Visual Assessment: Histograms and Density Plots
What you'll learn: How to use histograms and density plots to quickly check if your data looks normally distributed.
Before running formal statistical tests for normality, it's smart to simply look at your data. Visual inspection is fast, intuitive, and often reveals obvious departures from normality that matter in practice.
The Two Main Tools
Histograms divide your data into bins and show the frequency of observations in each bin as bars. If data is normally distributed, the histogram should resemble a symmetric, bell-shaped curve centered around the mean.
Kernel Density Plots (or density curves) smooth out the histogram into a continuous curve. Think of it as drawing a smooth line over the bumpy histogram bars. This makes it easier to see the overall shape without being distracted by bin choices.
What to Look For
When inspecting these plots, ask yourself:
- Is it symmetric? Normal distributions are mirror images around the center.
- Is it bell-shaped? There should be one central peak with frequencies tapering off smoothly on both sides.
- Are there outliers or multiple peaks? These suggest the data isn't normal.
- Are the tails too heavy or too thin? Normal distributions have specific tail behavior (recall the 68-95-99.7 rule).
Why This Matters
Visual assessment is your first line of defense. If a histogram shows severe skewness, multiple peaks, or strange gaps, you know immediately that normality assumptions don't hold—no calculation needed. This guides whether you can use methods that assume normality (like certain hypothesis tests) or need alternatives.
Key Takeaway: Histograms and density plots let you visually check if data appears normally distributed by revealing its shape, symmetry, and any obvious departures from the bell curve.