PSYC 2030 Lecture Notes - Lecture 3: Scatter Plot, Factor Analysis, Partial Correlation
Document Summary
This method does not control or manipulate variables, instead it examines the relationship b/t variables as they naturally occur. Also correlation helps to make prediction of behaviors. Stats that changes in the value of one variable are accompanied by predictable changes in a second variable. There are only 2 variables to see if they correlate. Ask people how much cigarette you smoke per day. Adapting the previous example, in a e, the researcher might ask some people to smoke 1 cigarette / day, 2 cigarette / day and then check their lungs. A correlation exists whenever 2 variables are associated or related to each other in some fashion. Two components: 1) number 2) sign (+ / -: number: Strength of the correlation, the greater the strength, the greater predictable they are. Range: 0 ~ 1 (not possible to have anything > 1) The value does not mean : sign: