Quest Ladder
TeachFORGE.ai
Quest Ladder / Blanks and cloze
Blanks and cloze: fill in the blank, multiple blanks, cloze passage
Three formats that remove the options, so a learner has to supply the word or the number themselves.
Take the options away and you find out what they know.
The three formats
| Format | Shape |
|---|---|
| Fill in the blank | One gap in a sentence or an expression. |
| Fill in the blank (multiple) | Several gaps in one prompt. Number of blanks runs 2 to 8, defaulting to 3. |
| Cloze passage | Gaps spread through a longer passage, so context carries across them. Same range, same default. |
The difference between the last two is the reading. Multiple blanks is a prompt with holes; a cloze passage is a text a learner works through, where an early gap changes what the later ones can be.
Writing the gaps
A gap is marked in the content with a blank placeholder, numbered in the order it appears. The same numbering is what the correct answer step uses, so blank 1 in the text is slot 1 in the answer.
With changing values in play, a blank can be the place a generated number lands — the sentence stays the same across every version while the missing value moves. That is the most useful thing these formats do and the reason they suit drill.
Keep one idea per gap. A blank that could reasonably be filled three ways is not testing knowledge, it is testing whether the learner guessed your phrasing — and you will spend the term adding alternatives to the acceptance list.
What counts as right
Marking is set on the answer policy, and blanks are where its settings do the most work.
- Text match mode chooses between lenient — case and space insensitive, with algebraic equivalence — and strict, which is exact and case sensitive.
- Other acceptable answers, one per line, is where synonyms and alternative spellings go.
- For a numeric gap, Numeric tolerance and Tolerance type decide how close is close enough.
Each blank is a slot on the correct answer step, with its own marks. Marking mode chooses between all-or-nothing across the blanks and weighted, where each contributes; Split evenly divides the total between them in one click.
The usual pitfall
Strict matching on a word answer fails a learner for a capital letter. Lenient matching on a word where spelling is the point passes one who cannot spell it.
Decide which of those you can live with per type, not per question, and write it in the description so the next person — including you in six months — knows it was a decision rather than a default.