SOCECOL 13 Lecture Notes - Lecture 11: Carcinogen, Gentrification, Dependent And Independent Variables
2 Nov 2016
School
Department
Course
Professor
Document Summary
Have to have linear relationship between the two variables. Y = ax + b (a = slope, b = y intercept) Intercept is value of regression line at the point it crosses the y axis. <> Problems with correlations: outliers can substantially inflate or deflate correlations, groups combined inappropriately may mask relationships. An outlier that is consistent with the trend of the rest of the data will inflate the correlation. An outlier that is not consistent with the rest of the data can substantially decrease the correlation. Always check for coding error: at least 5% of all data points are corrupted when initially recorded or entered into a computer the voice of experience. Bias issue: estimates of age biased toward older (young workers taking the us entry permits of workers retiring to china) Simpson"s paradox*: when groups merged, the previously separate correlations cancel out, resulting in low-no correlation in the merged group.
