Health Sciences 3801A/B Lecture Notes - Lecture 3: Scatter Plot, Categorical Variable, Standard Error
26 Feb 2018
School
Department
Course
Professor
Document Summary
Scatter plot: visual representation of the relationship between 2 variables. Fitting a straight line into a scatter plot: y"=a + (cid:271)x (predicted, y=a + bx + e (e = residual/error) In long run, e"s will (cid:272)an(cid:272)el each other out. The issue of goodness of fit is concerned with how well the regression line fits the data: assessed by comparing the predicted scores (on the regression line) with the observed scores. Slope of a regression line is tied to the units in which the independent variable is measured. Uses z scores rather than units in which the variables were originally measured zy" = zx. Standardized slope ( ) is calculated from the unstandardized slope (b) Continuous variable involves different amounts of something i. e. scores on vocabulary test, blood pressure reading. Categorical variable involves different kinds of something i. e. type of blood pressure monitor, type of teas.
