6.4 Normal Distribution (Pinkie Length)

Normal Distribution and Pinkie Length

Normal distributions are one of the most common patterns in statistics. They describe how data naturally spreads out around an average value. In this section, pinkie finger length serves as a concrete example for learning how to calculate probabilities, interpret z-scores, and apply the empirical rule to any normally distributed data.

Probabilities in Normal Distributions

A normal distribution is a continuous probability distribution that forms a symmetric, bell-shaped curve. Two numbers completely define it:

• Mean ($\mu$): the center of the distribution
• Standard deviation ($\sigma$): how spread out the data is from the center

The total area under the curve equals 1 (or 100%), representing all possible outcomes. Any slice of that area represents the probability of a value falling in that range.

To find the probability of a specific pinkie length (or range of lengths), you convert the raw measurement into a z-score and then look up the corresponding area:

1. Identify the mean ($\mu$) and standard deviation ($\sigma$) of the pinkie length distribution.
2. Convert the pinkie length to a z-score using:

$z = \frac{x - \mu}{\sigma}$

where $x$ is the pinkie length you're interested in.

1. Look up the z-score in a standard normal table (or use a calculator) to find the area to the left of that z-score. This area is the probability that a randomly selected pinkie is shorter than $x$.

Once you have that left-side area, you can answer different types of questions:

• "Less than" a value: Use the area to the left of the z-score directly.
• "Greater than" a value: Subtract the left-side area from 1.
• "Between" two values: Find the left-side area for each z-score, then subtract the smaller from the larger.

For example, if the mean pinkie length is 6.5 cm with a standard deviation of 0.5 cm, and you want the probability of a pinkie shorter than 7.0 cm: $z = \frac{7.0 - 6.5}{0.5} = 1.0$. A z-score of 1.0 corresponds to about 0.8413, so roughly 84.13% of pinkies are shorter than 7.0 cm.

Interpretation of Z-Scores

A z-score tells you how many standard deviations a particular value sits from the mean.

• A positive z-score means the pinkie length is above the mean (longer than average).
• A negative z-score means it's below the mean (shorter than average).
• A z-score of 0 means the pinkie length equals the mean exactly.

The size of the z-score tells you how unusual a measurement is. A z-score of 1 means the length is 1 standard deviation above average, while a z-score of $-2$ means it's 2 standard deviations below average. The farther a z-score is from 0 in either direction, the more unusual that measurement is within the distribution.

Z-scores also connect directly to percentiles. A percentile tells you what percentage of values fall below a given data point. If a pinkie length has a z-score of 1.0, the standard normal table gives about 0.8413, meaning that pinkie is longer than roughly 84% of all pinkies in the distribution. Comparing two z-scores is a quick way to see which measurement is more extreme relative to the group.

Empirical Rule for Data Ranges

The empirical rule (also called the 68-95-99.7 rule) gives you a fast way to estimate how data spreads in a normal distribution without looking anything up:

• About 68% of data falls within 1 standard deviation of the mean ($\mu \pm 1\sigma$)
• About 95% of data falls within 2 standard deviations ($\mu \pm 2\sigma$)
• About 99.7% of data falls within 3 standard deviations ($\mu \pm 3\sigma$)

To apply this to pinkie lengths:

1. Determine how many standard deviations your range covers from the mean.
2. Match that range to the empirical rule percentage.

Using the earlier example ($\mu = 6.5$ cm, $\sigma = 0.5$ cm): the range 6.0 cm to 7.0 cm is $\mu \pm 1\sigma$, so about 68% of pinkie lengths fall in that range. The range 5.5 cm to 7.5 cm is $\mu \pm 2\sigma$, capturing about 95%.

This rule is great for quick estimates and sanity checks. If you calculate a probability and it contradicts the empirical rule, that's a sign to double-check your work.

Distribution Characteristics

A few additional properties help describe how a distribution compares to a perfect normal curve:

• Skewness measures asymmetry. A perfectly normal distribution has a skewness of zero. Positive skew means the right tail is longer; negative skew means the left tail is longer.
• Kurtosis describes the shape of the tails and peak. Higher kurtosis means heavier tails (more extreme values); lower kurtosis means lighter tails.
• The central limit theorem states that when you take repeated samples from any population and calculate their means, the distribution of those sample means approaches a normal distribution as sample size increases. This is why the normal distribution shows up so often in statistics, even when the original data isn't perfectly bell-shaped.