psych 346 exam 2
Terms
undefined, object
copy deck
- What was Ebbinghaus's experiment?
- He created a number of lists, each containing 16 nonsense syllables. He would select a list and recite it rapidly for various #s of repetitions. A day later, he would find out how much of the list he learned by seeing how many additional trials he needed to re-learn it.
- What was the "total time hypothesis"?
- Learning obeys the rule that the amount learned depends on the time spent learning. "You get what you pay for".
- How did Ebbinghaus account for the fact that sometimes conscious recall fluctuates even though you may know something?
- He measured "savings in re-learning". This was the amount of time needed to re-learn material, rather than focusing on simple recall, because sometimes you know it but you just can't recall it.
- Describe how "constraint" helps memory and learning.
- Two important factors that help learning: redundancy and predictability. - Provides organization - Facilitates connections between what you already know and new info - Reduces the amount of "new" learning required - Saves on time and energy
- Describe the "cloze technique" and its significance.
- This involves giving sentences to subjects with missing words, and requiring them to complete it with first word that comes to mind. This is a way of measuring redundancy and predictability. High-cloze = most people give the same response (high constraint) Low-cloze = many different responses (low constraint)
- Describe higher vs. lower order constraints.
- Higher order constraints = structure or rules that reduce the number of possibilities for how information can be arranged, usually making it easier to remember.
- Explain the effects of distributed practice on explicit vs. implicit memory.
- Experiment involved doing abstract/concrete judgments for 182 words on Day 1. Then on Day 2, do a/c judgments on 26 NEW words. Then, also on Day 2, get into scanner and do a/c judgments for: - 26 from Day 1 (distributed) - 26 from Day 2 (massed) - 26 new words (no practice). Priming effects were better with massed practice (words learned that same day). However, explicit memory was better with distributed practice.
- What is the best way to distribute practice? (3 types of intervals, which are the best?)
- Inter-trial rest interval -- going through all material, taking a break, and repeating Inter-item repetition interval -- varying amount of time between chunks of info Inter-test interval -- varying cue-only judgments (weekly tests). The last two seem to influence rates of recall later on the most, because they lead to expanded retrieval.
- Explain the concept of "micro-distribution of practice".
- The best way to distribute items is such that you allow for varying gaps depending on how well specific information is recalled. More gaps should exist in between subjectively more difficult info, and less gaps in between easier concepts. This can be achieved by inter-item repetition lag.
- What are the possible reasons for the effects of distribution of practice?
- 1) Consolidation -- biological reason, maybe takes longer for LTP to occur 2) Multiple instances/contexts -- the way you learn it may be different from time to time (and makes more connections) 3) More attention -- easier to give more attention to shorter bursts than for constant repetition at increasingly longer periods of time
- What is one hypothesis that does NOT seem to account for the effects of distributed practice, and why?
- The inter-trial rest interval -- the idea is that maybe distributed practice works by reducing fatigue between trials. It was thought that fatigue has an effect, but it was proved that this is not true. Also, regarding verbal learning, inter-trial rest intervals have produced inconsistent results. BOTTOM LINE: This does not seem to be a strong contributor to distributed practice effects.
- What is the best way to distribute practice for: - Only studying - Only doing repeated testing - Doing both Why?
- If ONLY studying -- MORE distance is better. Distributed practice is better than massed practice. For testing -- successful retrieval practice is better. For long-term memory, testing yourself is better.
- What is the BEST way to distribute practice?
- Expanded retrieval seems to be the best way.
- What does Baddeley say about levels of processing?
- He says that LOP don't matter for implicit memory. He is only partially right.
- Describe the experiment involving priming/implicit memory.
- In task 1: - Words were categorized by 3 different types of judgments: size, shape, or in/organic judgments. In task 2: - All words required size judgments. - Used task 1 words or new words. Results showed that new words had slowest reaction times. Same kind of processing for same words was the best, and different type of judgment for same words was slightly better. Initial processing of stimulus DOES seem to matter.
- Describe classical conditioning and its effect on learning.
- Classical delay conditioning is effective and can even occur in amnesics. Note that backwards and simultaneous classical conditioning have NO PREDICTIVE VALUE.
- Describe the types of operant conditioning.
- Pos. reinforcement - giving someone a reward Neg. reinforcement - removing an undesired stimulus Pos. punishment - making someone do something they don't want to do Neg. punishment - removal of a desired object Note: Pos = adding something (good or bad) Neg = taking something away (good or bad)
- Describe the relationship between explicit vs. implicit systems.
- Sometimes they cooperate: - When at the beginning stages of learning a new complicated skill Sometimes compete: - When you get to the point where you have "automated" the skill and it is difficult to put into words.
- Explain the effects of arousal on memory (explicit or implicit?).
- There is an optimal level of arousal, and it most affects EXPLICIT memory. Explicit memory may be more likely to fail at EXTREME ends (very low or very high arousal). BOTTOM LINE -- Moderate arousal is generally the best.
- Explain the circadian influences on memory.
- Children and older adults are usually morning types. Young adults are usually neutral or evening types. This may matter more for some abilities (fluid, controlled) than others.
- Specifically, how do circadian influences affect different abilities?
- For fluid abilities (digit span and block design): - Makes a difference if you are a morning or evening person. For more crystallized abilities (like vocabulary): - Does not make much of a difference
- Describe the concepts of scripts and schemas.
- Schemas ar generalized pieces of conceptual knowledge used in understanding certain situations. - They meaningfully organize concepts. - They tell us what to expect and what we can infer. SCRIPTS - schemas for a particular event
- Describe the balloon study.
- The study involved giving a story involving balloons, which only made sense when accompanied by an appropriate illustration. ONLY the coherent illustration (as opposed to a partial illustration or no illustration) helped comprehension, and only when presented before the text. This indicates the need for organization and schemas in understanding otherwise potentially complicated concepts.
- What's in a script? What do they contain and what do we remember?
- Scripts and schemas contain info that is COMMON to most members of the category, but that is NOT shared with non-members. We remember those things shared by all members of the category, and for individual instances, we remember major deviations.
- What are 3 characteristics of skilled memory (based on PI guy, running times, and chessmasters)?
- 1) Encode meaningfully, using pre-existing knowledge. 2) Attach retrieval cues to structure based on pre-existing knowledge. 3) Process becomes faster with practice. These are not necessarily semantic tactics; can include chunking and rhythm.
- Describe the "War of the Ghosts" study, and its relation to schemas.
- Although schemas can be helpful, can also be misleading. People reading an odd story (War of the Ghosts) tended to make their recall of the story more coherent, shorter, and fit it more with their individual schemas.
- Organized material is ___________ but can also lead to ______.
- a) Easier to learn b) Errors Lists of words that are all related to one central word are recalled better than a list of unrelated words. However, the list with related words can lead to false recall of other similar words.
- Describe Type I vs. Type II processing.
- Type I -- Repeated processing at same level, can maintain things in attention, but not very good for LTM Type II -- Intentional and incidental learning - Incidental learning = not TRYING to remember it, but some processing happens, and type of processing influences whether it goes into memory - Intentional learning = learning knowing you will be tested on it, and using ANY coding structure that you want
- How does Craik criticize the claim that STM and LTM depend on the type of coding that is done (acoustic vs. semantic)?
- He says that you can show influence of semantic coding on STM: 1) Coffee, tea, juice, milk, soda 2) Shirt, note, window, pencil, hat 3) SHX, JWR, IFG, LMO, QPE These lists are harder to remember as you move down, showing that meaning/semantic coding are involved somehow.
- What are 3 possible reasons for why certain information fails to reach a semantic level of processing?
- 1) Nature of material - Many working memory experiments use materials that don't have much meaning inherent in them in the first place 2) Limited processing capacity available - You get lots of info quickly, and may not have time to do semantic analysis 3) Task demands - People don't usually try to do deep analysis for STM tasks
- What are Craik's views on capacity?
- He says the term "capacity" is unclear and not easy to define. Also, does the limitation refer to the storage capacity of memory or the rate at which the processing can occur? There are differences in capacity that need to be accounted for, but these differences are too flexible to be evidence for a multi-store model.
- The case in favor of multi-store models: dichotic listening studies.
- Subjects have two different streams of info coming in through each ear. Person had to pay attention to one ear, and experiment involved determining which info from the other ear was still processed. Broadbent proposed that info must be held transiently before entering limited-capacity processing channel. Items can be held over short term by recycling them through the same transient storage system. From there, info can be transferred into a more permanent LTS.
- What are the distinctions between sensory store, STS, and LTS?
- Sensory store -- "pre-attentive" since stimuli can enter without being paid attention to. It is modality-specific with a large capacity and very transient. Short-term store -- info is moved here from sensory store upon being paid attention to. Has a limited capacity, and a slower rate of forgetting than sensory store. Long-term store -- has no known limit, and forgetting does not occur or is very slow.
- The case against multi-store models: forgetting characteristics.
- If the different stores are real, each should have a particular length or duration, but many different experiments have shown different retention intervals for STM. Alternative: Aspects of the paradigm such as study time, amount of material presented, mode of test, etc influence retention intervals.
- What is Craik's idea about levels of processing and trace persistence?
- At preliminary stages of processing --> only perform analysis of physical/sensory features. At later stages --> you are matching input against stored abstractions from past learning. He suggests that trace persistence is a function of depth of analysis, and more meaningful and familiar stimuli are processed to a deeper level more quickly.
- Retention is a function of _____? What factors determine level of processing of a stimulus?
- a) depth 1) Amount of attention devoted to the stimulus 2) Its compatibility with the analyzing structures 3) Processing time available
- Describe the relationship between selective attention and memory storage (experiment that demonstrated this relationship?).
- If stimuli are only PARTIALLY analyzed or attended to, or processed only to peripheral levels, their record in memory is extremely fleeting. The shadowing experiments: involved playing the same prose passage to both ears dichotically, but staggered in time. When the unattended ear was leading, lag had to be reduced to 1.5s before subjects realized that they were the same. When attended ear was leading, lag could be 4.5s for subjects to realize similarity.
- What is Craik's explanation for the apparently obvious distinction between the STS and LTS?
- He says a "flexible central processor" can be deployed in different encoding dimensions, and can only deal with a limited number of items at a time. Items are kept in primary memory by continuous rehearsal at a fixed level of processing. The depth at which PM operates depends on the usefulness to the subject of continuing to process at that level.
- Acoustic errors will only persist if _______?
- a) analysis has not proceeded to a semantic level.
- How and why do "processing rate" and "unfamiliar words" have effects on ST vs. LT memory?
- Processing rate -- increasing processing rate prevents processing to those levels necessary to support LT retention, but does not affect coding operations of the kind necessary for ST retention. Unfamiliar words -- using unfamiliar words also inhibits processing rate from reaching a level necessary for LT retention, but does not affect ST retention.
- What is Craik's explanation for the serial position effects (primacy and recency)?
- Primacy: - Craik says that people realize that they must stop attending to the earlier words in order to perceive and rehearse subsequent items, so they subject these earlier items to Type II processing. - This reflects why primacy was associated with LTM. - Also explains why slowing presentation rate increases primacy Recency: - We're lazy! We can just rely on phonological info, so we'd rather do that. - We subject these words to Type I processing, explaining why recency was associated with STM.
- What is negative recency, and how does it affect Craik's explanation?
- Negative recency - if you come back the next day and ask people to recall the words, terminal items are recalled the worst. - This supports the prediction that terminal items are less deeply processed.
- What is the evidence for the fact that words early in the list undergo Type II processing? (think about interpolated tasks)
- If processing is interrupted, primacy is only disrupted if presentation rate is fast. If presentation rate is slow, interpolated tasks do not disturb primacy as much, because subjects still have time to engage in type II processing.
- Describe imagery/mnemonics and why it may be helpful?
- Some types of mnemonics are helpful: - method of loci - speeches - pegword system May be helpful because of DISTINCTIVENESS - More processing usually leads to better memory
- What are some problems with the LOP approach?
- There is some circular logic with "deep processing". Processing also isn't necessarily in sequential stages - Fluent readers seem to skip phonologic phase - Deep dyslexics can extract meaning but not sound
- Describe "elaborative processing".
- Elaborative processing is a better way of describing Craik's theory than "deep processing". It involves making more connections to semantic networks and schemas, and linking episodic details to each other and to abstract representations.
- Describe "transfer appropriate processing" (and encoding specificity).
- TAP -- processing a stimulus at the same level that will be required during recall. Encoding specificity -- memory is better the more that the information at test "matches" what is encoded Experiment: Showed 2 sets of word pairs: 1) Are they related? 2) Do they rhyme? - Then showed new words and asked if they rhymed with ANY of the words seen before - Performance was better for group who was asked if the original word pairs rhymed or not.
- Describe the idea of "consolidation".
- Consolidation is the process by which memories become more "solid" or long-lasting. - It is an idea that Craik is missing in his LOP theory.
- Describe the effects of alcohol on memory? (How is this related to consolidation?)
- Experiment: - All subjects listened two 2 stories - One group was then given vodka and OJ, and another group was given just OJ (with vodka odor) - Wait 10 minutes and then learn and recall 4 new stories ---For info that was both LEARNED and RECALLED after drinking, placebo group had advantage - Next day, both groups had to recall original 2 stories (learned before drinking) ---Alcohol group performed better ---This may be because the alcohol prevented intervening info from being consolidated, and therefore reduced interference
- What are methods for scanning through memory?
- Parallel -- scan each word at same time with same strength Serial self-terminating -- scan each item in succession, stopping when you reach the target Serial exhaustive -- scan entire list to the end even if you have already located the word
- Describe Tulving's concept of "encoding specificity".
- Encoding specificity: - What is stored depends on: --- What is perceived --- How it is attended to What is stored also determines what retrieval cues will be effective.
- Describe the "generate-recognize models" for retrieval.
- In these models: - A cue will activate many connections in your network - You mentally generate responses to the provided cues - You then recognize the response that was on the list Example: - If asked for fruit, generating "apple" will inhibit "pear" or "orange" - However, sometimes a "study list context" helps identify "orange" as the correct word, and then THAT inhibits "apple" and "pear"
- Describe "context effects" as they relate to cues and retrieval.
- If learning occurs in a distinctive context, recall is enhanced if it occurs in the same context. - Example -- If learning words underwater, recall is better underwater, and vice versa
- What are some problems with the conclusion that "serial exhaustive" search methods are employed when scanning memory?
- Experiments tend to indicate that subjects do not respond as soon as match is detected, but wait until the search ends. However, other features of performance on speed of retrieval tend not to fit with this idea: - When probe corresponds to last digit, subjects respond particularly rapidly
- Explain the difference between "direct access" vs. "retrieval by inference".
- In direct access --> the information is directly available - Ex: "What man's wife was turned into a pillar of salt?" - This type of info is likely to either be known automatically or not known at all In retrieval by inference --> subjects use other information they might have to infer the answer, even if it is not directly available - Ex: "What southern U.S. city is named after an ocean?"
- Describe state-dependent memory and mood effects.
- State-dependent memory: - A person's "state" can act as a cue for remembering information - It biases the search through a network; if you are mad, you are more likely to recall things that make you mad Mood dependency -- mood's effects on neutral material - Difficult to assess Mood congruency -- refers to material that is not neutral, but is emotionally toned. - Subjects that are depressed have difficulty retrieving pleasant memories.
- What is the dilemma that is created regarding distribution of practice?
- Obtaining the maximum spacing between presentations of material, to the extent that it will still be recalled, will enhance learning. However, successfully coming up with the correct answer on your own also strengthens learning and recall. The implications of this are the opposite of the above conclusion: the sooner an item is tested, the greater the probability that it will be correctly recalled, and hence greater probability that recall will be strengthened.
- What is the compromise solution for the distribution of practice dilemma.
- Solution is to use a flexible strategy in which a new item is initially tested after a short delay. Then, as the item becomes better learned, the practice interval is gradually extended, the aim being to test each item at the longest interval at which it can be reliably recalled.