PSYC 3380 Lecture Notes - Lecture 10: Scatter Plot, Homoscedasticity, Regression Analysis
Document Summary
Standardized variables in a standard metric, mean = 0, sd = 1, variance = 1: best way to standardize = make into z-scores (subtract the mean and /sd, standardized estimates: all interpreted the same way interpretation is standardized, all in the same units, so you can compare easily strengths and magnitudes. Unstandardized variables in original raw score units, depends on variable and how it was measured, mean and sd are determined by original units of variables: e. g. Time can be measured in hours, minutes, seconds, milliseconds. Each variable has been turned into a z-score (standardized) makes the mean, sd, and variance the same, values will be smaller, can accurately assess magnitude (diference in scores, low vs. high) Unstandardized estimates derived from unstandardized variables, original units not z-scores, can"t be directly compared across variables with diferent scales: covariance unstandardized form of correlation, magnitudes of covariance"s cannot be directly compared.