2.3.1 - Baddeley (1966b): influence of acoustic and semantic similarity on long-term memory for word sequences Flashcards
what was Baddeley’s aim?
to investigate the influence of acoustic and semantic word similarity on learning and recall in short- and long-term memory - specifically, to see whether people make acoustic errors when recalling information over long periods of time
why did Baddeley want to conduct this experiment?
earlier studies had shown him that when people store words over short timescales, they make errors based on how the words sound, suggesting they store them using acoustic encoding
therefore, he wanted to see whether people also made acoustic errors when recalling information over long time periods of time or if they made a different type of error, to establish whether there were one or two memory systems working
existing researchers who looked at STM and LTM did so in different ways, but he worked out how to do it in the same way
what was the independent variable?
whether words were semantically similar or dissimilar
what was the dependent variable?
the number of errors pps made while recalling the list
what was the sample?
men and women from the Applied Psychology Research Unit subject panel
they were assigned to one of the 4 list conditions as an independent groups design
how did Baddeley investigate his aim?
using a lab experiment which tested sequential recall of acoustically and semantically similar word lists
what was the procedure?
- 4 lists of 10 words were created - list A was acoustically similar, list B was acoustically dissimilar, list C was semantically similar and list D was semantically dissimilar
- pps first had a hearing test where they listened and copied each word presented in a random order from the list
- pps were then presented visually with the word list in the correct order
- after the visual presentation, pps completed six tasks which involved memory for digits, before recalling the word list by writing the sequence in the correct order (they could see the list in a random order as it was a test of learning the sequence order, not the words)
- the visual presentation process was repeated over 4 learning trials
- following these trials, groups completed a 15 minute interference task where they copied 8 digit sequences
- after this, a retest was given of the word list sequence
- Baddeley recorded the number of correct and incorrect answers from pps in each trial
why did Baddeley choose to test recall using participant’s memory of the position of the words rather than what the words were?
because their existing knowledge of the words would have affected their ability to remember what the words were, but not the position of the words
what were the results for lists A and B (acoustically similar and dissimilar)?
the pattern across the five trials was the same - across the first 3 trials, recall accuracy increased from 30%-40% up to 70% and didn’t improve beyond that (but overall it improved)
however, recall of the similar list was worse than the dissimilar list for the first 3 trials, and then better for the 4th and the retest (suggests information stored acoustically in short term)
what were the results for lists C and D (semantically similar and dissimilar)?
recall accuracy was around 30% for both lists in the first trial
for the semantically dissimilar list, accuracy rose over the 4 trials to about 85%, and remained there for the retest
for the semantically similar list, accuracy peaked at around 55% after 3 trials and didn’t improved from there - had the most errors but they didn’t appear until the 3rd/4th trial (so semantically similar list harder to remember over longer periods)
what conclusions did Baddeley come to?
as the acoustically similar list was harder to recall in the initial trials, short-term memory must encode information acoustically
a long-term memory store must be encoding information semantically, because pps in the semantically similar group made errors when trying to to retrieve the correct word in later trials, as they retrieved the incorrect word due to it having a similar meaning
this means human memory must at least consist of short-term memory which uses acoustic encoding and long-term memory which uses semantic encoding
strengths - replicability and reliability?
conducted in a controlled laboratory environment and used standardised procedure eg. same 4 word lists - allows other researchers to replicate and establish reliability of results
strengths - internal validity?
many control variables used eg. all pps went through same process of trials and interference task - allows Baddeley to establish cause and effect relationship between how information is stored and how well it is remembered in the short- and long-term
weaknesses - generalisability?
task had low mundane realism as we wouldn’t normally learn lists of random monosyllabic words
we also wouldn’t normally rehearse information in such a contrived manner
means findings can’t be generalised to everyday situations where we remember stuff
