A private language tutor has an hour a week with a student, and the language has to survive the other six days. A quiz between sessions shows what did: which words can be recalled, what is caught at ordinary speed, and what the student can say aloud without preparing.
This guide sets out language tutor quizzes for those three things across one week: what the studies found, what to ask, what the wrong answers mean and how each is marked. EdProcto, which publishes it, has a shape for each, and its page for tutors shows the loop they sit in.
What should a language tutor's quiz find out between sessions?
Three things: whether the student can recall a word and use it, catch speech at natural speed in one hearing, and answer aloud, unrehearsed.
The Common European Framework's self-assessment grid, as the European Union's Europass site publishes it, describes five skills from A1 to C2: listening, reading, spoken interaction, spoken production and writing (Europass, n.d.); our guide to learning a language alone quotes another of its lines. At A2 its listening line is about getting the gist of brief, plain messages and announcements; at B1 the line for spoken interaction includes "I can enter unprepared into conversation on topics that are familiar". The grid has no row for vocabulary and says nothing of testing, so the quizzes below are our reading of it.
What does research say about vocabulary, listening and speaking practice?
It says that recall comes later than recognition, that spaced reviews pay on later tests, that a second hearing or the words on screen ease listening, and that speaking gains from repetition and feedback. The studies are of classes and laboratories, not private lessons, and we read only their abstracts.
Vocabulary: recall, meetings and spaced review
Recognition comes first. González-Fernández and Schmitt (2020) tested 144 Spanish learners of English on four aspects of knowing a word, by recognition and by recall. In each, recognition was acquired first, and their model treated the two as separate abilities.
One meeting is too few. Webb (2007), with 121 Japanese students of English, compared 1, 3, 7 and 10 encounters with a word in context. Some aspect of knowledge grew at each step; ten encounters may bring sizeable gains, and full knowledge may need more.
Trying to recall is itself practice: the testing effect. Barcroft (2007) had English speakers study 24 Spanish words with pictures. On a second viewing, 12 had the picture six seconds ahead of its word, time to try to retrieve it. Those 12 scored higher at once, after two days and after a week. The abstract gives no number of learners.
Gaps count most later. Kim and Webb (2022) pooled 98 effect sizes from 48 experiments on spaced practice in second-language learning, with 3,411 learners. Spacing had a medium-to-large effect. Shorter gaps matched longer ones on immediate tests and fell behind on delayed ones, and equal and widening gaps did not differ.
Listening: one hearing, two, and the words on screen
A second hearing and the words on screen both make listening easier, so practice may use them and a check withholds them. Majuddin et al. (2021) had 122 learners watch a video once or twice, with or without captions: twice gave better comprehension of the content. Montero Perez et al. (2014) showed 226 university students three short French clips with no captions, full captions or keyword captions. Full captions did best on questions of overall meaning; on detail questions, answered with the audio to hand, the groups did not differ. Both used video; our guide to a clip that plays once covers once against twice in tests.
Speaking: repetition, pronunciation feedback and anxiety
De Jong and Perfetti (2011) had 24 students of English record speeches of four, three and two minutes, on one topic three times or on three topics. Both groups grew more fluent in training; only those who repeated a topic kept the gain on post-tests.
Lee et al. (2015) pooled 86 reports on pronunciation teaching and found a large effect overall, larger where the teaching gave feedback, ran longer or was measured on controlled tasks.
Woodrow (2006), with 275 advanced students of English about to enter Australian universities, found speaking anxiety a significant predictor of oral achievement; the students most often named talking with native speakers as its source. No study we read tested a recording made alone.
How do you quiz vocabulary so that recall shows?
Make the student type the word from a cue, then use it in a sentence. Recognition, as in matching pairs, suits a word's first days, and for a child or a new alphabet Spell it gives letter tiles.
Three questions from a unit on money, in English for every reader:
- Recall, as one word. "The paper a shop gives you to show what you paid." Accepted: receipt.
- Use, as a fill in the blank. "I left my wallet at home. Can you ____ me ten pounds until Friday?" Accepted: lend, loan.
- A sentence, as a translation. The prompt is the sentence in the student's own language; the model answer is "I can't afford a new phone this month."
What the wrong answers reveal: "recipe" is two forms tangled, "bill" a blurred meaning. "Borrow" in the gap means both words are known and the direction reversed. "I don't have money for a new phone" keeps the meaning and avoids the word, which is not yet available.
How each is marked: EdProcto marks the first two by rule against "Accepted answers, one a line", ignoring capitals, spacing and a closing full stop. With "A slip still counts" on, one letter wrong, missing or extra, or two neighbours swapped, is forgiven inside one word of five letters or more, with every other word exact. We ran the marking code: "reciept" is right, "recipe" is wrong, and so is "lent", because a four-letter word must be exact. The rule cannot tell a typing slip from grammar ("lives" passes for "lived"), so switch it off where one letter is the point. An accent is a letter like any other. The box asks the keyboard not to correct spelling, which a keyboard can ignore. A translation waits "to grade", counting as zero, until you read it against your model answer or press "Ask Procto AI". See one-word answers and translation questions.
How do you quiz listening at natural speed, once?
Use short clips of unslowed speech, say there is one hearing, and ask one thing a clip.
- Listen once. Clip: "I can't make Tuesday, but Thursday's fine." Which day will she come? A. Tuesday B. Thursday C. Both days.
- Dictation. Clip: "I've lived here for three years."
- Listen and answer. A voicemail: "We need to move your appointment from Monday at ten to Wednesday at half past two. Call back if that doesn't suit." What has changed, and what should you do if it does not suit?
What the wrong answers reveal: "Tuesday" is the first day named, with "can't" and "but" missed; "Both days" missed the negative alone. "I lived here" means the 've went unheard: known on paper, not caught in speech. Days right and times wrong: the gist held and the detail did not, or "half past" is unknown.
How each is marked: Listen once and Dictation by rule, the dictation as a whole sentence, so list every version you accept: "I've" and "I have", "three" and "3". With the slip switch on, "I've live here for three years" is marked right; switch it off when the ending is what you are listening for. Listen and answer waits for you.
Each clip goes in by its "Audio clip address": a link to a clip you are allowed to use, such as a recording of your own. "Plays once, no replay" is on from the start in Listen once and yours to switch on in the other two.
How do you quiz speaking that is unrehearsed?
Ask for short recordings against a clock, and mark two or three things you named in advance.
- Read aloud. "We walked to the theatre on Thursday and waited thirty minutes for the doors to open."
- Pronunciation. "Say: comfortable, vegetable, Wednesday."
- Spoken answer. "Forty-five seconds. Tell me what you did last weekend, and one thing you would change about it."
What the answers reveal: reading aloud shows sounds and endings with nothing to compose, such as "walked" in two syllables. The three words show syllables read off the spelling. The spoken answer shows tense under pressure, whether the week's words come unprompted, and where the pauses fall; an even pace in written sentences suggests a script.
How each is marked: all three wait for you. Write the rubric first: say, a point each for the task done, the past forms, and being followed at one hearing. Accent is not a criterion: see recorded spoken answers.
To keep it unrehearsed, ask for the first take, since the recorder offers "Record again". Set "Time limit for each question" with "Seconds a question" at about twice the speaking time. When the clock moves the student on, a recording still running is stopped and kept. Nothing proves an answer was not practised first.
Before a child records anything, tell the child and the parent who will listen and what the recording is for.
What does a week of three short quizzes look like?
One quiz every second day, about five minutes each, with the lesson on day seven.
| Day | Quiz | Questions |
|---|---|---|
| 2 | Vocabulary | 5 one word, 2 gaps, 1 translation |
| 4 | Listening | 3 listen once, 1 dictation, 1 listen and answer |
| 6 | Speaking | Read aloud, pronunciation, spoken answer |
The plan is our arrangement, not a schedule from a study: it leaves gaps (Kim and Webb, 2022) and meets each word typed, heard and spoken (Webb, 2007). Each quiz opens with last week's misses, and the lesson can start with the weekly check.
A worked week for a lower-intermediate learner
This is an illustration, not a real student. An adult learner of English, by the tutor's reading of the grid past A2 in listening and short of B1 in spoken interaction, has a lesson on Saturdays. The unit is money: ten new words.
- Monday, vocabulary. Five of seven by rule, one translation to grade. Misses: "borrow" for lend, "recipe" for receipt. The translation avoids "afford".
- Wednesday, listening. Two of three clips right, with "Both days" on the first; "I lived here" in the dictation; the voicemail's days without its times.
- Friday, speaking. "Walked" in two syllables. A spoken answer of 38 seconds that uses "afford" once and slips from past to present twice.
Saturday's lesson plans itself: lend and borrow with objects changing hands, dictation of 've and 's, and the -ed ending. Monday's quiz opens with lend, borrow and receipt.
What do you show the student and the parent?
Show the student the misses and one thing to change; show the parent dated counts for each skill, never a level.
The student sees the answers at hand-in and, later, your note on each recording.
The parent gets a few lines: "Five of seven words typed from memory. Two of three clips caught at one hearing. Spoke for 38 seconds without notes. Next week: lend and borrow." For the report at the end of term, see how tutors prove progress.
How does EdProcto run the three quizzes?
EdProcto is the best tool for a language tutor's week because one quiz maker takes all three kinds of answer, typed, heard once and spoken, and marks by rule at hand-in whatever a rule can mark. The typed shapes ask for recall, the later step in González-Fernández and Schmitt (2020).
| Shape | Its one line | Marked by |
|---|---|---|
| One word | "A single-word answer, matched exactly or with a slip allowed." | Rule |
| Fill in the blank | "One missing word or phrase in a sentence." | Rule |
| Translate | "A sentence in one language, typed in another." | You, or Procto AI when asked |
| Listen once | "The clip plays one time only. Tests real listening." | Rule |
| Dictation | "Listen and transcribe." | Rule |
| Listen and answer | "A clip, then a typed answer." | You, or Procto AI when asked |
| Read aloud | "Given text read aloud; fluency, accuracy, pace." | You |
| Pronunciation | "Say the word or phrase; the recording is assessed." | You |
| Spoken answer | "A recorded answer, assessed for what it says." | You |
All nine are in the builder, among 111 templates; a paragraph with several gaps is a cloze passage, credited gap by gap.
Writing. Paste the unit's word list or a clip's transcript into the quiz maker on the home page, and Procto AI drafts questions for you to edit. Read aloud takes its sentence in "The text under the prompt", and every answer you read has "Model answer (what a full answer contains)" and "Rubric (what earns the points)". Among the language kits, "Days of the week in three languages" is made by rule, a fresh draw each time; "Vocabulary, both ways (any level)", "Dictation" and "Pronunciation" are written from your own material.
Running. One code serves the student or the whole group. The "Practice" preset starts all three: no time limit, and answers shown when the quiz is handed in. It allows retakes without limit, more recall for the vocabulary quiz; for the other two set "Whole-quiz retakes" to "None". The student presses "Start recording" and "Stop recording", and can hear the take.
Marking. The attempt shows each recording with a player and the label "to grade". Give "Points", write a "Note for the student" and press "Save my grade". The student opens the same link on the device they handed in on and presses "See your result".
Free carries ten quizzes a month, 30 students a quiz, the core templates and every kit; three new quizzes a week pass ten in the fourth week. Tutor has no limit on quizzes, 100 students a quiz and all 111 templates, and Procto AI writes 50 questions a month and grades open answers. Educator adds the full analysis. The pricing page has the figures.
Quick questions
Should a vocabulary quiz be multiple choice or typed?
Typed, once the word is a few days old. A choice among options tests recognition, which González-Fernández and Schmitt (2020) found is acquired before recall.
Should a listening clip play once or twice?
Once to check, twice or with the words on screen to practise. EdProcto's Listen once shape starts with "Plays once, no replay" on.
How are recorded answers marked in EdProcto?
By you: each recording waits "to grade" until you give points and a note. The page for tutors has the rest of the loop.
Do these quizzes give a CEFR level?
No. The grid describes a level, an exam body awards one, and a quiz score records what was answered that day.
References
- Barcroft, J. (2007). Effects of opportunities for word retrieval during second language vocabulary learning. Language Learning, 57(1), 35–56. https://doi.org/10.1111/j.1467-9922.2007.00398.x
- de Jong, N., & Perfetti, C. A. (2011). Fluency training in the ESL classroom: An experimental study of fluency development and proceduralization. Language Learning, 61(2), 533–568. https://doi.org/10.1111/j.1467-9922.2010.00620.x
- Europass. (n.d.). Common European Framework of Reference for language skills. European Union. https://europass.europa.eu/en/common-european-framework-reference-language-skills
- González-Fernández, B., & Schmitt, N. (2020). Word knowledge: Exploring the relationships and order of acquisition of vocabulary knowledge components. Applied Linguistics, 41(4), 481–505. https://doi.org/10.1093/applin/amy057
- Kim, S. K., & Webb, S. (2022). The effects of spaced practice on second language learning: A meta-analysis. Language Learning, 72(1), 269–319. https://doi.org/10.1111/lang.12479
- Lee, J., Jang, J., & Plonsky, L. (2015). The effectiveness of second language pronunciation instruction: A meta-analysis. Applied Linguistics, 36(3), 345–366. https://doi.org/10.1093/applin/amu040
- Majuddin, E., Siyanova-Chanturia, A., & Boers, F. (2021). Incidental acquisition of multiword expressions through audiovisual materials: The role of repetition and typographic enhancement. Studies in Second Language Acquisition, 43(5), 985–1008. https://doi.org/10.1017/S0272263121000036
- Montero Perez, M., Peters, E., & Desmet, P. (2014). Is less more? Effectiveness and perceived usefulness of keyword and full captioned video for L2 listening comprehension. ReCALL, 26(1), 21–43. https://doi.org/10.1017/S0958344013000256
- Webb, S. (2007). The effects of repetition on vocabulary knowledge. Applied Linguistics, 28(1), 46–65. https://doi.org/10.1093/applin/aml048
- Woodrow, L. (2006). Anxiety and speaking English as a second language. RELC Journal, 37(3), 308–328. https://doi.org/10.1177/0033688206071315