Testing Normality: Visual Methods
You'll learn how to use three simple visual tools—histograms, Q-Q plots, and density plots—to check if your data looks normal enough for t-tests.
Why Visual Checks Matter
Before running a t-test, we need to confirm our data approximately follows a normal (bell-curve) distribution. While formal statistical tests exist, visual methods give you an intuitive, immediate sense of whether normality is reasonable. Think of it like checking if a puzzle piece fits by eye before forcing it in.
The Three Visual Tools
Histogram
A histogram shows the frequency of values in bins. For normal data, you should see a symmetric, bell-shaped mound centered around the mean. Look for severe skewness (long tail on one side) or multiple peaks—both suggest non-normality.
Q-Q Plot (Quantile-Quantile Plot)
A Q-Q plot compares your data's quantiles against a theoretical normal distribution. If your data is normal, points should fall roughly along a straight diagonal line. Curves or systematic departures indicate problems. This is the most sensitive visual test.
Density Plot
A density plot smooths out the histogram into a continuous curve. Like the histogram, you want a single, symmetric bump. It's easier to spot asymmetry or unusual tails here than in a choppy histogram.
What to Look For
- Symmetry: Data balanced around the center
- Single peak: One clear mode, not multiple humps
- Thin tails: No extreme outliers pulling away
- Q-Q linearity: Points hugging the diagonal reference line
Small deviations are usually fine—t-tests are robust to minor violations, especially with larger samples.
Key Takeaway: Histograms, Q-Q plots, and density plots let you quickly eyeball whether your data is "normal enough" for t-tests by checking for symmetry, single peaks, and alignment with theoretical patterns.