Sternberg scanning task
- Time taken to compare one mental comparison
- Given set of numbers, short pause
- Given probe number, does it match one of the numbers in the set
- Dependent variable is response time
Simple model of scanning task
- Probe digit (number to find in set)
- Encoding stage
- Memory comparison stage
- Response output stage
Hypotheses regarding search in short term store
- Parallel
- Serial exhaustive
- Serial self-terminating
Parallel hypothesis regarding search in short term store
- Compares all items of memory set at same time
- The number of items in the memory set doesn’t have an impact on response time
Serial exhaustive hypothesis regarding search in short term store
- Compares all items of memory set in sequence
- As items in the memory set increases, so does response time
- Where the probe is (first or last item in set) does not impact response time
Serial self-terminating hypothesis regarding search in short term store
- Compares items of memory set in sequence until finds probe
- Response time depends on how quickly you reach probe number in memory set
- If probe not found, response time depends on number of items in memory set
Results for short term store searching
- Matches serial exhaustive hypothesis
- Reaction time = 397.2 + 37.9i, (i = items in memory set)
Additive factors logic and modelling
- Sternberg scanning task
- Clear vs degraded probe
- Encoding and memory comparison now?
- Adjusts for degraded probe during encoding stage therefore slope of response line is the same, just moved higher
Advantage of additive factors logic
Simple logic used to isolate stages in a serial process
Disadvantages of additive factors logic
Only useful when examining discrete stages
Levels of processing framework
- Emphasis on what you do with the material
- Type of encoding important
- Orienting tasks: physical, phonemic, semantic
Asumptions of memory
- Not affected by rehearsal (Type 1)
- Depends on encoding level (Type 2)
Evidence of rehearsal not having the impact we thought
Can change amount we rehearse specific item but has no influence on number of items recalled
Evidence that encoding level matters
- Varied type of orienting task: counting # letters, check for letter e, pleasantness ratings
- Incidental vs intentional learning
- Shows support for level of processing increasing recall
Recognition test for recall
Support for using semantics increases recall amounts
More evidence for semantics increasing recall
Judging whether a word fits into a sentence doubles recall over rewriting the words
Nonverbal test for semantics increasing recall
- Sex of person
- Judge face likableness
- Rate honesty
- Rate honestly had highest recall amounts and requires more deeper level processing
Variance in complexity of sentence frames on recall
More complex sentence frames increases recall
Impacts on type of test
Standard recognition - semantic > phonemic
Rhyme recognition - phonemic > semantic
