Grant et al. (1998): Context-dependent memory

What you'll learn

• What context-dependent memory means and why it matters for students.
• How Grant et al. (1998) tested memory using silence and background noise.
• The key method, results and conclusions you need for AO1.
• How to evaluate the study and compare it with Loftus and Palmer (1974).

First, the basic idea: memory uses cues

When you learn something, you do not only encode the “main content”. You may also encode extra information from the situation around you: the room, sounds, smells, lighting, or even your internal state.

A cue is anything that helps trigger retrieval of information from memory. For example, hearing a song might help you remember a holiday because the song was present when the memory was formed.

In Grant et al. (1998), the key external context was auditory context: silence or background cafeteria noise.

The background: encoding specificity

Grant et al. were influenced by the encoding specificity principle, associated with Tulving. This principle suggests that memory is best when cues present at encoding are also present at retrieval.

A famous earlier example is Godden and Baddeley’s diver study, where divers remembered words better when learning and recall both happened in the same environment: either on land or underwater. Grant et al. wanted to test whether this effect also happens with more realistic student learning material.

Aim of the study

Grant et al. (1998) aimed to investigate whether context-dependent memory occurs for meaningful material: a two-page article, rather than just word lists.

They also wanted to see whether the effect appeared in two types of memory test:

• Recall, where participants have to produce the answer themselves.
• Recognition, where participants identify the correct answer from options.

Method: “tell the story”

Design

Grant et al. used a laboratory experiment, meaning the researchers manipulated variables in a controlled setting.

There were two independent variables:

1. Whether participants studied in silence or with background cafeteria noise.
2. Whether participants were tested in silence or with background cafeteria noise.

This created four conditions:

• Silent study / silent test.
• Silent study / noisy test.
• Noisy study / silent test.
• Noisy study / noisy test.

The dependent variables were:

• Score on a short-answer recall test.
• Score on a multiple-choice recognition test.

The study used an independent measures design, meaning each participant took part in only one condition.

Sample

The sample was 56 undergraduate students: 39 female and 17 male. They were from an American university setting.

Participants were randomly allocated to one of the four conditions, with 14 participants in each condition.

Materials

The main materials were:

• A two-page article about psychoimmunology, chosen as meaningful but unfamiliar material.
• A recording of background cafeteria noise.
• Headphones, so the noise or silence could be controlled.
• A short-answer test.
• A multiple-choice test.

Procedure

Participants wore headphones during both the study phase and the test phase.

They read the article either in silence or with cafeteria noise. Later, they were tested either in silence or with the same type of cafeteria noise.

The short-answer test came before the multiple-choice test. This mattered because multiple-choice options could give clues that might help with later recall.

Results

The key result was that participants performed better when the study and test contexts matched.

For short-answer recall, the commonly reported mean scores were approximately:

• Silent study / silent test: 6.7
• Silent study / noisy test: 5.4
• Noisy study / silent test: 4.6
• Noisy study / noisy test: 6.2

For multiple-choice recognition, the pattern was also better in matching conditions:

• Silent study / silent test: 14.3
• Silent study / noisy test: 12.7
• Noisy study / silent test: 12.7
• Noisy study / noisy test: 14.3

Conclusions

Grant et al. concluded that context-dependent memory can occur with meaningful learning material, not just artificial word lists.

The study suggests that environmental cues present during learning may become linked to the information being learned. When the same cues are present at retrieval, they help trigger memory.

For students, this has a practical message: if your exam will be in silence, revising in silence may be useful because the learning context is more similar to the retrieval context.

Evaluation: strengths and weaknesses

Strength: high control

Because this was a laboratory experiment, Grant et al. could control key variables. For example, all participants used headphones, and the same article and noise recording were used.

This improves reliability, meaning the procedure could be repeated in the same way.

Weakness: independent measures and participant variables

Because each participant only took part in one condition, individual differences may have affected results. Some participants may have had better reading ability, stronger motivation, or more prior knowledge.

Random allocation helps reduce this problem, but it does not remove it completely.

Strength: more ecological validity than many memory studies

The study used meaningful written material, similar to what students might read when studying. This gives it more ecological validity than studies using random word lists.

Weakness: still artificial

Although the article was meaningful, the setting was still controlled and short-term. Real revision may happen over days or weeks, and real exams involve pressure, timing, and more complex knowledge.

So, the study is useful, but it does not perfectly recreate real education.

Sampling bias

The sample consisted of American undergraduate students, with more females than males. This limits population validity, meaning the findings may not generalise fully to all ages, cultures, or educational backgrounds.

Ethnocentrism

The study was conducted in a Western university context. The basic cognitive process may be universal, but the specific setting — students reading an academic article and taking a test — is culturally specific.

So, the study is not strongly ethnocentric in its theory, but its sample and educational context are culturally narrow.

Ethics

The study has relatively few ethical concerns. Participants were adults, the task was low risk, and the background noise was not harmful.

Using the BPS Code as a guide, the study appears fairly strong on:

• Respect, because participants should be treated with dignity and allowed to withdraw.
• Responsibility, because the noise should not cause distress or harm.
• Integrity, because any deception or withholding of the full aim would need to be justified and followed by debriefing.

Comparison with Loftus and Palmer (1974)

Grant et al. is the contemporary study in the Memory theme. Its paired classic study is Loftus and Palmer (1974).

Loftus and Palmer investigated how leading questions affect eyewitness memory. Participants watched film clips of car crashes and then answered questions. Changing the verb, such as “hit” or “smashed”, affected speed estimates and later reports of broken glass.

Similarities

Both studies are in the Cognitive area because they investigate internal mental processes through observable behaviour.

Both used controlled experimental methods and collected quantitative data. This makes them more scientific and easier to replicate.

Both also show that memory is not like a perfect recording. Memory can be influenced by other factors: wording in Loftus and Palmer, and environmental context in Grant et al.

Differences

Loftus and Palmer focused on reconstructive memory, especially how post-event information can distort recall. Grant et al. focused on retrieval cues, showing that matching context can improve access to stored information.

Loftus and Palmer has stronger application to the legal system and eyewitness testimony. Grant et al. is more directly useful for education and revision.

Grant et al. has fewer ethical issues because reading an article with background noise is unlikely to cause distress. Loftus and Palmer involved crash films, which could be mildly upsetting, although still low risk compared with real accidents.

Link to the Cognitive area

The Cognitive area studies mental processes such as memory, attention, perception, and thinking. These processes cannot be directly observed, so psychologists infer them from behaviour.

Grant et al. measured memory using test scores. The researchers could not see memory directly, but they could compare performance across conditions.

This fits the Cognitive area because it treats memory as an information-processing system affected by encoding and retrieval cues.

This reductionism is useful because it allows scientific testing, but it may oversimplify real learning, where motivation, emotion, sleep, stress, and teaching quality also matter.

AO2: applying Grant et al. to a new situation

If a student revises every evening with loud music and then sits an exam in silence, Grant et al. would predict a possible disadvantage because the study and test contexts do not match.

However, this does not mean the student should play music in the exam. The more sensible application is to practise retrieval in conditions similar to the exam: quiet, timed, and without extra cues.

Final judgement

Grant et al. is a well-controlled and useful study showing that context can affect memory. Its main strength is that it tested a real-world issue using meaningful material while still maintaining experimental control.

Its main limitation is generalisability: the sample was narrow, and the task was still simpler than real long-term study and examination. For essays, a balanced conclusion is best: the study is scientifically useful, but its application to everyday learning should be made carefully.