Homework 5

#1. Define a binary outcome of your choosing

Have you been told in the past you have depression? addepev3 > depression 1: yes, 2: no

hardtoget<-haven::read_xpt("/Users/christacrumrine/Downloads/LLCP2020.XPT ")

names(hardtoget)<-tolower(gsub(pattern = "_", replacement = "",x=names(hardtoget)))

#tolower to make everything lowercase #names used in order to not change dataset, but to the right use the dataset name #can highlight part of chunk and hit command/enter

#1. Define a binary outcome of your choosing

1 Define an ordinal or multinomial outcome variable of your choosing and define how you will recode the original variable.

Provided regular care for family or friend: 1=1 up to 8 hours per week 2=2 9 to 19 hours per week 3=3 20 to 39 hours per week 4=4 40 hours or more per week else=NA

library(dplyr)
sub<-hardtoget%>%
  select(depression, hoursofcare, marst, race, ADLcare, IADLcare, llcpwt, ststr)%>%
  filter( complete.cases(.))

#First we tell R our survey design
options(survey.lonely.psu = "adjust")
des<-svydesign(ids=~1,
               strata=~ststr,
               weights=~llcpwt,
               data =sub )

State a research question about what factors you believe will affect your outcome variable.

Research question

Is there an association between the depression and hours of care provided to a family member for friend

Other Predictor Variables that I believe will affect the outcome are as followed:

1. marital status

2. race

3. assist personal care

4. assist c home tasks

3. Fit the ordinal or the multinomial logistic regression models to your outcome.

#Multinomial Model

mfit<-svy_vglm(depression~hoursofcare+marst+race+ADLcare+IADLcare,
           family=multinomial(refLevel = 1),
           design = des)

mfit%>%
  tbl_regression()

## ! `broom::tidy()` failed to tidy the model.

## x No tidy method for objects of class svy_vglm

## ✓ `tidy_parameters()` used instead.

## ℹ Add `tidy_fun = broom.helpers::tidy_parameters` to quiet these messages.

## x Unable to identify the list of variables.
## 
## This is usually due to an error calling `stats::model.frame(x)`or `stats::model.matrix(x)`.
## It could be the case if that type of model does not implement these methods.
## Rarely, this error may occur if the model object was created within
## a functional programming framework (e.g. using `lappy()`, `purrr::map()`, etc.).

Characteristic	Beta	95% CI1	p-value
(Intercept)	-2.2	-3.5, -0.79	0.002
hoursofcare2	0.85	0.19, 1.5	0.012
hoursofcare3	-0.29	-1.2, 0.60	0.5
hoursofcare4	0.32	-0.34, 1.0	0.3
marstdivorced	-0.01	-1.0, 1.0	>0.9
marstmarried	-0.57	-1.5, 0.32	0.2
marstnm	0.74	-0.22, 1.7	0.13
marstseparated	0.81	-0.47, 2.1	0.2
marstwidowed	-0.87	-2.4, 0.71	0.3
race77	-15	-16, -14	<0.001
race99	0.86	-0.58, 2.3	0.2
raceAsian	-1.2	-3.5, 1.0	0.3
raceblack	-0.70	-1.7, 0.28	0.2
racenhwhite	-0.17	-1.0, 0.67	0.7
raceother	-0.30	-2.9, 2.3	0.8
raceother_race	0.39	-1.1, 1.9	0.6
racePacific_Islander	-15	-16, -14	<0.001
ADLcare	1.0	0.43, 1.5	<0.001
IADLcare	0.89	0.22, 1.6	0.010
1 CI = Confidence Interval

In terms of depression, those who not married are more likely than married couple to report depression while providing for someone. People who are divorced are more likely than married individuals to report feeling depressed.

In regards to race, Asians are less likely than NH whites to feel depressed while caring for someone. Pacific Islanders have a significant P value, which a high beta. This could be because of a small sample. Blacks are more likely than NH whites to report depression while caring for someone.

The proportional odds assumption is violated because the p-value is significant.