P2 Lecture 7: Levels of processing Flashcards
What is Sternberg’s Scanning Task?
- -how long it takes to do certain things in STS (time to make a mental comparison)
- -scanning task: people were given a set of numbers that they had to keep in STS for a short period of time. After a 2 second pause, they were shown a probe item that was a number. This probe item either matched one of the items in the memory set or did not match. The participants had to respond with a yes/no button (memory search task). Different memory sets would have different number of items (6 item list or 1 item list or 3 item list) for every trial, but always only one probe item
- independent variable is memory set size and dependent variable is RT
What are the stages of a memory search/scanning task?
- Time between a stimulus (ex. probe item) and a response is occupied by a series of DISCRETE STAGES (meaning first stage does its full job before passing the information to the next stage)
1) Probe digit
2) Encoding stage: the system first has to encode the probe stimulus, it does its full job of encoding before passing the information to the memory comparison stage
3) deciding whether probe matches the memory set, then passing the information (yes or no) into the response output stage
4) motor response where you press the yes/no button
- Time between a stimulus (ex. probe item) and a response is occupied by a series of DISCRETE STAGES (meaning first stage does its full job before passing the information to the next stage)
- thus the RT (dependent variable) is how long it takes to go through all these three discrete stages
What are the three hypothesis regarding the memory comparison stage?
1) Parallel
- - very efficient
- - compare the probe stimulus (ex. 3) to all of the items in the memory set at the same time
- - the memory search time and RT will not change depending on how many items there are in the memory set
- - horizontal line in a graph with RT in y-axis and # items in memory set
2) Serial Exhaustive
- - you compare each one of the items in the memory set to the probe one at a time all the way to the end of the list (doesn’t matter if probe is there in the list or not, you go though every item all the way to the end)
- - the search/RT will be longer depending on the number of items in the memory set
- - positive slope line in a graph
3) Serial Self-Terminating
- - search every item in the list one at a time (just like the serial exhaustive), but unlike it if you get a match, you terminate the list (you stop checking). If there is no match, you keep going through every item until the last item in the list
- - in both the yes and no trial, the time it would take to match the probe with the first item is the same. But to match the probe with the subsequent items, in the yes trial, the amount of time it takes will be an average of the number of items in the list
- - two lines in a graph with the yes trial line lower (less RT) than the no trial line
What hypothesis does the Sternberg scanning task support?
– supports the serial exhaustive list
How long does it take to encode and respond in a scanning task?
- Without a mental comparison (no memory set), it takes 397.2ms to encode and then respond
- The time it takes to make one mental comparison (match probe to one item in the list) is 37.9ms
What is the Additive Factors Logic & Modelling
- same task as Sternberg
- -the probe item will be presented to people in one of two ways; clear or degraded (the quality of the information coming in is altered)
- -figure out if the degraded probe item is dealt with in the initial encoding, or during the memory comparison stage
- If the degraded item is dealt with in the encoding stage (stimulus is ”cleaned up” to get a clear picture and then sent to memory comparison stage), then the time it takes to encode the stimulus will be longer but all the other stages shouldn’t change; adds a constant to the data (the time it takes for memory encoding stage to ”clean up” probe stimulus)
- -If the degraded probe stimulus is not dealt with in the encoding stage, then the degraded probe is sent to the memory comparison stage where it takes the degraded item and compares it against the memory set
- Sternberg found that it takes 37.9ms to compare one probe to one item in the list, so if the memory comparison stage had to compare a degraded probe to an item, then each comparison stage would take slightly longer than 37.9ms
- if dealt with in the memory comparison stage, the graph will have two lines (one clear one degraded) that diverge out (unlike parallel in encoding graph); this is known as an interaction therefore stimulus quality is interacting with set size and this is because the two factors (stimulus quality and number of items in set) are affecting the same stage, not different stages (where the number of items in list affect memory stage and degraded item affects encoding stage); If they don’t they just add a constant and affect different discrete stages
- The study found that the encoding stage actually deals with/cleans up the degraded item
What are the advantages and disadvantages of the Additive Factors Logic?
– Advantages:
Simple logic used to isolate stages in a serial process.
– Disadvantage:
Only useful when examining a simple serial process. Assume discrete stages
What is the Craik & Lockhart levels of processing study?
- type of encoding important (storage is a by-product of encoding)
- – one group has to rhyme the word and the other group has to say a word that is semantically related to the word said
- they then have to write down the words that was said to them (the word they had to rhyme or come up with a semantically related word)
- on average, the phonetic (rhyme) group recalled 4.9 words and the semantic group recalled 8.4 words; this is because the semantic group had to do a deeper information processing while the phonetic group did a shallow level of information processing
- -memory of information is affected by the way you encode it
What are the two assumptions of memory persistence?
- -Memory persistence NOT affected by rehearsal (called Type I); rehearsing is just a shallow way of processing information and is not a way to get it into LTS
- Memory persistence depends on encoding level (called Type II); affects memory
What is the Craik & Watkins study on type I memory persistence?
- varied amount of rehearsal
- gave people a list of words and in some trials they had the participants remember the most recent word that began with a particular letter, such as “b”.
- -they could vary the rehearsal cycles of the word that started with “b” by varying the number of items between the “b” words
- -if “bag” had a lot of rehearsal cycles they would not remember it better than the word “bath” that didn’t have many rehearsal cycles
- Results: No relationship between probability of recall and # of rehearsal cycles.
What is the Hyde & Jenkins study on type II memory persistence (pleasantness)?
–oriented people towards different levels of processing; a very simple processing was asking the participants to count the number of letters that are in the words (“bath” has 4), another one was told to cross out the letter “E” from the words and the last group had to rate the pleasantness of the word (“hit” is not a pleasant word)
–therefore, two shallow groups and one deep-processing group
–they used two different memory tasks, incidental memory task (participants don’t know it is a memory task) and intentional memory task (the participants know it’s a memory task)
RESULTS
– the people in the two shallow processing groups didn’t do as well as the deep-processing group (pleasantness)
– the shallow processing groups did better when they knew it was a memory task (intentional) because they related the words personally while trying to remember them (a little bit of a deeper processing) but still not as good as the deep-processing group
– for the deep processing group, there was no difference in the # items recalled between incidental/intentional memory task
What was the Craig & Tulving recognition test on type II memory persistence (fit in sentence)
- one group was asked to judge whether the word presented to them was in capital letters (physical, shallow level)
- another group was asked to rhyme with the word presented (phonemic, shallow processing)
- the last group was asked if the word fit into the sentence (semantic, deep level processing)
- -this was a recognition task; all three groups had to determine if they had seen the word during the learning phase or not
- -recognition better for semantic than phonemic and phonemic better than physical (the deeper the processing the better recognition/memory of it)
In a separate study;
- varied complexity of the sentence in which the participants had to judge if word fit
- the more complex sentences had better recall; all of the sentences are semantic level but in the complex sentence, its extensive semantic levelling
- incidental memory task
What was the Tresselt & Mazner study on type II memory (word in category)?
- recognition task and incidental recall
- one group had to cross out vowels, another group had to simply copy the words (write them), and the third group had to judge the degree to which the word fit a category they were given (does the word “money” fit into the category “economics”)
- Results: recall for judge category twice as high as just coping the words!
What was Bower & Karlin Nonverbal material study on type II memory (face likeness)
- showed people a picture of a person and one group had to note the sex of person, another group had to judge how well they liked the face of the person and another group had to judge whether the person looked honest
- used recognition task and asked participants if they recognized the face they were shown (pictures of people they judged and pictures of people they had not seen before)
- likeability and honesty had better recognition rate because it was a deeper processing (note: chance = 50%)
What was the Borris, Bransford, & Franks study on type II memory task and memory test?
- had one semantic group and one phonemic group, but during the testing phase they had people do the standard recognition (was word on list or not) or had people do a rhyming test (does this word rhyme with one of the words presented before –word presented was boat and they were asked if the word goat rhymed with the word presented)
- with the standard recognition test the semantic group did better than phonemic but with rhyming test, the phonemic group did better than semantic group; this tells us that the type of test you do matters (we store information differently depending on how we encode it)
- suggests that type of test is important
