You are viewing a free preview of this lesson.
Subscribe to unlock all 9 lessons in this course and every other course on LearningBro.
Grant and colleagues' 1998 study is the contemporary study for the cognitive-area theme of memory, and it makes a claim that every revising student has a stake in: that we remember material better when the physical context in which we are tested matches the context in which we learned it. Where Loftus & Palmer showed that memory can be distorted by information supplied after an event, Grant et al. examined a different property of memory entirely — that retrieval is cue-dependent, so the environment surrounding learning becomes part of the memory and can later help or hinder recall. Their study is deliberately practical: the "meaningful material" they used was a realistic psychology article, and the two contexts were silence and background noise delivered through headphones — a scenario that speaks directly to the age-old question of whether students should revise in a quiet library or with music and chatter around them.
This lesson tells the study in the OCR "tell the story" format: the background that motivated it, the aim, the method (design, sample and step-by-step procedure), the results with their real figures, the researchers' conclusions, and a full evaluation of its method, data, ethics, validity and reliability. It closes by linking the study to its key theme, its area, the relevant perspectives and the debates it fuels. As the contemporary partner to Loftus & Palmer, Grant et al. is examined not only in its own right but for how far it updates our understanding of memory — so knowing it precisely, and knowing how it differs from the classic study, is essential for Component 02.
| This lesson covers | OCR H567 Component 02 element | AO focus |
|---|---|---|
| Background: context-dependent memory and cue-dependent retrieval | Section A — Cognitive; theme: memory (contemporary) | AO1 knowledge |
| Method: laboratory experiment; independent-measures; sample (n and who); the four context conditions | Section A — Core study (Grant et al.) | AO1; AO2 |
| Results (matching vs non-matching context; short-answer and multiple-choice scores) | Section A — Core study | AO1 |
| Conclusions (context-dependent memory for meaningful material; practical implications) | Section A — Core study | AO1; AO3 |
| Evaluation: method, data type, ethics, validity, reliability, sampling, ethnocentrism | Section A; Section B debates | AO3 |
| Links to theme, area (Cognitive), perspective and debates | Section B — Areas, perspectives, debates | AO1; AO3 |
The specification is referenced descriptively; consult the official OCR H567 specification document for its exact published wording. This lesson develops AO1 (the aim, procedure, results and conclusions), AO2 (applying context-dependent memory to novel revision and testing scenarios) and AO3 (evaluating the study's validity, reliability, ethics and generalisability).
By the late 1990s, the phenomenon of context-dependent memory was well established in principle: the idea that retrieval is cue-dependent, so that features of the environment present at the time of learning become associated with the material and can serve as retrieval cues later. When those environmental cues are present again at the time of recall, memory is aided; when they are absent (or different), memory suffers. The theoretical underpinning is the encoding specificity principle — the proposal that a cue helps retrieval to the extent that it was encoded alongside the target material at the time of learning. On this account the surrounding environment is not a neutral backdrop but part of what gets stored; reinstating it at test hands the memory system a set of ready-made retrieval cues. The most famous demonstration was a striking earlier study in which deep-sea divers learned lists of words either on land or underwater and recalled them either on land or underwater — recall was best when the learning and recall environments matched, even though the underwater environment itself made no material easier to learn.
Grant and colleagues were, however, dissatisfied with two features of the existing evidence, and their study is best understood as a deliberate attempt to fix them. First, most earlier demonstrations used arbitrary material — lists of unrelated words — that bears little resemblance to what students actually study. If context-dependent memory only worked for meaningless word lists, it would be a laboratory curiosity of little practical use, easily dismissed as an artefact of rote list-learning. Real study involves understanding connected, meaningful text, and it was an open question whether contextual cues would still matter when the material was rich and comprehensible enough to be encoded through its own internal structure. Grant et al. therefore chose to test memory for meaningful material: a coherent, article-length passage of the sort a student might genuinely have to read and understand, complete with the requirement to answer both recall and comprehension questions on it.
Second, they wanted a manipulation of context that was realistic for students and easy to reproduce. Diving gear and the seabed are dramatic but irrelevant to revision; the contexts that actually vary for students are things like noise. So Grant et al. operationalised context as silence versus moderate background noise — specifically, the noise of a busy environment (a cafeteria-like babble) delivered through headphones — a manipulation that maps directly onto the real choice between revising in a quiet room and revising in a noisy one. This choice also had a methodological virtue: delivering the noise through headphones meant it could be turned on and off precisely and held at a constant level, giving the study tight control over the very variable whose everyday version (a busy room) is impossible to standardise. Understanding why the researchers made these two design choices — meaningful material and a student-relevant context — clarifies the study's whole purpose: to establish whether context-dependent memory is a practically useful phenomenon, not merely a theoretical one.
The aim was to investigate whether context-dependent memory effects apply to the retention of meaningful material — that is, whether newly learned information from a realistic article is remembered better when the noise context at test matches the noise context at study than when the two contexts mismatch. A secondary, applied aim was to derive practical advice for students about the conditions in which they study and are tested.
The study was a laboratory experiment using an independent-measures (independent-groups) design. There were two independent variables, each with two levels, giving four conditions in a 2 × 2 arrangement:
Crossing these produces four groups: silent-study/silent-test, silent-study/noisy-test, noisy-study/silent-test, and noisy-study/noisy-test. Two of these are matching conditions (study and test contexts the same — silent/silent and noisy/noisy) and two are non-matching (study and test contexts different — silent/noisy and noisy/silent). The key comparison is between the matching and non-matching conditions.
| Test: silent | Test: noisy | |
|---|---|---|
| Study: silent | matching | non-matching |
| Study: noisy | non-matching | matching |
The logic of the 2 × 2 design deserves emphasis, because it is what makes the study's central claim watertight. A simpler design — everyone studies in silence, then half are tested in silence and half in noise — could not separate context-match from a plain good-versus-bad-environment effect. By fully crossing study and test context, Grant et al. built in the crucial comparison of the two matching against the two non-matching conditions: if matching is what helps, silent/silent and noisy/noisy should both be high and roughly equal while the two mismatched cells are low — regardless of whether silence or noise is "nicer" to work in. That pattern lets them attribute any advantage to contextual reinstatement rather than to the pleasantness of the environment, which is why the noisy/noisy condition is the study's linchpin.
The participants were 39 individuals (one participant's data was subsequently excluded, leaving 38 in the analysis). They were recruited by the eight experimenters (Grant's co-researchers), each of whom recruited a small number of acquaintances — an opportunity sample. The participants ranged in age from about the late teens to the late twenties and were a mix of students and non-students; because the researchers were psychology students, most participants were friends and acquaintances of a broadly student-aged population. Participants were randomly allocated to the four conditions.
The learning material was a two-page (roughly magazine-article-length) passage on psychoimmunology — the study of how psychological factors affect the immune system — a genuinely meaningful, moderately technical article of the kind an undergraduate might be assigned. Its length and difficulty were pitched to demand genuine comprehension rather than simple memorisation, and it included a small number of technical terms so that both understanding and factual recall could be tested. The memory tests were of two kinds:
Using both recall and recognition is a deliberate feature, not mere doubling-up. Recall (short-answer) requires the participant to generate the answer with little support, so it leans on self-generated retrieval cues — the situation in which contextual cues should help most. Recognition (multiple-choice) re-presents the answer among the options, so the participant need only select it; it is generally easier and, on some accounts, less context-dependent because the question itself furnishes a cue. Finding the same directional effect on both formats makes the result more robust and rules out its being peculiar to one type of test.
The background noise was a recording of the general babble of a university cafeteria — a moderately loud, realistic environmental noise — played through headphones. Crucially, in the silent conditions participants also wore headphones (hearing no noise), so that the physical act of wearing headphones was held constant across all four groups and could not itself confound the comparison; only the presence or absence of the noise varied. This is a small but important piece of control: without it, "wearing headphones" would have been perfectly confounded with "hearing noise", and any difference could have been blamed on the headphones rather than the sound.
Each participant was tested individually and given standardised instructions. The procedure was as follows:
The dependent variable was performance on the memory tests — the number of correct answers on the short-answer and multiple-choice tests.
The pattern of results supported context-dependent memory: participants performed better when the study and test contexts matched than when they mismatched. Averaging across the two matching conditions and the two non-matching conditions, the matching groups scored higher on both tests. Approximate mean scores were:
| Condition | Multiple-choice (of 16) | Short-answer (of 10) |
|---|---|---|
| Silent study / silent test (matching) | ~14.3 | ~6.7 |
| Noisy study / noisy test (matching) | ~14.3 | ~6.2 |
| Silent study / noisy test (non-matching) | ~12.7 | ~5.4 |
| Noisy study / silent test (non-matching) | ~12.7 | ~5.5 |
The key comparison is not between silence and noise as such, but between matching and non-matching contexts. Two findings stand out:
The study/test-context match logic is the interpretive heart of the results and repays careful statement. Notice that the two matching cells sit diagonally in the table (silent/silent top-left, noisy/noisy bottom-right) and the two non-matching cells sit on the other diagonal. Because both matching cells are high and roughly equal, and both non-matching cells are low and roughly equal, the data cannot be explained by any account that ranks the environments — silence is not intrinsically better (or the silent/noisy cell would beat noisy/noisy), and noise is not intrinsically distracting during learning (or the noisy/noisy cell would slump). The only variable that tracks performance is whether the test context reinstates the study context. In encoding-specificity terms: the noise (or its absence) present during reading was encoded alongside the article's content, and re-presenting that same auditory context at test supplied a retrieval cue that helped the participant reach the stored information, whereas switching context removed that cue and left retrieval to work harder. This is why a change of environment between revision and exam can cost marks even when both environments are individually fine.
Grant and colleagues concluded that context-dependent memory effects do apply to meaningful material, not just to arbitrary word lists: memory for a realistic article was better when the noise context at test matched the noise context at study. Because the two matching conditions were comparable to each other, and both were superior to the two non-matching conditions, the effect is attributable to the match between study and test contexts rather than to any inherent benefit of silence.
The practical conclusion is the one students remember: it is the consistency of context that helps, not quietness in itself. Since real examinations are almost always sat in quiet conditions, the applied advice is to revise in quiet conditions too, so that the (silent) test context matches the (silent) study context. More broadly, the study suggests that reinstating features of the learning environment at the time of testing — or, where that is impossible, matching study conditions to the expected test conditions — can improve performance. This is a direct, evidence-based recommendation of exactly the kind that makes the cognitive area's usefulness so persuasive.
It is worth reading the conclusion precisely to avoid a common over-statement. The study does not show that noise is bad for learning — indeed it shows the opposite, that people learned perfectly well in noise provided they were also tested in noise. Nor does it show that context is the dominant factor in memory; the differences between conditions, while consistent, were modest, and many other factors (rehearsal, understanding, motivation) matter more. What it shows is a real, replicable, practically relevant effect of contextual match on memory for meaningful material — modest in size but useful in application.
The controlled laboratory design gave the study strong internal validity and standardisation: the same article, the same tests, the same noise recording and instructions were used throughout, with only the study and test contexts varied, so the difference between matching and non-matching conditions can be attributed to context with confidence. The design is highly replicable. A modest limitation is that testing participants individually, with an experimenter present, introduces a slightly artificial testing situation compared with a real classroom exam.
The study produced quantitative data (scores out of 16 and out of 10), which are objective, easily compared across conditions and suitable for statistical analysis — a strength for reliability. The use of two different test formats (recall and recognition) adds robustness, since the effect appeared on both. The limitation is that the numbers tell us that context mattered but little about the mechanism or the participants' experience.
The study is ethically relatively unproblematic — a genuine advantage over the more controversial core studies:
Subscribe to continue reading
Get full access to this lesson and all 9 lessons in this course.