homework05_old.Rmd

---
title: "Homework 05"
---

## Homework 05 - due 11/01/2017

### ANCOVA - Analysis of Covariance Exercise

For Homework 05, you will be using the HELP dataset, learn more at:

* [https://melindahiggins2000.github.io/N736Fall2017_HELPdataset/](https://melindahiggins2000.github.io/N736Fall2017_HELPdataset/) &
* [https://github.com/melindahiggins2000/N736Fall2017_HELPdataset](https://github.com/melindahiggins2000/N736Fall2017_HELPdataset)

Complete the following for these variables:

* OUTCOME VARIABLE (Y): `indtot`
* INDEPENDENT VARIABLE (X): `mcs`
* COVARIATES (other X's): `pss_fr` or `female`

1. Run ANCOVA _(using a regression, ANOVA, or GLM approach - your choice)_ for the association between the SF36 Mental Component Score (`mcs`) and Inventory of Drug Use (`indtot`) adjusting for perceived social support from friends (`pss_fr`). Remember to:
    * mean center continuous variables before computing the interaction term _(i.e. create a new mean-centered variable by subtracting the mean)_
    * check for the assumption of homogenity of variance _(i.e. is the interaction term significant?)_
    * make an "effects plot" plot of the interaction between `mcs` and `pss_fr` 
    
2. Run ANCOVA _(using a regression, ANOVA, or GLM approach - your choice)_ for the association between the SF36 Mental Component Score (`mcs`) and Inventory of Drug Use (`indtot`) adjusting for gender (`female`). Remember to:
    * mean center continuous variables before computing the interaction term _(i.e. create a new mean-centered variable by subtracting the mean)_
    * check for the assumption of homogenity of variance _(i.e. is the interaction term significant?)_
    * make an "effects plot" plot of the interaction between `mcs` and `female` 

## Variables in HELP dataset to be used for Homework 05

```{r, echo=FALSE, message=FALSE, warning=FALSE}
helpdata <- readRDS("helpmkh.rds")
library(tidyverse)
sub1 <- helpdata %>%
  select(indtot,mcs,pss_fr,female)

# create a function to get the label
# label output from the attributes() function
getlabel <- function(x) attributes(x)$label
# getlabel(sub1$age)

library(purrr)
ldf <- purrr::map_df(sub1, getlabel) # this is a 1x15 tibble data.frame
# t(ldf) # transpose for easier reading to a 15x1 single column list

# using knitr to get a table of these
# variable names for Rmarkdown
library(knitr)
knitr::kable(t(ldf),
             col.names = c("Variable Label"),
             caption="Use these variables from HELP dataset for Homework 05")

```