Organising, Visualising and Describing Data

Question 1

3 Classes of Data Types

Answer 1

1. Numerical (quantative) vs Categorical (qualitative)
2. Time series vs X-Sectional
3. Structured vs Unstructured

Question 2

What are the 2 types of Categorical Data

Answer 2

Nominal - no logical order
Ordinal - logical order

Question 3

Can you perform mathematical operations on categorical data?

Answer 3

No

Question 4

What is the difference between Time Series and x-sectional data?

Answer 4

Time series is a set of many observations.
X-Sectional- one specific point in time, a set of comparable observations are made.

Question 5

What is Panel Data?

Answer 5

Combine x-sectional and time.

Question 6

Unstructured data can be classified according to how the data is generated. Give an example

Answer 6

Individuals (social media post)
Processes (withdrawal)
Sensors (Camera)

Question 7

Define the following:
1.Absolute, Relative and Cumulative Frequency

Answer 7

nominal, % and adds up to 100
Absolute- histogram

Question 8

What is a joint frequency and what is a marginal frequency ?

Answer 8

Joint - data cell of a contingency table (two-dimensional array = a normal table with columns and rows). Basically when the 2 variables (row and column label) occur simultaneously.
Marginal - Total frequency for a row or column.

Question 9

What is a contingency table and what is a confusion matrix?

Answer 9

Table to analyse 2 variables
A confusion matrix is an example of a contingency table. One variable is predicted…. and the other variable is actual… . so shows actual vs predicted

Question 10

Benefits of a : Histogram, Frequency Polygon, Cumulative frequency distro chart, Bar Chart, grouped bar chart or clustered bar chart, a stacked bar chart

Answer 10

1. Quickly see where the concentration lies
   2.Joins the midpoints of the histogram intervals
2. can be either relative or absolute
   4.Illustrate RELATIVE sizes/degrees/magnitudes.
3. can illustrate 2 categories at once (adds another variable)
4. shows both the cumulative and joint frequency in the same bar

Question 11

Benefits and features of: A Tree Map; word cloud; line charts; bubble line chart; scatter plot + scatter plot matrix (3 variables); Heat Map

Answer 11

1. Visualise relative size of categories
2. Visualise text - categorical data
3. illustrate time series data. Can plot multiple lines if scale is comparable
4. adds another dimension to a line chart, each point has a bubble that is in proportion to its variable
5. shows the relation between 2 variables and the strength of it.
6. Is drawn off of a contingency table and uses colour to visualise the concentration of data.

Question 12

Place all charts you can think of into the following 3 categories: Relationships, Comparisons and Distributions. Can be more than one

Answer 12

Relationships: Scatter/scatter plot matrix, heat maps
Comparisons: Bar chart, tree maps, heat maps, dual line charts, bubble line charts
Distributions: Histogram, frequency polygon, cumu distro charts, bar charts, tree maps, heat maps for categorical data, word clouds for text data

Question 13

Formula for u(population mean) and X(sample mean) and What is an Arithmetic Mean and its 4 properties?

Answer 13

u=sum of all observations/no. of obs
X= same but for sample
An arithmetic mean = sum of observations/no. of obs
1.All interval and ratio data sets have an arithmetic mean
2.all data values are considered and included in the arithmetic mean
3. a data set has only one arithmetic mean
4. the sum of the deviations of each data point in the set will sum to 0. so: sum of data points (Xi-X)=0

Question 14

2 techniques to deal with the pitfalls of the arithmetic mean

Answer 14

Trimmed Mean- a 1% trimmed mean would exclude the top and bottom 1/2%
Windsorized Mean - substitue out data rather than exclude it (doesn’t sound very good)

Question 15

How do you calculate Weighted Mean

Answer 15

Xw=(w1X1+w2X2+….+wnXn)

e.g. A portfolio consists of 50% common stocks, 40% bonds, and 10% cash. If the return on
common stocks is 12%, the return on bonds is 7%, and the return on cash is 3%, what is the
portfolio return?

Question 16

When is the geometric mean used? give a general formula for the geometric mean for calculating the returns of a data set

Answer 16

To calc investment returns over multiple periods or when mearing compounding growth rates.
1+G=^n√(x1+1).(x2+1) …. (xn+1)

Question 17

What is a harmonic mean and when is it used

Answer 17

N/ (sumof) 1/Xi, where there are n values of Xi.
Good for calculating average cost of shares
eg An investor purchases $1,000 of mutual fund shares each month, and over the last three
months, the prices paid per share were $8, $9, and $10. What is the average cost per share?

Question 18

Harmonic<Geometric<Arithmetic (in terms of the result). This fact has resulted in investors claiming benefit of … ?

Answer 18

Cost Averaging: buying shares on a regular basis