5 Evaluation and Continuous Improvement
Evaluation of my teaching relies on data triangulation: quantitative institutional evaluations, detailed qualitative reports, and direct feedback on pedagogical innovations.
5.1 Quantitative Synthesis of Teaching Evaluations
My portfolio shows consistent student appreciation, recognized through repeated receipt of the “Enseignant Etoile” award (2014, 2023, 2024, 2025) at the Faculty of Science and Engineering.
5.1.1 Overall Trend
Scores remain high and often above the departmental average. The figure below normalizes results (as a percentage of the maximum) to enable cross-comparison despite changes in scoring scales (3, 4, or 100).
5.1.2 Course-Level Detail (Most Recent Session Available)
| Course | Title | Semester | Score | Scale |
|---|---|---|---|---|
| MAT-1905 | Arithmetic (Elementary Education) | Summer 2024 | 2.98 | /3 |
| MAT-1906 | Geometry (Elementary Education) | Winter 2025 | 2.82 | /3 |
| STT-1100 | Intro. Statistical Software | Winter 2024 | 2.91 | /3 |
| STT-2200 | Data Analysis | Fall 2024 | 2.96 | /3 |
| STT-4230 | R for Scientists | Fall 2024 | 2.97 | /3 |
| 8INF416 | Visualization (UQAC) | Winter 2025 | 3.81 | /4 |
| 8STT108 | Massive Data (UQAC) | Winter 2025 | 3.86 | /4 |
(Consolidated table from CV and institutional reports)
5.2 Qualitative Analysis and Feedback on Innovation
Beyond scores, I focus on understanding the learning experience, especially when introducing new methods.
5.2.1 Case Study: “Mandatory AI” Assessment (STT-4230)
Within the LLM innovation described in Section 4, I conducted a specific survey to measure perceived pedagogical impact.
Perceived success: students report that the exercise strongly helped them understand object-oriented programming (score 4.75/5).
Technical friction: qualitative comments highlighted the difficulty of installing local models (Ollama), creating technical inequality.
Corrective action: this feedback directly led to creating a starter template for the next iteration, to reduce setup-related cognitive load.
5.2.2 Case Study: Cognitive Load in STT-1100
The simultaneous introduction of Git, R, and Quarto was identified as a barrier for complete beginners.
- Corrective action: the course plan was adjusted to introduce tools more sequentially (GitHub only from module 2 onward).
5.3 Philosophy of Learning Assessment
My approach to student assessment mirrors my approach to my own teaching: formative and iterative.
Automated feedback: in programming courses, unit tests provide immediate feedback, enabling trial-and-error without penalty.
Peer assessment: used in final projects, it forces students to adopt a critical stance and read code other than their own.