Acid–Base Lab
Strong and weak acids, dynamic equilibrium, titration curves, and the difference between equivalence point and end point.
Kompas Semester gives students a weekly rhythm of practice quizzes, typed notes, guided study prompts and interactive labs. It gives their lecturer an early view of who is keeping up and who has gone quiet, while there is still time to help.
It works alongside the institution's official LMS, not instead of it.
Figures exclude test accounts. They describe use, not learning outcomes; see Evidence for what is and isn't known yet.
Kompas Semester was built by a pharmacy lecturer for his own courses, with no dedicated developer support, to tackle two problems the usual systems leave invisible until results are in.
Comparing continuous assessment results with final exam performance showed a consistent gap: students could clear coursework without the mastery that holds up under exam conditions. Weekly, topic-scoped practice tied to the same learning outcomes lets students find and close gaps continuously, not at the final exam.
One lecturer with a hundred students can't tell who has fallen behind until a test has already been failed. Attention goes to whoever asks. Kompas surfaces engagement as it happens, so outreach can reach the quiet student early.
Students see where they stand each week: points, badge, streak and a clear target for full marks this week.
They act on it with short, specific study tasks: a practice quiz, a typed note, a guided prompt, an interactive lab.
A new topic opens every week, and the habit carries the semester. Past weeks stay open for revision.
What a student does becomes points in the database. Points turn into feedback for the student and an early signal for the lecturer, and a new topic each week starts the loop again.
Students sign themselves up, pick their courses and can install the app to their home screen. No app store is needed.
Five standalone chemistry labs show what particles are doing inside a flask, and why a trend runs the way it does. Each is layered: core text at secondary-school level, with "Go deeper" and "Into the maths" for pre-university and degree level. The same lab serves a first-year student and final-exam revision.
Each lab below is recorded from the real thing. "Try the live lab" opens the actual lab right here, no sign-in needed.
Strong and weak acids, dynamic equilibrium, titration curves, and the difference between equivalence point and end point.
Acid–base, precipitation, redox and EDTA titrations, and gravimetric analysis. Students add the drops themselves.
Effective nuclear charge, atomic and ionic size, ionisation energy, electron affinity, electronegativity and polarity.
s, p and d orbitals in 3D from hydrogen to krypton, with energy levels, nodes and orbital shapes.
The lecturer's view answers that one question first, then shows how the whole course is doing.
Each panel answers a decision the lecturer actually makes:
Points are calculated in the database from the raw record of what each student did. Each behaviour counts once per topic, so repeating it or leaving the app open earns nothing.
| Behaviour, per topic | Points |
|---|---|
| Take the practice quiz | 10 |
| Quiz score, first attempt | up to 15 |
| Write a note | 10 |
| Note length (full at 150 words) | up to 15 |
| Copy a guided study prompt | 8 |
| Open the weekly study notebook | 7 |
| Topic maximum | 65 |
The largest share rewards retrieval practice, then writing in one's own words, then guided self-study. Retrieval practice and distributed practice are among the best-supported strategies in the research on how students learn.
Engaging with a topic within a week of it opening earns full points; later engagement earns 80%. The penalty is deliberately mild, so a student who has fallen behind still has every reason to catch up.
A number between 0.00 and 1.00, using each course's own maximum so every course is on the same scale. Badge tiers sit on the same scale: Wayfinder at 25%, Navigator at 55%, Voyager at 85%. The weightage is shown to students every day and can feed continuous assessment under the course's assessment plan. Official marks always stay in the institution's LMS.
Four weeks into its first full semester, Kompas Semester has early evidence of use. Evidence about learning outcomes comes at the end of the semester. Here is the honest split.
71% of enrolments have earned points. Students have taken 217 topic quizzes, written notes on 139 topics and opened the study notebooks on 248 topics. Weekly active students rose through weeks 2 and 3. Practice quizzes earn the largest share of points in every course.
Practice-quiz first attempts average 86–91%, too close to full marks to tell students apart, and preparing beforehand hasn't yet shown a difference in quiz scores. At the end of the semester each student's engagement will be compared with their test and exam marks.
The scoring rules were reviewed in September: the late-engagement rule was softened, only first attempts now count, and badge thresholds were rescaled to each course's own maximum. The Insights page now flags easy quiz banks and click-heavy point patterns, so the next round of changes can be based on evidence.
Signals about a student exist to help the lecturer reach out. Nudge emails say "we're here to help", never "you are falling behind".
A bad week can't take away a badge a student earned. Going quiet shows as a separate, temporary status, and progress is kept.
"Are you keeping up right now?" and "What have you built so far?" are different questions, so they are answered separately.
Students read a plain-language notice and tick to agree before an account exists. No institutional student ID is required.
Designed around Malaysia's Personal Data Protection Act 2010, as amended in 2024. Every table is locked so a student can only read their own rows.
Lecturer accounts are granted on the server from an allowlist. There is no signup path or switch a student could use to claim the role.
Kompas Semester is a single static HTML file and a Supabase backend. There is no framework and no build step. A lecturer can run their own copy, and a developer can read the whole app in an afternoon. The code is MIT-licensed.
Built for one lecturer: courses aren't yet scoped to an educator, so a second lecturer needs a schema change first. There is no in-app quiz editor yet; question banks are edited in the database. There is no offline mode. The repository README documents these and every other trade-off in full.
Compare each student's engagement weightage with their test and final-exam marks at the end of this semester.
Mark quiz answers on the server as well as scoring them there, ahead of any use of the weightage in formal assessment.
Rework the banks the Insights page flags as too easy, so practice scores tell students apart.
Scope courses to their lecturer, add an in-app question editor, and make setup routine for a colleague's course.
A light, competitive daily game to bring students back between topics.
Bring the bilingual treatment already used in nudge emails to the interactive labs.
Whether you lead a programme, teach a course that could use it, or want to build on it, Auf would like to hear from you.
A walkthrough of the evidence so far, the scoring rules and the data-protection approach.
Ask about setting up a copy for your own course, or running a pilot next semester.
Open an issue or pull request on GitHub, or get in touch about the roadmap above.