correlations Flashcards
(12 cards)
What is a correlation in psychology?
A statistical technique used to assess the strength and direction of a relationship between two co-variables. Unlike experiments, no IV is manipulated.
What are co-variables?
The two variables measured in a correlation to see if there is an association between them. They must be quantitative.
What are the three types of correlation?
• Positive: both variables increase together.
• Negative: one increases, the other decreases.
• Zero: no consistent relationship.
Strengths of using correlations?
• Useful for identifying relationships that may warrant further experimental study.
• Allows research where manipulation would be unethical or impractical (e.g. smoking & cancer).
Weaknesses of correlations?
• Cannot establish cause and effect.
• May be affected by third variables (extraneous/confounding).
What is a correlation coefficient?
A number between +1 and -1 indicating the strength and direction of a correlation.
• +1 = perfect positive
• -1 = perfect negative
• 0 = no correlation
What graphical method is used to show a correlation?
A scattergram (scattergraph), which plots each co-variable on one axis to show the relationship visually.
What does a scattergram show that a correlation coefficient does not?
• Outliers
• Distribution pattern
• Visual representation of the relationship’s direction and spread
When might a psychologist choose a correlation over an experiment?
• When it’s unethical/impractical to manipulate variables
• When exploring natural relationships (e.g. stress & illness)
Why is it incorrect to say “correlation proves causation”?
Because correlations do not involve manipulation of variables or control over extraneous factors, so third variables may explain the relationship.
What statistical test is typically used for correlations in psychology?
Spearman’s rho (used for ordinal data or when assumptions for Pearson’s are not met).
What ethical issue may arise when using correlations?
Correlations can be misinterpreted or overgeneralised, especially in sensitive areas (e.g. mental health). Researchers must avoid making unsupported causal claims.
