Correlations

• Correlations


  method of data analysis

  - measures strength and direction of relationship between 2 or more variables - see a trend or pattern.

  - calculate a correlation co-efficient.

  - ranges from +1.0 (perfect positive correlation), to -1.0 (perfect negative correlation)

  - no correlation = 0.0
• Positive Correlation


  one variable increase, the other variable increases
• Negative Correlation


  one variables increases, the other variable decreases
• Positive Correlation example


  people rated as less attractive tend to choose less attractive dates

  
• Negative Correlation example


  the higher the number of vaccinations, the lower the occurrences of illness

  
• No Correlation example


  there's no relationship between intelligence and amount of ice cream eaten
• Curvilinear


  as temperature increase so do level of aggression, but as temperature continues to increase, levels of aggression decrease
• correlation co-efficient


  0- 0.33/-0.33 - weak

  0.33/-0.33 - 0.66/-0.66 - moderate

  0.66/-0.66 - 1/-1 - strong
• Strengths


  allows predictions to be made.

  allows quantification of relationships -> correlations can show strength of relationship between 2 variables.

  no manipulation -> don't require manipulation of behaviour

  quick and ethically sound.
• Weaknesses


  cause and effect can't be inferred -> no IV or DV.

  extraneous relationships -> other variables may influence the measured variables.

  problems in reading the result -> lack of correlation may indicate no relationship

  only works for linear relationships -> e.g. relationship between arousal + performance - curvilinear.
• Differences between experiments and correlations

  correlational study looks at relationship between 2 co-variables whereas an experiment point to cause and effect (manipulation of IV caused a change in DV)

  an alternative (correlational) hypothesis will predict a relationship.

  an experimental hypothesis will predict a differences
• Examples of Correlational Hypothesis

  non-directional - there will be a correlation between the amount of revision and test score.

  directional - there will be a positive correlation between amount of revision and test score.

  null - there will be no relationship between revision and test score.