Type: Assignment | Subject: Psychology | Level: Undergraduate | Word Count: ~2200 words
This model assignment was produced by an Essays UK specialist as reference material for learning purposes only. For support in this field, see our psychology assignment specialists.
You are an undergraduate Psychology student. Select one published experimental study from the core cognitive or social psychology literature and produce a structured critique of its methodology and ethical conduct, evaluating the study against contemporary research design and ethics standards, including the British Psychological Society’s Code of Human Research Ethics.
Critiquing a published experiment is a core skill in undergraduate psychology, requiring the researcher to move beyond simply describing a study’s findings and instead to evaluate the adequacy of its design, the validity of the conclusions it supports, and the ethical standards to which it was conducted. This assignment critiques Loftus and Palmer’s (1974) classic study on the interaction between language and memory, widely regarded as one of the foundational demonstrations that eyewitness memory can be systematically distorted by the way questions are phrased after an event. The study was selected because it remains one of the most frequently cited experiments in the eyewitness testimony literature, has been highly influential in shaping legal guidance on interviewing witnesses, and illustrates a set of methodological and ethical issues that recur throughout experimental psychology more broadly, from sampling and demand characteristics to the boundaries of acceptable deception.
The critique is organised into three sections. First, it provides a brief, purely descriptive overview of the study’s design and principal findings, to establish the basis for the evaluation that follows. Second, it critically evaluates the methodology in detail, considering the experimental design, sampling strategy, measures used, and the internal and ecological validity of the conclusions drawn. Third, it evaluates the study against contemporary ethical standards, principally the British Psychological Society’s (2021) Code of Human Research Ethics, while acknowledging that the study must also be judged in the context of the norms prevailing in experimental psychology at the time it was conducted.
Critique of this kind serves a purpose beyond the individual study under consideration. Undergraduate psychology curricula place considerable weight on the ability to evaluate published research critically, precisely because published findings are frequently applied well beyond the specific conditions under which they were originally generated — in this case, the finding that leading questions distort memory has informed real-world police interviewing protocols and courtroom guidance on the reliability of eyewitness testimony, making a careful assessment of exactly how far the original evidence can be trusted especially consequential (Wells and Olson, 2003). A rigorous critique therefore has to hold two things in mind simultaneously: an honest assessment of the study’s genuine limitations, and a fair recognition of what it nonetheless demonstrates convincingly, without collapsing into either uncritical acceptance or dismissive scepticism.
Loftus and Palmer (1974) reported two linked experiments investigating whether the wording of a question could alter a witness’s memory for an event. In Experiment 1, forty-five student participants were shown seven short films of traffic accidents and then asked to describe what they had seen and answer a series of specific questions, the critical one asking, ‘About how fast were the cars going when they [verb] each other?’, with the verb varied between five independent groups: smashed, collided, bumped, hit, and contacted. Mean estimated speed varied systematically with the verb used, from the highest estimate in the ‘smashed’ condition down to the lowest estimate in the ‘contacted’ condition, despite every participant having watched an identical film.
In Experiment 2, a new sample of one hundred and fifty participants watched a single film of a car accident that in fact contained no broken glass. Participants were then divided into three groups: one group was asked the ‘smashed’ version of the speed question, a second group was asked the ‘hit’ version, and a third, control group was not asked a speed question at all. One week later, all participants were asked a further set of questions, including whether they recalled seeing any broken glass in the film. A substantially higher proportion of participants in the ‘smashed’ condition reported having seen broken glass than participants in either the ‘hit’ or control conditions, despite no broken glass having appeared in the film at all. Loftus and Palmer interpreted this pattern as evidence that the wording of a question can not only bias a witness’s immediate verbal report but can actually become integrated into the underlying memory representation of the event itself, a phenomenon they termed the reconstructive nature of memory. Notably, Experiment 2 was specifically designed to distinguish between two competing explanations for the Experiment 1 finding: a relatively superficial response-bias account, in which participants simply adjust their verbal answer to match the perceived expectation of the question without any change to the underlying memory, and a more far-reaching memory-reconstruction account, in which the biased information becomes incorporated into the memory trace itself and can therefore influence recall of entirely different, ostensibly unrelated details a week later. The broken-glass finding was intended to support the latter, stronger interpretation.
The independent-groups design used across both experiments has clear strengths: because each participant experienced only one level of the verb manipulation, the design avoids order effects and demand characteristics that could arise if the same participant were asked about the same film using multiple different verbs (Howitt and Cramer, 2017). However, the design also has an important limitation common to all independent-groups research: because different participants were allocated to each condition, individual differences between groups — such as pre-existing differences in familiarity with vehicle speeds, general estimation ability, or attentiveness while watching the films — cannot be fully ruled out as alternative explanations for the observed group differences, and the original report provides limited detail on how allocation to conditions was carried out (Coolican, 2018).
A second concern relates to the ecological validity of the experimental stimulus. Participants viewed short, staged films of traffic accidents in a controlled laboratory setting, an experience that differs substantially from witnessing a real accident in terms of emotional arousal, physical involvement, and the consequences attached to accurate recall. Real eyewitnesses to genuine accidents typically experience a level of stress and personal relevance entirely absent from watching a brief film, and there is reason to think that memory under high arousal may behave differently from memory formed under the comparatively low-stakes conditions of a laboratory viewing (Wells and Olson, 2003). This does not invalidate the study’s core finding that question wording can influence recall, but it does mean the specific magnitude of the effect, and possibly its underlying mechanism, may not generalise directly to real forensic settings without further validation.
Sampling is a further limitation. Both experiments recruited university students, a sample that is convenient for researchers but is neither representative of the general population of potential eyewitnesses nor necessarily representative even of typical drivers, since students as a group tend to be younger, more highly educated, and may have less real-world driving experience than the general adult population from which actual eyewitnesses are drawn (Rosenthal and Rosnow, 2008). This limits the confidence with which the findings can be generalised to, for example, older adults, professional drivers, or individuals with limited exposure to estimating vehicle speeds, all of whom might respond differently to the same verb manipulation.
Finally, the speed-estimation measure itself raises questions about demand characteristics. Because the critical verb was embedded within an otherwise ordinary-seeming question, and because participants were aware they were taking part in a psychology study, it is plausible that at least some participants inferred the expected direction of response from the intensity of the verb used and adjusted their answers accordingly, independent of any genuine change in memory (Orne, 1962). Loftus and Palmer’s Experiment 2 partially addresses this concern, since the broken-glass question was asked a week after the original manipulation and concerned a different, ostensibly unrelated detail, making a simple demand-characteristics explanation for that specific finding less plausible; however, the concern remains more difficult to rule out for the immediate speed-estimate findings of Experiment 1 considered on their own.
A further point concerns the reliance on a purely verbal, self-report measure of memory throughout both experiments. Numerical speed estimates and yes/no recognition judgements are convenient to collect and score, but they are indirect measures of an underlying cognitive process that cannot be observed directly, and they conflate two theoretically distinct possibilities that a behavioural or physiological measure might have helped to separate: a genuine alteration of the stored memory trace, and a purely verbal decision made at the point of report without any change to memory itself (Howitt and Cramer, 2017). Contemporary replications of this line of research have sometimes supplemented verbal report with recognition-memory testing using forced-choice image arrays, or with measures of confidence and response latency, precisely in order to gain more direct evidence about whether the underlying memory trace has actually changed, rather than relying on a single verbal outcome measure as the sole window into the process under investigation. The absence of any such supplementary measure in the original 1974 study is a reasonable methodological limitation to flag, even though it reflects entirely standard practice for experimental cognitive psychology of that period.
Judged against the British Psychological Society’s (2021) Code of Human Research Ethics, several aspects of the study warrant scrutiny, while others were relatively unproblematic even by contemporary standards. The core procedure — watching short films of traffic accidents and answering questions about them — carried minimal risk of physical or psychological harm, and is unlikely to have caused meaningful distress to participants, particularly compared with more ethically contentious studies of the same era, such as Milgram’s (1963) obedience experiments. Nonetheless, the study did involve an element of deception, in that participants were not told the true purpose of the research was to test whether question wording could distort memory; they were led to believe the study concerned memory for accidents in a more general sense. Under contemporary BPS (2021) guidance, such deception would need to be justified as necessary to the validity of the research, since informing participants in advance would almost certainly have altered their responses, and would need to be followed by a full debrief explaining the true purpose of the study and offering participants the opportunity to withdraw their data.
The original published report provides only limited detail on debriefing procedures, informed consent, and the right to withdraw, which is common for experimental papers published in this period but falls short of the explicit, itemised reporting expected of contemporary ethics applications and publications. It is likely, though not explicitly confirmed in the paper itself, that some form of debrief was provided given the standards of the psychology departments in which the research was conducted, but the absence of documented detail makes it difficult to evaluate this aspect of the study’s ethical conduct with confidence today. This illustrates a broader, generalisable point about critiquing older published studies: the absence of information in a published report is not equivalent to evidence that the corresponding ethical safeguard was absent in practice, and critique should distinguish carefully between the two.
Confidentiality also appears to have been handled appropriately in that no individually identifiable data were reported, consistent with standard practice; and the level of risk involved was low enough that, even under current, more demanding review standards, the study would likely have received ethical approval, provided a full debrief was offered and participants were given a genuine and clearly communicated opportunity to withdraw their data after learning the study’s true purpose. Comparing the study to contemporaries such as Milgram (1963) is instructive: Loftus and Palmer’s study involved far less potential for psychological distress, reflecting a considerably more conservative ethical approach even within the norms of its own era, and its lasting influence on legal and forensic practice arguably outweighs the comparatively modest ethical concerns it raises.
A formal harm–benefit analysis, as would now be expected at ethical review, would also need to weigh the mild deception and the one-week delay before debriefing against the study’s scientific and applied value. The deception was necessary to the validity of the design, since informing participants in advance that the study concerned the effect of question wording on memory would have made the manipulation transparent and almost certainly changed how participants responded; the risk of harm from the deception itself was low, given the innocuous nature of the films and questions involved. The one-week gap between the original film viewing and the follow-up memory questions, while methodologically necessary to demonstrate a lasting effect on memory rather than a momentary response bias, does mean that any participant who found the deception objectionable was not given the opportunity to object until a full week after their initial participation; a contemporary ethics committee would likely require this delay, and the reasons for it, to be explained clearly to participants in advance as part of the consent process, even if full detail about the verb manipulation itself could not be disclosed until the final debrief.
Loftus and Palmer’s (1974) study remains a landmark demonstration that the wording of a question can shape eyewitness memory, and its influence on the development of the cognitive interview technique and on legal guidance for interviewing witnesses is considerable. This critique has identified genuine methodological limitations relating to sample representativeness, ecological validity, and the possible role of demand characteristics, alongside ethical questions concerning the reporting of consent, deception, and debriefing procedures that would need to be addressed more explicitly under contemporary BPS (2021) guidance. None of these limitations, however, substantially undermines the study’s central contribution. A contemporary replication addressing these concerns might employ more ecologically valid, immersive stimuli such as virtual-reality accident simulations, recruit a more demographically diverse community sample beyond the student population, pre-register hypotheses and analysis plans to guard against demand-characteristic concerns, and report informed consent and debriefing procedures in full detail consistent with current publication standards. Such a replication would allow the field to determine whether the original effect, discovered fifty years ago, still generalises as robustly to memory formed under the conditions real eyewitnesses actually experience.
Need a Model Assignment Written to Your Exact Brief?
Our 350+ UK-qualified writers deliver referenced model documents from £15 per 250 words, with free plagiarism and AI-detection reports.
You May Also Like
All work is written by human writers. 100% AI free, guaranteed.
100% money back guarantee if you find plagiarism in our work.
COMPANY DETAILS
