Methods of Representing Quantitative Data

Question 1

What different formats can you use when encoding data?

Answer 1

Stack data

Give nominal data categories

Question 2

What to histograms show?

Answer 2

Distribution

Question 3

What is positive skew, in terms of mode, median and mean.

Answer 3

Mode

Question 4

What is a bimodal distribution?

Answer 4

2 modes

Question 5

What are the origins of Exploratory Data Analysis?

Answer 5

1970s - John Tukey

Question 6

What is EDA about?

Answer 6

Understanding, rather than confirming results

Idea that there’s no point employing complex techniques on poorly understood data

Question 7

What are the features of EDA?

Answer 7

• Emphasizes importance of repeated examination of data
• Uncovers data structure and enhances pattern recognition
• Uses simple numerical summary (dispersion + centrality)
• Imaginative Graphical Display - clarify data variability
• Resistance (summaries not affected by anomalies)
• Robustness (not unduly effected by assumptions)
• More significance to extreme values - interesting
• More cautious approach
• Fits into mixed methods approaches

Question 8

What are the uses of EDA?

Answer 8

Helps formulate hypotheses & evaluate models
Identify major features of data
Useful when data is inaccurate
Might rule out doing confirmatory analysis

Question 9

What are the 5 steps involved in Aggregate Data Analysis?

Answer 9

1) Set context and Research Questions
2) Collect Data and Calculate Rates (turn nominal into ratio)
3) Data Processing (manipulate in minitab/excel)
4) Checking and describing data (obtain summary statistics)
5) Drawing Pictures (Boxplots/Histograms/Advanced features)