POS 3713 Lecture Notes - Lecture 12: Squared Deviations From The Mean, Standard Deviation, Statistical Inference
Document Summary
Outliers cases where the value is extremely high or extremely low compared to the rest of the values for that variable. Ex; 4, 5, 7, 7, 8, 9, 66: dispersion ex; obama is 6"1"". Problem the sum of deviations from a mean is always zero. The variance is the mean of the squared deviations. Greater values = greater dispersion around the mean. Variance of y = sum of squared deviations / n-1. Sd (y) = square root of (equation for variance) Kernel density plot takes a rolling average, eliminates binning problem. Histograms variable values appear on the x-axis. y-axis represents a similar count or proportion that value occurs in the data. The binning problem 5 bars vs. 8 bars (53) Skewness extent to which distribution of data deviates from symmetric distribution. Right (positive) skew: several extremely high values (outliers) are pulling the mean upward in value.
