A short week for IT research methods in terms of new material. Due to the literature review presentations we did not have a tutorial and only half a lecture. The topic of the lecture was ‘Correlation Analysis’, presented by Joze Kuzic.
Lets start with the simple definition of correlation analysis, ‘A statistical investigation of the relationship between one factor and one or more other factors’.
One point that I need reminding on was correlation vs regression (source: http://www.psych.utoronto.ca/courses/c1/chap9/chap9.html):
Correlation – both variables are random variables, and 2) the end goal is simply to find a number that expresses the relation between the variables
Regression – one of the variables is a fixed variable, and 2) the end goal is use the measure of relation to predict values of the random variable based on values of the fixed variable
The topic of causality and correlation was approached quite carefully in the lecture notes citing that correlation can be used to look for causality but does not infer causality.
Methods of correlations:
Pearson’s correlation coefficient – for parametric (randomized, normally distributed data).
Spearman rank order correlation coefficient – for non-parametric data, [-1.0 , 1.0]
Significance of correlations was the next logical point covered, not much mathematical reasoning was covered apart from p < 0.05 is good :).