Tables
Use graph to represent data
| Graph Type | Variables | Purpose | Pros | Cons |
|---|---|---|---|---|
| Bar Graph | 1 categorical | Display frequency/counts of categories | Simple, intuitive for comparing groups | Limited to counts; hides distribution details |
| Histogram | 1 quantitative | Show distribution shape, center, spread, outliers | Reveals patterns (skew, modes) | Bin size choice affects interpretation |
| Boxplot | 1 quantitative | Summarize center (median), spread (IQR), and outliers | Robust to outliers; compact visualization | Hides multimodality; less detail than histograms |
| Stacked/Dodged Bar | 2 categorical | Compare proportions/counts across categories | Shows subgroup breakdowns | Stacked: Hard to compare subgroups; Dodged: Space-intensive |
| Scatterplot | 2 quantitative | Explore relationships (form, direction, strength) | Reveals trends, clusters, outliers | Overplotting with large datasets |
| Side-by-Side Boxplots | 1 quantitative + 1 categorical | Compare distributions across categories | Highlights differences in medians/IQR | Simplifies distribution details |
| Faceted/Colored Plot | 3+ variables (mix of types) | Multivariate analysis (e.g., relationships across subgroups) | Incorporates multiple variables in one plot | Complexity; may become cluttered |
Conditional Statistics
A conditional statistic is a statistic derived from one or more variables for all observations sharing a value of another variable