Difficulty: Intermediate | Prerequisites: Introductory psychology or research methods coursework. Familiarity with basic statistical concepts (mean, ANOVA) is helpful.
This topic sits at the intersection of research methodology and social psychology. Evaluating the strengths and weaknesses of published studies is a core skill in any psychology programme, and exams frequently ask you to critique a study's design, sampling or statistical approach. The source material centres on Puente and Cohen's research into jealousy and intimate partner violence, which makes it a useful case study for practising methodological critique. You should already be comfortable with the difference between independent and dependent variables, and have a rough sense of what ANOVA does.
Puente and Cohen studied physiological responses to jealousy-related versus non-jealousy-related conflict scenarios, using an ANOVA design with multiple conditions. Their study's main weaknesses are the impression management effect (participants editing their own responses) and convenience sampling (intro-psych students only). Its strengths lie in the choice of statistical test and the systematic use of multiple experimental conditions.
Impression management effect
The tendency for research participants to adjust their responses to present themselves in a socially desirable way, rather than answering truthfully. In simple terms, people tell the experimenter what they think sounds good, not what they really feel.
Convenience sampling
A non-probability sampling method in which participants are selected because they are easy to access (e.g. students in a class the researcher teaches), rather than being randomly drawn from the broader population. Think of it as picking whoever is nearby rather than who is representative.
ANOVA (Analysis of Variance)
A statistical test that compares the means of two or more groups to determine whether the differences between them are statistically significant. In simple terms, it tells you whether the groups in your experiment are performing differently enough that the gap is unlikely to be down to chance.
Just-world belief
The cognitive bias that people get what they deserve, which can lead observers to blame victims for their own misfortune. Think of it as the assumption that bad things only happen to people who brought it on themselves.
Physiological response
Measurable bodily reactions (heart rate, skin conductance, blood pressure) used as indicators of emotional or psychological arousal. In simple terms, it is the body's involuntary reaction to a stimulus, which researchers can record even when a participant is not consciously aware of it.
Cross-sectional study
A research design that collects data from different groups (e.g. different ages or marital statuses) at a single point in time, rather than tracking the same people over months or years. Think of it as a snapshot comparing several groups side by side.
Independent variable (IV)
The variable a researcher manipulates or categorises participants by in order to observe its effect. In Puente and Cohen's proposed follow-up, the IVs would be age and marital status.
Dependent variable (DV)
The variable being measured as the outcome of the experiment. In the proposed follow-up, the DV is the level of jealousy-related response.
Impression management effect (self-report bias)
Self-reported studies rely on participants telling the truth, but participants often adjust their answers to look socially acceptable.
This bias can inflate or deflate the data in ways the experimenter cannot control.
Example from the study: when presented with a scenario in which a wife was physically abused after flirting, participants who hold a just-world belief were likely to blame the victim and endorse the husband's behaviour. Their self-reports may reflect that belief rather than a considered judgement, skewing the data.
The core problem is that the researcher cannot separate the participant's true reaction from the version they chose to present.
Convenience sampling
Puente and Cohen drew their sample from students in an introductory psychology class, a textbook case of convenience sampling.
Intro-psych students tend to be young, university-educated, and Western. Their physiological and attitudinal responses to jealousy-violence scenarios may not generalise to older adults, non-students, or people from different cultural backgrounds.
Convenience samples are cheaper and faster to recruit, but the trade-off is a serious threat to external validity: you cannot confidently say the findings apply beyond the sample.
The biased responses this produces are not random noise; they are systematically skewed toward whatever the convenient population has in common.
Use of ANOVA as the statistical test
ANOVA was well suited to the research question because the study needed to compare mean physiological reactions across several conditions (jealousy-related versus non-jealousy-related conflicts).
It allowed the researchers to test whether the observed differences between conditions were statistically significant, rather than relying on descriptive comparisons alone.
Choosing the right statistical test strengthens the internal validity of the conclusions.
Multiple experimental conditions
The study did not rely on a single comparison. It included several conditions: sexual versus emotional abuse, jealousy-related conflicts versus non-jealousy-related conflicts.
Varying the conditions in this way triggered participants' non-conscious physiological responses across different types of scenario, reducing the risk that results were driven by one particular framing.
This systematic approach to hypothesis testing makes the findings more robust, because positive results across multiple conditions are harder to explain away as artefacts of a single experimental setup.
Research question
Do older, married partners respond differently to jealousy-related events compared with younger, unmarried partners?
Hypothesis
Older and married couples, who tend to have more relationship experience and emotional maturity, are less likely to engage in aggressive behaviours during jealousy-related situations than younger and unmarried couples.
Design
A cross-sectional study comparing two groups at the same point in time.
Independent variables: age and marital status.
Dependent variable: the effects of jealousy (measured through physiological or behavioural indicators, following Puente and Cohen's approach).
The first wave tests young, unmarried participants. The second wave tests older, married participants under the same conditions.
By holding the experimental conditions constant and varying only the sample demographics, the study can isolate whether age and marital status are significant predictors of jealousy-related aggression.
What this would add
If the follow-up finds a significant difference, it would suggest that Puente and Cohen's original findings are moderated by life stage and relationship status, which would also highlight the limits of drawing conclusions from a convenience sample of young students.
Research on jealousy and intimate partner violence informs domestic violence prevention programmes, court assessments in abuse cases, and couples therapy approaches. Understanding how people's physiological responses differ across types of conflict helps clinicians design better interventions.
The methodological critique itself is transferable: every time you read a psychology paper, knowing how to spot self-report bias and sampling limitations lets you judge whether the findings are worth applying.
Students often think convenience sampling is always bad. It is not inherently invalid; the problem is that it limits generalisability. A convenience sample can still produce internally valid results within its own population.
Students sometimes confuse the impression management effect with demand characteristics. Impression management is about the participant's desire to look good in general; demand characteristics are about the participant figuring out the study's hypothesis and responding accordingly.
A common error is assuming that physiological measures are perfectly objective. While they bypass self-report bias, they can still be influenced by extraneous factors (caffeine, stress from the lab environment, individual differences in baseline arousal).
Students occasionally treat ANOVA as though it tells you which groups differ. ANOVA tells you that at least one group mean differs significantly from the others. To find out which specific groups differ, you need post-hoc tests (e.g. Tukey's HSD).
Critiquing a study's methodology is one of the most common exam question formats in social psychology. Be ready to name specific weaknesses and strengths with examples from the study.
Expect questions that ask you to identify the independent and dependent variables in a proposed follow-up study. Make sure you can do this quickly.
ANOVA comes up frequently in methods-focused exam questions. Know what it tests (differences between group means) and what it does not tell you (which specific groups differ).
Questions about external validity often use convenience sampling as their example. Be prepared to explain why a student sample limits generalisability, and to suggest a more representative alternative.
True or False: Convenience sampling always produces invalid results. (False. It limits generalisability but does not automatically invalidate the findings within the sampled population.)
Fill in the blank: The impression management effect occurs when participants adjust their responses to appear more __________ __________. (socially desirable)
True or False: ANOVA can tell you exactly which groups are significantly different from each other. (False. It tells you that at least one group differs; post-hoc tests identify which ones.)
Fill in the blank: In the proposed follow-up study, age and marital status are the __________ variables. (independent)
True or False: Physiological measures are completely free from bias. (False. They bypass self-report bias but can still be affected by extraneous factors.)
Q: Identify two methodological weaknesses of Puente and Cohen's study on jealousy and violence, and explain how each could affect the results.
A: First, the impression management effect: because the study relied on self-report measures, participants may have altered their responses to appear more socially acceptable, introducing systematic bias. Second, convenience sampling: drawing participants exclusively from an introductory psychology class means the sample is not representative of the general population, limiting the external validity of the findings.
Q: What statistical test did Puente and Cohen use, and why was it appropriate for their research design?
A: They used ANOVA (Analysis of Variance). It was appropriate because the study compared mean physiological responses across multiple conditions (jealousy-related versus non-jealousy-related conflicts), and ANOVA is designed to test whether group means differ significantly.
Q: Describe a follow-up study that could extend Puente and Cohen's findings. Include the independent and dependent variables.
A: A cross-sectional study comparing older, married couples with younger, unmarried couples on their responses to jealousy-related scenarios. The independent variables would be age and marital status. The dependent variable would be the level of jealousy-related response (measured physiologically or behaviourally). This design could reveal whether the original findings are moderated by life stage.
Q: Why might the just-world belief be relevant to interpreting participants' responses in a study on intimate partner violence?
A: Participants who hold a just-world belief tend to assume that victims are at least partly responsible for what happens to them. In a jealousy-violence study, this could lead participants to blame the victim and endorse the aggressor's behaviour, skewing self-report data in ways that do not reflect a measured judgement about the scenario.
This material connects directly to research methods in psychology: sampling techniques, internal versus external validity, and choosing appropriate statistical tests are covered in nearly every introductory methods unit. The just-world belief ties into attribution theory and cognitive biases, which appear throughout social psychology. The proposed follow-up study introduces cross-sectional design, which connects to developmental psychology's methods for studying age-related differences.
Impression management, social desirability bias, self-report bias, convenience sampling, non-probability sampling, ANOVA, analysis of variance, F-test, just-world hypothesis, just-world belief, victim blaming, physiological arousal, skin conductance, cross-sectional design, between-subjects design, internal validity, external validity, generalisability, Puente and Cohen, jealousy and violence, intimate partner violence, IPV, research methods in psychology, experimental conditions, post-hoc tests, Tukey HSD