CFA Level One: Quantitative Methods - Organizing, Visualizing, and Describing Data

Overview

• This is a new reading introduced in 2020 by two PhD researchers.
• Part of the CFA program to help with data collection, management, and usage for decision-making.

Learning Objectives

• Interpret, describe, and identify data types.
• Prepare for data collection, management, and usage in a decision-making framework.

Types of Data

1. Numerical Data

• Continuous Data: Measured on an infinite scale (e.g., temperature, golf scores).
• Discrete Data: Can be counted with a finite number of values (e.g., days of the week, touchdown passes).

2. Categorical Data

• Nominal Data: Represents qualitative outcomes without inherent numerical value (e.g., bond ratings).
• Ordinal Data: Rank ordered data according to some characteristic (e.g., parenting skills rated by children).

Data Analysis

• Cross-sectional Data: Observations at a single point in time (e.g., stock returns for a specific year).
• Time Series Data: Observations over time (e.g., monthly price changes of the Pink Panther diamond).
• Panel Data: Combines cross-sectional and time series data.
• Variables: Characteristics or quantities that can be measured, counted, or categorized.

Data Structuring

• Structured Data: Organized in a predefined manner, often visualized in spreadsheets.
• Unstructured Data: Not organized, includes data from social media, credit card transactions, etc.

Analyzing Data

• Frequency Distribution: Simplifies analysis by grouping data into intervals.
• Contingency Table: Represents categorical data and shows joint and marginal frequencies.
• Confusion Matrix: Compares actual vs. predicted outcomes, identifies type 1 and type 2 errors.

Data Visualization

• Histogram: Graphical representation of data distribution.
• Frequency Polygon: Line graph connecting the midpoints of the class intervals.
• Bar Chart: Displays data using rectangular bars (vertical or horizontal).
• Pareto Chart: Bar chart ordered by frequency in descending order.
• Tree Map: Shows hierarchical data using nested rectangles.
• Word Cloud: Visualizes the frequency of words in different sizes.
• Heat Map: Uses color coding to display data values.
• Scatter Plot: Shows relationship between two variables, important for understanding correlation.

Choosing Visualization Types

• Consider which type best represents the data and the analysis intent.
• Different types include histograms, bar charts, line charts, scatter plots, etc.

Conclusion

• Mastering the identification and comparison of data types helps simplify further analysis steps.
• Visualization is key in making data comprehensible and actionable.