{rsetup,include=FALSE}

knitr::opts_chunk$set(echo=TRUE)

Instructions

.alert .alert-danger

Getting started

1. You have been provided a data file, colleges.csv and this R Markdown file to complete the exam.

2. Read the data briefing. It will provide useful information about the variables in the dataset.

3. Attempt all questions. You have 100 minutes, including reading time of 10 minutes.

4. All answers must be presented in the appropriate sections and numbered according to the question.

5. All plots must be in ggplot2. Note that diagnostic plots (Residual, QQ plots) that are used to test for data assumptions may remain as base R plots.

6. Do not edit this template. Only provide answers in the appropriate blocks (see below).

7. For all hypothesis testing questions, unless stated otherwise, assume that

   $\alpha =0.05$

   .

   
   

Writing your answers

• All answers should be written within the answer alert blocks provided. These alert boxes look like below in the source document:

Submission

1. When the time is up (100 min), you will have 15 minutes to submit your assignment.

2. Only .html file submissions are accepted.

3. You will not be able to submit an incomplete .Rmd file.

4. Rename your final file to XXXXXXXXXX_MATH1115_Exam.html where XXXXXXXXXX is your SID.

Acknowledgement

Please acknowledge that you have read the instructions above by replacing XXXXXXXXXX with your SID.

.alert .alert-info

Briefing on Data

The given file, data.csv, contains information about universities and colleges in the U.S.A. Most of the data concerns the 1993-94 academic year. The data consists of the following variables,

• University: name of the university or college.

• Private: variable which equals 1 if the university or college is private and 0 if it is public.

• Apps: number of applications received.

• Accept: number of applicants accepted.

• Enrol: number of new students enrolled.

• Top10pct: percentage of new students from top 10% of high school class.

• Top25pct: percentage of new students from top 25% of high school class.

• FullTime: number of full-time undergraduates.

• PartTime: number of part-time undergraduates.

• RoomBaord: room and board costs.

• Books: estimated book costs.

• Personal: estimated personal spending.

The source of the data is the 1995 U.S. News & World Report's Guide to America's Best Colleges. This dataset may be obtained from the StatLib Library at Carnegie Mellon University,

http://lib.stat.cmu.edu/datasets/

The given file, data.csv, is an abridged version of the original dataset.

Research Task

Hoachen, your boss, is interested in understanding trends in the living expenses of university students in 1993, as a comparison to current reports. Your task is to help Haochen analyse the data.csv using this template provided.

The code chunk below will load the tidyverse for you.

{r,include=FALSE}

library(tidyverse)

Initial data analysis

Read colleges.csv into your R Markdown file such that the code below works. Your data object should be called crime. Set eval=TRUE once this is done.

{r,eval=FALSE}

colleges<-read.csv("colleges.csv")

Question 1: Write a short paragraph on at least one limitation and at least one ethical consideration associated with this data.

Question 2: Provide both a numerical summary and graphical summary for the variable "Books". Write a short paragraph on your findings for both the numerical summary and the graphical summary.

Question 3: For each of the variables "Private" and "RoomBoard", provide a graphical summary using ggplot. Write a short paragraph summarising your findings.

.alert .alert-info

Simple linear regression

Research Question: Haochen claims that there is a positive linear relationship between the number of applications received and the number of applications accepted. Evaluate the evidence in favour of this claim by performing a simple linear regression.

.alert .alert-info

Chi-squared test

Research Question: It is claimed that, on average, Chatham College and Lesley College have equal numbers of new enrolments every year, while Stephens College has twice as many yearly enrolments as the two other colleges. Perform a chi-squared test with

.alert .alert-info

T-test

In general, were the cost of books more expensive in private colleges compared to non-private colleges?

1. Visualise the relationship between Private and Books.

2. Perform a t-test comparing the mean cost of books between the two types of colleges. Make sure to use H A T P C. Note: if the assumptions were not met, interpret the results anyway.

.alert .alert-info

Data wrangling

Document the code below using comments (i.e. #) and explain what each line is doing. You may add comments above or beside each line.

{r,echo=T,warning=F,message=F,eval=F}

library(tidyverse)

boxplot(data$Enrol)

df<-data(data$Apps>50,)

data%>%

filter(data$Apps<100)%>%

dim%>%

nrow

data%>%

select(Private,Accept)%>%

group_by(Private)%>%

summarise(Accept=sum(Accept))

data$Enrol%>%

max(.,na.rm=T)

End

This is the end of the exam.