Exercise $1(50$ points$)$

Create the following visualizations using $R($ggplot$2)$ and dataset $($acs$\_$ny$.$csv$),$ which is a subset of the

$2010$ American Community Survey $($ACS$)$ for the New York state.

a$)$ Create a bar chart for income levels $($ $$50,000$ or less, $$50,001-$ $$100,000,$$$100,001-$$$150,000,$

and more than $$150,000)$ and language spoken at home.

b$)$ b$)$ Create a visualization to show the relationship between familyIncome $($ $$50,000$ or less,

$$50,001-$ $$100,000,$$$100,001-$ $$150,000,$ and more than $$150,000)$ and average HouseCost.

c$)$ Create scatterplots using $R$ to visualize relationship between the dependent variable

$($FamilyIncome$)$ and each independent variable.

d$)$ Using the dataset, create a multiple regression model in $R.$ The purpose of this model is to

predict the household income.

Create and compare at least four models.

Perform Stepwise regression to find an optimal set of predictors.

Explain your results including $r$ squared and $p-$values.

Which model provides better accuracy? Why?

Does your model conform to the assumptions of linear regression?

Exercise $2(50$ points$)$

Using the provided subset $($acs$\_$ny$.$csv$)$ of the $2010$ American Community Survey $($ACS$)$ for New York

state, create a logistic regression model in R$.$ The purpose of this model is to predict whether a

household has an income greater than $$150,000.$

a$)$ Explain your results including deviance residuals, coefficients, and AIC.

b$)$ Create coefficient plot for logistic regression on family income greater than $$150,000.$

Hint: You will need to create a new binary variable with TRUE for income above $$150,000$ and FALSE for

income below. It will be helpful to create density plot of family income to see distribution.