Cognitive Psychology - Memory Studies

Bahrick et al (1975) - Duration of the LTM

Year Book Study

Procedure: 374 participants aged between 17-74, asked to do: Free Recall, Name recognition, and the Matching condition with year book photos and names.

  • Free Call - Given a page from a year book and asked to name the pictures
  • Name Recognition - Given a list of names, some from their year and others were random photos and then asked to pick out names of the people in their year
  • Matching Condition - Names and face match up  

Findings: For participants that had left school up to 34 years previously the accuracy on the face name recognition task was still 90%, and even with participants who left school up to 40 years previously was still 80% however free recall steadily declined over time. 

Conclusion: Recall can be accurate over a very long period of time, leading to what researchers called the very long term memory (VLTM) 

1 of 22

Bahrick et al (1975) - Evaluation


  • This task is high in mundane realism as remembering names is a daily task
  • Wide range of ages and a large sample size, so the sample is more representative of the general population and we can therefore generalise the study
  • A large sample size means the study is less effected by anomalies or individual differences
  • Is easy to replicate with different samples 
  • Other factors other than time might affect recall for example the size of the year, or if you don't know the people it isn't testing recall 
2 of 22

Baddeley (1966) - Encoding into the STM & LTM

Semantic and Acoustic Similarity Study 

Procedure: Service men were given a list of semantically disimilar(Big,Little,Dog),/similar  (Big,Massive,Enourmous) words or acoustically dissimilar(Cat,Apple,Car)/similar,(Cat,Mat,Bat) words and were then asked to recall the list in no particular order. Those who were being tested on the STM were asked to recall the list straight away and those who were tested on the LTM were asked to recall the list after a timed delay

Findings: He found that participants had difficulty remembering acoustically similar words in the STM but not in the LTM, whereas for semantic encoding words posed little problems in the STM recall but led to muddled LTM recall. 

3 of 22

Baddeley (1966) - Evaluation


  • Some studies have shown that visual codes are also used in the STM. Brandimote (1992) found that if participants used visual encoding in the STM when they were given visual tasks and prevented from doing any verbal rehearsal 
  • Lacks population validity because sample used is not representative of everyone because they only used service men of a certain age
  • Participant variables could have affected results because different participants in each conditions (one group could have worse memory than the others) 
  • Low in mundane realism as we don't usually have to recall a random list of words, normal list would mean something 
  • High in control as the same words lists were used with neutral words 
  • No order effects with independent groups as they are reduced and no practice or boredom effects
4 of 22

Jacobs (1887) - Capacity of STM

Number and Letter Recall 

Procedure: Presented a series of numbers/letters to participants which they had to recall in the correct order

Findings: He found that there was an average span of 9.3 items for numbers and only a 7.3 span for letter. Jacobs suggested this was because there are only 9 digits but are 26 letters. He also found that capacity increases with age and put this down to potential strategies that people gain as the get older to increase this such as chunking. 

5 of 22

Jacobs (1887) - Evaluation


  • STM capacity may be affected by how the  information is presented (visually or acoustically) 
  • This task is low in mundane realism as the task is not reflective of everyday tasks
  • As this study was conducted so long ago it lacks historical validity 
  • Individual differences such as age and education level may affect the results of the study 
  • Real life applications of the findings for example number plates and post codes 
6 of 22

Peterson and Peterson (1959) - Duration Of the STM

The Trigram Study               

Procedure: University students were presented with nonsense trigrams. From the numbers the participants were given they had to count backwards out loud in threes at different time intervals to prevent any rehearsal as counting back from 3's requires a certain amount of thought. They were asked to repeat the letters at the beginning of the trigram in the correct order. They had 2 practice trials and then 8 trials each

Independant variable: Time interval (being manipulated/changed)

Dependant variable: Correct recall of the trigram (being measured) 

Findings: The accuracy of the recall slowly decreased from 80% to 10% over time. 

7 of 22

Peterson and Peterson (1959) - Evaluation


·         Age specific as only university students so we cannot generalise it to other ages

·         Highly controlled through standardised procedures as all the participants had to count down out loud. This prevented rehearsal which could have affected results as an extraneous variable

·         All the particpants were university students which mean they are likely to have better memories due to a higher level of education and we cannot generalise to non university students

·         Repeated measures used each participant in each condition therefore reducing participant variables 

·         Some trigrams may be significant to some participants but not to others and this extraneous variable cannot be controlled

·         Lacks mundane realism as not an everyday task

·         By the 8th trial participants may suffer order effects 

8 of 22

Yuille + Cutshall - Anxiety and Eye Witness Testim

Robbery Study


·         Studied witnesses of a robbery of a gun store in Vancouver where the robber was shot by the owner of the store.

·         After 48 hours police interviewed the 21 witnesses to gather evidence

·         approximately 4/5 months later 13 people agreed to be interviewed by the researchers.

·         This event was a real event that occurred.  


·         Researchers found that anxiety made the participants recall better

·         Therefore providing evidence that anxiety increases accuracy of EWT

9 of 22

Christianson + Hubinette - Anxiety and Eye Witness

Swedish Robberies 


·         58 witnesses were interviewed from 22 robberies in Stockholm

·         This event was real 


·         Self reported anxiety had no effect on the accuracy of EWT

10 of 22

Anxiety and EWT - Evaluation


  • These studies were real life events and therefore the anxiety would have been genuine and there would be no demand characteristics
  • Low control of extraneous variables, therefore we cannot identify cause and effect factors other than anxiety that may have affected recall accuracy
  • High in ecological validity as the study is a real life event in a natural environment meaning we can generalise them to more real life situations
  • Self report anxiety may be effect by social desirability bias - more upset for sympathy or less effected to make them seem llike they can handle a situation
  • Christianson + Hubinette had a high number of events studied more witnesses and a wider range of people and therefore we can generalise it to more people
  • Cultural bias as people in certain countries may be more or less affected by anxiety and therefore can not make accuarate comparisons between the two studies. 
  • Interviews may cause through reliving and talking about traumatic events however the participants gave informed consent
11 of 22

Loftus - Anxiety and EWT

Weapon Focus Experiment


  • Two conditions of independent groups
  • Participants overheard a heated arguement and then a man emerged
  • Condition 1 the man emerged with a pen covered in grease 
  • Condition 2 the man emerged with a knife covered in blood


  • In condition 2 with high anxiety there was an accuracy recall of 33%
  • In condition 1 with low anxiety there was an accuracy recall of 49%
  • Showing that the more anxious the less accuracy 
12 of 22

Loftus - Evaluation


  • High in control of extraneous variables using standardised procedure, for example participants heard the same scripted conversation and the same man emerged resulting in more accurate results
  • Independent groups means theres different participants in each condition and therefore participant variables may affect the outcome and make the results less reliable 
  • Participants were deceived sp not told the true outcomes of the study but this decreases the demand characteristics however there was no informed consent
  • If participants experienced anxiety they are causing psychological harm however they were debriefed 
13 of 22

Loftus and Palmer - Misleading information and EWT

Misleading Information Study 

Procedure: 45 university students were put into 5 groups of 9 and were shown 7 films of car accidents and asked to estimate the speed, Each group was asked a different leading question. They were asked how fast the car was going when the car collided (39mph), smashed (41mph), bumped (38mph), hit (34mph), or contacted the other car (32 mph) 

Findings: The word smashed was the highest speed estimate  and the word contacted was the lowest speed estimate. They concluded from this that EWT were generally inaccurate and therefore unreliable. 

Second Experiment

Procedure: 150 participants using independent groups were shown a film of an accident and asked to estimate the speed of the cars when hit/smashed into each other. In the control group where they weren't asked a leading question 6 people said they saw broken glass when there was none. In the smashed group 16 people said there was glass and the hit group 7 people saw glass.

Findings: Suggests that misleading information affects the way in which memory is stored

14 of 22

Loftus and Palmer - Evaluation


  • Uni students are young and inexperienced drivers and may be more influenced by the leading questions and therefore may not be representitive of the general population so the study lacks population validity 
  • Low in mundane realism as watching the video is not the same as experiencing the accident in real life. There would've been no emotional repsonse, there would have been high levels of arousal and shock but the participants were expecting it and therefore had low anxiety levels. They may not take the video as seriously as a real life accident
  • Standardised procedure as the same videos are shown to all which means there are high levels of control of extraneous variables which makes the results more reliable and increases replicability. 
  • Some extraneous variables cannot be controlled because participant variables might affect results as different participants in different conditions. Some participants may have experiencedd a car crash and find it more distressing than others (ethical issue of psychological harm) 
15 of 22

Age and EWT - Multiple Studies

Parker and Carranza (1989) - Compared primary school children and college students in their ability to correctly identify a target after following a slide sequence of mock crime. In photo identification children witnesses has a higher rate of choosing than the adult witnesses, although they were also more likely to make errors of identification than college students.

Yarmey (1993) - Stopped 651 adults in public places and asked them to recall physical characteristics of a young women they had previously spoken to 2 minutes earlier. Young (18-29) middle aged (30-44) and older (45-65). There were significant differences in the accuracy of recall also, the oldest group were inferior to the other two groups in terms of accuracy of recall. 

Memon et al (2003) - Studied the accuracy of young (16-33) and older (60-82) eyewitnesses. When the delay between incident and identification was short (approximately 35 minutes) there was no difference in the accuracy of the two age groups. However when there was more a time delay (1 week) the older witnesses were significantly less accurate.

16 of 22

Age and EWT - Evaluation


  • What is considered young/middle age/old is simply down to ones judgement and is therefore subjective so we are unable to accurately compare the studies as different researchers use different ages to classify these age groups. 
  • Most of the studies are also conducted on college students who are asked to identify someone of a similar age which could be why they do better than different ages. This is call own age bias and makes the studies less reliable
  • The differential hypothesis (Brigham and Malpass) suggests that the more time we spend with individuals the of a particular age or ethnicity the better our memory would be for identifying such individuals. Consequently the less time spent with a particular age group the greater the own age bias. 
  • Individual differences may effect how people identify people as a study by Clifasefi said that level of intoxication should be noted on the accuracy of EWT.  They suggested the more intoxicated a person becomes the less attention they can allocate to peripheral tasks. 
17 of 22

Kohnken et al (1999) - Effectiveness of the CI

Meta - Analysis of the Cognitive Interview

Procedure: A meta analysis of 53 studies 

Findings: They found on average an increase of 34% in the amount of correct information generated in the cognitive interview compared with the standard interviewing techniques. 


  • Most of the studies analysed tested volunteer witnesses in a laboratory and therefore make the results less valid as lab studies make the results less true to real life
  • A big sample of studies makes the findings more reliable 
  • Unable to generalise to the general population as most were conducted on college students
  • Volunteer test subjects may be more inclined to make a study work and more willing to do things that others wont
18 of 22

Stein and Memon (2006) - Cognitive Interview

Real World Application


  • Tested the effectiveness of the cognitive interview in Brazil, the first study done in a developing country
  • In Brazil the current way of police investigation is interrogation, torture and other forms of ill treatment 
  • Women recruited from the cleaning staff of a university watched a tape of an abduction (real life event) 
  • They were then asked to identify the criminal 


  • The use of the cognitive interview meant the police was able to gain more information such as that the criminal was in fact armed 
  • These results suggested that the cognitive interview may be the new approach in developing countries in order to reduce mistreatment of witnesses 
19 of 22

Atkinson and Raugh 1975 - Memory Improvement

Keyword Method

  • This is used when trying to associate two pieces of information
  • For example when learning a foreign language and wanting to remember a foreign word and its English equivalent
  • You would think of an image to link the two words together 
  • For example: the spanish word for horse is "caballo" pronounced "cob-eye-you" the keyword may be eye so you might visualise a horse with a large eye.
  • Conjuring the visual image should trigger the word recall 
20 of 22

Verbal Mnemonics - Studies

Verbal Mnemonics


Procedure: A survey was taken from psychology students, revising for final examinations.

Findings: Gruenberg discovered that about 30% of these students used mnemonics in their revision with first letter verbal mnemonics, such as acronyms and acrostics, For example (ROYGBIV) Red,Orange,Yellow,Green,Blue,Indigo, Violet (Acronyym for the colours of the rainbow), or My Very Easy Method Just Speeds Up Naming Planets (An acrostic to help remember the name of the planets) 

Glidden et al(1983)

Findings: Verbal mnemonics were effective in children learning with disabilities, althought their effectiveness over a control group was no longer evident after 12 months. But Broadly and Macdonald demonstrated the value of visual mnemonics for 63 down syndrome children. (Real world application) Their training program had significantly improved memory skillls among the group of children with down syndrome

21 of 22

Bower - Organisation

Evidence for organisation

Participants were given 112 words to learn

If the words were organised into conceptual hierarchies, recall was 2-3 times better than if words were presented in a random order

This shows evidence for organisation

22 of 22


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Memory resources »