Do male geography students drink more than females? Use an appropriate plot to visualize the differences and an appropriate test to test their significance.

 Assessments

Task: To produce an A3 An infographic containing text, tables, and figures to visually represent information on the lifestyle and values of the GEO1018 class collected from the class questionnaire.

Definition:   

An Infographic, or Information Graphic is a graphic visual representation of information, data or knowledge intended to present complex information quickly and clearly.

How to do it:

You should use R and ArcMap to produce tables of summary statistics, figures and maps showing graphical and geographic visualisations of the data from the GEO1018 survey. These should be mixed with short blocks of text to explain some of the key findings from the survey. Your final infographic should be both informative and visually appealing. Try to keep text to a minimum - let the graphics tell the story and use the text to explain the graphic and highlight any interesting / controversial findings. You can work in groups to analyse the survey data but should work individually to create your infographic.

You will produce tables of summary statistics and graphics in R along with maps in ArcMap and assemble these together with linking / explanatory text in PowerPoint (or any graphics package you know well). You should generate summary statistics using R but can then use Excel/Powerpoint (or any appropriate software) to format these into a visually attractive table. You can use any of the data in the survey and should think of the best way to represent it (table or graphic, what type of graphic?). Your infographic should use data different to the examples you used in the practicals where possible.

To start, create a new, blank PowerPoint presentation, then click Design and Slide Size, and Custom Slide Size, and choose A3, Landscape.

Your infographic should contain at least:

· one table,

· one simple bar chart or pie chart,

· one stacked or side-by-side bar chart or mosaic plot,

· one coded scatter plot or bubble plot to illustrated differences between responses, and associations / relationships between variables (you can use the Life Expectancy bubble plot you did in Practical 2 and describe how the UK compares to other regions)

· one boxplot to illustrate differences between groups

· two maps (one of which should be a spider map and one a regional data map)

The survey data for this task is in the file GEO1018_Survey2019.csv on the blackboard and there is a corresponding file (Key_To_Survey_Questions.docx) that explains what each column in the data represent.

Google “infographics” and refer back to the lecture slides/youtube video for examples.

Tips for improving graphics in R

You can cut and paste figures from RStudio to PowerPoint using the Export button in RStudio. To remove unwanted borders around figures use Irfanview or Microsoft Paint. These are simple graphics editing programs installed on the University computer clusters. They are free. You can also download and install Irfanview on your own computer from http://www.irfanview.com/.

To use, simply copy a figure from RStudio then paste into Irfanview or Paint. Then use the mouse to mark a rectangle around the part of the figure you want and Copy (CTRL-C), omitting any whitespace around the edges. Then paste into PowerPoint (CTRL-V).

Finally, there are a number of additional arguments we can use in the plotting functions to enhance various aspects of our plots. To modify graphs produced with ggplot, it is often useful to change the underlying theme to remove the grey background. To do this just place the following line before your ggplot command:

theme_set(theme_bw(base_size=16))

Try adjusting the value of the base_size argument to change the size of the axis labels (some trial and error may be required). Note that this command changes the theme and font sizes of all subsequent graphics produced using ggplot.

The following website has lots of additional tips and examples for modifying graphs created with ggplot: http://www.cookbook-r.com/Graphs/

The modify graphs produced with base graphics (base graphics are any plot functions that do not start with ggplot, e.g. functions to produce pie charts and mosaic plots) you can provide additional arguments to the plot function

·  las: orientation of axis labels (e.g. las=1)

·  xlab, ylab: labels for x, y axes (e.g. xlab=”Number of Students)

·  cex.axis, cex.lab: text size for axis values and labels (e.g. cex.lab=1.5 for bigger labels)

·  xlim, ylim: axis limits (e.g. xlim=c(0, 10) to scale the axis form 0 to 10. 

There is a guide in the Practicals folder on Blackboard entitled

“GEO1018_1103_Guide_to_Modifying_Graphics.pdf”. This contains more details and

examples of modifying figures for your infographic.

Assessment 1 Part 2: Statistical analysis in R

Introduction

Assessment 2 asks you to complete several statistical analyses using RStudio, based on techniques we have covered in the lectures and practical. The hand-in for the practice will consist of your annotated R output. Please follow the instructions below to complete the assessment.

Setting up your script

Create a script for the assessment. Put your student number and degree programme on the first 2 lines of your script, with the comment symbol (#) at the start of the line. Below this you should set the working directory and read in the data files you will need for this practical. Your script should look like this (but change line 4 to point to your working directory). Remember that you DO NOT type the prompts (>) shown below.

 > # 12345678

> # F800

> surv <- read.csv("GEO1018_survey2019.csv")

Task 1: t-test

Do male geography students drink more than females? Use an appropriate plot to visualize the differences and an appropriate test to test their significance.

Using the data from the first GEO1018 survey (in file GEO1018_Survey2019.csv), test the hypothesis that male geography students consume more alcohol than female students.

You should do the following:

Qn 1.Produce a boxplot of Alcohol as a function of Gender. Hint, use Units of Alcohol

Qn 2. Describe in one sentence what the boxplot shows. To answer this, start the line with a comment and question number, then add your text answer as follows:

ggplot(data=???, aes(x=???, y=???) + geom_boxplot() # Qn 2: The boxplot above shows that boys drink...

Qn 3.Use the function t.test to test the hypothesis that male students drink more than female students. Below the R command for the t.test you should answer the following, inserting your answers as comments, starting each line with the # symbol followed by question number. Your answers should consist of no more than one line per question.

Qn 4.What is the null hypothesis?

Qn 5.What is the alternative hypothesis?