Download Midterm 2 Study Guide for PSYC 325 and more Study Guides, Projects, Research Cognitive Psychology in PDF only on Docsity!
Study Guide Answers Chapter 6 Behavioural Processes Define generalization and discrimination. What are the potential effects of each on our real-life decisions and behaviours? โ Generalization: transferring past experiences to new situations โ Can make us blind to circumstances that have changed โ Discrimination: perception of difference btwn stimuli โ Can waste time and effort learning new things we shouldโve already known โ Opens us up to new experiences Draw a sample generalization gradient graph (e.g., speculate the responses of a rat to a tone of a particular frequency to obtain a food reward). How does this graph represent a balance between generalization and discrimination? If the curve is narrow, is this more like generalization or discrimination? What if the curve is broad (wide)? โ The middle zone is what is generalized - the rat will respond to obtain the reward โ The tails are discriminated, so response declines โ If the curve is narrow, less is generalized so itโs more like discrimination โ If the curve is broad/wide, more is generalized so itโs more like generalization โ Adapt your generalization gradient graph in #2 into a discrete-component representation model. Does this model allow for generalization? Why or why not?
โ This model does not allow for generalization because each stimulus is presented individually, without the relationships between them Adapt your generalization gradient graph in #2 into a distributed representation network model. Does this model allow for generalization? Why or why not? โ โ This allows for generalization because the similarity between stimuli is explicitly represented by shared nodes Define discrimination training and learned specificity. Give an example that illustrates these concepts. โ Discrimination training: providing two difference consequences for stimuli initially treated the same โ Reward for 1000Hz and no reward for 950Hz โ Learned specificity: narrowing of generalization gradient โ Learning to differentiate between 1000Hz and 950Hz - the outcome of discrimination training
โ Acquired equivalence is when you make novel similar predictions based on prior similar consequences โ When 2 stimuli predict same outcome (become generalized), even if stimuli are dissimilar โ ex) girl, boy, man, woman, and the kind of fish they like โ This is a better example of generalization Define category learning and prototype. We can make inductive inferences based on prototypes: how can these be useful, but also sometimes inaccurate? See textbook pages 249-252, section โConcept Formation, Category Learning, and Prototypes.โ โ Concepts: psychological representation of a category of objects, events, or people in the world โ Dog - what we understand about dogs โ Categories: division or class of entities in the world โ Dog - subdivision of mammals โ Category learning: establishing memory trace that improves efficiency of assigning novel objects to contrasting groups โ Categories enable us to make inferences about members of categories and guide us in predicting the future โ Prototype: the typical or model/ideal representation/version of a category or concept โ Concept formation vs discrimination learning โ Discrimination learning: paradigms where there are a relatively small number of stimuli, stimuli are simple idealized and often unnatural images/sounds in lab, and discrimination well defined โ Concept formation: many stimuli, naturalistic, varying/complex, ill-defined distinction โ A category can help us make inductive inferences about things belonging to the category โ Can be inaccurate due to faulty inverse reasoning โ All zebras are black and white; are all black an dwhite animals zebras How are generalization and specificity related to the idea of stereotypes? How can stereotypes be useful? How can they be misused and potentially harmful? โ Stereotypes: generalizing information about a group to an individual โ Stereotypes can be useful in determining important info you may need to know, but they can be potentially harmful if the person does not match up to the โprototypeโ of the category Brain Substrates Explain how neuron receptive fields typically allow for some degree of generalization. โ neuron receptive fields are tuned "broadly" - responses are graded โ shape of response curve similar to that of generalization gradients observed in behaviour
โ distributed representation models can explain broadly tuned receptive fields (in addition to behavioural generalization) โ bottom layer: sensory neurons in ear, middle: auditory cortex neurons, output: motor cortex neurons โ โ Neuron receptive fields: sensory cortices can reorganize in response to training with particular subset of stimuli Describe how discrimination training can affect neuron receptive fields โ and what conditions are necessary for changes to occur. How are the nucleus basalis and amygdala involved? What are the general effects of habituation and classical conditioning on sensory representations in the cortex? โ Discrimination training can lead to cortical remapping โ Habituation to a tone contracts cortical space dedicated to pitch โ Fear conditioning to tone expands cortical space devoted to that pitch โ Conditions - receptive fields only altered if input is โ meaningful โ tone predicts food or shock โ no altered fields w mere exposure โ Acetylcholine release (from nucleus basalis) occurs when stimulus is salient (determined by amygdala) โ Determines whether stimulus will cause cortical remapping Describe experimental evidence that the hippocampus is necessary for appropriate generalization to relevant stimuli and relational memory. How does the role of the hippocampus in sensory preconditioning and acquired equivalence demonstrate its capacity for relational memory? โ Hippocampus is active in early training, but then its role declines โ Hippocampus damage disrupts performance in โ Sensory preconditioning tasks (two stimuli occur together) โ Acquired equivalence tasks (generalizing between similar people)
โ After 16 days, had no effect โ Rats learn first using place strategy (hippocampus), then shift to response/habit learning (caudate/striatum) โ If one type of info is lost, animal can use the other (parallel storage) Draw the taxonomy of long-term memory, including both declarative and nondeclarative memory systems and the brain regions that support them. โ Explain what McDonald and White (1993) found about the roles of the dorsal striatum, hippocampus, and amygdala in learning and memory using rats and the 8-arm radial maze. Describe the win-shift, win-stay, and conditioned place preference tasks. โ Win-shift (place/relational learning - hippocampus) โ Rats placed in maze w food hidden at ends of all arms โ At peak learning, rat enters each arm once to obtain food - no repetitions โ Error = entering arm from which food already eaten โ Flexible place learning impaired by fornix lesions (disconnects hippocampus) โ Info about relationships among stimuli and events โ Win-stay (response learning - dorsal striatum) โ Lights signal locations of 4 food rewards โ Rats can enter twice to get food, then light turns off โ Rats only need to learn light-food association โ Response learning impaired by dorsal striatum lesions โ Formation of reinforced stimulus-response associations โ Conditioned place preference (emotional learning - amygdala) โ Only 2 out of 8 arms open for use - one lit, other not โ Either lit of dark always had food, other arm not baited โ Test: no food in either arm, how much time rats spend in each arm โ Preference for baited (paired) arm impaired by amygdala lesions โ Learning based on biologically significant events w affective properties - liking, hedonics
Why is function of the striatum altered in patients with Parkinsonโs disease? โ Degeneration of substantia nigra - loss of input to striatum (altered function) Describe how the weather forecasting game (Knowlton et al., 1996) works and for what type of memory it tests. How can declarative memory also be tested using this paradigm? โ Different cards are associated w sunshine percentage โ Participants shown 1-2 cards at a time, made prediction abt rain or shine (avg probability of all cards determined outcomes) โ Participants given feedback - right or wrong โ Controls and amnesiacs had better nondeclarative memory than those w Parkinsonโs disease โ There was also a debriefing test โ Everyone except amnesiacs had good declarative memory โ Medial temporal lobe memory system (declarative memory) and striatum/caudate + putamen (gradual learning of associations - habit learning) separate and parallel memory systems in humans โ Can learn smth without being aware you learned it What deficits were found in amnesic patients and Parkinsonโs patients in the weather forecasting game (Knowlton et al., 1996) and what does this show about the functions of the medial temporal lobe and the striatum? How do these data support the surprising conclusion that one can learn without being aware of it? โ Parkinsonโs patients didnโt have good nondeclarative memory โ Medial temporal lobe memory system (declarative memory) and striatum/caudate + putamen (gradual learning of associations - habit learning) separate and parallel memory systems in humans โ Can learn smth without being aware you learned it โ People with Parkinsonโs had loss of input to striatum, but not to medial temporal lobe โ Impairments in nondeclarative memory but not declarative โ Learn using nondeclarative memory, not declarative Describe the patients studied by Bechara et al. (1995) using auditory and visual classical conditioning. What were the neutral stimuli, unconditioned stimulus, unconditioned response, conditioned stimuli, and conditioned response in this study? How was declarative memory also tested? What did the findings show about the functions of the medial temporal lobe and amygdala in separate memory systems? โ NS: coloured cards and tones โ US: boat horn sound โ UR: skin conductance response โ CS: specific coloured card and tone โ CR: skin conductance response โ Debriefing to test for declarative memory
consolidation, declarative memory, nondeclarative memory, anterograde amnesia, retrograde amnesia, conscious awareness. โ He had severe epilepsy, so the surgery was for bilateral removal of medial temporal lobe structures (hippocampus, amygdala, and entorhinal cortex) โ Only bilateral lesions result in severe amnesia โ He remembered everything up until ~2 years prior to surgery โ Could follow convos and watch TV shows, but couldnโt encode new long-term memories โ Short-term memory was intact โ Because he couldnโt transfer new info from short-term memory to long-term memory, his consolidation process was impaired โ However, some types of lasting memories could be formed โ Priming โ Procedural โ Skills โ Done without conscious awareness โ His declarative memory was impaired (impaired consolidation), but nondeclarative was still functional โ He had a combination of anterograde and retrograde amnesia โ Couldnโt remember things before the surgery (2 years) - retrograde โ Couldnโt form new memories - anterograde Define these memory types: declarative , nondeclarative , explicit , implicit , episodic , semantic. Provide two examples of each โ Declarative/explicit/verbal โ events (episodic memory - I remember) โ facts (semantic memory - I know) - no associated context of learning โ relationships (relational memory, being able to describe the relationship) โ Nondeclarative/implicit/nonverbal โ procedures/skills โ priming โ conditioning โ habituation โ sensitization Compare skill (procedural) memory with declarative memory with regard to typical means of communication/demonstration, conscious accessibility, and practice/exposure required for acquisition. Then, compare episodic with semantic memory with regard to content for context, type of information stored, and optimal number of exposures/events for memory strength. Lastly, comment on the relationship between episodic and semantic memory โ does one precede the other in all situations? โ Skill memories โ Difficult to convey except by direct demonstration โ May be acquired without awareness
โ Require several repetitions โ Declarative memories โ Can be communicated flexibly, in different formats โ Have content that is consciously accessible โ Can be acquired in single exposure โ Episodic - I remember โ Memories for specific autobiographical events โ Can be communicated flexibly - in a format other than that in which it was acquired โ Consciously accessible (you know that you know) โ Tagged with spatial and temporal context โ You must have experienced the event personally โ Learned in single exposure: can be weakened by exposure to similar events โ Semantic - I know โ Memories for facts and general knowledge โ Can be communicated flexibly - in a format other than that in which it was acquired โ Consciously accessible (you know that you know) โ Not necessarily tagged with spatial or temporal context) โ Can be personal or general information โ Can be learned in a single exposure but can also be strengthened by repetition โ provide context necessary for episodic memories to form โ can't remember graduation ceremony if you don't know what graduation is โ can develop from repeated episodic memories - encounter same info across multiple contexts, the episodes blend together for semantic memory, lose info that differentiates btwn episodes/learning events How long can long-term memory potentially last? Briefly describe the findings of a study that investigated memory for names/faces of high-school classmates. โ Long-term memory can potentially last around 40 years or more โ The yearbook study found that visual info is retained at least 35 years, and there was no decline in verbal info for first 15 years โ However, after 47.5 years, there was a big drop in memory of classmates - probably due to degenerative changes Compare episodic with semantic memory with regard to content for context, type of information stored, and optimal number of exposures/events for memory strength. And, comment on the relationship between episodic and semantic memory โ does one precede the other in all situations? โ Semantic memories provide context necessary for episodic memories to form โ can develop from repeated episodic memories - encounter same info across multiple contexts, the episodes blend together for semantic memory, lose info that differentiates btwn episodes/elearning events
โ Difficulty remembering old number you once knew well after learning new number โ False memory (false confidence, false detail) โ Memories can be modified and manipulated after encoding โ False confidence: most are confident of false memory โ False detail: report specific details about false memory (position on word list) โ Tip-of-the-tongue (textbook p. 288) โ Challenged to retrieve a bit of info (name of film star, answer on test question) and feel like we know the answer, even if we canโt retrieve it at the moment โ When we hear answer, we recognize it โ Different from automatically knowing you donโt know the answer, like atomic number of Boron โ Metamemory โ Feeling of Knowing (formal name) โ Judgment of Learning: judgment during learning of whether info has been successfully acquired โ Also inaccurate, overconfident โ 4 common problems of memory failure โ Forgetting โ Interference โ Misattribution โ False memory Memory consolidation โ explain: โ Ribotโs law โ Temporal gradient โ Amnesia tends to be strongest for most recent events โ Amnesia can differentially affect STM, working memory, or LTM abilities โ Recent memories = more chance of issues with encoding and storage (forgetting curve) = more likely to be lost โ Effect of electroconvulsive therapy/brain injury on recent vs. distant memory โ Items that are lost appear to be those that have undergone an initial rapid consolidation but havenโt yet completed slower permanent consolidation process โ Impacts memories (retrograde for months before or anterograde months after)
What is memory reconsolidation? What are its implications for eyewitness testimonies? How might it suggest new forms of therapy to help deal with the psychological effects of traumatic events? โ Each use of memory may make it fragile again, requiring reconsolidation to recreate permanent storage โ Allows updating memory w new info, which may or may not be accurate (eyewitness testimony implications) โ Suggests reactivating memories may provide a chance to alter them (phobias, PTSD) Brain Substrates Describe Penfieldโs findings on memory using electrical stimulation as described by Milner (1977). โ Cortical stimulations can evoke simple sensations and complex memories Is it likely that you have just one neuron to store your memory for each celebrity you are aware of, or each person in your life? Explain. โ No - likely, there are other neurons that fire to that person, and the neuron may also fire to more than just the person Describe how memory for an object such as a watermelon or a car might be stored across a number of different cortical areas. โ The memory of a watermelon or car can be stored over many specialized processing centres in the cortex โ Visual components (colour of watermelon or car) in visual cortex โ Olfactory components (smell) in olfactory lobe โ Associations with other fruits or car models/vehicles in associative cortex What brain areas does it seem consolidation relies upon? How do we know this? What brain area seems to be involved in sorting out what should instead be forgotten? โ Consolidation seems to rely upon the medial temporal lobes (hippocampus and surrounding cortex - entorhinal, perirhinal, and parahippocampal)
Textbook question Discriminating between generalization paradigms
- Elizabeth impressed by men who bring either candy or flowers. If men bring both on first date, she is turned off, feeling he is coming on too strong. a. Negative patterning b. Candy -> good c. Flowers -> good d. Candy + flowers -> bad
- As a child, Samson learned people who have deep voices also tend to have beards. He later becomes convinced that men with beards are strong, and he inferred that a deep voice was also likely a sign of strength. a. Sensory preconditioning b. Phase 1: deep voice + beard c. Phase 2: beards -> strong d. Phase 3: infer deep voice -> strong
- By playing snippets of music by Brahms, then Schubert, then Brahms again, a music teacher is able to teach class how to recognize style of each. a. Discrimination training b. music1 -> Brahms c. music2 -> Schubert
- Mark and Kaori enjoy many same foods and people. Based on this observation, Mark guesses Kaori will like particular song that he also likes. a. Acquired equivalence b. Mark -> food c. Kaori -> food d. Mark -> music e. Infer Kaori -> music
