Revised City of South Bend Disparity Study Report

Home Browse Search

City of South Bend Disparity Study 2020 We employ a multiple regression statistical technique to process this data. This methodology allows us to perform two analyses: an estimation of how variations in certain characteristics (called independent variables) will impact the level of some particular outcome (called a dependent variable), and a determination of how confident we are that the estimated variation is statistically different from zero. We have provided more detail on this technique in Appendix A. With respect to the first result of regression analysis, we will examine how varia- tions in the race, gender, and industry of individuals impact the wages and other economic outcomes received by individuals. The technique allows us to determine the effect of changes in one variable, assuming that the other determining vari- ables are the same. That is, we compare individuals of different races, but of the same gender and in the same industry; or we compare individuals of different gen- ders, but of the same race and the same industry; or we compare individuals in d"rf- ferent industries, but of the same race and gender. We are determining the impact of changes in one variable (e.g., race, gender or industry) on another vari- able (wages), "controlling for' the movement of any other independent variables. With respect to the second result of regression analysis, thistechnique also allows us to determine the statistical significance of the relationship between the depen- dent variable and independent variable. For example, the relationship between gender and wages might exist but we find that it is not statistically different from zero. In this case, we are not confident that there is not any relationship between the two variables. If the relationship is not statistically different from zero, then a variation in the independent variable has no impact on the dependent variable. The regression analysis allows us to say with varying degrees of statistical confi- dence that a relationship is different from zero. If the estimated relationship is sta- tistically significant at the 0.05 level, that indicates we are 95 percent confident that the relationship is different from zero; if the estimated relationship is statisti- cally significant at the 0.01 level, that indicates we are 99 percent confident that the relationship is different from zero; if the estimated relationship is statistically significant at the 0.001 level, that indicates we are 99.9 percent confident that the relationship is different from zero. 147 In the next section, we report: data on the share of a demographic group that forms a business (business formation rates); the probabilities that a demographic group will form a business relative to White men (business formation probabili- ties); the differences in wages received by a demographic group relative to White men (wage differentials); and the differences in business earnings received by a demographic group relative to White men (business earnings differentials). The 147. Most social scientists do not endorse utilizing a confidence level of less than 95 percent. (Another way of stating a confi- dence level of 95 percent is to state the results are statistically significance at the 0.05 leveQ Appendix C explains more about statistical significance. 79 ©1010 Colette Holt & Associates, All Rights Reserved