Descriptive Statistics

목차

Descriptive (exploratory) Methods

• organize and summarize data
• has visual displays of data & calculation of numerical quantities (e.g., mean or median)

Variables

categorical vs. numerical

• categorical variables: few values, may or may not have numeric properties (e.g., male or female)
• numerical variables: many values, recorded as numbers (e.g., time in seconds)

---

OR,

 

qualitative vs. quantitative

• qualitative variables: no numeric/quantitative information (e.g., positive, negative)
• quantitative variables: numeric/quantitative information (e.g., temperature) 

discrete vs. continuous

• discrete variable: small number of possible values (e.g., score of 0~4, class of freshman~senior)
• continuous variable: large (or theoretically infinite) number of possible values (e.g., body temperature) 

---

independent vs. dependent

• independent variable (IV): manipulated by researcher, cause of result (e.g., study time)
• dependent variable (DV): outcome of IV (e.g., test score)

Level of Measurement

• Nominal
• Ordinal
• Interval
• Ratio

Level of Measurement	Variable	Example/notes
Nominal	categorical	cities you live in
Ordinal	categorical	ranking
Interval	numerical	temperature F/ no meaningful 0
Ratio	numerical	weight/ has meaningful 0

Sampling

• simple random sampling: every observation in the population have equal chance of being selected
• systematic sampling: random individuals are selected and then every nth individual is selected
• stratified sampling (proportional sampling): population is divided into strata (layers or groups)and random sample is taken from each stratum (subpopulation)
• cluster sampling: random cluster is selected  

Data Analysis

• Tabulating data: tallying 
• ordering (sorting) data: good for ordinal data and discrete variables
• grouping data: frequency distribution, relative frequency distribution

Measure of Center

Types of Variable (for DVs)	Measure of Center	Measure of variability
Nominal	Mode	Entropy
Ordinal	Median (=Q2)	IQR (=Q3-Q1) or range
Numeric (interval or ratio)	Mean	Variance