Content
Summary
This lesson explains how distribution curves visually represent data, where area corresponds to observation counts and width reflects standard deviation. It distinguishes between symmetric distributions, which are balanced mirror images, and skewed distributions, where the tail direction indicates the presence of extreme values.
Intuition
Think of a distribution curve not just as a line, but as a physical boundary containing stacked dots representing your data points. The total area under this curve equals the total number of observations. The width of this shape is controlled by the standard deviation: a small standard deviation means the dots are tightly clustered, creating a narrow curve, while a large standard deviation spreads the dots out, widening the curve. You can shift the entire shape left or right by changing the mean without altering its width. Finally, check for symmetry; if the left and right sides are mirror images, the data is balanced. If one side stretches out into a long 'tail,' the data is skewed in that direction, driven by extreme outliers rather than the bulk of the data.
Derivation
- Visualize the distribution curve as a boundary containing stacked data points.
- Recognize that the area under the curve represents the total count of observations.
- Identify that the standard deviation determines the width or spread of the curve.
- Observe that changing the mean shifts the curve horizontally without changing its shape.
- Assess symmetry by checking if the curve halves are mirror images.
- Determine skewness by identifying the direction of the elongated tail.
Worked example
Compare two income distributions with the same mean. Distribution A has a long tail extending to the right (higher values), while Distribution B is symmetric. Which distribution has a positive skew, and what does this imply about the data?
- Identify the tail direction for Distribution A: it extends to the right, toward higher values.
- Recall that positive skew is defined by a tail extending to the right.
- Note that Distribution B is symmetric, meaning it has no skew.
- Conclude that Distribution A exhibits positive skew, likely due to a few extremely high incomes pulling the tail outward.
Answer Distribution A has a positive skew. This implies that while most incomes are average, a few extremely high incomes create a long tail on the right side of the distribution.
Common mistakes
- Confusing the direction of skewness with the bulk of the data; skewness is defined by the tail, not where most points cluster.
- Assuming that different means always result in different shapes; distributions can have different means but identical widths if standard deviations are equal.
- Thinking that area under the curve represents probability density only; in this context, it physically represents the count of observations.