PSYC208 cognitive psychology exam

Mental imagery

The processing of perceptual information in the absence of an external source for the perceptual information

Pylyshyn - propositional representations

Pylyshyn argued that depictions in the head are verbal/sentence-like/meaning-based, rather than having pictures in the head.

Proposition

A language based representation of imagery

Depiction

A perception based representation of imagery (endorsed by Kosslyn)

Kosslyn - boat experiment

RT when people had to move attention from one part of imagined boat to another - RT increased as a function of distance between boat features (however propositions can also explain this bc of time taken to traverse links)

Santa (1977) shapes and words expt

Identification quicker in shape condition when shapes preserved original identical layout. Identification quicker in word condition when words had a linear layout. This shows verbal and visual images are processed differently.

Brooks (1968) corner scanning expt

Mentally scanning a visual array was quicker when P did it vocally rather than by pointing at a physical array. Mentally identifying nouns in a sentence did not produce this result. This supports the idea that mentally scanning a picture uses similar processes to scanning a physical array

Shepard & Metzler - mental rotation

He proposed that it is an analogue to perceived rotation because there was a linear increase in RT with increasing angular disparity for both picture and depth planes.

Motor areas become active in the frontal lobe in mental image rotation when...

...you imagine manipulating the objects yourself, or you imagine moving part of yourself. Parietal regions also activated.

Imagery

Imagining a mental image

Perception

Seeing an image externally

Imagination and perception both...

...activate similar regions in the brain (92% overlap). These areas include the frontal/parietal area. Perception activated more of the temporal/occipital area.

O'Craven & Kanwisher (2000) had people...

...alternatively view or imagine houses and faces. Parahippocampal place area and fusiform face area activated in both conditions, however weaker activation in imagery condition.

Deactivating imagery areas with TMS...

...impairs information processing, showing that imagery does play a causal role in mental imagery.

What happens when people try to reinterpret mental images?

They cannot reinterpret them (e.g. duck-rabbit) unless prompted.

Estimated size comparisons between objects...

...increase in RT as the size difference between objects decreases. True for imagined objects (animals expt) and perceived objects (line expt)

Imagery activates...

...the early topographically organised areas of the visual cortex (e.g. the fovea). More activation when people imagine a large letter A compared to a tiny letter A.

Spatial images

Represent where an object is, activates parietal lobe

Visual images

Represent the details of an object e.g. colour, shape, activates temporal lobe

Neuropsychological findings which suggest it is necessary to distinguish visual from spatial images

Parietal lobe processes spatial images, whereas temporal lobe processes visual images. Damage to parietal lobe meant p's couldn't identify where object was but could describe its features. Damage to temporal lobe meant p's couldn't describe features but could identify where it was.

Location mistakes

P's use info about large landmarks to figure out where 2 smaller locations are. Relative positions of large landmarks lead to errors in geographical judgments (e.g. no North island towns are west of CHCH)

Route maps

Action information (e.g. turn left onto Birdwood Ave, turn right onto Malcolm Ave)

Survey maps

Spatial information (e.g. my neighbourhood)

Egocentric representation

Space as we perceive it (parietal)

Allocentric representation

Physical maps (hippocampus)

Sperling's partial report method

P had to remember one row of letters out of 3 rows (cued afterwards with a sound as to which row to remember). P's could recall all 4 items immediately, but as time lengthened, people recalled less items. This shows visual info is stored in our STM, however decays within 1s.

Visual sensory memory/iconic memory

A system which holds all visual information

Auditory sensory memory/echoic memory

A system which holds all audio information

Goldman-Rakic single cell recording of location perception. What do single cells do during cue, retention, and recall of a location?

Different single cells are activated during cue, retention, and recall. During the delay/retention period, the cell associated with the location will fire rapidly, however will pass activation to another cell when location must be recalled.

Oculomotor delayed response task

Name of response task used by Goldman-Rakic involving location cuing and single cell recording in monkeys

Single neuron recording

The only method currently capable of dissecting the neural elements involved in WM processes

What brain region is consistently activated in WM tasks?

The prefrontal cortex.

Neural correlate of WM

Some neurons in PFC maintaining activity during delay period (allowing organism to keep info in memory)

Oculomotor delayed response task - successful saccades to a target location correlates with activity of neurons. Evidence?

Correct response shows more activity in neuron which corresponds to location, whereas incorrect responses show little activity in this particular neuron

Activity of brain cell that shows preference for a particular item (e.g. cup) that had been cued on the target trail, during the delay period when it is no longer present

Neurons that initially responded to the cup show less activity, but some activity is maintained indicating cup is preserved in memory. Furthermore, if cup becomes ineffective stimulus, a spike of activity in the neuron is still observed however quickly decays so a correct saccade can be made

Atkinson & Shiffrin's multi-store model of memory/modal model of memory

Sensory memory, STM/WM, LTM. Proposes that to get info into LTM, rehearsal in WM is needed. More rehearsal = higher chance of getting info into LTM

Levels of Processing theory

Suggests that the deeper/more meaningful the rehearsal of info is, the more likely it is to get into LTM.

Precise elaboration

These interrelate concepts in sentences, making them easier to remember, e.g. the tall man bought the crackers that had been lying on the top shelf

Imprecise elaboration

These leave concepts in sentences arbitrary making them harder to remember, e.g. the tall man bought the crackers from the clerk in the store.

Meaning determines...

...memorability of info

Outcome for training "less successful" children

"Less successful" children were less likely to generate precise elaborations, thus couldn't remember as much. However, when trained to produce precise elaborations, there was a rapid increase in generation and recall

Serial exhaustive account of STM scanning

Predicts that RT for identifying a stimulus increases as the size of the memory set increases regardless of Y/N response, bc all items are considered

Serial terminating account of STM scanning

Y responses produce faster RT than N because search is terminated as soon as item is identified

Unlimited capacity parallel processing account of STM

Everything is considered at once therefore RT should be fast

Sternberg STM scanning expt results

It took an additional 40ms to make response for each item in memory set, meaning that each item must've been individually scanned before making a decision, supporting serial exhaustive search.

Limited capacity parallel processing

Everything in STM considered at once, however more items: more load, therefore more items should produce slower decision time.

Difference between 2 sets of words used in Sternberg's STM scanning expt that supported ACT*.

One listed contained arbitrary words e.g. Jet, sled, cat, pineapple. Other list contained related words e.g. study, bored, stress, thisismycryforhelp. Related words should produce faster RT according to ACT whereas both lists should produce same RTs according to serial exhaustive. ACT supported because related words produced faster RTs.

Proposition

The smallest unit of knowledge that has a truth value

Fan effect

The increase in RT related to an increase in the number of arbitrary facts associated with a concept. Due to the limited capacity nature of spreading activation. E.g. "the lawyer is in the church" and "the lawyer is in the park" = lawyer should produce longer RT as it has more associations and spread of activation is going to 2 locations therefore is weaker (and greater activation = faster RT). Since park only has one association, RT will be faster.

Memory unit that serves as the interface between declarative knowledge and procedural knowledge according to ACT*

Working memory

Episodic memory

Knowing how - episodes/events

Semantic memory

General knowledge/facts (knowing that)

Info remembered as a function of retention duration

We lose studied info over time (well shit), however testing gives us a retention advantage (good thing I'm making these flashcards then)

Spreading activation in relation to memory function

The process by which currently attended items can make associated memories more available

Decay theory of forgetting

The idea that memory traces decay with time

The interference theory of forgetting

The idea that maintaining multiple associations is difficult and will interfere with other items in memory

Phonemic restoration effect

An unknown word in a sentence becomes disambiguated depending on later context in the sentence

Episodic memory

The memory for episodes and events

Semantic memory

The memory for facts, ideas, and concepts

Maintenance rehearsal

Involves repetition of info w/o thinking about it

Elaborative rehearsal

Remembering info in such a way that is meaningful and relates to/expands on other info already in LTM

Variable that different affects explicit and implicit memory

The context of a word. If presented with an antonym, it is remembered better in explicit memory, especially if participant generated it. However implicit memory better when word presented w/o context/antonym

Elizabeth Loftus' expt for constructive (erase and update) memory

P's shown video of car stopping at stop sign. Later were asked if car stopped at give way sign. P's ended up remembering give way sign. Misinformation about something that occurs can therefore change our memory of it.

Implicit memory

Unconscious

Explicit memory

Conscious

Follow up study to Loftus - chronological order of slides made misleading effects disappear. This supports...

...the encoding specificity principle

Encoding specificity principle

Successful retrieval depends on the degree to which the retrieval environment re-instates the encoding environment.

State-dependent learning

Memory performance better when we are tested in the same physical and emotional state as when we learned info

Word recall and word completion in Amnesic patients

Word recall poor (explicit memory), word completion good (implicit memory)

Retrograde amnesia

An Amnesic patient who cannot recall memories prior to the trauma

Anterograde Amnesia

An Amnesic patient who cannot learn new memories but can recall old ones

Features of HM

Stuck in time - memories from up to 16 years of age. Could not form new memories or remember people he had been working with for years. However, intelligence intact and had WM. Was about to implicitly learn things like the mirror tracing task.

What happens to brain activity as one becomes better at something?

Less metabolic expenditure - brain activity decreases

Behavioural plasticity

A remarkable and unique feature of the human species - our ability to acquire expertise at things which are not anticipated in our evolutionary history

Cognitive stage

- Declarative encoding of the skill

- Reliance on retrieval of verbal instructional information

- Slow, error prone

- Effortful extensive brain activation

Associative stage

- Errors are gradually detected and eliminated

- Connections among elements required for successful performance are strengthened

- Development of a successful procedure for the task which

is relatively free of declarative facts.

Autonomous stage

- Performance becomes faster with fewer errors

- More automated requiring fewer processing resources

- Task sharing and multitasking possible

- Recovery from disruption

- At high levels of skill may lose the ability to verbalise

Parahippocampus

Does not respond to false memories because it preserves original sensory info

Hippocampus

Responds to both false and true memories with high activation

New knowledge stored in...

...hippocampus

Old knowledge stored in...

...cortical structures

Basal ganglia

Supports procedural learning

The Power Law of Learning

Describes the diminishing returns relationship between performance time and practice (at first there is rapid learning, but it takes more practice to get better.)

Procedural knowledge retention

Skills show high retention over time, however some skill is lost. Retraining fits power law of learning to fit pre-break levels, but it takes 1/4 of the time to relearn skill than when it was initially learned

Proceduralised knowledge

Knowledge which enables one to perform a task without thinking about it directly (no declarative knowledge needed)

Tactics/tactical learning

A method which accomplishes a particular goal, the process by which people learn specific procedures for solving specific problems

Strategies/strategical learning

A process by which people learn to organise their problem solving (to capitalise on the general class of a structure of problems)

Perception of problems

Novices and experts perceive problems differently. Experts perceive them in a way that allow deeper/more effective procedures to apply, whereas novices just tend to map surface features

Pattern recognition in experts

Experts tend to have chunks of meaningful relations remembered (e.g. 50,000 positions in chess masters) and they know what to do in the presence of such patterns, allowing for quick judgments. This shows we are good at pattern learning

"I see only one move ahead, but it is always the correct one"

Chess master said this. Refers to how he perceives meaningful patterns on the board and knows how to act on them based on lots of previous experience.

LTM in chess masters

Expert chess players have an increased capacity to store info about the domain (but only for chess). They store larger patterns and more of them due to the development of their retrieval structures which allow for superior recall of past patterns

Characteristics of deliberate practice

- Motivation to learn vs play around

- Informative feedback is supplied

- Performance is continually monitored against correct performance

- Learners focus on eliminating differences between current performance and the ideal

Positive transfer

Occurs to the extent that domains use the same facts, rules, and knowledge. Does not often occur

Negative transfer

When learning one skill makes a person worse at another skill. This is rare.

Theory of identical elements - Thorndike

Skills transfer to the extent that domains have identical situation-response elements. Too specific.

Intelligent tutor

Guarantee deliberate practice by carefully monitoring individual components of a skill and provide feedback on learning.

Productivity in linguistics

The infinite number of utterances that are possible

Regularity in linguistics

The symbolic nature of utterances

In terms of language, behaviourists could not explain

1. Departure from contiguity (embedded sentences) such as "the girl, the boy kissed, ran away". 2. The existence of grammatical but meaningless sentences "colourless green ideas sleep furiously"

Chomsky's phrase structure grammar

A descriptive account of language that represents people's implicit or tacit knowledge of language - it's goal is a complete grammar which must generate all possible sentences & no non grammatical sentences