Correlation

• Correlations are used to test a relationship between 2 variables & experiments are a test of difference.
• Correlational analysis- used to investigate the relationship between 2 variables, eg obesity & heart attacks.
• Carrying out correlational analyses:

  1. Decide which variables you want to study, eg height & shoe size.
  2. Set a hypothesis- directional or non-directional hypothesis, if it's directional you are predicting the direction of the relationship.
• • Positive correlation= as one increases, so does the other
  • Negative correlation= as one increases, the other decreases
  • No correlation= no relationship/ association shown
• Calculating the correlation co-efficient:

  • This is a statistical technique which allows you to see if your relationship is strong enough to be significant, ie worth basing a conclusion on.
  • It gives the correlation a value from +1 to -1.
  • The closer the value is to +1 or -1, the stronger the relationship.
  • Zero indicates no relationship (& a positive isn't better than a negative, they just show different relationships).
• Correlational analysis Strengths:

  • Easy, quick, economical method, just need two sets of data (often collected by someone else like the government), no manipulation of variables, no need for controlled environment.
  • May form the basis for further research.
  • Can study relationships that it would be unethical to purposely manipulate (eg links between stress & illness).
• Correlational analysis Weaknesses:

  • Impossible to infer cause & effect, as IV is not manipulated.
  • Other variables could cause the relationship measured.
  • Only measures linear relationships- such a relationship would have a low correlation co-efficient but it is actually very strong.