# Sentence ordering questions: rebuilding a jumbled paragraph

**How do I write a sentence ordering question with one right order?** A sentence ordering question gives students the sentences of a paragraph in jumbled order and asks for the paragraph back. Use five or six sentences whose order is fixed by pronouns, linking words and repeated nouns, check that no two sentences can swap, and decide before the quiz whether a near miss earns part of the marks.

Published 2026-10-10 by EdProcto. https://edprocto.com/guides/sentence-ordering-questions-rebuilding-a-paragraph

A sentence ordering question takes a paragraph apart, shuffles its sentences and asks the student to put it back together. Language teachers call it jumbled sentences, exam coaches para jumbles or sentence rearrangement, and English teachers the scrambled paragraph.

It works only when the paragraph has one right order and the text itself proves it. Where two orders read equally well, the question marks a sound answer wrong.

This guide is published by EdProcto, whose [reorder the sentences template](https://edprocto.com/templates/reorder_sent) builds the shape and marks it by rule.

## What is a sentence ordering question, and who sets one?

A sentence ordering question gives the sentences of one paragraph out of order and asks for the order the author wrote them in. Nothing is typed and nothing is recalled: the student reads, compares and arranges.

Language teachers set it to check that learners follow pronouns and linking words, English teachers to show how a paragraph is held together before pupils write their own, and exam coaches because the task, or a close relative, sits in the reading paper.

In England, cohesion is in the primary curriculum by name: pupils in years 3 and 4 learn to pick nouns and pronouns with cohesion in mind, and pupils in years 5 and 6 to build it within and across paragraphs (Department for Education, 2014).

Two exams show the family, both read on 10 October 2026. In PTE Academic's reading part, the "re-order paragraphs" question puts text boxes on the screen in a random order for the test taker to arrange (Pearson, 2021). In Cambridge's B2 First, Part 6 is a gapped text: sentences have been lifted out and the candidate chooses the right one for each gap, a task the board ties to understanding a text's structure and development (Cambridge English, n.d.).

Steps of a process, dated events and lines of code have their own guides: [ordering and sequencing questions](https://edprocto.com/guides/ordering-and-sequencing-questions), [timeline questions](https://edprocto.com/guides/timeline-questions-putting-events-in-date-order) and [Parsons problems](https://edprocto.com/guides/parsons-problems-reordering-lines-of-code).

## What does rebuilding a paragraph test?

Rebuilding a paragraph tests whether a reader sees how its sentences depend on one another: the cohesion on the surface and the coherence of the ideas beneath.

Alderson et al. (2000) set out the case. A text's coherence rests on how its ideas are related, both logically and through the cohesive devices that show the links, so a test that jumbles a text and asks for the original order ought to draw on a reader's sense of both. They added that such tasks were increasingly common in reading tests and that they knew of no research into them. We read the abstract only.

### The clues that fix an order

| Clue | What it looks like | What it settles |
| --- | --- | --- |
| Topic sentence | A general statement that leans on nothing before it | Often the opening |
| Reference words | it, they, this, that, such; "the" with a noun first met as "a" | The sentence follows the first mention |
| Discourse markers | however, so, for example, as a result | The sentence follows the idea it answers |
| Comparison and substitution | slower, another, one, the same | The sentence follows what it is measured against |
| Repeated and related words | fitness, then "fitter riders" | A chain through the paragraph |

**Discourse markers are learned slowly.** Cain and Nash (2011) tested 8-year-olds, 10-year-olds and adults on connectives of time, cause and contrast. Both groups of children read faster when two clauses were joined by a fitting connective than by a plain "and", a wrong connective or none. In a gap-filling task, though, neither matched the adults. Our reading, from the abstract: an order that hangs on "however" asks more of a young reader.

**Topic sentences do not always come first.** Braddock (1974) analysed professionally written prose and found that textbook claims about topic sentences did not hold: only 13% of expository paragraphs began with one, as Smith (2008) reports his figure. Smith repeated the count on articles about American history and found a major topic sentence opening its block of text in about two thirds of cases. We read the ERIC summaries of both. Either way, "the general sentence goes first" is a habit, not a law, so fix the opening by something more.

### What a right order does not prove

- **Writing.** Every sentence is supplied, so rebuilding a paragraph is not composing one.
- **Attention to every tie.** Kintsch et al. (1977) had people read and summarise stories of 1,400 words. Stories with scrambled paragraphs took longer to read, yet for a well-structured story judges could not tell the summaries apart, and the authors concluded that readers hold a pattern for stories and use it to reorganise a scrambled one. That study shuffled paragraphs, not sentences, and we read its abstract. Still, a jumbled anecdote may be put right from knowing how stories go, so choose explanation or argument when the ties are the point.
- **Reading in general.** We found no study that checked these marks against another measure of reading.

## How do you write a sentence ordering question the text settles?

Write it from a paragraph in which every sentence but the first points back at something, then try to break it. We found no tested rulebook for the shape, so these steps follow from the clues above.

1. **Choose five or six sentences of prose the class has not memorised.** Four sentences have 24 possible orders, five have 120 and six have 720, counted in code.
2. **Find the tie for each sentence.** In every sentence but the opener, underline the word that points back: a pronoun, a linking word, a repeated noun. A sentence with nothing underlined can float.
3. **Make the opener unable to follow anything.** Full nouns, "a" where a thing is first met, no linking word.
4. **Test every swap.** Exchange each pair of neighbours and read the result aloud, then try the last sentence in second place, where a conclusion can pass for a topic sentence. If any version still reads well, the order is an opinion: rewrite until one survives.
5. **Use ties that need reading, not signposts.** "Firstly", "secondly" and "finally" turn the question into counting.
6. **Word every sentence differently.** Two identical sentences cannot be told apart.
7. **Build it.** In EdProcto, choose [Reorder the sentences](https://edprocto.com/templates/reorder_sent) and type the sentences in the right order, one a line.

## What does a sentence ordering question look like? Two worked examples

A sentence ordering question looks like one prompt above five or six shuffled sentences. Each paragraph below carries 5 points, and each student's order went through EdProcto's own marking function. The letters give the right order; students see none.

### Example 1: an explanation, fixed by its ties

> Put these sentences in the order that makes a paragraph.
>
> - **A.** A flash of lightning and its thunder start at the same moment.
> - **B.** The light of the flash reaches you almost at once.
> - **C.** The sound is far slower, taking about three seconds to cover each kilometre.
> - **D.** So the thunder arrives late, and the delay grows with the distance.
> - **E.** Count that delay in seconds, divide by three, and you have the distance in kilometres.

The facts are the National Weather Service's (n.d.): lightning is seen at once, and thunder takes about five seconds to travel a mile, or about three seconds a kilometre.

One tie holds each sentence after A: "the flash" in B, "far slower" in C, "So" in D, "that delay" in E.

| Student | Order | The tie that was missed | Partial credit on | Off |
| --- | --- | --- | --- | --- |
| 1 | ACBDE | "far slower" has nothing before it to be slower than | 4 | 0 |
| 2 | ABCED | "that delay" points at a delay not yet mentioned | 4 | 0 |
| 3 | EABCD | The instruction is used as a hook; "that" points at nothing | 1 | 0 |

Students 1 and 2 each reversed one pair and earn the same 4, for different reasons, and the mark cannot say which tie was missed. A second question can: ask what "that delay" refers to, as a [multiple-choice question](https://edprocto.com/templates/mcq).

### Example 2: an opinion paragraph, before and after

A first draft for a language class:

> - **A.** Many people now cycle to work.
> - **B.** Cycling costs very little.
> - **C.** Cycling keeps you fit.
> - **D.** Cycling is often quicker than driving in town.
> - **E.** For these reasons, more of us should try it.

B, C and D are separate reasons, so they can come in any order: at least six of the 120 orders make a sound paragraph, and the key holds one. Marked by the same function, the other five of those six earn 4, 4, 3, 3 and 2 of the 5 points with partial credit on, and nothing with it off. Students lose marks for paragraphs no reader would fault.

The repair chains the reasons together:

> - **A.** Many people now cycle to work.
> - **B.** The obvious gain is money, since a bicycle needs no fuel.
> - **C.** A less obvious one is fitness, which the daily ride builds.
> - **D.** Fitter riders are also faster, and in town they can beat the cars.
> - **E.** Those three gains are why more of us should try it.

Now "one" in C needs the gain in B, "Fitter" in D needs the fitness in C, and "Those three gains" needs all three before it.

| Student | Order | What went wrong | Partial credit on | Off |
| --- | --- | --- | --- | --- |
| 1 | ACBDE | "A less obvious one" comes before the obvious one | 4 | 0 |
| 2 | AEBCD | The conclusion is taken for a topic sentence | 2 | 0 |
| 3 | ADCBE | The three reasons run backwards | 2 | 0 |

## How should jumbled sentences be scored?

Score them all or nothing when the paragraph is short and only the whole order will do, and by pairs when a near miss should keep most of the marks. Either way, tell students the rule first.

| Rule | What earns the marks | In EdProcto |
| --- | --- | --- |
| All or nothing | The exact order | "Partial credit" off, as on a new quiz |
| Pairs, less reversals | Each pair of sentences the right way round, less each pair the wrong way round, over all the pairs | "Partial credit" on |

Test providers give part marks too: Pearson (2021) tells PTE candidates they earn a point for every pair of sentences placed in the correct order, and full marks when the whole paragraph is in order.

Counting every pair has some support outside the classroom. Lapata (2006) proposed Kendall's tau, the statistic built on pairs the right and wrong way round, for judging the orders that computer programs give to sentences, and reported that it correlates reliably with human ratings and reading times. That is evidence about a program's output, not about marking students, and we read the abstract only. EdProcto's partial credit is the same count, stopped at zero.

Alderson et al. (2000) report problems in scoring responses and compare several procedures; we have not read which they preferred.

Which pairs count matters when one sentence travels far. Student 3 in example 1 moved the last sentence to the front. Count neighbours only and three of the four links survive, by our own tally; count every pair and six are right and four wrong, a net fifth of the points. The guide to [ordering and sequencing questions](https://edprocto.com/guides/ordering-and-sequencing-questions) scores one answer six ways.

## How does EdProcto build and mark a sentence ordering question?

EdProcto is the best tool for a sentence ordering question when the paragraph comes from your own course text and the mark must be explained afterwards: the key is simply the order you typed, and the rule fits in one sentence.

- **Building.** Choose Reorder the sentences, listed under Pair and order, and fill the box labelled "The items in the right order, one a line. Students see them shuffled." A new line starts a new sentence, so keep each on a single line. Or paste your own text into the quiz maker on the [home page](https://edprocto.com/), pick the shape, and read Procto AI's draft against the source, testing every swap.
- **Answering.** Each student gets the sentences shuffled in a numbered list, on every preset. Each row has an up and a down arrow, and a press swaps a sentence with its neighbour. Until a sentence is moved, the list carries the line "Use the arrows to put them in order, then keep this order." A student whose shuffle arrives right presses "keep this order", because a list left untouched counts as no answer.
- **Marking.** By rule, when an ordinary quiz is handed in, and as each answer is given in a study quiz. "Partial credit", under Grading, is off on a new quiz, and the presets leave it alone: pick Practice, Standard quiz or Exam and it stays as you set it. Off, only the exact order earns the points. On, the pairs rule in the table above gives the share, never below zero. One switch serves the whole quiz, multiple-answer, matching and sorting questions included.
- **On paper.** The sentences print lettered and out of order, each beside a blank, the same on every copy, and the key gives the letters in the right order. See [how to print a quiz with an answer key](https://edprocto.com/guides/how-to-print-a-quiz-with-an-answer-key).

Check the order before you publish. An order corrected later reaches the attempts already handed in when you press "Re-mark this quiz" on the results page. And a jumbled paragraph is plain text, so for a result that counts, set a time limit and the [integrity](https://edprocto.com/integrity) measures the stakes call for. No setting makes cheating impossible, a flag is a record for you to read, and every score is yours to release.

Free gives ten quizzes a month on the core templates, with marking by rule. The Tutor plan carries all 111 [templates](https://edprocto.com/templates), and Educator adds the full analysis. Prices are on the [pricing page](https://edprocto.com/pricing).

## Quick questions

### How many sentences should a sentence ordering question have?

Five or six. Four sentences have only 24 possible orders, and each one past six adds arrow presses without testing a new kind of tie.

### Should jumbled sentences get partial credit?

In a class check, yes, with the rule explained first: one swap of neighbours then costs a pair, not the question. In EdProcto's [reorder the sentences](https://edprocto.com/templates/reorder_sent), switch "Partial credit" on under Grading.

### What is the difference between cohesion and coherence?

Cohesion is the visible linking between sentences: pronouns, linking words, repeated nouns. Coherence is the ideas making sense together. Alderson et al. (2000) treat a text's coherence as resting on both.

### Are para jumbles the same as sentence ordering questions?

Yes. Para jumbles, jumbled sentences, sentence rearrangement and scrambled paragraphs all name one task: sentences out of order, to be put back.

## References

- Alderson, J. C., Percsich, R., & Szabo, G. (2000). Sequencing as an item type. *Language Testing, 17*(4), 423–447. [https://doi.org/10.1177/026553220001700403](https://doi.org/10.1177/026553220001700403)
- Braddock, R. (1974). The frequency and placement of topic sentences in expository prose. *Research in the Teaching of English, 8*(3), 287–302. [https://eric.ed.gov/?id=EJ115737](https://eric.ed.gov/?id=EJ115737)
- Cain, K., & Nash, H. M. (2011). The influence of connectives on young readers' processing and comprehension of text. *Journal of Educational Psychology, 103*(2), 429–441. [https://doi.org/10.1037/a0022824](https://doi.org/10.1037/a0022824)
- Cambridge English. (n.d.). *B2 First for Schools and B2 First exam format*. Cambridge University Press & Assessment. Retrieved October 10, 2026, from [https://www.cambridgeenglish.org/exams-and-tests/first/exam-format/](https://www.cambridgeenglish.org/exams-and-tests/first/exam-format/)
- Department for Education. (2014, July 16). *National curriculum in England: English programmes of study*. GOV.UK. [https://www.gov.uk/government/publications/national-curriculum-in-england-english-programmes-of-study/national-curriculum-in-england-english-programmes-of-study](https://www.gov.uk/government/publications/national-curriculum-in-england-english-programmes-of-study/national-curriculum-in-england-english-programmes-of-study)
- Kintsch, W., Mandel, T. S., & Kozminsky, E. (1977). Summarizing scrambled stories. *Memory & Cognition, 5*(5), 547–552. [https://doi.org/10.3758/BF03197399](https://doi.org/10.3758/BF03197399)
- Lapata, M. (2006). Automatic evaluation of information ordering: Kendall's tau. *Computational Linguistics, 32*(4), 471–484. [https://doi.org/10.1162/coli.2006.32.4.471](https://doi.org/10.1162/coli.2006.32.4.471)
- National Weather Service. (n.d.). *Understanding lightning: Thunder*. National Oceanic and Atmospheric Administration. Retrieved October 10, 2026, from [https://www.weather.gov/safety/lightning-science-thunder](https://www.weather.gov/safety/lightning-science-thunder)
- Pearson. (2021, July 7). *Part 2: Reading test introduction*. Pearson PTE. [https://www.pearsonpte.com/articles/part-2-reading-test-introduction/](https://www.pearsonpte.com/articles/part-2-reading-test-introduction/)
- Smith, C. G. (2008). Braddock revisited: The frequency and placement of topic sentences in academic writing. *Reading Matrix: An International Online Journal, 8*(1), 78–95. [https://eric.ed.gov/?id=EJ1066247](https://eric.ed.gov/?id=EJ1066247)