Treisman's attenuation theory

Meaning: Treisman's attenuation theory is a model of selective attention. Like Broadbent's filter model, it puts the filter early, sorting input by physical features such as pitch or location. But the filter turns down unattended input instead of blocking it. The weakened signal still goes on to analysis for meaning. Each stimulus has a recognition threshold; personally important items, such as your own name, have low thresholds, so even an attenuated signal can reach awareness.

On the MCAT: Passages use dichotic listening and shadowing: a person repeats the message in one ear while a different message plays in the other. Noticing your own name in the ignored ear supports attenuation over Broadbent's model. Expect comparisons of three models: Broadbent (early, all-or-none filter), Treisman (early, partial filtering), and Deutsch and Deutsch (late selection, all input analyzed for meaning first).

Common confusion: Students label attenuation a late-selection theory because the meaning of unattended input can matter. Selection still starts early, on physical features, but the filter is leaky. The cocktail party effect is the observation the theory helps explain, not a competing model.