Driver + Sternberg 2

Cocktail party problem and how it was studied

Picking out one relevant sound source (a conversation) among many simultaneous sounds; studied by playing two messages at once and having listeners repeat ('shadow') only one

After selective shadowing, what can and can't people report about the non-shadowed message?

Can report simple physical properties (sudden pitch change, a loud tone, message ending); can't report content, topic, repeated words, or a change of language

What does Broadbent's filter theory (1958) propose?

Stage 1 extracts physical properties of all inputs in parallel; a selective filter passes only inputs with a chosen physical property to a limited-capacity Stage 2 that extracts abstract properties (identity, meaning)

What is the methodological problem in assessing processing of unattended information, and what is the solution?

Problem: surprise retrospective questions may reflect forgetting, not lack of processing; asking directly makes people attend. Solution: indirect, on-line measures (effects on responses to related attended info, or autonomic/neural responses)

What is Treisman's attenuation theory, and what is the key evidence?

The filter weakens (attenuates) rather than blocks unattended input; weak input usually can't support identification but can for words with a low threshold (primed by context, or personally significant like your own name). Evidence: in shadowing with swapped messages, people switched ears to follow the sentence context ('posted the... LETTER')

What did Underwood (1977) add about priming from unattended messages?

Single unattended words can prime, but priming from extended sentence context occurs only for attended messages, so whole sentences are integrated only when attended

What are the parallels to early vs. late selection in visual attention?

Sperling's partial report (iconic memory) resembles a filter passing relevant items, working best with simple physical cues; Rock & Gutman found poor memory for unattended overlapping shapes (early); Tipper's negative priming and flanker/Stroop-type interference suggested unattended items are processed (late)

How does Lavie's perceptual load theory reconcile early and late selection?

Perception is automatic (like late selection) but capacity is limited (like early selection). Low load: spare capacity spills over to distractors, so more distractor processing/interference (late-like). High load: capacity is exhausted by the target task, so little distractor processing (early-like)

What is extinction in neglect, and why is it linked to attention?

Patients with parietal damage detect a single contralesional stimulus but miss it when a stimulus appears simultaneously on the ipsilesional side; it resembles an exaggerated normal 'two-target cost' (biased competition for attention)

What is known about processing of neglected information?

Neglected stimuli can still influence responses to ipsilesional stimuli without awareness (their presence, nature, even semantics), mirroring indirect-measure findings in normal attention; this may inform rehabilitation and the neural basis of awareness

How does attention modulate sensory neural responses (animal single-cell and human ERP/fMRI evidence)?

Responses to a stimulus are larger when attended than ignored, across much of the visual hierarchy, mostly attenuation rather than elimination, which is problematic for strict late selection and consistent with Treisman. ERPs show attention effects within ~100 ms at occipital sites; fMRI/PET show task-specific modulation (e.g. colour vs motion areas)

Define knowledge representation, declarative knowledge, and procedural knowledge.

Knowledge representation = the form in which what you know about the outside world is stored in the mind; declarative = knowing THAT (facts that can be stated); procedural = knowing HOW (procedures, e.g. tying shoelaces)

Compare analog (depictive) vs. symbolic (propositional-style) codes.

Analog: resembles what it represents, conveys concrete/spatial info all at once, no arbitrary rules (e.g. a picture of a cat). Symbolic: arbitrary link to what it represents, follows rules, conveys info sequentially, good for abstract/categorical info (e.g. the word 'cat', 'justice')

What is the functional-equivalence hypothesis?

Visual imagery is not identical to visual perception but is functionally equivalent: the two use about the same operations to serve about the same purposes, so images are analogous to the percepts they represent

What neuroimaging evidence suggests mental imagery/rotation uses the same areas as visual perception?

fMRI (Cohen et al., 1996) found the same brain areas involved in perception are involved in mental rotation. Early visual cortex is activated when imagery requires high-resolution 'seeing' (e.g. shape of an animal's ears) but not for spatial tasks, and motor areas are involved when you imagine moving objects yourself

What does lateralization of function suggest about analog vs. symbolic codes?

Left hemisphere (right-handers): language/verbal/symbolic knowledge. Right hemisphere: visuospatial/analog knowledge, manipulated much like perception (from split-brain and lesion studies). Corballis (1997) qualified this: both hemispheres contribute, and apparent right dominance may reflect left-hemisphere functions being overshadowed by language

What is the dissociation between visual and spatial imagery

Visual imagery represents appearance; spatial imagery represents location, distance, orientation. L.H. could see and copy drawings but couldn't recognize them and was very poor on visual imagery tasks, yet was normal on spatial imagery tasks. So the two kinds of imagery are dissociable.