Matching Quiz Maker: Build Match-the-Column Items From Your Pairs
A matching quiz maker turns paired items into a match-the-column question: students connect each item in one column to its correct partner in the other — a term to its definition, a country to its capital, a date to its event. It is one of the fastest ways to check factual recall at scale, and it is distinct from multiple choice (one-to-many) and from fill-in-the-blank (recall a word). This guide shows how to build matching items from pairs you curated, not from an AI that assembles pairs you never taught.
Common matching pairs
| Left column | Right column |
|---|---|
| Term | Definition |
| Country | Capital |
| Date | Event |
| Symbol | Element name |
| Author | Work |
| Word | Synonym |
The fixed-column pair sheet
Two columns, one correct pair per row:
| left | right |
|---|---|
| Photosynthesis | Conversion of light to chemical energy |
| Mitosis | Cell division into two identical cells |
| Osmosis | Water moving across a membrane |
The converter sets the left items as prompts and the right items as a shuffled response pool, grading by correct pairing. If you want several small sets in one file, add an id column so pairs stay grouped.
From pairs to a matching question
List the pairs
Put each correct pair in one row (left + right). These are the associations you actually taught.
Aim for 4–8 pairs
Fewer is trivial; more than eight makes the grid unwieldy. Split large lists into several matching questions.
Convert
Emit a matching item in your platform's format. Where native matching isn't supported, deliver as MCQ (each left item asks 'which right item matches?'). QTI 2.1 carries matching natively.
Shuffle the response pool
Randomize the right column order so the answer isn't always '1→A, 2→B'. Most platforms do this automatically at delivery.
Assign and review
Push to Canvas, Moodle, Blackboard, or Google Forms, then review which pairs students most often mismatched — that is your re-teach list.
Matching vs the other item types
A matching item is one-to-one association; multiple choice is one-to-many recognition; fill-in-the-blank is free recall. Use matching when you have clean pairs and want fast factual checks. It grades automatically and is far quicker to answer than writing definitions by hand.
Why your own pairs matter
A matching question is only correct if every pair is correct. Letting a model assemble pairs risks planting a wrong association that students then memorize. Keep the pairs authoritative and use the tool only to format, shuffle, and grade. More guides at examparser.com/blog/en.
Frequently asked questions
What is a matching quiz maker?
A matching quiz maker turns paired items into a match-the-column question: students connect each item in one column to its correct partner in the other (term to definition, country to capital, date to event). The safe version pairs items you curated; the risky version lets AI assemble pairs, which can mismatch things you never taught.
What are common matching pairs?
Term ↔ definition, country ↔ capital, date ↔ event, symbol ↔ element name, author ↔ work, word ↔ synonym. Any two related columns of equal length make a valid matching set.
How is a matching item stored in a fixed-column sheet?
Two columns: left and right. Each row is one correct pair. The converter sets the left items as prompts and the right items as the response pool (shuffled), grading by correct pairing. Add an 'id' column only if you want to keep pairs grouped across multiple sets.
Does every LMS support matching questions?
Most do, but not all. Canvas, Moodle, and Blackboard have native matching item types. Where a platform lacks it, the same pairs can be delivered as multiple choice (each left item asks 'which right item matches?'). QTI 2.1 carries matching natively, which helps on platforms that import it.
How many pairs per question is reasonable?
Four to eight pairs works best. Fewer is trivial; more than eight makes the two columns unwieldy and raises the chance a student guesses by elimination. Split large lists into several matching questions rather than one giant grid.
Build a matching quiz from your pairs
Stop letting a model pair things you never taught. Upload the pairs where your associations already live, and turn them into a match-the-column quiz on any LMS — with your pairings intact.
Try ExamParser →