A fully labeled normal distribution curve needs six elements: a title, an x-axis with a variable name and scale, a y-axis labeled “Probability Density,” a marked center point at the mean, standard deviation markers on either side, and percentage labels showing how much data falls within each section. Getting each of these right turns a generic bell shape into a precise, informative graph.
The Two Axes
The x-axis represents the values of whatever variable you’re plotting, such as test scores, heights, or weights. You can label it with raw data values (like 60, 70, 80, 90 for exam scores) or with z-scores (standard deviations from the mean, like -2, -1, 0, 1, 2). Many textbook figures show both scales stacked together: raw scores on top, z-scores below, so a reader can translate between the two at a glance.
The y-axis is labeled “Probability Density,” not “Probability.” This distinction matters. The height of the curve at any single point does not give you the probability of that exact value. Instead, probability comes from the area under the curve between two points. For example, the area under a standard normal curve between x = 2 and x = 2.5 is about 0.09, meaning there’s roughly a 9% chance of observing a value in that range. The probability of any single exact value is always zero, even where the curve is tallest.
Marking the Mean
The mean (μ) sits at the exact center of the curve, directly under its peak. Draw a vertical dashed line from the x-axis up to the top of the curve and label it with the Greek letter μ or the actual numerical value. On a standard normal distribution, this is 0. On a real-world distribution, it’s whatever the average of your data happens to be. Every normal distribution is symmetric around this center line, so it also marks the median and the mode.
Standard Deviation Markers
Place tick marks along the x-axis at each standard deviation (σ) away from the mean: μ − 3σ, μ − 2σ, μ − σ, μ + σ, μ + 2σ, and μ + 3σ. On a standard normal curve where μ = 0 and σ = 1, these simply become -3, -2, -1, 1, 2, and 3.
The marks at μ − σ and μ + σ deserve extra attention because they correspond to the curve’s inflection points. These are the spots where the curve changes from bending inward (concave down near the peak) to bending outward (concave up in the tails). If you’re drawing the curve by hand, placing these inflection points at exactly one standard deviation from the mean on each side gives the bell its correct shape.
Percentage Labels Using the Empirical Rule
The most recognizable feature of a labeled bell curve is the set of percentage bands that show how data clusters around the mean. These come from the empirical rule, sometimes called the 68-95-99.7 rule:
- 68% of values fall within one standard deviation of the mean (between μ − σ and μ + σ).
- 95% of values fall within two standard deviations (between μ − 2σ and μ + 2σ).
- 99.7% of values fall within three standard deviations (between μ − 3σ and μ + 3σ).
To label these on the curve, draw vertical lines at each standard deviation boundary and write the percentage above the corresponding section. A common approach is to use brackets or horizontal arrows spanning each range, with the percentage centered above. You can also break the 68% region into two halves of 34% each (one on either side of the mean), the next band into two sections of about 13.5% each, and the outermost band into two sections of about 2.35% each. This level of detail helps readers see exactly how probability distributes across each slice.
Shading Regions and Tail Areas
When your curve illustrates a specific probability, shade the area under the curve that represents it. A left tail is the shaded region to the left of some cutoff value, and a right tail is the region to the right. Label the shaded area with its probability, either as a decimal (0.025) or a percentage (2.5%).
For hypothesis testing or confidence interval diagrams, you’ll often shade both tails symmetrically. In that case, label each tail with its area and mark the cutoff values on the x-axis. For instance, a 95% confidence interval leaves 2.5% in each tail, with cutoff z-scores at -1.96 and 1.96. Write the tail area inside or just above the shaded region, and note the z-score (or raw value) on the axis directly below the cutoff line.
Choosing Between Z-Scores and Raw Values
If you’re drawing a general-purpose normal curve for a statistics class, label the x-axis with z-scores. The standard normal distribution has a mean of 0 and a standard deviation of 1, so the axis reads -3 through 3. This is the universal version of the bell curve that z-tables and most statistical software reference.
If you’re plotting actual data, label the x-axis with raw values and optionally include a second row of z-scores underneath. For example, if you’re graphing adult male heights with a mean of 70 inches and a standard deviation of 3 inches, the center mark reads 70, one standard deviation out reads 67 and 73, two out reads 64 and 76, and so on. A z-score row below would show -1 and 1, -2 and 2, making it easy to convert between the two scales.
Putting It All Together
Here’s a checklist for a complete label set on any normal distribution curve:
- X-axis: Variable name, unit of measurement, and evenly spaced tick marks at each standard deviation from the mean.
- Y-axis: “Probability Density” with a numerical scale starting at 0.
- Center line: A vertical line at μ, labeled with its value.
- Standard deviation marks: Vertical lines or tick marks at μ ± 1σ, μ ± 2σ, and μ ± 3σ.
- Percentage bands: 34% + 34% in the center, 13.5% in each next band, 2.35% in each outer band, and 0.15% in each extreme tail.
- Shaded areas (if applicable): Regions of interest filled in and labeled with their probability.
- Parameters: A notation box or caption stating the values of μ and σ so the reader knows which specific normal distribution the curve represents.
The two parameters μ and σ fully define any normal distribution. Every other label on the curve is derived from them. Once you place the mean at the center and mark the standard deviations outward, the percentages, inflection points, and tail areas all fall into place automatically.