FOR INSTRUCTORS · EVIDENCE OF LEARNING

Is the course working?

Seven measures, computed in your browser from files you already have, with confidence intervals and effect sizes ready for a course report or a paper. Nothing is uploaded or stored. The design, research questions and ethics steps are in the evaluation plan.

1. Concept gain. The 12-item concept inventory on Chapters 13–20 (kept privately with the instructor), taken before the Chapter 13 lecture and again after Chapter 20, both as Blackboard tests. Chapter 12 was taught before the pre-test and is not included. Reports the normalized gain ⟨g⟩ = (post − pre) / (100 − pre).
2. Peer-instruction gain. The class votes in each lecture. On the episode close, presenter mode offers Export class evidence (CSV). Reports % correct before and after discussion, per act.
3. Ownership. The micro-viva record sheet, downloaded as CSV and completed. Reports how often the oral explanation agrees with the written work.
4. Engineering build. The students’ final exports: course checks, own tests, known bugs caught, and REFIT choices, per assignment.
5. Assessment trajectory. The Grade Center: means per assignment, the within-course comparison of Assignments 1–2 with 3–9, and the correlation with the exams.
6. Perception survey. Twelve items on four scales, with Cronbach’s α.
7. Research dataset. One de-identified row per consenting student.

Reference points (not targets): traditional lecture courses commonly report ⟨g⟩ around 0.2 and interactive-engagement courses 0.35–0.5 (Hake, 1998). In peer instruction, % correct usually rises after discussion when the first-round result is between about 35% and 70% (Crouch & Mazur, 2001). Effect sizes follow Lakens (2013); reliability follows Cronbach (1951). Small sections give wide uncertainty; read the N and the interval, not only the point estimate.

1 · Concept inventory: pre/post gain

Download each test's column from the Blackboard Grade Center (Work offline → Download, selected column, comma-delimited). Choose the two files. Students are matched by the identifier column.

2 · Peer instruction: votes before and after discussion

Choose one or more class-evidence CSV files exported from the lectures (several sections or chapters at once is fine).

3 · Micro-viva: does the explanation agree with the written work?

Choose the completed record sheets (CSV from the micro-viva page). The agreement column should say agrees, partly or does not agree.

4 · Engineering build: what the code and tests show

Choose the students’ final exports: the final editable .md files, or the submitted PDFs (text is read in your browser). Each file gives one build record: course checks, own tests, known bugs caught (mutation test, Assignments 3–9), and the boundary status and REFIT choice after the new evidence.

5 · Assessment trajectory and within-course comparison

Download the full Grade Center (Work offline → Download → Full Grade Center, comma-delimited). Columns named like “Assignment 3”, “Midterm” and “Final” are found automatically; maximum points are read from “Total Pts” in the header (otherwise Assignment 1 = 4, Assignments 2–9 = 5).

Design note: Assignments 1–2 (Chapters 10–11) keep the earlier format; Assignments 3–9 use the story lecture, the Python build and the new-evidence cycle. The same students therefore provide a within-course comparison. It is quasi-experimental: chapter content and timing also differ, so report it with that limitation.

6 · Student perception survey (four scales, reliability)

Run the 12-item survey below as an anonymous Blackboard survey after Assignment 9, then download its results as CSV. Item codes (E1–B3) must appear in the question text. Reports each scale’s mean, SD and Cronbach’s α.

7 · De-identified research dataset

For publication, analyse only students who consented, and never share Student IDs. This builds one CSV per student from sections 1, 4 and 5, with each ID replaced by a keyed hash (HMAC-SHA-256 with a secret phrase that you keep). The same phrase gives the same codes later, so data from the two exams can be added. Nothing leaves your browser.

Before publishing: obtain approval from the KAU research ethics committee, tell students in writing how course data may be used, and let them opt out without any effect on grades. See the evaluation plan.

Summary for the course report

Updated as you load files. Copy it into the course report.