After the weird results of the first round of analyses, where more inequality seemed to predict better psychological health (http://rpubs.com/mts/18108), I conducted revised analyses.
I have backwards eliminated from this beast of a term: Gini2 x BelowMedianIncome x PovStat x Race
We aim to examine the relation of neighborhood-level income inequality with psychosocial factors related to CVD within a racially and socioeconomically diverse cohort of adults in Baltimore, Maryland.
Psychosocial factors include:
Covariates include: demographic (age, sex, race, education), medical conditions (CVD, DMM, BMI, SF12PCS), medications (psych meds), neighborhood factors (median Income, racial segregation)
Examined:
Below are the final models and graphs after backwards elimination.
Anger: Race
Discrim01d (Discrimination due to income): BelowMedianIncome, Race
Discrim01b (Discrimination due to race): Race:BelowMedianIncome:I(Gini100C2 )
Discrim01c (Discrimination due to ethnicity/culture): PovStat:BelowMedianIncome, Gini100C:PovStat:Race
UnfairTreatment: Multiple Regression: PovStat:I(Gini100C2 ), Gini100C:Race:BelowMedianIncome
Logistic Regression: ??
Instrumental support: Race:BelowMedianIncome, Gini100C, PovStat
Emotional Support: PovStat:Race:BelowMedianIncome:I(Gini100C2)
##
## ************************************************************
## ************************************************************
## BEGIN: CESimp
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = "CESimp ~ Gini100C*PovStat*Race*BelowMedianIncome + I(Gini100C^2)*PovStat*Race*BelowMedianIncome + Age0 + Sex + Race + HseHldEducationDich + PovStat + CVDclusterdich + DMMclusterdich + PhysBMI + HseHldSF12PCSimp + PhysBMI + rxPsych + MedianHouseholdIncome2C + PercentBlackOnly0 + PercentBlackOnly1",
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -22.02 -7.31 -1.85 5.72 42.78
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 26.46535 2.73587 9.67
## Gini100C -0.12106 0.12472 -0.97
## PovStat 5.71629 2.18566 2.62
## Race 2.01327 1.33379 1.51
## BelowMedianIncome 1.54013 1.43954 1.07
## I(Gini100C^2) 0.00158 0.01338 0.12
## Age0 -0.10275 0.02819 -3.65
## Sex -1.02732 0.50653 -2.03
## HseHldEducationDich -3.51068 0.56619 -6.20
## CVDclusterdich -0.89000 1.30861 -0.68
## DMMclusterdich 0.59487 0.58999 1.01
## PhysBMI -0.07786 0.03430 -2.27
## HseHldSF12PCSimp -0.10151 0.02614 -3.88
## rxPsych 7.05367 0.71855 9.82
## MedianHouseholdIncome2C -0.57604 0.28613 -2.01
## PercentBlackOnly0 -0.54784 0.88667 -0.62
## PercentBlackOnly1 -1.93618 1.08458 -1.79
## Gini100C:PovStat -0.05627 0.30312 -0.19
## Gini100C:Race 0.11453 0.18396 0.62
## PovStat:Race -4.16960 2.47192 -1.69
## Gini100C:BelowMedianIncome -0.16953 0.19080 -0.89
## PovStat:BelowMedianIncome -2.83111 2.74870 -1.03
## Race:BelowMedianIncome -1.59893 1.84907 -0.86
## PovStat:I(Gini100C^2) -0.04633 0.03606 -1.28
## Race:I(Gini100C^2) -0.02589 0.01946 -1.33
## BelowMedianIncome:I(Gini100C^2) 0.00685 0.02142 0.32
## Gini100C:PovStat:Race -0.01392 0.38654 -0.04
## Gini100C:PovStat:BelowMedianIncome 0.45979 0.37855 1.21
## Gini100C:Race:BelowMedianIncome 0.08703 0.26786 0.32
## PovStat:Race:BelowMedianIncome 5.20307 3.19896 1.63
## PovStat:Race:I(Gini100C^2) 0.07967 0.04132 1.93
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.03573 0.04376 0.82
## Race:BelowMedianIncome:I(Gini100C^2) 0.04344 0.03092 1.40
## Gini100C:PovStat:Race:BelowMedianIncome -0.31775 0.48296 -0.66
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) -0.10573 0.05229 -2.02
## Pr(>|t|)
## (Intercept) < 2e-16
## Gini100C 0.33185
## PovStat 0.00899
## Race 0.13137
## BelowMedianIncome 0.28482
## I(Gini100C^2) 0.90584
## Age0 0.00027
## Sex 0.04270
## HseHldEducationDich 0.0000000007
## CVDclusterdich 0.49652
## DMMclusterdich 0.31346
## PhysBMI 0.02333
## HseHldSF12PCSimp 0.00011
## rxPsych < 2e-16
## MedianHouseholdIncome2C 0.04425
## PercentBlackOnly0 0.53675
## PercentBlackOnly1 0.07440
## Gini100C:PovStat 0.85276
## Gini100C:Race 0.53365
## PovStat:Race 0.09182
## Gini100C:BelowMedianIncome 0.37438
## PovStat:BelowMedianIncome 0.30316
## Race:BelowMedianIncome 0.38731
## PovStat:I(Gini100C^2) 0.19902
## Race:I(Gini100C^2) 0.18365
## BelowMedianIncome:I(Gini100C^2) 0.74902
## Gini100C:PovStat:Race 0.97127
## Gini100C:PovStat:BelowMedianIncome 0.22468
## Gini100C:Race:BelowMedianIncome 0.74530
## PovStat:Race:BelowMedianIncome 0.10402
## PovStat:Race:I(Gini100C^2) 0.05398
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.41439
## Race:BelowMedianIncome:I(Gini100C^2) 0.16027
## Gini100C:PovStat:Race:BelowMedianIncome 0.51067
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) 0.04333
##
## Residual standard error: 10.2 on 1762 degrees of freedom
## Multiple R-squared: 0.16, Adjusted R-squared: 0.144
## F-statistic: 9.89 on 34 and 1762 DF, p-value: <2e-16
##
## ***************** END: CESimp *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: acasiPDSQanx
## ************************************************************
## ************************************************************
## Warning: All variables have been thrown out of the model.
## No model could be specified.
##
## Call:
## lm(formula = acasiPDSQanx ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1, data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.376 -2.258 -0.998 2.045 8.405
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.89689 0.76845 8.98 < 2e-16
## Gini100C -0.01930 0.01171 -1.65 0.0996
## PovStat 0.46029 0.15948 2.89 0.0039
## Race -0.01529 0.21769 -0.07 0.9440
## BelowMedianIncome 0.47242 0.15388 3.07 0.0022
## Age0 -0.04321 0.00824 -5.24 1.8e-07
## Sex -0.55178 0.14786 -3.73 0.0002
## HseHldEducationDich -0.74790 0.16468 -4.54 6.0e-06
## CVDclusterdich -0.12896 0.38084 -0.34 0.7349
## DMMclusterdich 0.22059 0.17242 1.28 0.2009
## PhysBMI -0.01374 0.01001 -1.37 0.1700
## HseHldSF12PCSimp -0.02118 0.00756 -2.80 0.0051
## rxPsych 1.56525 0.20895 7.49 1.1e-13
## MedianHouseholdIncome2C -0.11087 0.07866 -1.41 0.1589
## PercentBlackOnly0 -0.32530 0.23627 -1.38 0.1687
## PercentBlackOnly1 -0.77146 0.30474 -2.53 0.0114
##
## Residual standard error: 3.01 on 1781 degrees of freedom
## Multiple R-squared: 0.113, Adjusted R-squared: 0.106
## F-statistic: 15.1 on 15 and 1781 DF, p-value: <2e-16
##
## ***************** END: acasiPDSQanx *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: acasiPerStr
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = "acasiPerStr ~ Gini100C*PovStat*Race*BelowMedianIncome + I(Gini100C^2)*PovStat*Race*BelowMedianIncome + Age0 + Sex + Race + HseHldEducationDich + PovStat + CVDclusterdich + DMMclusterdich + PhysBMI + HseHldSF12PCSimp + PhysBMI + rxPsych + MedianHouseholdIncome2C + PercentBlackOnly0 + PercentBlackOnly1",
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -7.773 -2.277 -0.125 2.170 10.455
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 8.240459 0.838793 9.82
## Gini100C -0.000826 0.038238 -0.02
## PovStat 2.251622 0.670102 3.36
## Race -0.120090 0.408927 -0.29
## BelowMedianIncome 1.221950 0.441349 2.77
## I(Gini100C^2) -0.000916 0.004102 -0.22
## Age0 -0.028534 0.008641 -3.30
## Sex -0.028091 0.155298 -0.18
## HseHldEducationDich -0.910659 0.173590 -5.25
## CVDclusterdich 0.437010 0.401208 1.09
## DMMclusterdich -0.075910 0.180884 -0.42
## PhysBMI -0.024034 0.010517 -2.29
## HseHldSF12PCSimp -0.013073 0.008014 -1.63
## rxPsych 1.132658 0.220302 5.14
## MedianHouseholdIncome2C -0.162938 0.087726 -1.86
## PercentBlackOnly0 -0.281054 0.271845 -1.03
## PercentBlackOnly1 -0.304090 0.332523 -0.91
## Gini100C:PovStat 0.079613 0.092934 0.86
## Gini100C:Race -0.008609 0.056401 -0.15
## PovStat:Race -1.564080 0.757868 -2.06
## Gini100C:BelowMedianIncome -0.013380 0.058498 -0.23
## PovStat:BelowMedianIncome -2.013568 0.842725 -2.39
## Race:BelowMedianIncome -0.402944 0.566909 -0.71
## PovStat:I(Gini100C^2) -0.020078 0.011055 -1.82
## Race:I(Gini100C^2) -0.005115 0.005967 -0.86
## BelowMedianIncome:I(Gini100C^2) -0.006626 0.006566 -1.01
## Gini100C:PovStat:Race -0.146736 0.118511 -1.24
## Gini100C:PovStat:BelowMedianIncome -0.084253 0.116061 -0.73
## Gini100C:Race:BelowMedianIncome 0.045744 0.082124 0.56
## PovStat:Race:BelowMedianIncome 1.726359 0.980773 1.76
## PovStat:Race:I(Gini100C^2) 0.029768 0.012668 2.35
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.026852 0.013417 2.00
## Race:BelowMedianIncome:I(Gini100C^2) 0.014893 0.009481 1.57
## Gini100C:PovStat:Race:BelowMedianIncome 0.153172 0.148071 1.03
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) -0.042851 0.016032 -2.67
## Pr(>|t|)
## (Intercept) < 2e-16
## Gini100C 0.98276
## PovStat 0.00080
## Race 0.76904
## BelowMedianIncome 0.00569
## I(Gini100C^2) 0.82341
## Age0 0.00098
## Sex 0.85648
## HseHldEducationDich 0.00000017
## CVDclusterdich 0.27620
## DMMclusterdich 0.67478
## PhysBMI 0.02241
## HseHldSF12PCSimp 0.10299
## rxPsych 0.00000030
## MedianHouseholdIncome2C 0.06343
## PercentBlackOnly0 0.30134
## PercentBlackOnly1 0.36058
## Gini100C:PovStat 0.39175
## Gini100C:Race 0.87870
## PovStat:Race 0.03918
## Gini100C:BelowMedianIncome 0.81911
## PovStat:BelowMedianIncome 0.01698
## Race:BelowMedianIncome 0.47732
## PovStat:I(Gini100C^2) 0.06950
## Race:I(Gini100C^2) 0.39150
## BelowMedianIncome:I(Gini100C^2) 0.31309
## Gini100C:PovStat:Race 0.21582
## Gini100C:PovStat:BelowMedianIncome 0.46798
## Gini100C:Race:BelowMedianIncome 0.57760
## PovStat:Race:BelowMedianIncome 0.07855
## PovStat:Race:I(Gini100C^2) 0.01889
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.04550
## Race:BelowMedianIncome:I(Gini100C^2) 0.11639
## Gini100C:PovStat:Race:BelowMedianIncome 0.30107
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) 0.00759
##
## Residual standard error: 3.14 on 1762 degrees of freedom
## Multiple R-squared: 0.105, Adjusted R-squared: 0.0874
## F-statistic: 6.06 on 34 and 1762 DF, p-value: <2e-16
##
## ***************** END: acasiPerStr *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: acasiAnger
## ************************************************************
## ************************************************************
## Warning: All variables have been thrown out of the model.
## No model could be specified.
##
## Call:
## lm(formula = acasiAnger ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1, data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -13.39 -4.87 -0.75 4.21 18.81
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 17.41899 1.56076 11.16 < 2e-16
## Gini100C -0.04332 0.02379 -1.82 0.06873
## PovStat 0.61499 0.32392 1.90 0.05778
## Race -1.81049 0.44214 -4.09 0.00004414
## BelowMedianIncome 0.58183 0.31255 1.86 0.06282
## Age0 -0.09437 0.01674 -5.64 0.00000002
## Sex 0.03743 0.30031 0.12 0.90083
## HseHldEducationDich -1.61975 0.33447 -4.84 0.00000139
## CVDclusterdich 0.45056 0.77350 0.58 0.56031
## DMMclusterdich 0.05866 0.35019 0.17 0.86698
## PhysBMI -0.00492 0.02033 -0.24 0.80878
## HseHldSF12PCSimp -0.03994 0.01535 -2.60 0.00934
## rxPsych 1.47135 0.42440 3.47 0.00054
## MedianHouseholdIncome2C -0.14522 0.15975 -0.91 0.36346
## PercentBlackOnly0 0.03048 0.47987 0.06 0.94937
## PercentBlackOnly1 0.63539 0.61894 1.03 0.30475
##
## Residual standard error: 6.11 on 1781 degrees of freedom
## Multiple R-squared: 0.0698, Adjusted R-squared: 0.062
## F-statistic: 8.91 on 15 and 1781 DF, p-value: <2e-16
##
## ***************** END: acasiAnger *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: Discrim01d
## ************************************************************
## ************************************************************
## Warning: All variables have been thrown out of the model.
## No model could be specified.
##
## Call:
## lm(formula = Discrim01d ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1, data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.700 -0.861 -0.330 0.801 2.617
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.324144 0.266609 8.72 < 2e-16
## Gini100C 0.002276 0.004063 0.56 0.575
## PovStat 0.046532 0.055332 0.84 0.400
## Race 0.468715 0.075527 6.21 0.00000000067
## BelowMedianIncome 0.214759 0.053389 4.02 0.00005998191
## Age0 -0.004582 0.002860 -1.60 0.109
## Sex 0.050068 0.051298 0.98 0.329
## HseHldEducationDich -0.025299 0.057134 -0.44 0.658
## CVDclusterdich 0.030093 0.132128 0.23 0.820
## DMMclusterdich 0.021994 0.059820 0.37 0.713
## PhysBMI 0.000422 0.003473 0.12 0.903
## HseHldSF12PCSimp -0.013993 0.002622 -5.34 0.00000010680
## rxPsych 0.151431 0.072495 2.09 0.037
## MedianHouseholdIncome2C 0.025383 0.027289 0.93 0.352
## PercentBlackOnly0 0.103429 0.081972 1.26 0.207
## PercentBlackOnly1 0.037889 0.105726 0.36 0.720
##
## Residual standard error: 1.04 on 1781 degrees of freedom
## Multiple R-squared: 0.0814, Adjusted R-squared: 0.0737
## F-statistic: 10.5 on 15 and 1781 DF, p-value: <2e-16
##
## ***************** END: Discrim01d *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: Discrim01b
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = Discrim01b ~ Gini100C + PovStat + Race + BelowMedianIncome +
## I(Gini100C^2) + Age0 + Sex + HseHldEducationDich + CVDclusterdich +
## DMMclusterdich + PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1 + Gini100C:Race + Gini100C:BelowMedianIncome +
## PovStat:BelowMedianIncome + Race:BelowMedianIncome + Race:I(Gini100C^2) +
## BelowMedianIncome:I(Gini100C^2) + Gini100C:Race:BelowMedianIncome +
## Race:BelowMedianIncome:I(Gini100C^2), data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.941 -0.644 -0.335 0.563 2.579
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.797362 0.245918 7.31 4.1e-13
## Gini100C 0.011773 0.010206 1.15 0.2488
## PovStat -0.194501 0.076312 -2.55 0.0109
## Race 0.781171 0.105296 7.42 1.8e-13
## BelowMedianIncome -0.113133 0.111509 -1.01 0.3105
## I(Gini100C^2) -0.002318 0.001107 -2.09 0.0364
## Age0 0.001257 0.002558 0.49 0.6231
## Sex 0.196155 0.046083 4.26 2.2e-05
## HseHldEducationDich 0.111923 0.051329 2.18 0.0294
## CVDclusterdich -0.114239 0.118748 -0.96 0.3362
## DMMclusterdich 0.039691 0.053597 0.74 0.4591
## PhysBMI 0.000703 0.003114 0.23 0.8215
## HseHldSF12PCSimp -0.007243 0.002369 -3.06 0.0023
## rxPsych 0.075094 0.065085 1.15 0.2487
## MedianHouseholdIncome2C 0.042276 0.025627 1.65 0.0992
## PercentBlackOnly0 0.085748 0.078950 1.09 0.2776
## PercentBlackOnly1 0.089899 0.097473 0.92 0.3565
## Gini100C:Race -0.028874 0.014435 -2.00 0.0456
## Gini100C:BelowMedianIncome 0.007931 0.014357 0.55 0.5807
## PovStat:BelowMedianIncome 0.212352 0.097131 2.19 0.0289
## Race:BelowMedianIncome 0.018732 0.129023 0.15 0.8846
## Race:I(Gini100C^2) 0.004275 0.001393 3.07 0.0022
## BelowMedianIncome:I(Gini100C^2) 0.003469 0.001547 2.24 0.0250
## Gini100C:Race:BelowMedianIncome 0.007798 0.019376 0.40 0.6874
## Race:BelowMedianIncome:I(Gini100C^2) -0.005318 0.001934 -2.75 0.0060
##
## Residual standard error: 0.933 on 1772 degrees of freedom
## Multiple R-squared: 0.2, Adjusted R-squared: 0.189
## F-statistic: 18.4 on 24 and 1772 DF, p-value: <2e-16
##
## ***************** END: Discrim01b *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: Discrim01c
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = Discrim01c ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1 + Gini100C:PovStat +
## Gini100C:Race + PovStat:Race + PovStat:BelowMedianIncome +
## Gini100C:PovStat:Race, data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.818 -0.521 -0.271 0.572 2.800
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1.4577792 0.2331813 6.25 0.00000000051
## Gini100C -0.0021514 0.0059202 -0.36 0.71635
## PovStat 0.0181586 0.0986695 0.18 0.85401
## Race 1.0497570 0.0750110 13.99 < 2e-16
## BelowMedianIncome -0.0602507 0.0595056 -1.01 0.31143
## Age0 0.0010756 0.0024699 0.44 0.66327
## Sex 0.2209964 0.0443518 4.98 0.00000068763
## HseHldEducationDich 0.1453140 0.0495107 2.94 0.00338
## CVDclusterdich -0.1453143 0.1141524 -1.27 0.20319
## DMMclusterdich 0.0598259 0.0516643 1.16 0.24703
## PhysBMI 0.0000936 0.0030081 0.03 0.97519
## HseHldSF12PCSimp -0.0082659 0.0022891 -3.61 0.00031
## rxPsych 0.1025345 0.0628559 1.63 0.10301
## MedianHouseholdIncome2C 0.0357345 0.0239611 1.49 0.13605
## PercentBlackOnly0 0.0836184 0.0724505 1.15 0.24860
## PercentBlackOnly1 0.0920359 0.0923072 1.00 0.31887
## Gini100C:PovStat 0.0244807 0.0126976 1.93 0.05402
## Gini100C:Race 0.0115380 0.0089876 1.28 0.19939
## PovStat:Race -0.1989574 0.0988460 -2.01 0.04429
## PovStat:BelowMedianIncome 0.2139603 0.0914607 2.34 0.01943
## Gini100C:PovStat:Race -0.0392373 0.0154532 -2.54 0.01120
##
## Residual standard error: 0.902 on 1776 degrees of freedom
## Multiple R-squared: 0.251, Adjusted R-squared: 0.243
## F-statistic: 29.8 on 20 and 1776 DF, p-value: <2e-16
##
## ***************** END: Discrim01c *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: UnfairTreatment
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = UnfairTreatment ~ Gini100C + PovStat + Race + BelowMedianIncome +
## I(Gini100C^2) + Age0 + Sex + HseHldEducationDich + CVDclusterdich +
## DMMclusterdich + PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1 + Gini100C:PovStat +
## Gini100C:Race + Gini100C:BelowMedianIncome + Race:BelowMedianIncome +
## PovStat:I(Gini100C^2) + Gini100C:Race:BelowMedianIncome,
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -26.67 -12.30 -5.28 14.33 38.38
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 15.1149 4.0121 3.77 0.00017
## Gini100C -0.3200 0.1394 -2.29 0.02187
## PovStat 2.6273 1.0459 2.51 0.01209
## Race 11.3059 1.3990 8.08 1.2e-15
## BelowMedianIncome 0.3301 1.2598 0.26 0.79334
## I(Gini100C^2) 0.0294 0.0119 2.47 0.01355
## Age0 0.0746 0.0424 1.76 0.07867
## Sex 2.8547 0.7627 3.74 0.00019
## HseHldEducationDich 1.0677 0.8515 1.25 0.21005
## CVDclusterdich -3.5059 1.9639 -1.79 0.07441
## DMMclusterdich 1.6079 0.8880 1.81 0.07034
## PhysBMI 0.0221 0.0517 0.43 0.66921
## HseHldSF12PCSimp -0.0913 0.0393 -2.33 0.02008
## rxPsych 2.4213 1.0824 2.24 0.02542
## MedianHouseholdIncome2C -0.3427 0.4251 -0.81 0.42029
## PercentBlackOnly0 -1.3191 1.2933 -1.02 0.30791
## PercentBlackOnly1 -1.4888 1.5998 -0.93 0.35218
## Gini100C:PovStat 0.1455 0.1585 0.92 0.35905
## Gini100C:Race 0.1656 0.1716 0.97 0.33466
## Gini100C:BelowMedianIncome 0.5089 0.1978 2.57 0.01016
## Race:BelowMedianIncome -0.6494 1.5831 -0.41 0.68171
## PovStat:I(Gini100C^2) -0.0368 0.0158 -2.33 0.01996
## Gini100C:Race:BelowMedianIncome -0.5338 0.2415 -2.21 0.02723
##
## Residual standard error: 15.5 on 1774 degrees of freedom
## Multiple R-squared: 0.131, Adjusted R-squared: 0.12
## F-statistic: 12.1 on 22 and 1774 DF, p-value: <2e-16
##
## ***************** END: UnfairTreatment *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: acasiInsSup
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = acasiInsSup ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1 + Race:BelowMedianIncome,
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -7.906 -2.987 -0.276 2.598 12.038
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 9.42968 0.99079 9.52 <2e-16
## Gini100C -0.03444 0.01500 -2.30 0.0218
## PovStat -0.42405 0.20421 -2.08 0.0380
## Race 0.69327 0.33803 2.05 0.0404
## BelowMedianIncome -0.95870 0.29709 -3.23 0.0013
## Age0 -0.03042 0.01054 -2.89 0.0039
## Sex -0.14448 0.18896 -0.76 0.4446
## HseHldEducationDich -0.45414 0.21094 -2.15 0.0315
## CVDclusterdich 0.36794 0.48688 0.76 0.4499
## DMMclusterdich -0.15143 0.22047 -0.69 0.4923
## PhysBMI 0.00143 0.01279 0.11 0.9108
## HseHldSF12PCSimp -0.00595 0.00967 -0.62 0.5381
## rxPsych -0.80575 0.26737 -3.01 0.0026
## MedianHouseholdIncome2C 0.06688 0.10053 0.67 0.5060
## PercentBlackOnly0 -0.35690 0.30233 -1.18 0.2380
## PercentBlackOnly1 -0.34053 0.38962 -0.87 0.3822
## Race:BelowMedianIncome 0.82940 0.37346 2.22 0.0265
##
## Residual standard error: 3.85 on 1780 degrees of freedom
## Multiple R-squared: 0.0378, Adjusted R-squared: 0.0292
## F-statistic: 4.37 on 16 and 1780 DF, p-value: 0.0000000154
##
## ***************** END: acasiInsSup *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: acasiEmoSup
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = "acasiEmoSup ~ Gini100C*PovStat*Race*BelowMedianIncome + I(Gini100C^2)*PovStat*Race*BelowMedianIncome + Age0 + Sex + Race + HseHldEducationDich + PovStat + CVDclusterdich + DMMclusterdich + PhysBMI + HseHldSF12PCSimp + PhysBMI + rxPsych + MedianHouseholdIncome2C + PercentBlackOnly0 + PercentBlackOnly1",
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -28.647 -7.606 -0.227 6.116 30.294
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 14.80452 2.58303 5.73
## Gini100C 0.10287 0.11775 0.87
## PovStat -2.14043 2.06355 -1.04
## Race -0.87558 1.25927 -0.70
## BelowMedianIncome 2.00764 1.35912 1.48
## I(Gini100C^2) -0.00471 0.01263 -0.37
## Age0 0.03812 0.02661 1.43
## Sex -0.75845 0.47824 -1.59
## HseHldEducationDich 1.25690 0.53456 2.35
## CVDclusterdich -0.83532 1.23551 -0.68
## DMMclusterdich -0.24143 0.55703 -0.43
## PhysBMI 0.03623 0.03239 1.12
## HseHldSF12PCSimp 0.05365 0.02468 2.17
## rxPsych 1.63853 0.67841 2.42
## MedianHouseholdIncome2C -0.07092 0.27015 -0.26
## PercentBlackOnly0 0.31967 0.83714 0.38
## PercentBlackOnly1 0.71364 1.02399 0.70
## Gini100C:PovStat -0.55050 0.28619 -1.92
## Gini100C:Race -0.18966 0.17368 -1.09
## PovStat:Race 5.42288 2.33383 2.32
## Gini100C:BelowMedianIncome 0.09963 0.18014 0.55
## PovStat:BelowMedianIncome 1.55972 2.59514 0.60
## Race:BelowMedianIncome -0.44815 1.74577 -0.26
## PovStat:I(Gini100C^2) 0.08928 0.03404 2.62
## Race:I(Gini100C^2) 0.01558 0.01838 0.85
## BelowMedianIncome:I(Gini100C^2) 0.03757 0.02022 1.86
## Gini100C:PovStat:Race 0.61017 0.36495 1.67
## Gini100C:PovStat:BelowMedianIncome 0.60491 0.35741 1.69
## Gini100C:Race:BelowMedianIncome 0.25208 0.25290 1.00
## PovStat:Race:BelowMedianIncome -3.69868 3.02025 -1.22
## PovStat:Race:I(Gini100C^2) -0.09545 0.03901 -2.45
## PovStat:BelowMedianIncome:I(Gini100C^2) -0.10656 0.04132 -2.58
## Race:BelowMedianIncome:I(Gini100C^2) -0.04005 0.02920 -1.37
## Gini100C:PovStat:Race:BelowMedianIncome -0.76908 0.45598 -1.69
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) 0.10630 0.04937 2.15
## Pr(>|t|)
## (Intercept) 0.000000012
## Gini100C 0.3825
## PovStat 0.2998
## Race 0.4870
## BelowMedianIncome 0.1398
## I(Gini100C^2) 0.7096
## Age0 0.1522
## Sex 0.1129
## HseHldEducationDich 0.0188
## CVDclusterdich 0.4991
## DMMclusterdich 0.6648
## PhysBMI 0.2634
## HseHldSF12PCSimp 0.0298
## rxPsych 0.0158
## MedianHouseholdIncome2C 0.7929
## PercentBlackOnly0 0.7026
## PercentBlackOnly1 0.4859
## Gini100C:PovStat 0.0546
## Gini100C:Race 0.2750
## PovStat:Race 0.0203
## Gini100C:BelowMedianIncome 0.5803
## PovStat:BelowMedianIncome 0.5479
## Race:BelowMedianIncome 0.7974
## PovStat:I(Gini100C^2) 0.0088
## Race:I(Gini100C^2) 0.3967
## BelowMedianIncome:I(Gini100C^2) 0.0633
## Gini100C:PovStat:Race 0.0947
## Gini100C:PovStat:BelowMedianIncome 0.0907
## Gini100C:Race:BelowMedianIncome 0.3190
## PovStat:Race:BelowMedianIncome 0.2209
## PovStat:Race:I(Gini100C^2) 0.0145
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.0100
## Race:BelowMedianIncome:I(Gini100C^2) 0.1703
## Gini100C:PovStat:Race:BelowMedianIncome 0.0918
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) 0.0314
##
## Residual standard error: 9.66 on 1762 degrees of freedom
## Multiple R-squared: 0.046, Adjusted R-squared: 0.0276
## F-statistic: 2.5 on 34 and 1762 DF, p-value: 0.00000442
##
## ***************** END: acasiEmoSup *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: MCLcountry
## ************************************************************
## ************************************************************
##
## Call:
## lm(formula = MCLcountry ~ Gini100C + PovStat + Race + BelowMedianIncome +
## I(Gini100C^2) + Age0 + Sex + HseHldEducationDich + CVDclusterdich +
## DMMclusterdich + PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1 + Gini100C:Race + Gini100C:BelowMedianIncome +
## Race:BelowMedianIncome + Race:I(Gini100C^2), data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5.009 -1.240 0.133 1.164 5.958
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.80901 0.50312 5.58 0.000000027
## Gini100C 0.03096 0.01646 1.88 0.06007
## PovStat -0.23419 0.10407 -2.25 0.02456
## Race 0.08841 0.20147 0.44 0.66085
## BelowMedianIncome -0.66483 0.15474 -4.30 0.000018332
## I(Gini100C^2) 0.00384 0.00171 2.24 0.02518
## Age0 0.01128 0.00532 2.12 0.03426
## Sex 0.09330 0.09607 0.97 0.33163
## HseHldEducationDich 0.33889 0.10683 3.17 0.00154
## CVDclusterdich 0.30686 0.24844 1.24 0.21694
## DMMclusterdich -0.07904 0.11166 -0.71 0.47912
## PhysBMI -0.00336 0.00648 -0.52 0.60385
## HseHldSF12PCSimp 0.02058 0.00495 4.16 0.000033344
## rxPsych -0.28045 0.13552 -2.07 0.03866
## MedianHouseholdIncome2C 0.08261 0.05428 1.52 0.12825
## PercentBlackOnly0 0.04084 0.16554 0.25 0.80515
## PercentBlackOnly1 0.11105 0.20474 0.54 0.58763
## Gini100C:Race 0.01598 0.02024 0.79 0.42995
## Gini100C:BelowMedianIncome -0.03470 0.01488 -2.33 0.01979
## Race:BelowMedianIncome 0.67498 0.19891 3.39 0.00071
## Race:I(Gini100C^2) -0.00720 0.00206 -3.50 0.00048
##
## Residual standard error: 1.91 on 1719 degrees of freedom
## (57 observations deleted due to missingness)
## Multiple R-squared: 0.0873, Adjusted R-squared: 0.0767
## F-statistic: 8.22 on 20 and 1719 DF, p-value: <2e-16
##
## ***************** END: MCLcountry *****************************
##
##
## ************************************************************
## ************************************************************
## BEGIN: MCLcommunity
## ************************************************************
## ************************************************************
## Warning: All variables have been thrown out of the model.
## No model could be specified.
##
## Call:
## lm(formula = MCLcommunity ~ Gini100C + PovStat + Race + BelowMedianIncome +
## Age0 + Sex + HseHldEducationDich + CVDclusterdich + DMMclusterdich +
## PhysBMI + HseHldSF12PCSimp + rxPsych + MedianHouseholdIncome2C +
## PercentBlackOnly0 + PercentBlackOnly1, data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.348 -1.140 -0.012 1.274 4.824
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.6137369 0.4871324 7.42 1.9e-13
## Gini100C 0.0040038 0.0074516 0.54 0.5911
## PovStat -0.2298422 0.1012489 -2.27 0.0233
## Race 0.6883865 0.1392390 4.94 8.4e-07
## BelowMedianIncome -0.2316010 0.0979459 -2.36 0.0182
## Age0 0.0276658 0.0052247 5.30 1.3e-07
## Sex -0.0571967 0.0939226 -0.61 0.5426
## HseHldEducationDich 0.3159652 0.1043172 3.03 0.0025
## CVDclusterdich 0.0000291 0.2415132 0.00 0.9999
## DMMclusterdich -0.0859681 0.1093870 -0.79 0.4320
## PhysBMI -0.0103652 0.0063424 -1.63 0.1024
## HseHldSF12PCSimp 0.0211978 0.0048094 4.41 1.1e-05
## rxPsych -0.1398716 0.1325433 -1.06 0.2914
## MedianHouseholdIncome2C 0.0326431 0.0504314 0.65 0.5175
## PercentBlackOnly0 -0.1954927 0.1503262 -1.30 0.1936
## PercentBlackOnly1 -0.2568833 0.1941454 -1.32 0.1860
##
## Residual standard error: 1.88 on 1725 degrees of freedom
## (56 observations deleted due to missingness)
## Multiple R-squared: 0.0725, Adjusted R-squared: 0.0644
## F-statistic: 8.99 on 15 and 1725 DF, p-value: <2e-16
##
## ***************** END: MCLcommunity *****************************
Insanity for 2 of the 4 psychological outcomes!! Let me know if there might be another way to frame these results that is more understandable.
Whites:
Whites above poverty
For Whites above poverty, with greater inequality, depressive symptoms decrease. For Whites above poverty, those above neighborhood median income (NMI) have fewer depressive symptoms than those below NMI.
Whites below poverty
For Whites below poverty and above NMI, depressive symptoms increase until about Gini=43, then dramatically decrease. For Whites below poverty and below NMI, depressive symptoms increase with greater inequality.
Blacks:
Blacks above poverty
For Blacks above poverty and above NMI, depressive symptoms initially increase until Gini= 43, then decrease. The opposite is depicted for Blacks above poverty and below NMI.
Blacks below poverty
For Blacks below poverty and above NMI, depressive symptoms appear to modestly increase with greater inequality.
For Blacks below poverty and below NMI, depressive symptoms initially increase until Gini= 43, then decrease.
The full interaction was NS (PovStat:Race:BelowMedianIncome:I(Gini100C2). This is what it would have looked like if it were significant. It was just barely significant in my last set of analyses, and its significance went away when I added PercentBlack as a covariate.
##
## Call:
## lm(formula = "acasiPDSQanx ~ Gini100C*PovStat*Race*BelowMedianIncome + I(Gini100C^2)*PovStat*Race*BelowMedianIncome + Age0 + Sex + Race + HseHldEducationDich + PovStat + CVDclusterdich + DMMclusterdich + PhysBMI + HseHldSF12PCSimp + PhysBMI + rxPsych + MedianHouseholdIncome2C + PercentBlackOnly0 + PercentBlackOnly1",
## data = analysisData)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.392 -2.276 -0.965 2.023 8.338
##
## Coefficients:
## Estimate Std. Error t value
## (Intercept) 6.744914 0.805872 8.37
## Gini100C -0.006311 0.036737 -0.17
## PovStat 1.416157 0.643802 2.20
## Race 0.151974 0.392877 0.39
## BelowMedianIncome 0.444289 0.424027 1.05
## I(Gini100C^2) -0.001121 0.003941 -0.28
## Age0 -0.043738 0.008302 -5.27
## Sex -0.562060 0.149203 -3.77
## HseHldEducationDich -0.729439 0.166777 -4.37
## CVDclusterdich -0.107336 0.385462 -0.28
## DMMclusterdich 0.231564 0.173785 1.33
## PhysBMI -0.014852 0.010104 -1.47
## HseHldSF12PCSimp -0.018557 0.007699 -2.41
## rxPsych 1.578424 0.211656 7.46
## MedianHouseholdIncome2C -0.104128 0.084283 -1.24
## PercentBlackOnly0 -0.249078 0.261176 -0.95
## PercentBlackOnly1 -0.741881 0.319472 -2.32
## Gini100C:PovStat 0.032789 0.089287 0.37
## Gini100C:Race 0.025665 0.054187 0.47
## PovStat:Race -1.082771 0.728124 -1.49
## Gini100C:BelowMedianIncome -0.083066 0.056202 -1.48
## PovStat:BelowMedianIncome -0.924257 0.809650 -1.14
## Race:BelowMedianIncome -0.141265 0.544659 -0.26
## PovStat:I(Gini100C^2) -0.018933 0.010621 -1.78
## Race:I(Gini100C^2) -0.004813 0.005733 -0.84
## BelowMedianIncome:I(Gini100C^2) 0.000627 0.006308 0.10
## Gini100C:PovStat:Race -0.053375 0.113860 -0.47
## Gini100C:PovStat:BelowMedianIncome 0.025049 0.111506 0.22
## Gini100C:Race:BelowMedianIncome 0.080783 0.078901 1.02
## PovStat:Race:BelowMedianIncome 1.091506 0.942280 1.16
## PovStat:Race:I(Gini100C^2) 0.025563 0.012171 2.10
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.016038 0.012890 1.24
## Race:BelowMedianIncome:I(Gini100C^2) 0.008485 0.009109 0.93
## Gini100C:PovStat:Race:BelowMedianIncome -0.035899 0.142259 -0.25
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) -0.026887 0.015403 -1.75
## Pr(>|t|)
## (Intercept) < 2e-16
## Gini100C 0.86362
## PovStat 0.02796
## Race 0.69893
## BelowMedianIncome 0.29488
## I(Gini100C^2) 0.77615
## Age0 1.5e-07
## Sex 0.00017
## HseHldEducationDich 1.3e-05
## CVDclusterdich 0.78069
## DMMclusterdich 0.18288
## PhysBMI 0.14177
## HseHldSF12PCSimp 0.01604
## rxPsych 1.4e-13
## MedianHouseholdIncome2C 0.21683
## PercentBlackOnly0 0.34038
## PercentBlackOnly1 0.02034
## Gini100C:PovStat 0.71349
## Gini100C:Race 0.63582
## PovStat:Race 0.13718
## Gini100C:BelowMedianIncome 0.13959
## PovStat:BelowMedianIncome 0.25380
## Race:BelowMedianIncome 0.79538
## PovStat:I(Gini100C^2) 0.07482
## Race:I(Gini100C^2) 0.40131
## BelowMedianIncome:I(Gini100C^2) 0.92085
## Gini100C:PovStat:Race 0.63929
## Gini100C:PovStat:BelowMedianIncome 0.82229
## Gini100C:Race:BelowMedianIncome 0.30605
## PovStat:Race:BelowMedianIncome 0.24687
## PovStat:Race:I(Gini100C^2) 0.03583
## PovStat:BelowMedianIncome:I(Gini100C^2) 0.21359
## Race:BelowMedianIncome:I(Gini100C^2) 0.35169
## Gini100C:PovStat:Race:BelowMedianIncome 0.80080
## PovStat:Race:BelowMedianIncome:I(Gini100C^2) 0.08105
##
## Residual standard error: 3.01 on 1762 degrees of freedom
## Multiple R-squared: 0.12, Adjusted R-squared: 0.103
## F-statistic: 7.09 on 34 and 1762 DF, p-value: <2e-16
Whites:
Whites above poverty
For Whites above poverty and above NMI, with greater inequality, PSS decreases slightly. For Whites above poverty and below NMI, PSS increases until about Gini=43, then dramatically decreases. (Different than CESD results for Whites above poverty)
Whites below poverty
For Whites below poverty and above NMI, PSS symptoms increase until about Gini=43, then dramatically decrease. (same as CESD results for this group) For Whites below poverty and below NMI, PSS symptoms decrease with greater inequality. (opposite of CESD result for this group)
Blacks:
Blacks above poverty
For Blacks above poverty and above NMI, PSS initially increases until Gini= 43, then decreases. For Blacks above poverty and below NMI, PSS increases with increasing inequality. (Similar to CESD and anxiety results for Blacks above poverty)
Blacks below poverty
For Blacks below poverty and above NMI, PSS decreases with increasing inequality until Gini=48, then levels off. (opposite of CESD result)
For Blacks below poverty and below NMI, PSS initially increases until Gini= 43, then decreases. (similar to CESD for this group)
(One through four scale, with higher numbers indicating more frequent discrimination)
Whites:
For Whites above NMI, discrimination due to race ratings increase until about Gini=45, then decrease. For Whites below NMI, discrimination ratings increase with greater inequality. (similar to CESD results for Whites below poverty)
Blacks:
For Blacks above NMI, discrimination ratings initially decrease till Gini=48, then increase. For Blacks below NMI, discrimination rates are more or less the same with increasing inequality.
(One through four scale, with higher numbers indicating more frequent discrimination)
Whites:
Whites above poverty report more or less the same levels of discrimination across inequality levels. Whites above poverty report higher discrimination with increasing inequality.
Blacks:
Blacks above poverty report greater discrimination with increasing inequality. In contrast, Blacks below poverty report lesser discrimination with greater inequality.
Whites:
Blacks:
Main effects:
Whites:
Whites below NMI report greater instrumental support than those above NMI.
Blacks:
Blacks below and above NMI report roughly the same amount of instrumental support. In general, Balcks report greater levels of instrumental support than Whites.
Generally, with greater inequality, if you're below NMI, then you report more emotional support.
Whites:
Blacks:
