Correlation

• correlation is the strength of the relationship
• regression is the nature of a relationship
• scatterplots can be used to visualise the relationship
• correlation coefficient has value range of -1 to +1
• correlation coefficient tells the direction and strength of the correlation
• if r= -1 or +1 then there is a linear relationship
• r values

  A) very weak

  B) weak

  C) moderate

  D) strong

  E) very strong

  
• r value closer to 0 is weaker
• strong r shows less likely due to chance
• Pearson's correlation is the significance test
• parametric test assumptions

  linear relationship, no outliers, normal distribution, no heteroscedascity
• if the parametric assumptions are not met you would use the Spearmans rank correlation
• What test do we use if the parametric assumptions are not met?

  Spearmans rank correlation