Correlational Analysis: Co-Variables, Correlation Coefficients and Scattergrams - Worksheets, Questions and Revision
12 original exam-style questions - 4 pages of questions with a full mark scheme - free printable PDF.
Correlational Analysis: Co-Variables, Correlation Coefficients and Scattergrams revision guideWrite-in editionMore A Level Psychology practiceMock exams
This topic is chapter 7 of A Level Psychology Topic Workbook 4 and chapter 26 of A Level Psychology 10-Minute Tests 2.
7.11 Correlational Analysis: Co-Variables, Correlation Coefficients and Scattergrams
Question 1
- defines co-variable as a measurable variable that can vary and be measured in a correlation study, not manipulated oeB1
- gives an appropriate example, e.g. hours of exercise per week and resting heart rate, or stress level and sleep duration oeB1
- Answer: A co-variable is a measurable variable that can vary and be measured in a correlational study rather than being manipulated. Example: hours of exercise per week and resting heart rate.
Question 2
- identifies the directionality problem and explains it weakens causal claims because correlation cannot show which variable came first oeB1
- identifies the third-variable problem and explains it weakens causal claims because an unmeasured factor may cause both correlated variables oeB1
- identifies measurement or reliability issues and explains that poor measurement can hide or create spurious correlations, reducing confidence in causal interpretation oeB1
- identifies that correlations cannot detect non-linear relationships if a linear coefficient is used, and explains this could lead to incorrect conclusions about absence of a relationship oeB1
- Answer: Limitations: (1) Directionality problem: correlations cannot show which variable causes the other, so causal claims are weakened. (2) Third-variable problem: an unmeasured factor may cause both variables, so the observed association is not causal. Additional limitations include measurement reliability issues which can obscure true effects, and linear correlation measures missing non-linear relationships.
Question 3
- (i) positive correlation, more revision associated with higher exam score, oeB1
- (ii) negative correlation, more cigarettes associated with lower lung capacity, oeB1
- (iii) zero correlation, number of fingers normally does not vary with stress level, or no systematic relationship oeB1
- Answer: (i) Positive correlation. (ii) Negative correlation. (iii) Zero correlation.
Question 4
- plots all five points in their correct positions on a labelled axes system (hours sleep on x axis, attention score on y axis) awarding the method markM1
- states plotted coordinates correctly: A (6,62), B (8,78), C (5,55), D (7,73), E (4,50) oeB1
- labels axes with units or scales for hours and attention score oeB1
- provides a brief comment on overall pattern, e.g. points show a positive association, attention tends to increase with hours sleep oeB1
- Answer: Plotted coordinates: A (6,62), B (8,78), C (5,55), D (7,73), E (4,50). Method mark awarded for plotting all five points correctly on labelled axes with a consistent scale; the plotted pattern shows a positive association where attention tends to increase with more hours of sleep.
Question 5
- states direction: r = 0.78 is a positive correlation, as r is greater than zero oeB1
- states strength: 0.78 indicates a strong relationship, fairly close to +1 oeB1
- gives limitation: correlation does not imply causation, or there may be a third variable affecting both exercise and wellbeing, oeB1
- Answer: r = 0.78 indicates a strong positive relationship, so higher weekly exercise is associated with higher wellbeing. A limitation is that the coefficient alone does not show causation and could reflect a third variable or directionality problems.
Question 6
- states that correlations measure association, not manipulation of an IV, so no variable has been experimentally manipulated to establish causation oeB1
- states directionality problem: cannot tell which variable, if either, causes the other oeB1
- states third-variable problem: a separate factor could be causing both correlated variables oeB1
- gives a plausible example of a third variable, e.g. socio-economic status affecting both exercise levels and wellbeing, or personality influencing both stress and sleep duration oeB1
- Answer: Correlations show associations but do not manipulate an independent variable, so they cannot prove cause and effect. There is a directionality problem, and a third-variable could cause both measured variables. Example third variable: socio-economic status could influence both exercise levels and self-reported wellbeing.
Question 7
- describes pattern: as hours of study increase, number of social activities tends to decrease oeB1
- identifies correlation as negative oeB1
- comments on strength: gives plausible qualifier, e.g. weak/moderate/strong negative association based on how tightly points cluster oeB1
- gives a practical implication, e.g. universities may consider that encouraging more study time could reduce social engagement, or they should design support so students can balance study and social life oeB1
- Answer: The scattergram shows a negative correlation: as study hours increase, social activities tend to decrease. Depending on clustering, this could be described as a moderate negative association. A practical implication is that support for student wellbeing should consider how increased study requirements may reduce social engagement, so timetabling could help students balance both.
Question 8
- plots all four points in their correct positions on labelled axes, awarding the method markM1
- states plotted coordinates correctly: P1 (2,8.0), P2 (4,7.0), P3 (6,6.0), P4 (8,5.0) oeB1
- explains that the points lie roughly on a straight line sloping downwards, showing that as x increases y decreases, which is a linear negative correlation oeB1
- Answer: Plotted coordinates: P1 (2,8.0), P2 (4,7.0), P3 (6,6.0), P4 (8,5.0). Method mark for plotting all four correctly with labelled axes. These points lie close to a straight line sloping downwards, so higher daily screen time is associated with fewer hours of sleep, illustrating a linear negative correlation.
Question 9
- states direction: r = -0.05 is a very small negative correlation, slightly below zero oeB1
- states strength: -0.05 indicates a negligible or almost zero relationship, very close to 0 oeB1
- gives two reasons not to conclude no effect: e.g. measurement error or low reliability of the happiness measure, or a third variable masking an effect, or a non-linear relationship that Pearson r would not capture oeB1
- Answer: r = -0.05 indicates a very small negative relationship, essentially negligible. However, this does not prove pets have no effect: measurement error or poor reliability could hide a real effect, a third variable could be masking a relationship, or the true relationship might be non-linear and so not well summarised by Pearson r.
Question 10
- useful for studying variables that cannot be ethically or practically manipulated, e.g. correlation between smoking and lung disease where experimental manipulation would be unethical oeB1
- useful for establishing relationships to generate hypotheses for later experimental testing, e.g. finding a correlation between social media use and anxiety could lead to experimental studies or longitudinal research oeB1
- Answer: Correlational studies are useful for examining relationships where manipulation is unethical or impossible, for example smoking and lung disease. They are also useful for identifying associations that suggest hypotheses for later experimental or longitudinal study, for example an observed link between social media use and anxiety.
Question 11
- states direction: r = -0.64 is a negative correlation oeM1
- states strength: -0.64 indicates a moderately strong to strong negative relationship, fairly far from 0 toward -1 oeA1
- explains meaning in context: higher stress scores are associated with fewer hours of sleep per night, on average oeA1
- evaluation point: practical significance, e.g. the strength suggests a meaningful association that might warrant workplace interventions to reduce stress to improve sleep oeB1
- evaluation point: limitation due to causality, e.g. directionality problem means it could be that poor sleep increases stress rather than stress reducing sleep oeB1
- evaluation point: third-variable or sampling issues, e.g. a health condition or shift-working status might cause both high stress and reduced sleep, or the sample may not generalise to all workers oeB1
- Answer: r = -0.64 indicates a moderately strong negative relationship: as stress increases, hours of sleep tend to decrease. Practically, this suggests a meaningful association that might justify interventions to reduce workplace stress to improve sleep. Limitations are that causality is not established, since sleep problems might cause stress (directionality), and a third variable such as shift work or health problems could underlie the association; the sample may also limit generalisability.
Question 12
- Level 4 Knowledge of correlational analysis is accurate and well detailed. The discussion is thorough and balanced, showing clear understanding of how correlations differ from experiments and giving a range of relevant strengths and limitations, developed with reasons and examples. Realistic applications or study examples are used to support points. The answer is coherent, well organised and uses appropriate specialist terms.7-8
- Level 3 Knowledge of correlational analysis is clear with some good detail. The discussion includes both strengths and limitations with some development and examples, though one area may be less fully developed. Differences from experimental methods are explained. The answer is mostly organised and uses specialist terms appropriately.5-6
- Level 2 Knowledge is present but limited. The answer includes basic points about correlational analysis and at least one strength or limitation, but explanations lack depth or development. Comparisons with experiments may be superficial. Organisation may be uneven and specialist terms used inconsistently.3-4
- Level 1 Knowledge is very limited. The answer may be a list of brief, undeveloped points. There is little or no effective evaluation and little comparison with experimental methods.1-2
- Level 0 No creditable content.0
- Indicative content:
- AO1: Correlational analysis measures the association between two co-variables, producing a coefficient (for example Pearson r) between -1 and +1 that indicates direction and strength of a linear relationship.
- AO1: Scattergrams are used to display paired data and to visualise positive, negative or zero correlations and the degree of scatter around a relationship.
- AO1: Correlations differ from experiments because no independent variable is manipulated and no random allocation to conditions occurs, so correlational designs do not establish cause and effect.
- AO3: Strength - ethical and practical usefulness: correlations allow study of variables that cannot be manipulated for ethical or practical reasons, for example the relationship between smoking and lung cancer, or trauma exposure and later mental health.
- AO3: Strength - useful for prediction and hypothesis generation: a reliable correlation, for example between exam attendance and grades, can be used for predictive purposes and to generate causal hypotheses to test experimentally or longitudinally.
- AO3: Limitation - directionality problem: a correlation between A and B does not show whether A causes B or B causes A, for example between stress and poor sleep.
- AO3: Limitation - third-variable problem: an unmeasured variable may cause the association, for example socio-economic status affecting both nutrition and cognitive test performance.
- AO3: Limitation - correlations do not capture non-linear relationships well when summarised by a single linear coefficient, and measurement error can obscure real relationships.
- AO3: Application and evaluation: consider real studies or plausible examples, weighing practical value against limitations and suggesting ways to strengthen evidence (triangulation, longitudinal designs, statistical control for third variables, or follow-up experiments).
Mark your answers
Check answers without an account. Some answers can be checked automatically; use the mark scheme to mark the rest.