|
- Broadbent (1958)
- Stimuli analysed in terms of coarse physical features
- These are low level auditory characteristics e.g. pitch, loudness...
- Stimuli filtered early on the basis of these.
- If it's not processed by now, won't be passed on any further.
- Processed info receives further processing to identify meaning and determine an appropriate response.
|
- Triesman (1960)
- Stimuli analysed in terms of coarse physical features
- Unwanted stimuli will be attenuated (turned down), but still processed for meaning
- All stimuli analysed semantically
- Stimuli matched to memory to determine priority.
- Permanent high-priorities: danger signals/own name
- Current high-priorities: friend's voice
- Stimuli filtered late on basis of semantic analysis (priority)
- Stimuli not passing through this not processed any further
- Stimuli passing through receives further processing to determine an appropriate response
|
|
- Evidence: Unattended speech (Cherry & Taylor, 1954) dichotic listening task with 2 different messages in ear, told to focus on one ear and repeat aloud. Reported only low-level info e.g. gender of speaker, suggesting filter is before information gets processed for meaning.
- Criticism:
- Cocktail party: Doesn't explain why selected attention can be suddenly interrupted by someone calling your name.
- Moray (1959) in 34% of trials, people noticed own name in unattended ear, meaning some information must go beyond simple analysis of physical features as ps responded to the meaning of a word. Wood and Cowan (1995) found the same (35% of trials).
|
- Evidence: cocktail party
- Triesman (1960) - story 1: attended ear, story 2: unattended ear. Story switched, story 1 into unattended ear and story 3 introduced to attended ear. Ps continued with story 1 even though it required them switching their attention, so must have had knowledge of info in unattended ear as otherwise wouldn't have known story was continuing in that ear. Therefore, both stories must have been processed for meaning, just info in unattended ear was attenuated ('turned down').
- Criticism:
- Doesn't explain how info is attenuated or process of semantic analysis.
- Dichotic task can't guarantee ps are listening to only one ear, could be switching attention between the two.
|
Comments
No comments have yet been made