Homework 6 Stats 2
Joshua Reyna
3/15/2021
library(car)## Loading required package: carDatalibrary(stargazer)##
## Please cite as:## Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.## R package version 5.2.2. https://CRAN.R-project.org/package=stargazerlibrary(survey)## Loading required package: grid## Loading required package: Matrix## Loading required package: survival##
## Attaching package: 'survey'## The following object is masked from 'package:graphics':
##
## dotchartlibrary(ggplot2)library(haven)
NSDUH_2019 <- read_sav("NSDUH_2019.SAV")
View(NSDUH_2019)## marital status
NSDUH_2019$marst<-Recode(NSDUH_2019$irmarit, recodes="1='married'; 2='divorced'; 3='widowed'; 4='separated'; else=NA", as.factor=T)
NSDUH_2019$marst<-relevel(NSDUH_2019$marst, ref='married')
## education recodes
NSDUH_2019$educ<-Recode(NSDUH_2019$ireduhighst2, recodes="1:7='LssThnHgh'; 8='highschool'; 9='someCollege'; 10='associates'; 11='colgrad';else=NA", as.factor=T)
NSDUH_2019$educ<-relevel(NSDUH_2019$educ, ref='colgrad')
## sexuality recodes
NSDUH_2019$sexuality<-Recode(NSDUH_2019$sexident, recodes="1='Heterosexual'; 2='Les/Gay'; 3='Bisexual';else=NA", as.factor=T)
NSDUH_2019$sexuality<-relevel(NSDUH_2019$sexuality, ref='Heterosexual')
## gender recodes
NSDUH_2019$male<-as.factor(ifelse(NSDUH_2019$irsex==1, "Male", "Female"))
## Race recoded items
NSDUH_2019$black<-Recode(NSDUH_2019$newrace2, recodes="2=1; 9=NA; else=0")
NSDUH_2019$white<-Recode(NSDUH_2019$newrace2, recodes="1=1; 9=NA; else=0")
NSDUH_2019$other<-Recode(NSDUH_2019$newrace2, recodes="3:4=1; 9=NA; else=0")
NSDUH_2019$mult_race<-Recode(NSDUH_2019$newrace2, recodes="6=1; 9=NA; else=0")
NSDUH_2019$asian<-Recode(NSDUH_2019$newrace2, recodes="5=1; 9=NA; else=0")
NSDUH_2019$hispanic<-Recode(NSDUH_2019$newrace2, recodes="7=1; 9=NA; else=0")
NSDUH_2019$race_eth<-Recode(NSDUH_2019$newrace2,
recodes="1='white'; 2='black'; 3='other'; 4='asian'; 5='mult_race'; 6='hispanic'; else=NA",
as.factor = T)
NSDUH_2019$race_eth<-relevel(NSDUH_2019$race_eth, ref='white')
NSDUH_2019$lst_alc_use2<-Recode(NSDUH_2019$iralcrc, recodes="1='last 30days'; 2='12>1month'; 3='>12months'; else=NA", as.factor=T)
NSDUH_2019$dep_year2<-Recode(NSDUH_2019$amdeyr, recodes="1=1; 2=0;else=NA", as.factor=T)
NSDUH_2019$age_cat<-Recode(NSDUH_2019$age2, recodes="7:8='18-19'; 9:10='20-21'; 11='22-23'; 12='24-25'; 13='26-29'; 14='30-34'; 15='35-49'; 16='50-64'; 17='65+'; else=NA", as.factor=T)Histogram
NSDUH_2019$alcyrtot2<-Recode(NSDUH_2019$alcyrtot, recodes = "985:998=NA")
hist(NSDUH_2019$alcyrtot2)
summary(NSDUH_2019$alcyrtot2)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 1.00 12.00 48.00 80.83 120.00 365.00 23006library(dplyr)##
## Attaching package: 'dplyr'## The following object is masked from 'package:car':
##
## recode## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionsub<-NSDUH_2019%>%
select(alcyrtot2, age_cat, race_eth,
marst, educ, lst_alc_use2, dep_year2, white, black, hispanic,
other, mult_race, asian, hispanic, income, male, sexuality, analwtc, vestr) %>%
filter( complete.cases(.))
options(survey.lonely.psu = "adjust")
des<-svydesign(ids=~1, strata=~vestr, weights=~analwtc, data =sub )
options(survey.lonely.psu = "adjust")
des<-svydesign(ids=~1, strata=~vestr,
weights=~analwtc,
data = NSDUH_2019[is.na(NSDUH_2019$analwtc)==F,])svyhist(~alcyrtot2, des)
svyby(~alcyrtot2, ~race_eth+educ, des, svymean, na.rm=T)## race_eth educ alcyrtot2 se
## white.colgrad white colgrad 110.01825 1.733805
## asian.colgrad asian colgrad 40.86604 13.174518
## black.colgrad black colgrad 74.13335 4.002535
## hispanic.colgrad hispanic colgrad 88.63893 9.033293
## mult_race.colgrad mult_race colgrad 58.56157 4.176777
## other.colgrad other colgrad 57.23540 8.537946
## white.associates white associates 99.52782 3.499248
## asian.associates asian associates 71.75474 23.619244
## black.associates black associates 74.84763 5.405869
## hispanic.associates hispanic associates 96.85974 15.445061
## mult_race.associates mult_race associates 75.74198 12.793567
## other.associates other associates 56.40716 18.401269
## white.highschool white highschool 94.33584 2.588339
## asian.highschool asian highschool 74.41367 16.201048
## black.highschool black highschool 84.50323 3.988172
## hispanic.highschool hispanic highschool 114.50072 16.537347
## mult_race.highschool mult_race highschool 91.27327 17.205912
## other.highschool other highschool 101.23943 12.643581
## white.LssThnHgh white LssThnHgh 69.69736 3.493612
## asian.LssThnHgh asian LssThnHgh 38.90550 10.571440
## black.LssThnHgh black LssThnHgh 82.48821 6.010704
## hispanic.LssThnHgh hispanic LssThnHgh 59.63737 10.219330
## mult_race.LssThnHgh mult_race LssThnHgh 57.93165 24.880795
## other.LssThnHgh other LssThnHgh 62.47405 16.481318
## white.someCollege white someCollege 94.20759 2.281043
## asian.someCollege asian someCollege 68.95634 12.631771
## black.someCollege black someCollege 83.41700 4.333830
## hispanic.someCollege hispanic someCollege 88.36257 10.059974
## mult_race.someCollege mult_race someCollege 67.49612 6.942847
## other.someCollege other someCollege 87.10928 12.856316##1) Define a count outcome for the dataset of your choosing ## A count outcome variable known in the codebook as alcyrtot, or total number of days alcohol was used in the past year. ## a. State a research question about your outcome ##Disadvantaged groups by demographic type and socioeconomic status are more at risk of more days of alcohol usage compared to those that are not disadvantaged. This is expected to be similarly found by racial category, at baseline whites should be at lower risk of using alcohol throughout the year than other racial categories. ##b. Is an offset term necessary? why or why not? ##An offset term is not neccessary, as every adult person in the study was asked about their alcohol usage using the same time period of 1 year or 365 days. As time went on as indicated by the histogram, less people used alcohol, a much larger concentration of people used it during the early period of the year. This is clearly not a normally distrbibuted variable.
##2) Consider a Poisson regression model for the outcome ## Poisson Regression Model
svyhist(~alcyrtot2, des)
fit1<-svyglm(alcyrtot2~factor(race_eth)+factor(educ)+factor(age_cat)+factor(marst)+factor(sexuality)+factor(male)+scale(income)+factor(lst_alc_use2)+factor(dep_year2), design=des, family=poisson)
summary(fit1)##
## Call:
## svyglm(formula = alcyrtot2 ~ factor(race_eth) + factor(educ) +
## factor(age_cat) + factor(marst) + factor(sexuality) + factor(male) +
## scale(income) + factor(lst_alc_use2) + factor(dep_year2),
## design = des, family = poisson)
##
## Survey design:
## svydesign(ids = ~1, strata = ~vestr, weights = ~analwtc, data = NSDUH_2019[is.na(NSDUH_2019$analwtc) ==
## F, ])
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.454456 0.068345 35.913 < 2e-16 ***
## factor(race_eth)asian -0.488750 0.116292 -4.203 2.65e-05 ***
## factor(race_eth)black -0.142283 0.027443 -5.185 2.18e-07 ***
## factor(race_eth)hispanic -0.011118 0.057021 -0.195 0.845412
## factor(race_eth)mult_race -0.371465 0.054244 -6.848 7.66e-12 ***
## factor(race_eth)other 0.008578 0.075423 0.114 0.909452
## factor(educ)associates -0.022904 0.032328 -0.708 0.478649
## factor(educ)highschool 0.012922 0.027595 0.468 0.639585
## factor(educ)LssThnHgh 0.030781 0.043268 0.711 0.476834
## factor(educ)someCollege -0.019943 0.025453 -0.784 0.433318
## factor(age_cat)20-21 0.282204 0.055705 5.066 4.09e-07 ***
## factor(age_cat)22-23 0.478663 0.054510 8.781 < 2e-16 ***
## factor(age_cat)24-25 0.527403 0.053895 9.786 < 2e-16 ***
## factor(age_cat)26-29 0.562635 0.053455 10.525 < 2e-16 ***
## factor(age_cat)30-34 0.577600 0.054626 10.574 < 2e-16 ***
## factor(age_cat)35-49 0.630430 0.053651 11.751 < 2e-16 ***
## factor(age_cat)50-64 0.715680 0.056638 12.636 < 2e-16 ***
## factor(age_cat)65+ 0.808953 0.059187 13.668 < 2e-16 ***
## factor(marst)divorced 0.085370 0.059462 1.436 0.151100
## factor(marst)separated 0.088914 0.025062 3.548 0.000389 ***
## factor(marst)widowed 0.088877 0.030004 2.962 0.003058 **
## factor(sexuality)Bisexual 0.095272 0.037729 2.525 0.011570 *
## factor(sexuality)Les/Gay 0.147879 0.053489 2.765 0.005702 **
## factor(male)Male 0.232782 0.019057 12.215 < 2e-16 ***
## scale(income) 0.022603 0.011618 1.945 0.051730 .
## factor(lst_alc_use2)last 30days 1.484232 0.040291 36.838 < 2e-16 ***
## factor(dep_year2)1 -0.003427 0.032058 -0.107 0.914874
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for poisson family taken to be 98.65696)
##
## Number of Fisher Scoring iterations: 6round(exp(summary(fit1)$coef[-1,1]), 3)## factor(race_eth)asian factor(race_eth)black
## 0.613 0.867
## factor(race_eth)hispanic factor(race_eth)mult_race
## 0.989 0.690
## factor(race_eth)other factor(educ)associates
## 1.009 0.977
## factor(educ)highschool factor(educ)LssThnHgh
## 1.013 1.031
## factor(educ)someCollege factor(age_cat)20-21
## 0.980 1.326
## factor(age_cat)22-23 factor(age_cat)24-25
## 1.614 1.695
## factor(age_cat)26-29 factor(age_cat)30-34
## 1.755 1.782
## factor(age_cat)35-49 factor(age_cat)50-64
## 1.878 2.046
## factor(age_cat)65+ factor(marst)divorced
## 2.246 1.089
## factor(marst)separated factor(marst)widowed
## 1.093 1.093
## factor(sexuality)Bisexual factor(sexuality)Les/Gay
## 1.100 1.159
## factor(male)Male scale(income)
## 1.262 1.023
## factor(lst_alc_use2)last 30days factor(dep_year2)1
## 4.412 0.997##a. Evaluate the level of dispersion in the outcome b. Is the Poisson model a good choice? ## According to the models level of dispersion, the poisson regression model is not a good fit for this particular model. This models dispersion being 184.60, which is severely above the 1.0 threshold, and indicates a high level of dispersion. From here another model is recommended in that allows for the more variability seen in this model.
Scaling for over dispersion work
fit2<-glm(alcyrtot2~factor(race_eth)+factor(educ)+factor(age_cat)+factor(marst)+factor(sexuality)+factor(male)+scale(income)+factor(lst_alc_use2)+factor(dep_year2), data=NSDUH_2019, family=poisson)
summary(fit2)##
## Call:
## glm(formula = alcyrtot2 ~ factor(race_eth) + factor(educ) + factor(age_cat) +
## factor(marst) + factor(sexuality) + factor(male) + scale(income) +
## factor(lst_alc_use2) + factor(dep_year2), family = poisson,
## data = NSDUH_2019)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -17.855 -7.187 -3.309 3.709 33.049
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.5546061 0.0052850 483.366 < 2e-16 ***
## factor(race_eth)asian -0.1992896 0.0105016 -18.977 < 2e-16 ***
## factor(race_eth)black -0.1079351 0.0021691 -49.761 < 2e-16 ***
## factor(race_eth)hispanic -0.0376994 0.0035315 -10.675 < 2e-16 ***
## factor(race_eth)mult_race -0.3242750 0.0037758 -85.884 < 2e-16 ***
## factor(race_eth)other 0.0191324 0.0060730 3.150 0.00163 **
## factor(educ)associates -0.0627613 0.0024031 -26.117 < 2e-16 ***
## factor(educ)highschool -0.0104790 0.0019387 -5.405 6.47e-08 ***
## factor(educ)LssThnHgh 0.0152747 0.0030463 5.014 5.33e-07 ***
## factor(educ)someCollege -0.0378893 0.0018746 -20.212 < 2e-16 ***
## factor(age_cat)20-21 0.2975109 0.0048148 61.791 < 2e-16 ***
## factor(age_cat)22-23 0.4784333 0.0045802 104.457 < 2e-16 ***
## factor(age_cat)24-25 0.5047068 0.0045248 111.541 < 2e-16 ***
## factor(age_cat)26-29 0.5407936 0.0045282 119.428 < 2e-16 ***
## factor(age_cat)30-34 0.5698645 0.0045375 125.590 < 2e-16 ***
## factor(age_cat)35-49 0.6266393 0.0043965 142.533 < 2e-16 ***
## factor(age_cat)50-64 0.7000581 0.0046555 150.372 < 2e-16 ***
## factor(age_cat)65+ 0.8050754 0.0048619 165.589 < 2e-16 ***
## factor(marst)divorced 0.0703371 0.0043426 16.197 < 2e-16 ***
## factor(marst)separated 0.1108477 0.0019284 57.480 < 2e-16 ***
## factor(marst)widowed 0.0957282 0.0022998 41.625 < 2e-16 ***
## factor(sexuality)Bisexual 0.1083087 0.0028224 38.375 < 2e-16 ***
## factor(sexuality)Les/Gay 0.0888386 0.0042132 21.086 < 2e-16 ***
## factor(male)Male 0.2412059 0.0013840 174.286 < 2e-16 ***
## scale(income) 0.0223243 0.0008098 27.567 < 2e-16 ***
## factor(lst_alc_use2)last 30days 1.3826601 0.0028727 481.312 < 2e-16 ***
## factor(dep_year2)1 0.0275527 0.0022154 12.437 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for poisson family taken to be 1)
##
## Null deviance: 2260301 on 24982 degrees of freedom
## Residual deviance: 1770061 on 24956 degrees of freedom
## (31153 observations deleted due to missingness)
## AIC: 1909340
##
## Number of Fisher Scoring iterations: 6scale<-sqrt(fit2$deviance/fit2$df.residual)
scale## [1] 8.4218331-pchisq(fit2$deviance, df = fit2$df.residual)## [1] 0##Quasi Model Work
fit3<-glm(alcyrtot2~factor(race_eth)+factor(educ)+factor(age_cat)+factor(marst)+factor(sexuality)+factor(male)+scale(income)+factor(lst_alc_use2)+factor(dep_year2), data=NSDUH_2019, family=quasipoisson)
summary(fit3)##
## Call:
## glm(formula = alcyrtot2 ~ factor(race_eth) + factor(educ) + factor(age_cat) +
## factor(marst) + factor(sexuality) + factor(male) + scale(income) +
## factor(lst_alc_use2) + factor(dep_year2), family = quasipoisson,
## data = NSDUH_2019)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -17.855 -7.187 -3.309 3.709 33.049
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.554606 0.047478 53.806 < 2e-16 ***
## factor(race_eth)asian -0.199290 0.094341 -2.112 0.03466 *
## factor(race_eth)black -0.107935 0.019486 -5.539 3.07e-08 ***
## factor(race_eth)hispanic -0.037699 0.031725 -1.188 0.23472
## factor(race_eth)mult_race -0.324275 0.033919 -9.560 < 2e-16 ***
## factor(race_eth)other 0.019132 0.054556 0.351 0.72582
## factor(educ)associates -0.062761 0.021588 -2.907 0.00365 **
## factor(educ)highschool -0.010479 0.017416 -0.602 0.54739
## factor(educ)LssThnHgh 0.015275 0.027367 0.558 0.57675
## factor(educ)someCollege -0.037889 0.016840 -2.250 0.02446 *
## factor(age_cat)20-21 0.297511 0.043254 6.878 6.20e-12 ***
## factor(age_cat)22-23 0.478433 0.041146 11.628 < 2e-16 ***
## factor(age_cat)24-25 0.504707 0.040649 12.416 < 2e-16 ***
## factor(age_cat)26-29 0.540794 0.040679 13.294 < 2e-16 ***
## factor(age_cat)30-34 0.569865 0.040763 13.980 < 2e-16 ***
## factor(age_cat)35-49 0.626639 0.039495 15.866 < 2e-16 ***
## factor(age_cat)50-64 0.700058 0.041822 16.739 < 2e-16 ***
## factor(age_cat)65+ 0.805075 0.043677 18.433 < 2e-16 ***
## factor(marst)divorced 0.070337 0.039012 1.803 0.07141 .
## factor(marst)separated 0.110848 0.017324 6.398 1.60e-10 ***
## factor(marst)widowed 0.095728 0.020660 4.633 3.61e-06 ***
## factor(sexuality)Bisexual 0.108309 0.025355 4.272 1.95e-05 ***
## factor(sexuality)Les/Gay 0.088839 0.037849 2.347 0.01892 *
## factor(male)Male 0.241206 0.012433 19.401 < 2e-16 ***
## scale(income) 0.022324 0.007275 3.069 0.00215 **
## factor(lst_alc_use2)last 30days 1.382660 0.025807 53.578 < 2e-16 ***
## factor(dep_year2)1 0.027553 0.019902 1.384 0.16624
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for quasipoisson family taken to be 80.70269)
##
## Null deviance: 2260301 on 24982 degrees of freedom
## Residual deviance: 1770061 on 24956 degrees of freedom
## (31153 observations deleted due to missingness)
## AIC: NA
##
## Number of Fisher Scoring iterations: 63) Consider a Negative binomial model
Negative Binomial Model
library(lmtest)## Loading required package: zoo##
## Attaching package: 'zoo'## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numericlibrary(sandwich)
coeftest(fit2, vcov=vcovHC(fit2, type="HC1",cluster="vestr"))##
## z test of coefficients:
##
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 2.5546061 0.0471625 54.1660 < 2.2e-16 ***
## factor(race_eth)asian -0.1992896 0.0912469 -2.1841 0.028957 *
## factor(race_eth)black -0.1079351 0.0196561 -5.4912 3.993e-08 ***
## factor(race_eth)hispanic -0.0376994 0.0327476 -1.1512 0.249646
## factor(race_eth)mult_race -0.3242750 0.0351083 -9.2364 < 2.2e-16 ***
## factor(race_eth)other 0.0191324 0.0530902 0.3604 0.718566
## factor(educ)associates -0.0627613 0.0218199 -2.8763 0.004023 **
## factor(educ)highschool -0.0104790 0.0177666 -0.5898 0.555314
## factor(educ)LssThnHgh 0.0152747 0.0288400 0.5296 0.596364
## factor(educ)someCollege -0.0378893 0.0167849 -2.2573 0.023986 *
## factor(age_cat)20-21 0.2975109 0.0418530 7.1085 1.173e-12 ***
## factor(age_cat)22-23 0.4784333 0.0401811 11.9069 < 2.2e-16 ***
## factor(age_cat)24-25 0.5047068 0.0402217 12.5481 < 2.2e-16 ***
## factor(age_cat)26-29 0.5407936 0.0399452 13.5384 < 2.2e-16 ***
## factor(age_cat)30-34 0.5698645 0.0403261 14.1314 < 2.2e-16 ***
## factor(age_cat)35-49 0.6266393 0.0392568 15.9626 < 2.2e-16 ***
## factor(age_cat)50-64 0.7000581 0.0418884 16.7124 < 2.2e-16 ***
## factor(age_cat)65+ 0.8050754 0.0441789 18.2231 < 2.2e-16 ***
## factor(marst)divorced 0.0703371 0.0442580 1.5893 0.112003
## factor(marst)separated 0.1108477 0.0171592 6.4600 1.047e-10 ***
## factor(marst)widowed 0.0957282 0.0214408 4.4648 8.015e-06 ***
## factor(sexuality)Bisexual 0.1083087 0.0255481 4.2394 2.241e-05 ***
## factor(sexuality)Les/Gay 0.0888386 0.0367351 2.4184 0.015591 *
## factor(male)Male 0.2412059 0.0124550 19.3662 < 2.2e-16 ***
## scale(income) 0.0223243 0.0073309 3.0452 0.002325 **
## factor(lst_alc_use2)last 30days 1.3826601 0.0259688 53.2432 < 2.2e-16 ***
## factor(dep_year2)1 0.0275527 0.0204328 1.3485 0.177512
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1library(MASS)##
## Attaching package: 'MASS'## The following object is masked from 'package:dplyr':
##
## selectfit.nb1<-glm.nb(alcyrtot2~factor(race_eth),
data=sub,
weights=analwtc/mean(analwtc, na.rm=T))
fit.nb2<-glm.nb(alcyrtot2~factor(race_eth)+factor(educ)+factor(age_cat)+factor(marst)+factor(sexuality)+factor(male)+scale(income)+factor(lst_alc_use2)+factor(dep_year2),
data=sub,
weights=analwtc/mean(analwtc, na.rm=T))
#clx2(fit.nb2,cluster =sub$ststr)
tests1<-coeftest(fit.nb1, vcov=vcovHC(fit.nb2, type="HC1",cluster="vestr"))
tests<-coeftest(fit.nb2, vcov=vcovHC(fit.nb2, type="HC1",cluster="vestr"))
library(stargazer)
stargazer(fit.nb1, fit.nb2,style="demography", type = "text", t.auto=F,p.auto=F,coef=list(tests1[, 1],tests[,1]), se =list(tests1[, 2], tests[, 2]), p=list(tests1[,4],tests[, 4]) )##
## -----------------------------------------------------------------
## alcyrtot2
## Model 1 Model 2
## -----------------------------------------------------------------
## factor(race_eth)asian -0.505*** -0.404**
## (0.145) (0.145)
## factor(race_eth)black -0.208*** -0.062
## (0.034) (0.034)
## factor(race_eth)hispanic -0.069 -0.053
## (0.053) (0.053)
## factor(race_eth)mult_race -0.462*** -0.396***
## (0.052) (0.052)
## factor(race_eth)other -0.189 0.044
## (0.103) (0.103)
## factor(educ)associates -0.031
## (0.036)
## factor(educ)highschool 0.040
## (0.032)
## factor(educ)LssThnHgh 0.045
## (0.051)
## factor(educ)someCollege -0.023
## (0.029)
## factor(age_cat)20-21 0.247***
## (0.069)
## factor(age_cat)22-23 0.463***
## (0.071)
## factor(age_cat)24-25 0.500***
## (0.069)
## factor(age_cat)26-29 0.525***
## (0.069)
## factor(age_cat)30-34 0.549***
## (0.069)
## factor(age_cat)35-49 0.556***
## (0.068)
## factor(age_cat)50-64 0.625***
## (0.072)
## factor(age_cat)65+ 0.727***
## (0.077)
## factor(marst)divorced 0.032
## (0.065)
## factor(marst)separated 0.090**
## (0.032)
## factor(marst)widowed 0.080*
## (0.035)
## factor(sexuality)Bisexual 0.103**
## (0.038)
## factor(sexuality)Les/Gay 0.110*
## (0.056)
## factor(male)Male 0.233***
## (0.021)
## scale(income) 0.004
## (0.013)
## factor(lst_alc_use2)last 30days 1.483***
## (0.037)
## factor(dep_year2)1 0.004
## (0.037)
## Constant 4.616*** 2.520***
## (0.080) (0.080)
## N 24,983 24,983
## Log Likelihood -138,637.200 -135,506.600
## theta 0.792*** (0.006) 0.966*** (0.008)
## AIC 277,286.300 271,067.200
## -----------------------------------------------------------------
## *p < .05; **p < .01; ***p < .001##4) Compare the model fits of the alternative models using AIC
AIC(fit1)## eff.p AIC deltabar
## 4942.1145 2216200.6628 190.0813AIC(fit2)## [1] 1909340AIC(fit3)## [1] NAAIC(fit.nb1)## [1] 277286.3AIC(fit.nb2)## [1] 271067.2