Mathwords logoMathwords

Location Parameter — Definition, Formula & Example

A location parameter is a number that sets where a probability distribution sits on the number line. Changing it slides the whole distribution left or right. The shape and spread stay exactly the same.

A parameter μ\mu is a location parameter for a family of distributions if there is a fixed density f0f_0 such that every member has density f(x;μ)=f0(x−μ)f(x;\mu) = f_0(x - \mu). Equivalently, X=μ+ZX = \mu + Z, where ZZ has distribution f0f_0 and does not depend on μ\mu. The set of all such distributions is called a location family.

Key Formula

f(x;μ)=f0(x−μ)f(x;\mu) = f_0(x - \mu)
Where:
  • f(x;μ)f(x;\mu) = Density of the shifted distribution, evaluated at the value x
  • f0f_0 = The fixed base density, which is the family member whose location is zero
  • μ\mu = The location parameter, which is the distance the base density is shifted to the right
  • xx = A possible value of the random variable

Worked Example

Problem: A machine fills bottles with an amount that is normally distributed with mean 16 ounces and standard deviation 0.5 ounces. A technician recalibrates the machine so the mean becomes 16.5 ounces. The spread does not change. Find the probability that a bottle holds more than 17 ounces, before and after the change.
Step 1: The mean is the location parameter here. Shifting it moves the curve 0.5 ounces to the right, and the standard deviation stays 0.5. So you can standardize both cases with the same spread.
z=x−μσz = \frac{x - \mu}{\sigma}
Step 2: Before the change, use a mean of 16. Then read the upper tail from a standard normal table.
z=17−160.5=2,P(X>17)=P(Z>2)≈0.0228z = \frac{17 - 16}{0.5} = 2, \qquad P(X > 17) = P(Z > 2) \approx 0.0228
Step 3: After the change, use a mean of 16.5. The cutoff of 17 ounces is now closer to the center of the curve.
z=17−16.50.5=1,P(X>17)=P(Z>1)≈0.1587z = \frac{17 - 16.5}{0.5} = 1, \qquad P(X > 17) = P(Z > 1) \approx 0.1587
Answer: The probability rises from about 0.0228 to about 0.1587. Only the location changed. The curve's shape and its 0.5-ounce spread are identical in both cases.

Why It Matters

Standardizing with a z-score subtracts the location parameter, and that is why one standard normal table works for every normal curve. In AP Statistics inference, a hypothesis test about the mean of a normal population asks whether its location parameter equals a claimed value.

Common Mistakes

Mistake: Assuming the location parameter is always the mean.
Correction: It is the mean for the normal distribution, but not in general. For a shifted exponential, X = μ + Y with Y exponential with rate λ, the location μ is the smallest possible value. The mean is μ + 1/λ.

Related Terms