correlations

• what is a correlation

  • a way of measuring the extent to which two variables are related
  • measures the pattern of responses across variables
  • no IVs and DVs
  • should generally have 100 + participants for reliable results
• Measuring correlation

  -       Pearson correlation coefficient ‘r’

  -       R = - 1 = perfect negative correlation

  -       R = +1 perfect positive correlation

  -       R = 0 no correlation
• What are we trying to look for in correlation and regression?

  Look for relations (usually linear) between variables
• What is the difference between correlation and regression??

  ·      Correlation looks for the (symmetric) relation between variables

  ·      Correlation assumes bivariate normality

  ·      Regression examines how one or more variables (Xs) can predict another variable (Y)
• What is partial correlation?

  ·      the correlation between variables 1 & 2, when the effects of variable 3 have been partialled out
• ■      This can be done by calculating the Covariance.

  –      We look at how much each score deviates from the mean.

  –      If both variables deviate from the mean by the same amount, they are likely to be related.

   
• Effect size Cohen (1988) [slide 25]

  •       Correlation strength varies between -1.0 and +1.0

         0 = no relationship,   ‘–’ negative, ‘+’ positive

  •       Correlation is an effect size measure itself!

  ± .10 to ± .29 = small effect

  ± .30 to ±.49 = medium effect    

  ± .50 to ±  1.0 = large effect