Does hearing an object's sound help you remember where you saw it?
Seeing an object together with its matching sound, like a frog with a ribbit, improved memory for which scene it appeared in, but did not help memory for other objects shown beside it.
Source
Multisensory processing impacts memory for objects and their sources
Study at a glance
- Design
- Human experiment — Within-subject manipulation of sound congruency at incidental encoding (congruent object sound vs white noise) followed by surprise memory tests: item recognition with confidence-based ROCs (Exps 1a, 1b) and 4-alternative environment source memory plus location memory in virtual scenes (Exp 2)
- N
- N=168 · 50 analysed in Exp 1a (from over a hundred tested), 50 in Exp 1b (of 83 tested) and 68 in Exp 2 (of 209 tested); each experiment used different undergraduates, and the high exclusion rates followed preregistered criteria.
- Population
- Undergraduate students at the University of California, Davis, tested online for course credit
- Outcome
- Recollection (ROC y-intercept), familiarity (d') and overall recognition (AUC) for congruent-sound, neighbouring and control objects; environment source memory, confidence and location accuracy
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
In Experiments 1a and 1b, students saw pairs of objects with a sound that matched one of them or with meaningless white noise, while doing a size judgement (1a) or rating how related the two objects were (1b). A surprise recognition test with confidence ratings let the authors separate recollection (remembering details) from familiarity. In Experiment 2, students searched for objects placed in scenes from four learned virtual environments while hearing either the object's sound or white noise, and were later asked which environment each object had been in and where.
What they found
In Experiment 1a, objects heard with their matching sound were better recollected than control objects, but the object sitting next to them gained no significant benefit, and familiarity and overall recognition were unaffected. In Experiment 1b, where the task made people relate the two objects, the sound benefit disappeared entirely. In Experiment 2, environment source memory was better for objects paired with matching sounds than with white noise (mean proportion correct 0.41 versus 0.38, a small effect), but memory for exact location did not differ.
The limits
What it doesn't show
The source-memory benefit was small and one-tailed, and the location task was so hard that performance was near guessing, so a location benefit could have been missed. Studies were run online with large preregistered exclusions (for example, well over half of Experiment 1a's participants were dropped for not understanding the 'recollect' response), which may limit generality. Experiments 1a and 1b tested memory for single objects, not for the pairing between objects, so they cannot say whether sounds help associative memory for object pairs. The study also cannot separate true audiovisual integration from the sound simply reinforcing the object's identity.
Key terms
- Recollection vs familiarity
- Two routes to recognising something: recollection retrieves specific details of the original event, while familiarity is a vague sense of having seen it before.
- Receiver operating characteristic (ROC)
- A curve plotting hits against false alarms across confidence levels; its shape can be used to estimate recollection and familiarity separately.
- Source memory
- Memory for the context in which something was encountered, such as where or when, rather than for the item itself.
- Semantic congruency
- A match in meaning between what is seen and what is heard, as when a dog is shown with a barking sound.
- Incidental encoding
- Learning that happens while doing some other task, without knowing a memory test will follow.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What control condition was used instead of silence?
Common questions
Why use white noise as the control rather than silence?
Any sound might boost memory just by making a trial more alerting or distinctive. White noise controls for that, so the comparison isolates the effect of a meaningful, matching sound.
Why did the sound benefit vanish in Experiment 1b?
When people had to think about how the two objects related, that deeper, relational processing boosted memory overall and seems to have swamped any extra benefit from the sound.
Does this mean multisensory learning always improves memory?
No. The benefit was specific to recollection of the sounding object and to its scene context; it did not spread to neighbouring objects and disappeared under relational encoding.
More on Long-term memory