Ear training quizzes: intervals, chords and rhythm

Ear training quizzes play a clip and ask a student to name it: a perfect fifth, a minor chord, a cadence, three beats in a bar. The answer is one choice, so the quiz marks itself and your time goes to the pairs the class confuses.

EdProcto, the quiz maker that publishes this guide, carries the shape as Recognise the sound: your clip, a prompt and options, marked by rule. Its page for teachers gives the route from a first question to a class code. The guide is for a school music teacher, an instrumental or singing teacher, or a lecturer on a first-year musicianship course.

Every source was opened on 10 October 2026: three as abstracts, marked below, the rest in full. Cutting and levelling a clip, and the right to use one, are in sound recognition questions, which reports other studies.

What are aural skills, and what does research say about learning them?

Aural skills are what a musician does by ear: hold a phrase in memory, name it, sing it back, write it down, hear a wrong note. In a curriculum survey that Chenette (2021) reports, 99% of respondents used sight singing and dictation in their aural classes and 61% error detection. His own argument is that a task is more purely aural the more it rests on memory and the less on theory labels.

Six findings bear on a quiz.

• Naming an interval alone is not hearing it in a key. Chenette (2021) cites an earlier review of the experiments, by Karpinski, as finding little connection between identifying intervals out of context and in a tonal one.
• Interval identification still goes with melodic dictation. Nichols and Springer (2022; abstract read) gave 35 college musicians an interval test and dictation tasks that opened by establishing the key. With musical experience controlled, the interval score accounted for 28.9% of the variance in dictation.
• And with hearing a wrong note. Stambaugh and Nichols (2020; abstract read) tested 50 music education students on 33 intervals from C2 to B5 and on 15 two-bar melodies with one pitch error, in piano and vocal timbres. The two scores correlated at r = .75, and right interval answers came faster than wrong ones. They report no main effects or interactions between a student's main performing area and the timbre of the item.
• A change of timbre inside a question gets in the way. Zarate et al. (2013) had 15 musicians and 14 non-musicians say which of two intervals was larger. Discrimination was best overall with pure tones throughout, and worse when one tone was a flute, a piano or a synthetic voice. The authors add that pure-tone trials were more frequent, so practice may explain part.
• Major or minor is no easy opener. Mednicoff et al. (2018) played 73 undergraduates random runs of pure tones, at four speeds, that differed only in a major or minor third, with feedback after every trial. A listener who did poorly at one speed did poorly at all four, and of the 25 with five or more years of training, 13 averaged under 70% correct. These were random sequences, not melodies, as the authors say.
• Singing back measures the voice too. In Pfordresher and Brown (2007; abstract read), 10% to 15% of participants sang four-note sequences back at least a semitone off, yet did not differ from good singers in telling pitches apart.

On mixing kinds of interval in one practice set, see interleaving.

Which aural tasks can a self-marking question ask?

Any task whose answer is a name or a place can be marked by rule; one whose answer is a sound or a page of notation needs a listener or a reader.

Same or different carries more than it seems: Law and Zentner (2012) built a nine-part test of musical hearing on such trials alone, with parts for melody, rhythm, accent, tempo and tuning.

Picking a notated rhythm is not writing one, so keep a task of each kind in the term. Pick the clip, where the options are clips, Compare two clips and Music notation, with notes placed on a staff, are in the builder.

What should a question on chords and cadences ask?

Ask for the bass note or the cadence before the full chord label. In Chenette's (2021) pilot, 74 participants answered an anonymous online questionnaire on the chords of three excerpts. Listening to the bass line was the strategy most reported, by 88%. Of the 16 who were nearly right, eight labelled one first-inversion supertonic chord correctly and seven called it a subdominant: the two share a bass note. He notes that the answers could not be verified.

How do you make a clip that tests the ear and not the file?

Give every clip in a set one instrument, register, loudness and length, so that only the musical idea differs.

1. One idea a clip. An interval, a chord, a two-bar rhythm.
2. One instrument across the set. Law and Zentner (2012) faulted older tests for human performers whose timing, timbre and loudness varied from trial to trial, and built theirs from sampled and synthetic sounds. Export every clip from one notation program or keyboard at one setting.
3. Move the starting note. Samplaski (2004) played each interval at ten transpositions. If every fifth starts on middle C, students learn the top note.
4. Rhythm without a tune. Law and Zentner sounded their rhythm patterns on one unpitched percussion voice, chosen as clear and crisp. Clap, or use a woodblock, so that no tune helps.
5. A count-in that tells nothing. One bar of even clicks before a rhythm; none before a metre question, where a count-in is the answer. For a question in a key, play the tonic chord first.
6. Neutral file names. q03, never major-third: the address reaches the student's browser as typed.
7. A recording you may use. Your own, or one whose licence you have read.
8. Build it and sit it. In EdProcto, paste the clip's address into the question and take the quiz on a phone with earphones.

How many hearings should a clip get?

One or two for a clip to be named, more when a phrase must be held to compare or to write down, and as many as the student likes in practice. No study we read set one hearing against two for an interval or a chord, so these are designs to borrow, not comparisons.

• Samplaski (2004) allowed 8.5 seconds for a snap answer to each interval, itself under two seconds long, in blocks of 20.
• Law and Zentner (2012) played the first phrase twice before the comparison, to help listeners take it in.
• Nichols and Springer (2025) gave 34 music undergraduates four hearings of an eight-bar melody to write down, with 30 seconds of silence between. The mean was 36 of 63.

The studies of one play against two, in language tests, are in the clip that plays once. EdProcto sets the hearings on each question: a player the student may replay, or a clip that plays once. For exactly two hearings, record the idea twice in one clip, two seconds apart, and set it to play once.

In what order should ear training quizzes go?

From wide contrasts to close pairs, one change at a time. This is our working order, drawn from the findings above; no study we read tested it as a curriculum.

1. Same or different, higher or lower, step or leap.
2. Intervals far apart in size and kind: a second, a fifth, an octave, rising, on one instrument.
3. The close pairs. In Samplaski's (2004) map of 27 music students' confusions, the major and minor forms of the second, the third, the sixth and the seventh clustered strongly. The tritone moved most between the rising, falling and sounded-together forms, so add those forms here. He calls for replication with more listeners.
4. Chord quality: major against minor, then diminished and augmented. A class may split on major and minor.
5. In a key: the bass note, the cadence, then chord labels.
6. New instruments and registers last.

Run rhythm beside it. The easy rhythm items of Law and Zentner (2012) added or removed a note on a downbeat, the moderate ones changed an upbeat note, and the hardest altered semiquavers.

How many questions, and how often?

Eight to ten clips, about ten minutes, once or twice a week: a working rule, since nothing we read fixes it for aural quizzes. Short sets are for practice. With 18 same-or-different trials a part, internal consistency in Law and Zentner's first study ran from .48 for melody to .78 for tempo, so eight questions cannot rank a class. How many questions a quiz needs has the arithmetic, and retrieval practice the case for little and often.

A worked example: eight questions on intervals and chords

The class is invented: a first-year musicianship group in its third week. Every clip is a piano, the idea played twice with two seconds between, set to play once.

Questions 2 and 3 ask the same pair either way round, so a student who answers "major third" to both is separated from one who hears the difference. Question 4 falls, and question 5 is the tritone against its neighbours. Each question's topic is "Melodic intervals", "Harmonic intervals" or "Chord quality". The explanation for question 3: "A minor third, three semitones. Sing both notes, then the major third above the same A, and hear the top note rise."

How do you read the results of an ear training quiz?

Read the pairs before the total.

• The share right on a pair. Where a question sets one close pair, its share right says how well the class separates that pair. EdProcto's results show "Got it" for each question on every plan.
• Slow answers. In Stambaugh and Nichols (2020) right answers were the faster ones. With one question a page, every plan shows the median time on each question.
• Which wrong option. On the Educator plan the full analysis counts the takers of each option under "Chose it" and scores each topic under "By topic".
• Two camps on major and minor. If half the class is nearly perfect and half near chance, the question may be sound. Mednicoff et al. (2018) describe that pattern in an earlier study from the same laboratory: about 70% of listeners near chance, the rest nearly perfect.

Then set the confused pair side by side in the next quiz.

How does EdProcto run ear training quizzes?

EdProcto is the best tool for ear training quizzes on your own clips, because the hearings are fixed on the question for every student, the answer is marked by rule, and each question's share right is shown, which is where a confused pair appears, as major and minor thirds did in Samplaski's (2004) data.

• The clip. Recognise the sound is in the builder now. With "It is" at "Audio", the box "Audio clip address" takes the web address of your own file, kept wherever you hold files that open by a link.
• The hearings. With "Plays once, no replay" off, the student has the browser's ordinary audio player and may replay. With it on, the player is one button, "Play the clip", and a line that fills; afterwards the page reads "The clip has played its one time." A clip that fails to load is not counted as heard.
• The question. "Prompt", then "Options · mark the right one", from two to eight, "Topic (for the results by topic)" and "Explanation shown to the student afterwards".
• A clip on another shape. A True or false question has the link "Add a picture or a clip" under its prompt, which opens "It is": choose "Audio". Listen once is the choice shape whose clip plays one time only.
• Singing and writing. Sing or play records a take in the browser for you to mark by ear, as in recorded singing and playing.
• Marks. A study quiz checks each answer as it is given; an exam marks at hand-in. A score can be held until you release it, and no score is a grade.

Free covers ten quizzes a month, thirty students a quiz and the core templates. Tutor has all 111 templates for 100 students, and Educator 500 students with the full analysis. See pricing.

Quick questions

How many times should students hear an interval in a test?

Once or twice: no study we read compared the two for intervals. For two hearings in EdProcto, record the interval twice in one clip.

Can a quiz mark melodic dictation or sight singing by itself?

A reader marks the notation and a listener the singing. A self-marking question can ask which notated tune was played.

Which intervals do students confuse?

Often the major and minor forms of one interval: in Samplaski (2004) seconds, thirds, sixths and sevenths each clustered so.

Where does a music teacher start with EdProcto?

With the eight questions above, on your own clips. The page for teachers sets out the route to a class code.

References

• Chenette, T. (2021). What are the truly aural skills? Music Theory Online, 27(2). https://www.mtosmt.org/issues/mto.21.27.2/mto.21.27.2.chenette.html
• Law, L. N. C., & Zentner, M. (2012). Assessing musical abilities objectively: Construction and validation of the Profile of Music Perception Skills. PLoS ONE, 7(12), Article e52508. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0052508
• Mednicoff, S., Mejia, S., Rashid, J. A., & Chubb, C. (2018). Many listeners cannot discriminate major vs minor tone-scrambles regardless of presentation rate. The Journal of the Acoustical Society of America, 144(4), 2242–2255. https://pmc.ncbi.nlm.nih.gov/articles/PMC6192795/
• Nichols, B. E., & Springer, D. G. (2022). Interval identification predicts success in melodic dictation. Journal of Research in Music Education, 70(1), 109–126. https://pure.psu.edu/en/publications/interval-identification-predicts-success-in-melodic-dictation/
• Nichols, B. E., & Springer, D. G. (2025). Piano history, aural skills, and working memory predict melodic dictation performance. Frontiers in Psychology, 16, Article 1579714. https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2025.1579714/full
• Pfordresher, P. Q., & Brown, S. (2007). Poor-pitch singing in the absence of "tone deafness". Music Perception, 25(2), 95–115. https://researchconnect.buffalo.edu/en/publications/poor-pitch-singing-in-the-absence-of-tone-deafness/
• Samplaski, A. (2004). Interval-classes and psychological space. Music Theory Online, 10(2). https://mtosmt.org/issues/mto.04.10.2/mto.04.10.2.samplaski.html
• Stambaugh, L. A., & Nichols, B. E. (2020). The relationships among interval identification, pitch error detection, and stimulus timbre by preservice teachers. Journal of Research in Music Education, 67(4), 465–480. https://pure.psu.edu/en/publications/the-relationships-among-interval-identification-pitch-error-detec/
• Zarate, J. M., Ritson, C. R., & Poeppel, D. (2013). The effect of instrumental timbre on interval discrimination. PLoS ONE, 8(9), Article e75410. https://pmc.ncbi.nlm.nih.gov/articles/PMC3774646/