Chapter 83: Grammar Testing and Assessment

ADVANCED ENGLISH GRAMMAR
PART 9: PEDAGOGY AND LEARNING
ADVANCED · 33 MINUTE READ
Grammar Testing and Assessment
A study of how grammar is tested and assessed: the purposes and types of grammar tests (discrete-point and integrative), multiple-choice items and how to write them, gap-fill and cloze tests, error correction tests, communicative testing of grammar in context, rubrics for scoring (analytic and holistic), self-assessment and can-do statements, the qualities of a good test (validity, reliability, authenticity, washback), item analysis, test design and blueprints, interpreting results, and common misconceptions.

Grammar testing is the measurement of what learners know and can do with grammar. A teacher gives a quiz to see whether a class has understood the past perfect; an examiner sets a section on grammar and usage; an institution places new students by a grammar test; a learner takes a practice test to find weak points. Each purpose calls for a different kind of test, and the choice matters, because a test that asks the wrong question gives a misleading answer: a learner who can choose the correct option in a multiple-choice item may still make the error in a letter, and a learner who writes fluently may fail a test of rules.

This chapter explains how to design, use, and interpret grammar assessment. You will study the types of tests (discrete-point and integrative), multiple-choice questions, gap-fill and cloze exercises, error correction tests, communicative testing in context, rubrics for scoring, and self-assessment. You will see the qualities of a good test (validity, reliability, authenticity, and washback), how items are analyzed, and how a test is planned. The emphasis is on testing grammar in meaningful context, and on matching the form of testing grammar to the purpose of the grammar evaluation, whether the grammar exams are classroom quizzes or high-stakes tests. Thirty exercises and a quiz follow.

Infographic. Grammar testing and assessment. Match the test to the purpose. It shows: Purposes ; placement · diagnostic · progress · achievement · proficiency ; Discrete-point formats ; multiple choice · gap-fill · error correction ; Quality ; validity · reliability · authenticity · washback ; In context ; test grammar in meaningful context.
Grammar tests differ in purpose and format and are judged by validity, reliability, authenticity, and washback. Tap or click to see it full size.
✦ In plain words

A test is only as good as the question it answers. Before you write a test, ask what you want to know: whether learners can recognize a correct form, produce it, find and fix errors, or use grammar in a real task.

Why it matters: A good test is valid (it measures what it claims to), reliable (it gives consistent results), practical (you can give and mark it), and fair. It should also have a helpful effect on teaching and learning.

★ In brief

Summary: Grammar tests have different purposes (placement, diagnostic, progress, achievement, proficiency) and formats. Discrete-point tests isolate one element at a time (multiple choice, gap-fill, error correction); integrative tests require several elements together (cloze, dictation, writing tasks); communicative tests ask learners to perform realistic tasks and are scored with rubrics. Multiple-choice items need a clear stem, one key, and plausible distractors drawn from real errors. Gap-fill tests production; error correction tests editing. Rubrics define criteria (accuracy, range, appropriateness) and levels. Self-assessment uses can-do statements and portfolios. The qualities of a test are validity, reliability, practicality, authenticity, and washback. Item analysis uses facility and discrimination. A test blueprint matches items to objectives. Grammar knowledge shown in a test does not guarantee use in communication, so tests should be balanced and interpreted with care.

  • Match the test to the purpose.
  • Formats: multiple choice, gap-fill, error correction, communicative task.
  • Quality: validity, reliability, authenticity, washback.
  • Test grammar in meaningful context.

About 33 minutes to read. Rules and examples last reviewed October 29, 2026.

AdvancedIntermediate readers should start with Intermediate Grammar Review Part 1, and beginners with Basic Grammar Review. This chapter assumes that you know common grammar areas and are interested in how they can be assessed.

1. Grammar Testing: Types of Grammar Tests

💡 Memory Trick: Test grammar in meaningful context!
Purpose of a test Question it answers Example
placement Which level should the learner join? a short grammar and vocabulary test at enrollment
diagnostic What are the learner’s strengths and weaknesses? an error analysis test; a profile by area
progress (formative) Is the learner learning what is being taught? a weekly quiz
achievement (summative) Has the learner mastered the course content? an end-of-term exam
proficiency What can the learner do in the language, regardless of a course? a standardized language exam
aptitude Can the learner learn a language? —
selection or certification Does the candidate meet a standard? a university entrance test
Type Description Examples
discrete-point tests one element at a time (a tense, an article, a preposition) multiple choice; gap-fill; error identification
integrative requires several elements and skills at once cloze; dictation; writing a paragraph; oral interview
selected-response the test taker chooses an answer multiple choice; true/false; matching
constructed-response the test taker writes or says an answer gap-fill; short answer; essay
direct the test taker performs the skill being tested writing an email
indirect the test measures an element that supports the skill a multiple-choice grammar test
norm-referenced scores are compared with other test takers a ranking
criterion-referenced scores are compared with a standard or objectives “can use the past perfect correctly”
formative used to guide learning a quiz with feedback
summative used to judge achievement a final exam
timed (implicit knowledge) speed prevents conscious rule use a timed grammaticality judgment task
untimed (explicit knowledge) allows reflection a rule explanation question
computer-adaptive items change according to responses some online placement tests
performance-based a realistic task scored with a rubric a role play
Format (a selection) What it tests Strength Limit
multiple choice recognition of a correct form quick and objective to mark guessing; tests recognition more than production
gap-fill (single word) production of a form in context tests production; easy to mark multiple acceptable answers
cloze grammar and vocabulary in a text integrative; context marking issues; difficulty of design
word formation derivation and form focused on morphology —
sentence transformation (key word) rephrasing with a given structure tests flexibility may test the trick more than grammar
error identification (underlined parts) recognition of an error quick only recognition
error correction (rewriting) finding and fixing tests editing marking; several fixes
sentence completion production in a frame — —
matching association simple —
true/false (grammaticality judgment) intuition of acceptability tests implicit knowledge if timed guessing
ordering words into a sentence syntax — one answer only if constrained
translation L1 to L2 or reverse — not authentic
dictation listening and form integrative marking
short answer production — —
paragraph writing all grammar in use direct and authentic needs rubric; subjective
email or letter writing task grammar for purpose authentic needs rubric
oral interview speaking grammar authentic time, rater training
picture description spoken grammar tests tenses and prepositions of place —
role play functional grammar communicative needs rubric
dialogue completion pragmatic grammar context —
summary writing complex structures integrative —
portfolio accumulated work shows growth time to mark
self-assessment checklist learner’s perception autonomy accuracy varies
peer assessment feedback by classmates social learning training needed
Quality (Bachman and Palmer, 1996: test usefulness) Meaning Question
validity the test measures what it is meant to measure (construct validity) Does it test grammar knowledge or reading skill?
reliability consistency of results across occasions and raters Would the same learner get the same score?
authenticity the tasks resemble real use Would a learner meet this task outside the test?
interactiveness the task engages the learner’s abilities —
impact (washback) the effect on teaching and learning Does it encourage the right kind of study?
practicality resources of time, money, and effort are adequate Can we give and mark it?
fairness no group is disadvantaged Is the language free of cultural bias?
security and ethics results are protected; the test is used properly —

The best tests combine formats, because each format has strengths and limits. A multiple-choice test is efficient, but a writing task is more authentic; a combination gives a fuller picture. The relation of testing to teaching is also important: tests should reflect what has been taught, and what has been taught should reflect the goals of the learners, as the chapter on grammar teaching methods explains.

2. Multiple Choice Questions: Strengths, Limitations, and How to Write Them

Part of an item Meaning Example
stem the question or the incomplete sentence She ___ here since 2019.
key the correct answer has lived
distractors plausible wrong options lived; is living; has been lived
options the key and the distractors (a) lived (b) has lived (c) is living (d) was living
Strength Limitation
quick, objective, and cheap to mark tests recognition, not production
can sample many areas in a short time guessing can give a score (25 percent chance with four options)
reliable scoring writing good items is difficult
suits large groups and computer marking does not reveal how the learner would use the form
permits item analysis may favor test-wise learners
can test grammar in context (a short text) can encourage rote learning of tricks
Guideline for writing items Poor Better
one correct answer She (a) go (b) goes (c) went — ambiguous without context Yesterday she ___ to school. (a) go (b) goes (c) went (d) going
put most of the wording in the stem Which of these is correct? (a) … She ___ her homework every day.
plausible distractors from common errors (a) goed (b) went (c) banana (d) table (a) goed (b) went (c) gone (d) going
options similar in length and form (a) went (b) has been going on for years (c) go (d) gone (a) went (b) has gone (c) goes (d) going
avoid negatives in the stem Which is NOT incorrect? Which sentence is correct?
avoid “all of the above” and “none of the above” (d) none of the above four distinct options
avoid clues (grammar of the stem matches only the key) an ___ (key begins with a vowel) use “a/an” in the options
test one thing at a time a stem that tests tense and vocabulary together focus on the tense
place the key randomly always (b) vary
avoid trick items trivial exceptions test important structures
use context isolated sentence with no time cue a short text with a time cue
avoid overlapping options (a) before 5 (b) before 3 clear ranges
keep the item independent item 4 depends on item 3 each item stands alone
check for bias a context unfamiliar to some groups a neutral context
edit for clarity a long, complex stem a short stem
Item (original) Target Key Distractor reasoning
By the time we arrived, the film ___. past perfect (earlier event) had started started (simple past), has started (present perfect), was starting (progressive)
If I ___ more time, I would learn Spanish. second conditional had have (first conditional form), would have (a common error), will have
She is ___ than her sister. comparative taller tall, more tall, tallest
I look forward ___ you. verb + to + gerund to seeing to see, for seeing, seeing
Neither the teacher nor the students ___ ready. agreement with the nearer subject were was, is, be
She asked me where ___. indirect question order I lived did I live, I did live, lived I
He suggested that she ___ a doctor. mandative subjunctive see sees, saw, would see
There ___ a lot of people at the party. agreement were was, is, has been (a lot of people is plural)
Please do not hesitate to contact ___. pronoun case me I, myself, my
The report, along with the charts, ___ ready. agreement (intervening phrase) is are, were, have been
Item analysis Meaning Interpretation
facility value (item difficulty) the proportion of test takers who answered correctly 0.30 to 0.80 is usually acceptable; below 0.20 is very hard; above 0.90 is very easy
discrimination index how well the item separates high scorers from low scorers (the proportion correct in the top group minus the bottom group) above 0.30 is good; near zero or negative means a poor item
distractor analysis how many test takers choose each distractor a distractor that nobody chooses is useless; one chosen more than the key signals a problem
example (invented) of 100 test takers, 60 answer correctly; 90 percent of the top group and 30 percent of the bottom group are correct facility = 0.60; discrimination = 0.90 − 0.30 = 0.60 (a good item)
reliability of the test as a whole an estimate of consistency (for example Cronbach’s alpha) about 0.80 or more is commonly sought for important tests

Multiple-choice grammar items are used in many standardized tests, including the conventions-of-English sections of some admission tests. The test taker must recognize standard conventions such as agreement, punctuation, and parallel structure, which is why the chapters on these topics are useful for test preparation. The use of corpora to base items on real errors and real usage is described in the chapter on corpus linguistics.

3. Gap-Fill Exercises: Testing Production

Type Description Example
single-word gap-fill one word completes the sentence She ___ (go) to school every day.
gap-fill with a given word to change the word in brackets is put into the correct form She ___ (not/see) him yesterday.
open cloze a text with words removed at intervals or by selection; no options Yesterday I ___ to the market and bought ___ apple.
rational (selective) cloze words removed deliberately to test certain items articles, prepositions, auxiliaries
fixed-ratio (pseudo-random) cloze every nth word is removed (often every 7th) —
multiple-choice cloze a text with gaps and options —
word formation a root word is changed to fit the gap She is a very ___ (care) driver.
key word transformation a second sentence is completed using a given word “I haven’t seen him for years.” → LAST The ___ him was years ago.
C-test the second half of every second word is deleted Yes_______ I w__ to t__ market (partial words)
sentence completion a sentence frame is completed If I won the lottery, I would ___.
dialogue completion a missing line is completed A: ___? B: Yes, I have.
table or form completion information is entered in grammatical form —
ordering words are put in order —
dictogloss (as a test) listen and reconstruct —
Gap-fill item (original) Target Acceptable answers
If I ___ (be) you, I would go. subjunctive in a conditional were (was in informal use, if the test allows)
She has lived here ___ 2015. since since
I look forward ___ (see) you. gerund after to to seeing
He asked me ___ I had finished. indirect question word if / whether
They ___ (build) the bridge in 1990. passive past built (active) — the item must specify the voice
We ___ (wait) for an hour when the bus finally came. past perfect progressive had been waiting
___ the rain, we went out. concessive preposition Despite / In spite of
She is the person ___ called me. relative pronoun who / that
I wish I ___ (can) swim. wish + past could
The more you practice, ___ you improve. correlative comparative the more / the better (item must specify)
Guideline for gap-fill Why
provide enough context to make the answer clear otherwise several answers are possible
use a text, not isolated sentences, where possible context is part of grammar
test one point per gap clear scoring
predict acceptable alternatives and put them in the key fair scoring
decide whether to score exact word or acceptable word exact-word scoring is stricter
decide how to treat spelling errors a grammar item may ignore spelling
avoid gaps that depend on knowing a rare word —
include a variety of areas —
mark the gaps with the same length no clues

Gap-fill and cloze tests are widely used, because they require learners to produce forms in context. The cloze procedure was developed in the 1950s and was studied as a measure of reading and of overall language ability; whether it tests grammar, vocabulary, or reading comprehension depends on the deletions and the text. A rational cloze, with deliberate deletions, gives more control over what is tested.

4. Error Correction Tests: Finding and Fixing Errors

Format Description Example
error identification a sentence has four underlined parts; one contains an error Neither of the students have finished their work yet.
error identification (no underlines) find the error in a sentence He go to school every day.
error correction (line by line) a text with one error per line (or none) Line 3: I am agree with him. → I agree with him.
error correction (editing a text) a short text with a set number of errors find and correct 10 errors
proofreading task correct a paragraph —
correct or incorrect decide whether a sentence is correct —
rewrite the sentence correctly supply the corrected sentence —
spot the extra word a text with an extra word in some lines The house was very big and and old.
error classification label the type of error article / tense / preposition
explain the error a metalinguistic task why is informations wrong?
Error correction item (original) Error Correction
She don’t like tea. agreement She doesn’t like tea.
I have 18 years. transfer I am 18 years old.
He suggested me to go. verb pattern He suggested that I go.
If it will rain, we will stay. tense in condition If it rains, we will stay.
I enjoy to swim. gerund I enjoy swimming.
We discussed about the plan. extra preposition We discussed the plan.
Where you are going? question order Where are you going?
The informations are useful. uncountable The information is useful.
She is more taller than me. double comparative She is taller than I am / than me.
Despite of the rain, we went. extra preposition Despite the rain, we went.
A man who he called me … resumptive pronoun A man who called me …
I have seen him yesterday. tense I saw him yesterday.
Design guideline Why
use errors that learners actually make validity
limit the number of errors per item clarity
avoid errors that are matters of style or dispute fairness
make clear whether the learner should correct or only identify scoring
include some correct lines prevent assuming every line has an error
specify one correct answer or accept alternatives scoring
test the error types taught in the course alignment
avoid trivial or typographical errors focus on grammar
score identification and correction separately if both are tested information

Error identification is the format of a number of older standardized tests, and is still used in many school exams; it tests the learner’s ability to see an error, which is a part of editing, but not the ability to write correct sentences. Learners can often identify errors that they would make themselves in writing, or the reverse. Combining error correction with writing tasks gives a fuller picture, and the analysis of the errors that the test reveals follows the procedure described in the chapter on error analysis and correction.

5. Communicative Testing: Grammar in Context

Principle Meaning Example
authentic tasks tasks resemble real-world language use write an email to a landlord about a repair
grammar for purpose accuracy is judged by its effect on communication and by appropriateness requests; advice; narrating
integration of skills reading, listening, writing, and speaking are combined listen to a message, write a reply
performance assessment the learner does a task and is scored by criteria a role play
holistic and analytic scoring overall impression or separate criteria —
contextualized items items are embedded in a text or situation a gapped email
focus on meaning and form tasks that require forms to be used accurately for meaning a story retelling to test past tenses
feedback results are useful for learning comments on the criteria
Task type Grammar assessed Scoring focus
write an email requesting information modals; polite forms; question forms accuracy and appropriateness
describe a picture present progressive; prepositions of place accuracy and range
narrate a sequence of pictures past tenses; linking words tense control; complexity
give advice in a role play modals; conditionals accuracy; appropriateness
summarize an article reported speech; nominalization accuracy; complexity
write a report on a survey present perfect; comparison accuracy; range
respond to a complaint politeness; apology; past and perfect tone; accuracy
oral interview all spoken grammar accuracy; range; fluency
write an opinion essay complex sentences; hedging accuracy; complexity; coherence
complete a dialogue functional grammar appropriateness
give instructions for a process imperatives; sequencers accuracy; clarity
write a short biography past simple; time expressions tense; accuracy
Measures of grammatical performance (CAF) Meaning Example
complexity the range and sophistication of structures (clauses per unit; variety of structures) use of subordination and varied tenses
accuracy the proportion of error-free units error-free clauses per hundred clauses
fluency speed and flow words per minute (speech)
range the variety of forms used several tenses, modals, clause types
appropriateness fit to the register and purpose a formal request in a formal register
  • Define the construct. What aspect of grammar ability is being assessed?
  • Give a clear task and audience.
  • Provide a rubric and share it with learners.
  • Train raters and use samples to calibrate.
  • Use more than one task for important decisions.
  • Check reliability (double marking).
  • Give feedback linked to the criteria.

Communicative tests are more authentic and more informative about use, but they are more expensive to mark and less reliable unless rubrics and rater training are good. A balance of discrete-point and communicative tasks is a common compromise, and the choice depends on the stakes of the test.

6. Rubrics for Assessment: Scoring Criteria

Type of rubric Description Use
holistic one overall score for the whole performance quick; a general impression
analytic separate scores for each criterion (accuracy, range, and so on) diagnostic; more reliable and informative
primary trait scores one key feature (for example, tense control in a narrative) a focused task
checklist yes/no for each feature simple tasks; self-assessment
scale (band descriptors) a series of bands with descriptions of performance standardized tests; the CEFR descriptors
weighted rubric different weights for criteria when some criteria matter more
Criterion (original rubric for a written narrative) 4 (strong) 3 (good) 2 (developing) 1 (limited)
accuracy of tenses tenses are accurate and consistent; errors are rare generally accurate; some errors that do not impede meaning frequent errors; some impede meaning pervasive errors; meaning is often unclear
range of structures a wide range of structures used appropriately, including complex sentences a good range, with some complex sentences limited range; mostly simple sentences very limited; fragments and memorized phrases
sentence structure and word order sentences are well formed and varied sentences are generally well formed frequent problems of order and boundaries sentences are often unclear
accuracy of articles, prepositions, and agreement almost always accurate mostly accurate frequent errors pervasive errors
linking and cohesion ideas are clearly connected with appropriate linkers mostly clear connections connections sometimes unclear little or no connection
appropriateness to the task register and tone fit the task mostly appropriate some mismatch inappropriate
Sample scoring (invented paragraph and scores) Observation
Last summer I went to Mysuru with my family. We visited the palace and it was very beautiful. The guide told us that the palace was built in 1912. In the evening we have eaten dinner in a restaurant, and my father pay the bill. I enjoyed very much the trip. —
accuracy of tenses 3: mostly accurate; have eaten (should be ate) and pay (should be paid) are errors
range 3: a passive (was built), reported speech, a time clause
sentence structure and word order 3: enjoyed very much the trip is a word-order error
articles, prepositions, agreement 4: accurate
linking 3: and, in the evening; adequate
appropriateness 4: appropriate
total 20 of 24 (a good performance with two tense errors and one word-order error to target)
Practice for fair scoring Why
train raters with sample scripts shared understanding
use anchor samples at each band calibration
double mark a sample check reliability
moderate disagreements fairness
mark criteria separately analytic information
decide in advance how to treat spelling and punctuation consistency
count errors by gravity, not only by number a global error matters more than a local one
give feedback linked to criteria learning
review the rubric after use improvement
do not let one error decide the score balance

Scales of this kind are used in the Common European Framework of Reference (CEFR) and its descriptors of grammatical accuracy and range at each level from A1 to C2. A rubric makes the standards public, which helps both markers and learners; it should be adapted to the task and the level, and its descriptors should describe performance in positive terms where possible.

7. Self-Assessment: Learner Autonomy

Method Description Example
can-do statements descriptions of what a learner can do; the learner marks each I can describe past events using the past simple and past continuous.
checklist a list of grammar points to tick I use articles correctly in most sentences.
portfolio a collection of the learner’s work with reflection writing samples; corrected drafts
learning log or journal regular reflection on what was learned a weekly entry
error log a record of errors and corrections —
rubric-based self-assessment the learner scores their own text against a rubric —
peer assessment classmates assess each other —
goal setting the learner sets targets for the next period This month I will reduce article errors.
self-test the learner takes a practice test and reviews —
conference with the teacher a discussion of self-assessment and progress —
Can-do statements for grammar (original) Level
I can use the simple present and present continuous to talk about habits and current events. beginner
I can use the simple past to tell a story. beginner to intermediate
I can use the present perfect to describe experiences. intermediate
I can use modals to give advice and make requests politely. intermediate
I can use the first and second conditionals. intermediate
I can use relative clauses to add information. intermediate
I can report what someone said using backshift. intermediate
I can use the passive voice appropriately in descriptions of processes. upper intermediate
I can use mixed conditionals and inversion for emphasis. advanced
I can vary sentence structure for style and use hedging appropriately in academic writing. advanced
  • Teach learners to self-assess. Show them the criteria and examples.
  • Compare self-assessment with teacher assessment. The comparison improves accuracy.
  • Use self-assessment formatively, for learning, rather than for grades.
  • Combine with error analysis. A learner who can name their errors can plan practice.
  • Be aware of limits. Learners at lower levels may over- or under-estimate their ability, and self-assessment varies by culture and personality.
  • Use it for motivation. Recording progress is encouraging.

8. Designing a Grammar Test: Specifications, Blueprints, and Interpretation

  1. Define the purpose and the test takers. Placement, progress, achievement? Which level?
  2. Define the construct. Which grammar knowledge or ability is to be measured?
  3. List the content. The structures taught or required.
  4. Choose formats to match the construct and the practical limits.
  5. Write a blueprint (specification). The number of items per area and format.
  6. Write the items following guidelines.
  7. Review the items with a colleague; check for ambiguity and bias.
  8. Pilot the test with a sample of learners.
  9. Analyze the results (facility, discrimination, reliability) and revise.
  10. Set the scoring and the cut scores.
  11. Administer under fair conditions and keep it secure.
  12. Report and use the results appropriately, with feedback.
Blueprint (invented 40-item achievement test for an intermediate course) Items Format
present perfect and simple past 6 multiple choice and gap-fill
conditionals (first and second) 6 gap-fill and sentence completion
relative clauses 4 sentence combining
reported speech 4 transformation
passive voice 4 gap-fill
articles and prepositions 6 cloze (a short text)
modals 4 multiple choice
error correction 6 a paragraph with errors
writing task (separate, scored by rubric) 1 an email (analytic rubric)
total 40 items plus a writing task —
Interpreting results Guidance
a score is an estimate, not a fact there is always error of measurement
look at patterns, not only totals which areas are weak?
use several sources of evidence tests, classwork, observation, portfolios
consider the test’s limits a grammar test does not measure communicative ability
avoid labeling learners scores describe performance on one test
give feedback results should lead to action
follow up with teaching reteach weak areas
check for bias and fairness —
keep records and compare over time progress
do not over-interpret small differences measurement error
Washback (the effect of tests on teaching) Positive Negative
content teaches useful skills narrows teaching to the test
methods encourages communicative practice if the test is communicative drilling of test formats
motivation goals and feedback anxiety
learner behavior regular study cramming
design response tests that reflect the course tests that distort the course

Test design is an iterative process. A teacher’s class quiz needs only a simple version of it (a clear purpose, a sample of items matched to the lesson, and a quick review of the results), while a high-stakes test needs the full process, and a team. In both cases the guiding question is the same: does this test give useful information about what I want to know? The relation of testing to the teaching described in the earlier chapters is that tests should reflect the goals and the methods of the course, and that the results should lead back into teaching. The grammar of the structures tested is described in the chapters of this course, such as editing and proofreading for the proofreading skills that error-correction tests assess.

9. Limits of Grammar Tests and Common Misconceptions

Misconception Reality
A high grammar test score means good communication. Knowing rules does not guarantee using them in real communication.
A low score means the learner cannot communicate. A learner may communicate well but not know the rules or the test format.
Multiple-choice tests are always unreliable. They can be highly reliable but measure recognition.
Open-ended tests are always more valid. They are more authentic but need rubrics and raters.
One test is enough for important decisions. Use several sources of evidence.
Tests are neutral. Tests reflect assumptions about what grammar is and how it is learned.
Test items from a textbook are good tests. Items need review and piloting.
Self-assessment is not reliable at all. It can be accurate when learners are trained.
The more items, the better. Quality and coverage matter.
Error counts show proficiency. The gravity and variety of the language matter too.
A test measures only what is taught. Learners also carry out other learning.
Every “rule” in a test is a rule. Tests should avoid contested points (see the chapter on grammar myths).
Test preparation is cheating. Familiarity with the format is fair; memorizing items is not.
✎ Worked example

Write five test items for the present perfect (original), in five formats, and judge them.

  1. Multiple choice: I ___ in this city since 2018. (a) lived (b) have lived (c) am living (d) live. Key: (b). Tests recognition; the time cue since makes it fair.
  2. Gap-fill: She ___ (finish) her report, so she can go home. Key: has finished. Tests production; accept has just finished.
  3. Error correction: I have seen that film last year. Key: I saw that film last year. Tests editing.
  4. Transformation (key word): “I started working here in 2020.” → BEEN I ___ here since 2020. Key: have been working. Tests flexibility; check that the key word forces one structure.
  5. Communicative task: Write a short email to a friend describing three things you have done this month. Scored with an analytic rubric (accuracy of the present perfect, range, appropriateness).
  6. Judgment: the first four items test form and use at sentence level and are quick to mark; the fifth tests the same grammar for a purpose, and shows whether the learner uses it spontaneously. A balanced test would use several of them.
▶

Recap in 12 seconds

Watch the lesson build itself: Grammar tests differ in purpose and format and are judged by validity, reliability, authenticity, and washback. The video has no sound.

Exercises

Part A: Types and Qualities (Exercises 1–7)

Exercise 1: Name three purposes of tests.

Show Answer

Answer: Placement, diagnostic, progress, achievement, proficiency.

Exercise 2: Define a discrete-point test.

Show Answer

Answer: A test of one language element at a time.

Exercise 3: Define an integrative test.

Show Answer

Answer: A test that requires several elements and skills at once.

Exercise 4: What is validity?

Show Answer

Answer: The test measures what it is meant to measure.

Exercise 5: What is reliability?

Show Answer

Answer: Consistency of results.

Exercise 6: What is washback?

Show Answer

Answer: The effect of a test on teaching and learning.

Exercise 7: What is authenticity?

Show Answer

Answer: Tasks resemble real use.

Part B: Multiple Choice (Exercises 8–13)

Exercise 8: Name the parts of a multiple-choice item.

Show Answer

Answer: Stem, key, distractors.

Exercise 9: What is a distractor?

Show Answer

Answer: A plausible wrong option.

Exercise 10: Name a guideline for writing items.

Show Answer

Answer: One correct answer; plausible distractors from common errors; options of similar length; avoid “all of the above.”

Exercise 11: What does the facility value show?

Show Answer

Answer: The proportion who answered correctly.

Exercise 12: What does the discrimination index show?

Show Answer

Answer: How well the item separates high and low scorers.

Exercise 13: Name a limitation of multiple choice.

Show Answer

Answer: It tests recognition, not production; guessing.

Part C: Gap-Fill and Error Correction (Exercises 14–20)

Exercise 14: Define a cloze test.

Show Answer

Answer: A text with words removed to be restored.

Exercise 15: What is a rational cloze?

Show Answer

Answer: A cloze with deliberately chosen deletions.

Exercise 16: What is word formation as a task?

Show Answer

Answer: Changing a root word to fit the gap.

Exercise 17: Name a feature of key word transformation.

Show Answer

Answer: A second sentence completed using a given word to keep the meaning.

Exercise 18: Correct: *She don’t like tea.

Show Answer

Answer: She doesn’t like tea.

Exercise 19: What is error identification?

Show Answer

Answer: Finding the error in a sentence or an underlined part.

Exercise 20: Why include correct lines in an error-correction test?

Show Answer

Answer: To prevent assuming every line has an error.

Part D: Communicative Testing and Rubrics (Exercises 21–26)

Exercise 21: Give a feature of a communicative test.

Show Answer

Answer: Realistic task; scored with a rubric; authentic purpose.

Exercise 22: What does CAF stand for?

Show Answer

Answer: Complexity, accuracy, fluency.

Exercise 23: What is an analytic rubric?

Show Answer

Answer: One that scores separate criteria.

Exercise 24: What is a holistic rubric?

Show Answer

Answer: One overall score.

Exercise 25: Name a way to improve rater reliability.

Show Answer

Answer: Training, anchor samples, double marking.

Exercise 26: Name a criterion for grammar in writing.

Show Answer

Answer: Accuracy, range, sentence structure, appropriateness.

Part E: Self-Assessment and Design (Exercises 27–30)

Exercise 27: What is a can-do statement?

Show Answer

Answer: A statement of what a learner can do, used in self-assessment.

Exercise 28: What is a test blueprint?

Show Answer

Answer: A specification of the number and type of items for each area.

Exercise 29: Name two steps of test design.

Show Answer

Answer: Define purpose; define construct; write items; pilot; analyze.

Exercise 30: Does a high grammar test score guarantee good communication?

Show Answer

Answer: No.

Test Your Knowledge

Ready to check what you learned? Answer all 10 questions, then submit. You need 80% (8 of 10) to pass. Your answers are not stored.

Question 1:

What is a discrete-point test?




Question 2:

What is a distractor?




Question 3:

What is a cloze test?




Question 4:

Which is a feature of a communicative test?




Question 5:

What does validity mean?




Question 6:

What does reliability mean?




Question 7:

What is washback?




Question 8:

An analytic rubric




Question 9:

What does the facility value show?




Question 10:

What is a can-do statement?




Frequently Asked Questions

What is grammar testing?

Grammar testing is the measurement of grammar knowledge and ability by tests such as multiple-choice items, gap-fills, error correction, and writing tasks. Grammar testing should match its purpose and reflect meaningful use.

What is grammar assessment?

Grammar assessment is the broader process of gathering evidence of grammar learning, using tests, tasks, rubrics, portfolios, and self-assessment. Grammar assessment can be formative or summative.

How do I go about testing grammar?

Define the purpose and construct, choose formats (multiple choice, gap-fill, error correction, tasks), write and review items, pilot, analyze, and interpret. Testing grammar well means testing it in context.

What is grammar evaluation?

Grammar evaluation is the judgment of the quality of a learner’s grammar, or of the effectiveness of a grammar test or course. Grammar evaluation uses criteria such as accuracy, range, and appropriateness.

What are grammar exams?

Grammar exams are tests, or sections of tests, that assess grammar, from school exams to sections of standardized tests. Grammar exams use formats such as multiple choice, error identification, and gap-fill.

How do I write a good multiple-choice item?

Use a clear stem, one key, plausible distractors based on real errors, options of similar length, and a context that makes the answer clear. Review and pilot the items.

Can a grammar test measure communicative ability?

Not by itself. Grammar tests measure knowledge or use of forms; communicative ability needs tasks in which grammar is used for a purpose, so a balance of formats gives a better picture.

Conclusion

You now have a study of grammar testing and assessment: the types and qualities of tests, multiple-choice items, gap-fill and cloze, error correction, communicative testing, rubrics, self-assessment, test design, and the limits of testing. You can design and interpret grammar tests with care.

When you design a test, ask what you want to know and what kind of evidence will tell you. Next, you will study grammar resources and references.

▣ Sources and further reading
  • Bachman, Lyle F., and Adrian S. Palmer. Language Testing in Practice. Oxford University Press, 1996.: the qualities of test usefulness.
  • Purpura, James E. Assessing Grammar. Cambridge University Press, 2004.: the assessment of grammatical ability.
  • Hughes, Arthur. Testing for Language Teachers. 2nd ed. Cambridge University Press, 2003.: a practical guide to test design.
  • Council of Europe. Common European Framework of Reference for Languages: Learning, Teaching, Assessment. Cambridge University Press, 2001.: descriptors of levels.
  • Merriam-Webster Dictionary: assessment: the dictionary entry.
ⓘ About this page

Last reviewed: October 29, 2026. How it was made: drafted with AI assistance, then checked with automated tests (structure, links, answer keys, spelling) and against the sources above. A person has not yet read this chapter line by line. Found a mistake? Report it.