Topic 2: Data & Graphical Summaries

Question 1

When you first look at the data to get the general snapshot without actually answering the research question, what process is it?

Answer 1

Initial Data Analysis (IDA)

Question 2

During the first snapshot of the data, what criteria are used to analyse it?

Answer 2

1. Data background (quality + integrity)
2. Data structure (characteristics of data including types, size, number)
3. Data wrangling (making changes like scraping, cleaning, tidying, reshaping, splitting, combining)
4. Data summaries (graphical & numerical)

Question 3

Why is having the first look at the data critical?

Answer 3

IDA helps to capture the main qualities of the data as well as suggesting about the population.
Also, helps to analyse whether the data can answer the research question and pose follow-up ones

Question 4

Which two aspects need to be considered when analyzing the data?

Answer 4

Size: how many variables/subjects (p/n)
Type: quantitative or qualitative

Question 5

How can size of data be described?

Answer 5

Multivariate (2+ variables)
+ Bivariate (2 variables)
+ Univariate (1 variable)

Question 6

Explain different types of data

Answer 6

1. Qualitative (category):
   + Ordinal (in order): binary/3+
   + Norminal: binary/3+
2. Quantitative (measurement):
   + Discrete (separated)
   + Continous (continuum)

Question 7

Explain data and variable

Answer 7

Data is information about sets of subjects being studied.
Variables are different measurement or categories describing attributes of the subjects.

Question 8

What type of graphical summary can be used for 1/2 qualitative variables?

Answer 8

1 qual variable: simple bar plot (table before plot)
2 qual variable: double bar plot (stacked or side-by-side)

Question 9

What type of graphical summary can be used for 1/2 quantitative variables?

Answer 9

1 quan variable: histogram/boxplot
2 quan variable: scatter plot

Question 10

What type of graphical summary can be used for 1 quan and 1 qual variables?

Answer 10

Comparative box plot

Question 11

Function of histogram

Answer 11

Present the overall distribution of the dataset
Hightlight the data percentage in 1 interval and compare to others

Question 12

How is block height in histogram calculated?
What is it called?

Answer 12

Density scale
Block height = % in the block/interval length

Question 13

What is end point convention? Give an example

Answer 13

To decide the inclusion of points that fall on the border
i.e: (20,45] 20 is not included, but 45 is

Question 14

What information is conveyed from a box plot?

Answer 14

Box plot compares different data sets by presenting “anchor” points including median, middle 50%, outliers

Question 15

In what case a comparative box plot can be used?

Answer 15

1 quan and 1 qual variable
The quan variable is split up by a qualitative variable

Question 16

Describe good, poor, and bad data viz

Answer 16

Good data viz: interesting story + visually appealing

Poor data viz: boring, distracting presentation (chartjunk)

Bad data viz: misleading story

Question 17

What is ggplot2 and why do we use it?

Answer 17

An intuitive way to produce products
Why: allow us to build different building blocks of the data, then combine to product a complex viz with layers

Question 18

What are the parameters of ggplot?

Answer 18

Aesthetic (aes): what we want to see in the viz (presented variable)

Geometric objects (geom_xxx): the type of graphical summary (actual marks)

Facet: subset of the data