Long lists are extremely costly in the auditory channel

What it is

Long lists are expensive in audio: reaching item n takes time proportional to n, and until the end a listener cannot tell whether something better comes later. Moving the same menu from screen to speech keeps the item count and changes the cost by orders of magnitude.

Why it happens

Three costs stack. Playback time is the floor: ten items at two seconds is twenty seconds. Memory load follows: options must be held in mind to be compared, and working memory holds very few. Interruption is last: a lapse means replaying, because audio has no coordinate to rewind to. A visual list, by contrast, affords parallel comparison and re-scanning at a cost nearly independent of length.

Studying it

Auditory menu studies characterize list length through selection time and accuracy, typically finding time rising linearly with target position; providing jumps or grouping flattens the slope. Variables include list length, grouping cues, skip allowance, speech rate, and whether items are numbered. Beyond time, measure wrong selections and perceived workload, since failure on long lists often looks like picking arbitrarily rather than choosing slowly.

Where it stops holding

Below working-memory capacity—roughly four to five items—the gap narrows, because the whole list can be held and compared. If items are grouped semantically and the user cares about one group, cost tracks group size rather than total size. "Long lists don't work in audio" is thus more precisely: when many options must be compared, cost rises linearly with the candidate count.

Applying it

• Cap spoken lists at what can be compared in one pass; push the rest into groups or onto the screen.
• Number and group items while speaking so the user can stop the moment the target appears.
• Support direct selection and keyword jump instead of forced traversal.
• Verification: measure the time difference between choosing item 1 and item 10; if it approaches the summed playback time, list length is unhandled.

Related

• Within the group: D2.07.1 Sound unfolds in time and cannot be scanned at a glance · D2.07.3 Auditory interfaces need stronger structural cues
• Adjacent: D2.06.2 Speech output occupies the auditory channel and cannot be browsed in parallel · D2.06.1 Speech conveys meaning well but structure poorly
• Search terms: auditory menu · serial search · list length