Death to Deadlines A 21st Century Look at the Use of Deadlines and Late Penalties in Programming Assignments Katrin Becker We Need Deadlines… Because….. • Must learn to budget time • Discipline • Instructor workload • Order • Because Not-quite-golden rule: Do unto others what was done unto you. WCCCE -2006 2 Are we really sure that strict deadlines help people learn time management? (after 25 years of teaching, I’m not…) WCCCE -2006 3 Research on Deadlines: • Distance education • Women and minorities • “at-risk” (disadvantaged, encumbered, older) Current Wisdom: “Extensions mitigate against students’ learning how to budget time.” But do they really? Evidence  deadlines = when to work *not* endpoint. WCCCE -2006 4 The Incredible Shrinking Enrollments…. What makes CS unattractive? • Excessive workload • Competitiveness • Dullness WCCCE -2006 5 The Incredible Shrinking Enrollments…. What helps? • Flexibility  autonomy • Learner control (ownership) • Well-defined expectations • Authentic tasks WCCCE -2006 6 What are programming assignments for? • Experience with language • Towards understanding fundamental concepts • Logical program structures • Sound design • Clear documentation • Correctness • Error checking & recovery Adherence to deadlines???? WCCCE -2006 7 Observations Class Demographics (over 7 years): CS101: 60->200 students CS 102: 29 - >150 students 6 assignments / term Worth 20-30% (last 2 X for CS102: 50%) WCCCE -2006 8 Variations on a Deadline 1. Electronically enforced (large class….) On time or don’t bother. Occasional extensions (NOT ideal) 2. 1 letter grade / day late 3. Bonus for “on-time, 3 days’ grace, then late 10-20% consistently earned bonus 4. NO deadlines 5. Bonus for on time + grace + late penalty WCCCE -2006 9 Submission Requirements MUST submit 4/6 - or – Doesn’t matter Made no difference in rates of submission: 1st assignment = 90% submitted Close to 100% of students who remained in class Last assignment = 60-70% submitted Deadlines – strict or non-existent made no difference. WCCCE -2006 10 What was different? GRADES Averages went from ‘C’ to ‘B’ on assignments. WCCCE -2006 11 Can we learn from games??? Quality of learning & will to continue… – Depends on what goals students bring into the classroom & prevailing in-class rewards structures. WCCCE -2006 12 Typical: Relative assessment (curving) leads to competitive ability games failure-oriented WCCCE -2006 13 Results: Failure avoidance, not mastery Other students become obstacles* instead of allies * high score = perfect score WCCCE -2006 14 What can we learn from games? Reward achievement Increasing penalties that start low Opportunities to try again Clear goals One’s achievement not tied to another’s failure WCCCE -2006 15 Putting Ideas to Use Descriptive explanations of requirements and how to meet them Rubrics A-B-C requirements Flexible Deadlines WCCCE -2006 16 Benefits Lowers stress – encourages risk-taking Places control with student Can work around life & other classes Fewer complaints Fix that last bug Gender & cultural equity Reduces risk of cheating? WCCCE -2006 17 Costs Grading process more complex (assignments don’t come in ‘batches’) Re-submission increases workload (only 1020% though) Markers need to be flexible (and competent) – questions get asked out of sequence & out of context Instructors must be attentive to students – learning time management becomes explicit WCCCE -2006 18 Best Practices Clear limits – When – How often Draw the line Bonus in favour of penalty – w/ upper limit for perspective Limited grace period w/ no penalties Increasing penalites Resubmission (once or twice) WCCCE -2006 19 Message Achievement vs. failure avoidance Learning vs. hoop-jumping Collaboration vs. competition Community of Learners Increased retention WCCCE -2006 20 Thanks WCCCE -2006 21