Modality effect

The modality effect in memory is a pattern found in immediate verbal-list experiments. People often recall the last items better when they heard the list than when they silently read the same list. The finding is narrower than the slogan “audio beats text,” and it does not show that speech creates a more durable memory.

Illustrative hypothetical: eight names before the game

At a fictional neighborhood game night, a host is about to start a cooperative star-map puzzle. Two teams receive the same eight invented planet names, one at a time. One team hears recordings through a speaker. The other sees each name printed alone on a card.

As soon as the eighth item ends, everyone writes down the names they remember.

This is an illustrative setup, not a documented experiment, and it supplies no invented score. Its purpose is to sharpen the question. Researchers would not need to ask only which team remembered more names overall. They could compare each position in the sequence. Did the groups perform similarly at the beginning but separate near the end?

That late-list comparison is where the classic modality effect appears.

What the modality effect means in memory research

In several experiments that held the verbal list constant and changed how it arrived, recall at the final positions was significantly higher after auditory presentation than after visual presentation. Early work connected input modality to the shape of the serial-position curve (Conrad and Hull, 1968; Craik, 1969). Modern studies have reproduced the auditory late-list advantage in immediate free recall and immediate serial recall (Grenfell-Essam, Ward, and Tan, 2017; Pazdera and Kahana, 2023).

That supports the locked Topic 191 claim: when matched lists are presented auditorily rather than visually, recall of the final items can be significantly higher. The qualification matters. This is an average difference under particular immediate-memory conditions, not a promise about every item, person, task or delay.

The rest of the curve also prevents an easy “hearing is better” conclusion. Grenfell-Essam and colleagues tested lists ranging from two to 12 words. For short lists, the auditory advantage was large and mainly confined to the final position. For longer lists, it was smaller and stretched across more late positions. The same experiments found a visual advantage at early positions. Pazdera and Kahana also found stronger auditory recency alongside stronger visual primacy in two large online experiments.

One channel can therefore lead at the end while the other leads earlier. An overall score can hide both patterns.

Why there is no single settled mechanism

The effect has outlasted several explanations of it. Crowder and Morton (1969) proposed a short-lived precategorical acoustic store. On that account, recent sounds retained extra information that was unavailable for silently presented visual items.

Later findings made the picture less tidy. Watkins and Watkins (1980) showed that an irrelevant auditory suffix or an auditory distractor could sharply reduce the advantage. Their suffix effects remained after delays of two and four seconds, longer than a very brief echo would predict. Other accounts have emphasized temporal distinctiveness (Glenberg and Swanson, 1986), changes in retrieved context (Pazdera and Kahana, 2023), or the way auditory and visual sequences form perceptual and speech-motor objects (Macken and colleagues, 2016).

These are competing explanations, not interchangeable names for a proven cause. It is misleading to assign heard words to a sound system, silently read words to a visual system, and declare the problem solved.

The boundaries change the result

The auditory recency advantage depends on more than whether a person heard or saw an item.

• List length and recall task: The size and spread of the difference change across short and long lists, free recall and ordered recall.
• Grouping: In one grouped-list experiment, the auditory advantage appeared at the first item of the final group, even though that was the third- or fourth-to-last item (Frick, 1989).
• Interference before recall: Auditory suffixes and distractors can weaken the advantage, while visual and auditory interruptions need not have the same effect.
• The material being remembered: Roberts found the auditory late-position advantage for linguistic material but not for melodic or harmonic sequences in the same research program (Roberts, 1986).
• The delay and what happens during it: Immediate auditory recency does not guarantee better long-term memory. Engle and Mobley found visual superiority at recent positions on a delayed test for lists that had not first been recalled (Engle and Mobley, 1976).
• Presentation rate: Older theories predict particular rate effects, but Pazdera and Kahana did not find a modality-by-rate interaction for primacy or recency in their experiments.

These limits are the useful part of the research. They turn a broad media preference into a question that can be tested: which items, in which task, measured when?

A different “modality effect” in multimedia learning

Educational research also uses modality effect for a different comparison. In that literature, learners usually see graphics or animation while related words arrive either as narration or as on-screen text. The outcomes are retention, explanation or transfer, not the final positions of a word-list recall curve.

For example, Mayer and Moreno (1998) compared animation with narration against animation with synchronized on-screen text. Their narration groups performed better on specified learning measures. A 2005 meta-analysis by Ginns synthesized 43 independent multimedia effects and found that pacing and the complexity of interacting elements moderated the results. The advantage was stronger in system-paced materials.

Learner control can change the direction. In a learner-paced web lesson, students given spoken rather than written explanatory text scored lower on retention and transfer (Tabbers, Martens, and van Merrienboer, 2004). In another learner-paced study, the audiovisual and visual-only groups did not differ immediately; after one day, the visual-only group scored higher on three of four outcomes (van den Broek, Segers, and Verhoeven, 2014). The visual-only learners could revisit slides and used that option more often.

The two literatures should not be collapsed. A late-list auditory advantage does not prove that a narrated course will teach better. A multimedia result does not explain the serial-position curve by itself.

Common misunderstandings

“Audio is more memorable than text.” Sometimes auditory presentation improves recall at recent list positions. Visual presentation can match or exceed it elsewhere, and delayed or learner-controlled tasks can reverse the comparison.

“The effect proves one working-memory model.” The empirical pattern has survived disagreement about acoustic persistence, temporal information, contextual retrieval and perceptual-motor organization.

“Reading the same words aloud will always help.” Self-production introduces another comparison, often studied as the production effect. It is related to modality research, but it is not identical to hearing a complete list from an external source.