurred about a week before the availability window closed (the deadline) and the associated exam was scheduled. By comparison, the median EOW indicates that half the submissions occurred within two days before the opportunity window closed.

Fig. 1.

Histograms representing no-stakes practice question set submissions (A,B) and access (C,D) with associated elapsed opportunity window. A) In-person (n = 570; ave = 81.2 ± 25.4 sd; med = 93.8) and B) online section (n = 4,610, ave = 81 ± 25.1 sd; med = 93.5) practice question set submissions. C) In-person (n = 1550; ave = 77.6 ± 28.3 sd; med = 91.7) and D) online section (n = 14,093, ave = 77.5 ± 29 sd; med = 92.5) practice question set access. Inset percentages reflect the proportion of submissions/access during the indicated week range.

Given that practice question sets were available over a four to five week period and that question sets could be repeated multiple times before final submission, it is not possible to interpret the submission data above without also understanding the timing and frequency in which the question sets were accessed. As such, the course logs were queried for all practice question set activity and an EOW was calculated and plotted. In total, the question sets were accessed 1550 and 14,093 times in the in-person and online sections, respectively (Fig. 1C and D). This activity, divided by the number of submissions (1550/570 for in-person; 14,093/4610 for online) reveals that each question set was accessed 2–3 times, including the final submission. Relative to submissions, the average EOW for practice question set activity was 77.6 % for in-person (Figs. 1C) and 77.5 % for online sections (Fig. 1D). This indicates that, on average, question sets were submitted within approximately 1–1.5 days after being first accessed. By comparison, median data indicate that half of the submissions were accessed and submitted within ∼8 h. Importantly, the data also reveal that between 69 and 71 % of students allowed >85 % of the opportunity window to elapse before accessing any of the multiple, yet voluntary, practice question sets.

3.2. Low-stakes reading quizzes

In total, 6433 reading quizzes were submitted (591, in-person; 5,842, online). Students enrolled in the in-person section submitted an average of 10.9 (591/54) out the 12 assigned quizzes, while online sections submitted an average of 10.7 (5842/548) quizzes out of the 13 assigned. This indicates an 82–90 % completion rate. However, this average value is artificially low given that the data includes students who withdrew from the course (5/54, in-person; 34/548, online) and parsing out data from withdrawn students was not possible given the anonymized nature of the dataset. However, like that for practice question sets, there were no early quiz submissions from the in-person sections, but 104 (1.8 %) quizzes were submitted early from the online sections (Table 1). In both course modalities, 97.4 % of quizzes were submitted on time, while 1.8–2.2 % of quizzes were submitted late.

To evaluate how students managed the flexibility offered for quiz completion, an EOW was calculated for each quiz submitted within the opportunity window. The data was then parsed and plotted as before. Importantly, although all quizzes were available starting the first day of class, the EOW was calculated using an opportunity window of a single week — the time in which the content was covered. For in-person sections, the average EOW was 97.3 % (median of 99.4 %, Fig. 2A), while online sections had an average EOW of 91.7 % (median of 98.0 %, Fig. 2B). For context, this means that the average quiz submission occurred around noon on the day it was due, while 50 % were submitted after 8 p.m. By comparison, 31.7 % (1967/6212) of quiz submissions occurred after 11 p.m. Unsurprisingly, assignment deadline flexibility did not eliminate late submissions (1.8–2.2 %; Table 1) by 77 students, which required a student-initiated extension request.

Fig. 2.

Elapsed opportunity window distribution of low-stakes quizzes submissions from the A) in-person (n = 578, ave = 97.3 ± 7.5 sd; med = 99.4) and B) online sections (n = 5,634, ave = 91.7 ± 15.7 sd; med = 98) where the inset day and percentages indicate the proportion of the total submissions within the specified range. C) Distribution of early quiz submissions from online sections (n = 104, ave = −108.1 ± 181.1 sd, med −29.1) where the inset negative numbers indicate the number of weeks, before the designated opportunity window, and the associated percentage of quizzes that were submitted.

Although the vast majority of students significantly delayed submitting quizzes, 39 (6.4 %; 39/604) students enrolled in online sections submitted quizzes early (Table 1). From this group, 22 students submitted 1 quiz, 7 students submitted 2–4 quizzes, 7 students submitted 5–7 quizzes, and 2 students 10–11 quizzes. The average EOW for quizzes submitted early was −108 % (median: 29.1 %; Fig. 2C). This data indicates that two students submitted quizzes as much as 9 weeks early, however the average submission occurred about a week in advance.

3.3. High-stakes exams

Greater than 90 % of students submitted all 3 exams in both in-person (155/162) and online sections (1525/1644). Like that for quizzes, average exam submission rates were artifactually low. The total number of possible exam submissions was determined using the total enrollment, which includes withdrawn students. As such, the actual submission rates are likely 96 % or greater. Establishing the exact rate is not possible without knowing when students withdrew from the course and this information was not present in the anonymized activity logs.

Like that for no-stakes and low-stakes assignments, an EOW was calculated for each submitted high-stakes exam. The average EOW was 68.9 % (median of 88.2 %, Fig. 3A) and 79.2 % (median of 91.7 %, Fig. 3B) for in-person and online sections, respectively. Between 18 and 31 % of exams were submitted on the first day the exam became available, while ∼69–82 % were submitted on the second day (Fig. 3A and B). Interestingly, only 3 (1.9 %) exams were submitted from the in-person sections during the normally scheduled class period. Of the exams submitted on the second day, 64.9 % (880/1363) were taken only after ≥90 % of the opportunity window elapsed. For context, this means that these exams were submitted within 3 h or less until the deadline. Critically, 40.0 % (546/1363) of exams were submitted within 75 min or less remaining (≥96 % EOW) before the deadline.

Fig. 3.

Elapsed opportunity window distribution for high-stakes exam submissions from the A) in-person (n = 155, ave = 68.9 ± 33.6 sd; med = 88.2) and B) online n = 1525 (ave = 79.2 ± 26.9 sd; med = 91.7) sections. Insets indicate the day and percentage of total submissions within the specified range.

3.4. Student activity and engagement

Overall, the preponderance of evidence reveals a common and consistent pattern. Despite having extended access and a broad window of time in which to complete assignments, the majority of students allowed >90 % of the opportunity window to elapse before completing assignments. This occurred independent of assignment type, length of opportunity window, or course modality (Fig. 1, Fig. 2, Fig. 3). However, interpreting these findings requires a better understanding of student activity and their availability. For example, despite broad opportunity windows, students might delay assignment submissions due to constrained schedules that limit course engagement to specific periods of time. Alternatively, delayed submissions may reflect students prioritizing other obligations, problems with time management, procrastination, or any number of other reasons.

To evaluate these possibilities, course logs were analyzed for all non-assessment-based activity. In total, 16,084 activity events were identified for the in-person sections where the average EOW was 61.2 % (median: 66 %, Fig. 4A). Not surprisingly, the least amount of activity occurred on Saturday (7.5 %) and Sunday (6.7 %). However, significant activity was observed throughout the week with the highest activity on Thursday (20.8 %) and Friday (20.7 %). Importantly, two peaks were observed on Tuesday and Thursday, which overlap with the regularly-scheduled course meeting time. By comparison, 155,070 events were logged for the online course where the average EOW was 60.5 % (median: 64 %, Fig. 4B). Like the in-person section, the least amount of activity was observed on Saturday (7.5 %) and Sunday (8.9 %). However, significant activity was also observed throughout the week with the greatest level of activity on Friday (25.1 %). Combined, these results indicate that activity is distributed throughout the week and suggests broad student availability for those enrolled in either the in-person or online sections.

Fig. 4.

Distribution of online access to non-assessment-based course content from the A) in-person (n = 16,084, ave = 61.2 ± 26.7 sd; med = 66) and B) online (n = 155,070, ave = 60.5 ± 28.1 sd; med = 64) sections. Insets indicate the day and percentage of total activity within the specified range. C) The percentage distribution of streamed videos (n = 13,682 total) and the associated time spent (2905 total hr) from 3 online sections (318 students total) for each indicated day of the week.

A limitation of the course activity logs is that the duration of activity recorded for each event may not be meaningful. For example, if a student accessed a course webpage, the activity was recorded as a single event. However, if the student leaves a course webpage open and walks away, the time duration recording continues. As such, activity duration could not be confidently used as a reflection of meaningful course engagement (data not shown). As an alternative, logs from the video storage platform, which records both the number and duration of videos streamed were used to evaluate engagement. This data was available for 318 students enrolled in three online sections of the course. While the anonymized video logs do no provide granular time-of-day access data, they do provide the day in which content was accessed. The video logs reveal student engagement occurs throughout the week (Fig. 4C), where the pattern of video views and time streamed are markedly similar to the course activity logs (Fig. 4B vs 4C). Given this, it is fair to conclude that the activity events serve as a reasonable proxy for student availability and engagement.

Collectively, the student activity and engagement data indicate broad student availability throughout the week. Yet most assignments were submitted shortly before the assignment availability window closed. As a consequence, student scheduling constraints are unlikely to account for the assignment submission patterns observed. Moreover, the delayed submissions do not appear to result from a general student adaptation to a ‘Friday’ deadline. 51.8 % (28/54) and 56.6 % (77/136) of the total number of assignments from the in-person and online sections, respectively, were due on a day other than Friday. Thus, regardless of the time of day or day of the week, most students maximized the time taken before submitting assignments. In essence, the data strongly suggest that the majority of students did not leverage the broad assignment submission flexibility to meet the needs of their individual schedule. Instead, the data supports the idea that students adapted their schedules to meet the assignment deadline.

4. Discussion

4.1. Flexibility to promote learning outcomes

Published studies draw a connection between the types of study approaches that positively correlate with course performance. Not surprisingly, students who invest in connecting ideas and conclusions to a particular subject matter developed competencies in that subject matter, and as a result, were found to perform better than those students who superficially memorized the material (Elias, 2005; Davidson, 2002). Moreover, other reports indicate that students who demonstrate good time management and distribute their effort over time perform better ([16]; Stewart et al., 2016). Thus, providing students with early and continued access to formative assessments like practice questions are an obvious opportunity for instructors facilitate student learning. However, the efficacy of any resource in promoting student learning will always be dependent on how students choose to leverage the opportunity offered.

The results presented here indicate that most students do not capitalize on the early availability to resources like practice questions, which were intended to expand their depth and breadth of understanding. Instead, the majority access the content shortly before the associated exam. In other words, students crammed multiple chapters of information within hours of the deadline, despite having unlimited access to the content for four to five weeks previous. These data not only provide direct and quantitative evidence supporting published survey studies suggesting that 85–90 % of students procrastinate [18], they reveal the extent to which students delay before choosing to engage. This delay was independent of teaching modality given that no difference between in-person and online sections was observed. At least in the context of this particular course, the data also seem to contradict the notion that providing extended assignment access and broad submission flexibility influences engagement patterns for the average student. Given this, it is hard to envision how the practice questions, as structured, could have met the goal of enhancing the depth of student understanding. While the student-requested resource may have helped a limited number a students, it is likely that the high levels of last-minute participation simply reflected a motivation for extra credit, similar to that previously reported by others [19].

4.2. Flexibility to reduce stress

Assignment submission flexibility has been argued to reduce stress by providing students greater autonomy and control over the learning process [5,11,12]. The idea being that flexibility allows students more power in balancing personal, academic, and professional obligations. The submission patterns observed indicate that most students routinely maximize the time taken before submitting quizzes. While delayed submissions were anticipated, the extent and magnitude to which students delayed was surprising given that half of the submissions occurred only after ∼98 % of the opportunity window elapsed. Given this, it is hard to reconcile how the manner in which most students leverage quiz submission flexibility could lead to reductions in stress. This later point is not to suggest that submission flexibility cannot be used as a mechanism to reduce stress. Instead, it is to illustrate that, despite the flexibility provided, most students did not appear to leverage the flexibility in a manner consistent with stress reduction. This is especially evident in the high-stakes exam data where nearly a third of all students delay starting the exam until an hour or so before the deadline.

5. Conclusions

The major finding of this investigation is that when provided assignment submission flexibility, most students maximized the time taken before submitting assignments. This occurred independent of assignment type, length of time students had access and opportunity to complete the assignment, or course modality. Moreover, the activity records fail to identify specific scheduling constraints that could account for assignment submission patterns. To be clear, assignment structure, availability, and the rationale behind the flexibility offered was explicitly described to students at the beginning of each semester. This information was also posted online in the course syllabus for continued access. However, the consistency of delayed submissions strongly suggests that assignment submission flexibility has little, if any, impact on the majority of student scheduling decisions. Thus, a reasonable conclusion is that despite broad flexibility, most students did not recognize the flexibility as such. Instead, students identify the end of the opportunity window as the ‘deadline’ and make a choice to act accordingly. In essence, the majority of students do not capitalize on assignment deadline flexibility to meet their unique and diverse scheduling constraints. Instead, they adapt their schedule to meet the ‘deadline’.

This raises an important consideration. How do students interpret assignment submission flexibility? A recent survey seems to suggest that students primarily consider assignment submission flexibility only in the context of a late submission [20]. In this limited view, front-loading flexibility and providing students with broad opportunity to manage their own schedules will unlikely provide the benefits the flexibility was intended to provide. It is important to note that a subset of students clearly leveraged the assignment flexibility offered. This subpopulation likely represented students with strong study and time management skills. Thus, it stands to reason that if meaningful benefit is to be realized from assignment submission flexibility, it must be coupled with increased student study and time management skills.

The extent to which the findings presented here apply to other courses across disciplines is unknown. However, given the long and extensive literature investigating student procrastination [3,18,[21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]], a reasonable prediction is that they are generalizable. If true, there may be unintended consequences if assignment submission flexibility is broadly implemented without first understanding how students will use that flexibility in any particular context. For instance, if the tendency of most students is to maximize the time before submitting assignments and recalibrate their schedules to meet a perceived ‘deadline’, what happens when assignment due dates from multiple courses begin to stack/overlap? The literature suggests increased stress [21,27,33]. This would completely undermine a compelling rationale for offering the flexibility in the first place. Importantly, this scenario reinforces the need for developing student study and time management skills so that students are better equipped to take advantage of opportunities as they arise.

Finally, one should not interpret these findings to argue for or against integrating flexibility in any particular course. The data only illustrate how students consistently used the assignment deadline flexibility offered in this specific context. Unfortunately, the quantitative data tend to reinforce a large body of qualitative literature related to student procrastination [34]. Whether or not the data reflect actual student procrastination, their prioritizing other events, and/or reveal issues with time management is a matter of debate. The important point is that incorporating flexibility as a strategy to improve educational outcomes should be informed by an understanding of how students will use the opportunity if any benefit is to be realized.

6. Limitations

This study relies solely on anonymized datasets. As such, factors that may influence how students navigate course flexibility (i.e. student mental health, motivation, student satisfaction, etc) cannot be taken into account. Moreover, this study does not directly address the relationship between how course flexibility may impact learning outcomes, whether the flexibility offered had any impact on student stress, or if the flexibility impacted student satisfaction. Understanding these relationships will be important in facilitating students success and will require additional study.

AI and AI-assisted technologies

No AI or related technologies were used at any step in the preparation of this manuscript.

Funding disclosure

This work was not funded by any external source.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.