Guiding question: Does the student analyse the information presented in the essay and produce a coherent line of argument?
If the work does not reach a standard outlined by the performance level descriptors, 0 marks are awarded for this criterion.
| Number of Marks | Analysis | Line of Argument |
|---|
| 1-2 | The essay is descriptive rather than analytical. | A partial line of argument is present. |
| 3-4 | The essay includes analysis that is partially effective and produces some relevant findings. | A partially consistent line of argument links the research question, research findings and conclusions. |
| 5-6 | Analysis in the essay is effective and consistently produces relevant findings | A clear, sustained line of argument links the research question, research findings and conclusions. |
The way you summarise and interpret that research will directly determine how examiners judge your achievement in Criterion C (analysis and argument) and Criterion D (discussion and evaluation).
Recording evidence is not about compiling every detail from every study. It is about showing that you understand which aspects of the research are central to your question.
For instance, if your essay explores the relationship between social media use and body image, you do not need to reproduce the entire method of each study you cite.
Instead, you need to highlight the parts of the study that are relevant:
Psychological research can be quantitative, qualitative, or mixed-methods.
Both forms of evidence can be included in an EE, but they must always be interpreted rather than listed.
In both cases, evidence is presented not as raw data but as interpreted information that advances your essay’s argument.
A common weakness in Psychology EEs is including either too much or too little detail.
This structure ensures that every study you record has both context and purpose in your essay.
To see the difference, compare these two ways of presenting the same study:
Weak (descriptive):
Sherman et al. (2016) studied adolescents using fMRI and found that participants’ reward centres were activated when their photos received more likes. The study concluded that social media likes influence neural responses.
Strong (analytical):
Sherman et al. (2016) used fMRI with adolescents and found increased activation in reward-related brain regions when participants’ photos received more likes. This shows that social approval on social media can engage neural mechanisms similar to those activated by primary rewards. This supports the view that digital feedback contributes to reinforcing behaviour, helping to explain why adolescents may be particularly sensitive to online approval in ways consistent with principles of reinforcement learning.
The strong version records the evidence but also interprets and links it to psychological theory, moving beyond description to analysis.
As the EE is not just a literature review, presenting evidence requires you to synthesise multiple studies rather than summarising them in isolation.
Becker et al. (2002) showed that the introduction of television to Fiji was associated with an increase in reported eating disorder symptoms among adolescent girls, suggesting that media exposure influences body ideals. Similarly, Fardouly et al. (2015) found that time spent on social media correlated with body dissatisfaction in Australian adolescents. Together, these findings suggest that both traditional media and newer digital platforms can amplify cultural pressures around body image, lending support to the sociocultural model of eating disorders.
Here, two studies are recorded, compared, and tied into a broader theoretical framework.
This kind of integration shows examiners that the student can process evidence into a line of argument.
Weak (descriptive):
Loftus and Palmer (1974) found that participants estimated higher speeds when asked how fast cars were going when they ‘smashed’ rather than ‘hit’ each other. This shows that memory can be influenced by wording of questions.
Strong (analytical):
“Loftus and Palmer (1974) demonstrated that eyewitness memory can be reconstructed based on leading questions, supporting the idea that recall is not a simple retrieval of stored events but a reconstructive process. However, demand characteristics may have influenced the results, as participants could have guessed the aim of the study. This raises questions about ecological validity, since memory for real accidents may not be as easily manipulated. Comparing these findings with Yuille and Cutshall’s (1986) real-life crime study highlights the importance of context: while Loftus and Palmer suggest memory is highly malleable, naturalistic evidence shows that emotional salience and real-world settings may produce more reliable recall. Taken together, these findings suggest that memory is reconstructive, but the extent of distortion depends on context.”
This shows how to move from reporting a finding to evaluating its strengths, weaknesses, and relevance to the RQ.
Every time you record evidence, ask:
Students who present evidence without linking it to theory risk producing an essay that feels like a list of findings.
Students who consistently tie findings to theory demonstrate the deep understanding expected for top marks in Criterion B and C.
Using case studies like these demonstrates evaluative skill.
For instance, saying that a study “proves” a theory is inaccurate; instead, you might write that a study “supports” or “provides evidence consistent with” a theory.
Reliability and validity in psychology play a role similar to accuracy and precision in the sciences.
When writing a study summary, make sure each paragraph includes:
If any of these elements are missing, your recorded evidence may come across as descriptive rather than analytical.
Done well, this approach ensures that your essay reads as a critical investigation rather than a descriptive report.
A strong EE constantly asks:
By weaving these questions into every section of analysis, the essay moves from description to critical engagement.
Exposure to idealised body images has consistently been linked to body dissatisfaction in adolescents. Becker et al. (2002) found that the introduction of television in Fiji was associated with increased eating disorder symptoms, demonstrating the powerful effect of media imagery. Fardouly et al. (2015) extended this finding to digital platforms, reporting a positive correlation between Facebook use and body dissatisfaction. However, Ferguson et al. (2014) found only a small effect in a large longitudinal sample, suggesting that individual vulnerability factors may moderate this relationship. Together, these studies support the sociocultural model but also highlight the need to account for individual differences when interpreting media effects.
This single paragraph shows synthesis: multiple sources are compared, theory is referenced, and nuance is introduced.
Theories prevent the essay from drifting into unrelated results and keep the discussion anchored in psychology.
A high-level EE demonstrates the ability to compare evidence. This can be done by:
While Sherman et al. (2016) found that social approval on Instagram activated reward-related brain regions in adolescents, Odgers and Jensen (2020) concluded that overall links between social media use and adolescent mental health are weak and inconsistent. The difference may lie in the level of analysis: neural responses may capture immediate reinforcement processes, whereas survey-based studies may underestimate these effects over time. This contrast illustrates the complexity of linking brain-level findings to broad behavioural outcomes.
Such comparisons demonstrate analytical maturity and are rewarded under Criterion C.
Weak paragraph (descriptive):
Loftus and Palmer (1974) investigated the effect of leading questions on memory. They showed participants a car crash video and asked about the speed of the cars. Depending on the verb used, participants’ speed estimates differed. This shows that memory can be distorted by language.
Strong paragraph (analytical):
Loftus and Palmer (1974) demonstrated that leading questions influence eyewitness memory, with participants giving higher speed estimates when the verb ‘smashed’ was used. This supports the idea that recall is not a direct record of events but is reconstructed based on schema-consistent cues. However, the artificial setting raises concerns about ecological validity, as real eyewitnesses may experience higher stress and different attentional processes. The study therefore provides strong experimental evidence for schema theory but has limited generalisability to real-world testimony.
The second version processes the evidence into argument, balancing strengths, limitations, and theoretical significance.
One of the biggest pitfalls is the “shopping list” essay (where each study is presented in sequence with minimal connection between them).
For instance:
“Study A found this. Study B found that. Study C found another thing.”
Such linking phrases demonstrate that the student is processing evidence, not just listing it.
Qualitative studies must also be analysed critically.
Willig (2000) interviewed women with eating disorders and identified themes of control and identity. These insights highlight lived experience that cannot be captured in quantitative surveys. However, qualitative methods rely heavily on interpretation, and researcher bias may shape the themes identified. Triangulating qualitative findings with quantitative data strengthens the overall argument, showing how subjective experience aligns with broader patterns.
This shows the examiner that qualitative evidence is not just quoted but evaluated in terms of credibility and relevance.
Balanced writing demonstrates awareness of limitations and earns higher marks in evaluation.
The examiner should feel, at every stage, that you are building toward a coherent, evidence-based conclusion.