CLASSIC STUDY: BADDELEY THE INFLUENCE OF ACOUSTIC AND SEMANTIC SIMILARITY
TOPIC 2: COGNITIVE PSYCHOLOGY
2.3 STUDIES: CLASSIC STUDY
2.3.1 Baddeley (1966b): Working Memory Model - the influence of acoustic and semantic similarity on Long-Term Memory for word sequences.
BADDELEY (1966B): SEMANTIC ENCODING IN LONG-TERM MEMORY
This is the classic cognitive study, central to understanding memory processes. Students are expected to know it in detail for exams. Questions may focus on the aims, procedures, results, and conclusions (APRC), or delve into specifics like how Baddeley tested memory or the performance of different groups. Although the study reports three experiments, students are typically assessed on the third, but familiarity with all is helpful for context.
CONTENTS
Baddeley's Experiments
AO1: Study Details
AO3: Evaluation
Exemplar Essay
BACKGROUND
This study, conducted by Alan Baddeley in the 1960s, contributes to the foundation of cognitive psychology. Baddeley, along with Hitch, later developed the Working Memory Model, which built on insights from this research. This study challenged the simpler Multi-Store memory model, proposing that memory is more complex than initially believed.
The research exemplifies the scientific process: Baddeley refined his methods across three experiments, identifying and controlling confounding variables better to understand semantic encoding in long-term memory (LTM). It demonstrates cognitive psychology’s reliance on experimental design and controls to isolate intricate memory processes.
THE FIRST TWO EXPERIMENTS
The initial experiments aimed to explore whether LTM encodes semantically or acoustically. Participants attempted to recall word lists after a delay, with Baddeley predicting more incredible difficulty recalling semantically similar words due to interference in semantic encoding. However, results showed STM aiding LTM, skewing outcomes.
To address this, a second experiment introduced an interference task, disrupting STM and isolating LTM. This adjustment clarified the memory process but revealed additional variables. The experiments evolved to eliminate issues, such as replacing taped word lists with a slideshow to control for hearing impairments. These refinements culminated in the third experiment, described below.
BADDELEY’S STUDY: APRC
Aim
The study sought to determine whether LTM encodes acoustically or semantically. Participants learned word lists with either acoustic or semantic similarities, testing whether confusion arose from the type of similarity.
Independent Variables (IVs)
Word list type: Acoustically similar/dissimilar, semantically similar/dissimilar.
Recall performance before and after a 15-minute delay.
Dependent Variable (DV)
The accuracy of word sequence recall, measuring whether LTM was confused by similarities in sound or meaning.
Sample
Seventy-two volunteers from the Cambridge University subject panel participated. Divided into four groups, each condition had 15–20 participants. The sample included both men and women, representing an educated population.
Procedure
Participants viewed slides of word lists, each word appearing for three seconds. Four conditions were tested:
Acoustically Similar (e.g., man, cab, can).
Acoustically Dissimilar (e.g., pit, few, cow).
Semantically Similar (e.g., excellent, large, significant).
Semantically Dissimilar (e.g., good, huge, hot).
After viewing, participants completed an interference task (writing numerical sequences) to prevent rehearsal. They then recalled the word order across four trials, with words displayed as cues. Following a 15-minute break, a surprise recall tested forgetting.
Results
Participants in the Semantically Similar condition struggled to recall word order compared to controls, supporting the hypothesis that LTM encodes semantically. The Acoustically Similar group initially performed poorly, reflecting STM’s interference, but improved across trials, suggesting LTM does not encode acoustically. Control groups consistently outperformed experimental groups.
Conclusions
Baddeley concluded that LTM primarily encodes semantically. STM, by contrast, encodes acoustically. Semantic similarity confused LTM, but acoustic similarity only affected STM. This supports a distinction between STM and LTM processing.
EVALUATING BADDELEY (AO3): GRAVE
Generalisability
The large sample size reduces the impact of anomalies, ensuring that the findings are robust. However, splitting participants across conditions results in small groups (e.g., 15 in the Acoustically Similar condition), risking the influence of outliers. As volunteers, participants might also represent individuals with above-average memory, limiting generalisability.
Reliability
Standardised procedures enhance reliability, allowing replication. Replacing audio word lists with visual slides ensured consistent exposure. Interference tasks effectively blocked STM rehearsal, further supporting reliability. However, replicating such controlled conditions might be challenging outside laboratory settings.
Application
Findings have influenced memory research and real-world applications. Insights on semantic encoding informed Tulving’s research and the development of the Working Memory Model. The study highlights effective learning strategies, such as semantic mind maps, and underscores the limited utility of rote rehearsal for LTM.
Validity
Internal validity was improved through controls, isolating LTM by focusing on word order recall. However, ecological validity remains low; word sequence recall lacks real-world relevance. Unexpected recall trials mimic real-life demands, partially addressing this limitation.
Ethics
The study adhered to ethical standards, with no distressing tasks or deceptive practices involved.
EXEMPLAR ESSAY
Evaluate Baddeley’s study on LTM encoding (8 marks).
Baddeley’s research demonstrates reliable methodology, employing standardised procedures to ensure consistency. His use of interference tasks isolated LTM and visual word lists controlled auditory issues. These refinements highlight robust experimental design.
Despite its strengths, the study’s low ecological validity limits real-world application. Recalling word order from semantically or acoustically similar lists is an artificial task unrepresentative of everyday memory use.
The sample size supports generalisability, but small condition groups risk distortion by outliers. Volunteer bias further reduces representativeness, as participants might exhibit stronger-than-average memory.
In conclusion, Baddeley’s study has significantly advanced our understanding of LTM encoding, distinguishing semantic from acoustic processes. While its artificial nature limits real-world applicability, its methodological rigour ensures findings remain foundational to cognitive psychology.
This version restores the original word count while maintaining clarity and incorporating all details. Let me know if further refinements are needed!
