How scoring actually works
- Item scoring: Reading answers are scored dichotomously → right or wrong. A blank counts as wrong. All Reading scoring is done by computer.
- Points & totals: There are 38 scored items, 1 point per correct. Some test forms include unscored items (for test development) that look identical and can appear anywhere—treat every question seriously. Pack.pdf)
- Equating (fairness across forms): CELPIP applies score equating so small difficulty differences across forms don’t advantage or disadvantage anyone. This is why raw-to-level bands are approximate and may slightly overlap.
Approximate raw-score bands → CELPIP Levels (Reading):
| CELPIP Level | Correct answers (out of 38) |
|---|---|
| 10–12 | 33–38 |
| 9 | 31–33 |
| 8 | 28–31 |
| 7 | 24–28 |
| 6 | 19–25 |
| 5 | 15–20 |
| 4 | 10–16 |
| 3 | 8–11 |
| M | 0–7 |
These bands are published by CELPIP for Reading as an approximate guide; final levels reflect equating.
Time and items by part (what the clock expects)
These are the official suggested per-part timings that align with the overall window. You’ll adopt and fine-tune them later in the course.
- Practice Task (unscored): ~1 min
- Part 1 — Reading Correspondence: 11 items, ~11 min
- Part 2 — Reading to Apply a Diagram: 8 items, ~9 min
- Part 3 — Reading for Information: 9 items, ~10 min
- Part 4 — Reading for Viewpoints: 10 items, ~13 min
What each part includes (clear previews you’ll build on next)
Part 1 — Reading Correspondence (11 items)
A short, practical email/message followed by a brief reply with blanks. Your job is to identify the sender, recipient, purpose, and any constraints (times, costs, availability), then complete the reply so it fits the original message in tone and facts.
- Traps to expect: polite paraphrases that hide the same meaning (e.g., “at your earliest convenience” vs. “as soon as possible”), softening words (“might”, “perhaps”), and small numbers (budget, time windows) that must be preserved.
- Where we go next: In Part 4 of this course, you’ll learn a 30-second who/why/tone map, then practice reply wording that remains accurate, polite, and natural under time.
Part 2 — Reading to Apply a Diagram (8 items)
A visual (schedule, brochure, map, guide) plus a short email with blanks. Start with the title, then the legend/notes, then trace how information is organized (rows/columns/sections). Match the question’s constraints precisely (e.g., “at least 3 lanes, after 7 pm, weekdays only”).
- Traps to expect: near-synonyms in options, label swaps in the diagram, tiny footnotes that change availability (“except holidays”, “reduced services 12–2”).
- Where we go next: In Part 5, you’ll get a reliable routine: Title → Legend/Notes → Locate → Check constraints → Choose, with drills against the most common misreads.
Part 3 — Reading for Information (9 items)
A four-paragraph informational text (process, place, biography, event). Use each paragraph’s main idea as a map. Jump to the likely paragraph, then verify the exact wording before selecting.
- Traps to expect: the classic “Not stated” misread (it’s only correct when the statement has no support anywhere in the passage), plus timeline distortions and cause/effect reversals.
- Where we go next: In Part 6, you’ll use a 4-step locate-and-verify routine to protect accuracy without over-reading: Find paragraph → Scan with keywords/paraphrases → Read target lines carefully → Confirm match (or confirm not stated).
Part 4 — Reading for Viewpoints (10 items)
An opinion article with multiple viewpoints, followed by a reader comment with blanks. Build a quick opinions map: who thinks what, and why? Track contrast markers (however, whereas, on the other hand) and hedges (may, could, tends to).
- Traps to expect: options that keep the topic but bend the stance, or that over-generalize a specific claim. Tone words (“cautiously optimistic”, “dismissive”) matter.
- Where we go next: In Part 7, you’ll drill fast stance-checks and master eliminating look-alike answers that don’t fully match the author’s or commenter’s position.
Interface & test-day rules that actually matter
- Same-screen layout: For each Reading part, the text and questions live together; you scroll.
- Locking: You may change answers within the current part; when its timer ends, it locks and you move on automatically.
- Test centre vs. at-home: At centres, you receive paper + pen for notes (collected afterward). At-home testing uses an on-screen scratchpad only.
- Practical tip: Because you can’t highlight on-screen, make quick pencil marks (centre) or scratchpad notes (online) for names, numbers, and constraints—the items that don’t paraphrase easily.