Health Sciences 3801A/B Lecture 4: Correlation & Regression
Document Summary
Scatterplots & direction: good place to start with correlational analysis - scatterplot. Three important pieces of information: directions of relationship, form of a relationship, degree of a relationship, correlations only appropriate for linear relationships. Curvilinear relationship : movement time speeds up until early 20"s, and then begins to slow down. Imagine a line drawn through the middle of an array of points as close as possible to all points. Line of best fit: draw an ellipse around outside of all points, degree of relationship increases as line becomes steeper, and as scatter decreases. Types of coefficients: pearson product-moment correlation, two interval/ratio variables, spearman rank order correlation, two ordinal variables, one ordinal variable, one interval/ratio variable, point-biserial correlation, one variable interval/ratio, one dichotomous, phi coefficient, both variables measures on a dichotomous scale. Dummy-coding means assigning numerical values to category values (e. g. , for sex, coding male as 0 and female as 1)