MEASURES OF DISPERSION - Division of Social Sciences：社会科学的分散措施分.pptVIP

MEASURES OF DISPERSION:SPREAD AND VARIABILITY DATA SETS FOR PROJECT NES2000.sav States.sav World.sav 2. Variance (continued) Thus: We want to understand variation in values of a variable—the so-called “dependent” variable. Question: Might the variation in the dependent variable be a function (or consequence) of variation in another variable—the “independent” variable? Key concept: Explanation of variance (in the dependent variable). 3. The Standard Deviation Measure of how individual observations deviate from or vary around the mean of the variable Allows comparison of variation: Standard deviation is 0 only if no variation The greater the spread, the greater the standard deviation of variable Two variables with similar means but different standard deviations differ in extent of variation around mean Example: Socio-economic Status Education: Mean = 10 years , sd = 2, observation X = 16, Zx = (16-10)/2 = + 3 Parents’ annual income: Mean = $ 50,000, sd = 5,000, observation X = $40,000, Zx = (40,000-50,000)/5,000 = -2 Postscript: On Skewness Sk = (X – Mo)/s = 3 (X – Md)/s Sk = 0 for a symmetrical (normal) curve, positive if skewed to the right, negative if skewed to the left If Sk 1, consider using median not mean If Sk/standard error 2, use median not mean * Outline: Key Measures of Dispersion (Interval Scale) Range Variance Standard Deviation Standard Score 1. Range of Values Xmax – Xmin = Range i i 2. The Meaning of Variance Formula: Variance = s2 = Σ (Xi – X)2/n [or n – 1] Definition: Mean squared deviation from the mean That is: Overall variation or cumulative “spread” of values around the arithmetic mean (or “average”) for a variable Relevance: A key goal of statistical analysis is to detect underlying patterns within the overall variance. Question: Why are some values below the mean and some values above? Can we find an explanation? Definition: Zi = (Xi – X)/s