For Students: Understanding Course Feedback#

What This Will Show You#

This feedback will be about one specific instance of a course (spring 2025, autumn 2024, etc), not the course in general. It will tell you concrete facts about how the course actually runs: whether attendance matters, what the exam looks like, how responsive the teacher is.

The feedback will come from students who took it. You’ll be able to see what surprised them, what they wish they’d known, and how they’d describe the experience.


Can You Actually Do This Course?#

Time & Attendance#

How important is showing up?

  • Mandatory for passing — You can’t pass without attending
  • Mandatory for high grade — You can pass without it, but top grades need attendance
  • Bonus — Helps your grade, but not required
  • Optional — You can skip it without penalty

How are lectures & exercises delivered?

  • Physical only — No recordings
  • Video recordings — Sessions are recorded

What Will Wreck Your Grade?#

When do you take the exam?

  • Fixed schedule — Everyone at the same time, like a traditional exam
  • Self-booked — You pick your exam slot

What’s the exam actually like?

  • Written short-answer
  • Multiple choice
  • Oral

Can you prepare from past exams?

  • Representative — Past exams are available and actually match what you’ll see
  • Available but not useful — They exist but don’t help much
  • Not available — You can’t access them

Can you bring notes?

  • Allowed — Cheatsheet is okay
  • Not allowed — You go in cold

What else counts toward your grade?

  • Presentations — One or more presentations are part of the course
  • Project work — Includes a substantial project component

Will You Get Help?#

How responsive is the teacher?

  • Never replies — Messages typically get ignored
  • Replies in days — You usually wait several days for a response
  • Fast replies — Responses come quickly and consistently

Is the course organized?

  • Messy — Instructions unclear, multiple platforms, deadlines confusing
  • Manageable — Some rough edges, but you can figure it out
  • Well structured — Clear instructions, consistent organization

What People Are Actually Saying#

Quick Read: Tags#

These will be the patterns students mention. They will often contradict each other, and that’s intentional. A course can be “Challenging but rewarding” and “Heavy workload.”

Teaching

  • Amazing teacher
  • Bad teaching
  • Lectures worth going to
  • Lectures not worth attending
  • No one shows up to lectures

Difficulty & Workload

  • Easy 5
  • Grade grinder
  • Heavy workload
  • Brutal exam
  • Challenging but rewarding
  • Painful and pointless

Exams & Assessment

  • Exam decides everything
  • Project carries the grade
  • Grading feels random
  • Impossible to get a high grade

Value & Outcomes

  • Actually useful
  • Builds real skills
  • Career-relevant
  • Feels like busywork
  • Too theoretical to matter

Overall

  • Hidden gem
  • Overhyped
  • Would take again
  • Regret taking this
  • Avoid at all costs

The Full Stories#

You’ll be able to read what students wrote. They’ll tell you what the boxes above don’t capture: the vibe, what surprised them, what they’d do differently.


Understanding Their Perspective#

What will they share about their background?

Each person will share their background going in. This helps you understand what their experience actually means for you.

They’ll enter it as: Area: Topic → Comfort Level

Examples:

  • Math: Epsilon-delta → Comfortable → “This course was harder than expected” (from someone comfortable with abstract math)
  • Programming: Version control → No experience → “Well-paced with clear instructions” (from someone new to Git/etc)
  • Writing: Academic → Struggle → “Really insightful on essay structure” (from someone who’s struggled with papers before)

The same course feels completely different depending on where you’re starting from. If your background is similar to theirs, their experience probably applies to you.


For Maintainers: Design Rationale#

Core Principles#

  1. No vague scales

    • No agree-disagree ranges, no numeric ratings
    • Every option must be concrete and interpretable without explanation
  2. Keywords for input, stories for texture

    • Structured fields are factual and scannable
    • Tags and free-form text capture personality and edge cases
  3. Structure over mandatory verbosity

    • Structured fields carry the signal
    • Free-form text is optional but welcomed
  4. Instance-level accuracy

    • Feedback always ties to a specific course instance (term + year)
    • Not a general “course code” rating that persists across changes
  5. Low cognitive load for contributors

    • Experience-based language
    • No interpretation required to submit
  6. Personality lives in tags and stories

    • Structured fields stay factual
    • Tags are allowed to be sharp, opinionated, emotional
    • Free-form text lets storytellers shine

Structured Fields Design#

All structured inputs are:

  • Mutually exclusive where applicable (exactly one choice per question)
  • Concrete and tied to observable reality
  • Keyword-style labels only
  • Free of hidden interpretation scales

Rationale for each category:

Attendance & Delivery — Directly affects planning. Students need to know if they can pass while working, if lectures are mandatory, if they can catch up via recording.

Exams & Assessment — Determines study strategy and stress level. Knowing “past exams exist but don’t help” is actionable in a way that “exam difficulty: 7/10” is not.

Teacher Responsiveness — Captures observed behavior, not intent. A teacher who never replies can’t be defended as “busy”—the data is the experience.

Organization — Affects mental load. “Messy” vs “well structured” maps to concrete friction (multiple platforms, unclear deadlines, etc).


Tag System#

Tags are:

  • Multi-select, non-exclusive — A course can be “Challenging but rewarding” and “Heavy workload” simultaneously
  • Curated, not user-generated — Prevents dilution and maintains signal
  • Intentionally sharp — “Brutal exam” is better than “difficult exam.” “Bad teaching” is more honest than “could improve.”
  • Expected to disagree — If all tags align, you’ve lost information
  • Allowed to highlight extremes — Not every course is average, and tags should reflect that

Why this works: Students recognize themselves in sharp language. “Overhyped” tells you something “Moderately popular” never could.


Free-Form Text#

Role: Explanation, edge cases, advice, vibe.

Characteristics:

  • Optional
  • No minimum length requirement
  • Reasonable maximum (3,000–5,000 characters)
  • Markdown optional

UX behavior:

  • Empty text is fully valid
  • Very short text may trigger a non-blocking hint (“Anything else future students should know?”)
  • No hard validation based on content length

Quality signals (for ranking/moderation, non-blocking):

  • Extremely generic content
  • Obvious spam
  • Copy-paste artifacts

Design note: Free-form feedback is where the texture and honesty lives. Don’t over-police it. A well-written paragraph from one student is often more useful than perfectly-structured data from ten.


Personal Background Context#

Purpose: Interpretability, not categorization.

Design intent:

  • Intentionally flexible
  • Free-form but guided
  • No enforced taxonomy
  • Examples show typical usage, not required entries

Structure: Multiple entries allowed. Each has:

  • Area (free text) — Broad domain (math, programming, writing, etc)
  • Topic (free text) — Specific concept or skill (calculus, version control, essays, etc)
  • Comfort Level (exactly one) — Comfortable / Struggle / No experience

Example:

  • Area: Math → Topic: Calculus → Struggle
  • Area: Programming → Topic: Version control → No experience

How readers use it: “This person struggled with calculus and they’re saying the course is math-heavy, so maybe that’s relevant to me.”


Aggregation & Future Work#

Current approach:

  • Structured fields are the primary aggregation source (countable, comparable)
  • Tags provide fast qualitative signals (frequency, clustering)
  • Free-form text remains supplementary (searchable, readable)

Expected behavior:

  • Disagreement is visible and expected
  • Feedback is always instance-specific
  • No averaging or smoothing that would hide real variation

Known gaps:

  • Temporal tracking: Need clear timestamps so students know which term/year they’re reading about
  • Course changes: If a course significantly changes format, old feedback should be clearly marked as out-of-date
  • Verification: Future versions might tie feedback to actual course enrollment (anonymous but verifiable)

Explicit Non-Goals#

  • Agree–disagree scales or Likert ratings
  • Numeric ratings of any kind
  • Mandatory long text
  • Polite but meaningless feedback
  • “Instructor quality” contests
  • Categorizing students by background (using it for profiling, not interpretation)

Status#

  • Current status: Design specification, not yet implemented
  • Target release: Post–v1.0.0 of Sisukas
  • Design stance: Opinionated, concrete, expressive where it matters. Honest over diplomatic.