Case 5 Case 5 Introduction to Demographic Introduction to Demographic Research Using Aggregated ACS Data Research Using Aggregated ACS Data for Ecological Regression: for Ecological Regression: Changes in County Poverty Changes in County Poverty Katherine Curtis Katherine Curtis Adam Slez Adam Slez Jennifer Huck Jennifer Huck University of Wisconsin – Madison University of Wisconsin – Madison Prepared for presentation at the Introduction to the American Community Survey workshop of the 2009 annual Prepared for presentation at the Introduction to the American Community Survey workshop of the 2009 annual meeting of the PAA, April 29 meeting of the PAA, April 29 th th , Detroit, MI. , Detroit, MI. Center for Demography & Center for Demography & Ecology Ecology Applied Population Applied Population Laboratory Laboratory
19
Embed
Case 5 Introduction to Demographic Research Using Aggregated ACS Data for Ecological Regression: Changes in County Poverty Katherine Curtis Adam Slez Jennifer.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Case 5Case 5
Introduction to Demographic Research Introduction to Demographic Research Using Aggregated ACS Data for Using Aggregated ACS Data for
Ecological Regression:Ecological Regression:Changes in County PovertyChanges in County Poverty
University of Wisconsin – Madison University of Wisconsin – Madison
Prepared for presentation at the Introduction to the American Community Survey workshop of the 2009 annual meeting of the PAA, April 29Prepared for presentation at the Introduction to the American Community Survey workshop of the 2009 annual meeting of the PAA, April 29thth, Detroit, MI., Detroit, MI.
Center for Demography & EcologyCenter for Demography & Ecology Applied Population LaboratoryApplied Population Laboratory
• Comparability of ACS with Census Long-FormComparability of ACS with Census Long-Form
• Focus on changes in relationships between Focus on changes in relationships between county poverty rates and structural covariatescounty poverty rates and structural covariates
• SE of an estimate (Y) is inversely related to R (sampling fraction) & N (total population), and positively related to D (design factor)– SE increases as R & N decreases and as D increases
• ACS is at a disadvantage for estimate reliabilityestimate reliability given the smaller sample size (compared to SF3)
Minus All Minus All Suppressed DataSuppressed Data
N = 708 N = 708
• Comparative analysis to examine the way Comparative analysis to examine the way differences in differences in survey designsurvey design influence results of influence results of a conventional ecological regression analysisa conventional ecological regression analysis
– American FactFinder > Download Center > Data ProfilesData Profiles– American FactFinder > Download Center > Selected Detailed TablesSelected Detailed Tables
• Variable CalculationVariable Calculation
– Use of different denominator (e.g., education)– Changing variable definitions (e.g., industry)– Create new variables (e.g., underemployment and commuter rates)
• Spatial Error ModelSpatial Error Model– yy is the county poverty rate – xx is the set of structural covariates associated with poverty– ββ is the set of effects associated with these factors– λλ measures the extent to which the spatial error in a county
tends to be correlated with the spatial error in neighboring counties
– WW is a row-standardized matrix depicting the spatial relationship between counties
– uu is a measure of spatial error– εε is a measure of non-spatial error
Population with income in the past 12 months below poverty level 5,256 ± 731Male: 2,132 ± 359
Under 5 years 346 ± 1555 years 66 ± 596 to 11 years 227 ± 9612 to 14 years 140 ± 7415 years 8 ± 1016 and 17 years 28 ± 3418 to 24 years 199 ± 14925 to 34 years 397 ± 14335 to 44 years 170 ± 7245 to 54 years 240 ± 9255 to 64 years 144 ± 8565 to 74 years 40 ± 2875 years and over 127 ± 70
Female: 3,124 ± 479Under 5 years 231 ± 1045 years 29 ± 266 to 11 years 340 ± 14912 to 14 years 142 ± 7615 years 35 ± 2916 and 17 years 184 ± 11918 to 24 years 409 ± 16025 to 34 years 434 ± 18035 to 44 years 395 ± 11845 to 54 years 237 ± 9155 to 64 years 122 ± 5265 to 74 years 114 ± 8675 years and over 452 ± 149
Total population 57,154 ± 124
Estimated proportion below poverty 0.092 ± 0.013
Table 1. Calculating a margin of error for a derived count and derived proportion, Sauk County, Wisconsin, ACS 2005-2007 Table B17001