Anda di halaman 1dari 3

Understanding Z-Scores and Relevant Excel Functions

Individual values taken from a distribution with known (population) mean x p and standard deviation s p can be standardized by conversion into Z-Scores. Conversion is done by taking the difference between each individual value and the mean, and dividing the resulting difference by the standard deviation.

zi =

xi - x p

sp

This process converts the original unit of measurement (seconds, grams, meters, milligrams per milliliter, etc.) into units of standard deviations from the mean. For example, given measurements of a substance in blood with population mean of x p = 5000 milligrams per milliliter, and population standard deviation of s p = 400 milligrams per milliliter, the following set of measurements: {5500, 4800, 3950, 5100, 5200, 4950} Can be expressed as Z-scores: {1.25, -.5, -2.65, .25, .5, -.125} In Excel, the function =Standardize(x,mean,standard_deviation) returns the Zscore for x. Converting Distributions into Standard Distributions If every value f(x) of a well-behaved1 random variable with distribution mean m and x -u standard deviation s is converted into a z-score, so that z(x) = f ( ) , the resulting standardized distribution has mean 0 and standard deviation 1.

Well-behaved distributions have a defined mean and finite standard deviation and moments.
1

When any Gaussian (normal) distribution is standardized, the result is a Standard Normal Distribution, with mean = 0 and standard deviation = 1, written f (0,1) . Converting Distributions into Probability Mass Functions To convert the distribution of any well-behaved random variable into a probability mass function, multiply each value of y by a normalizing factor k, so that the resulting integral sums to 1.

k f (x)dx = 1.
-

For example, the Standard Normal probability mass function has the form

1 -2 f (x) = e . 2p
Z-scores are values of x on a Standard Normal probability mass function. Converting Probability Mass Functions to Cumulative Distribution Functions The cumulative distribution function (CDF) of a probability mass function f(x) is a function that returns the area of the integral to the left of x, known as the cumulative distribution. It is equal to the probability that a value selected at random is less than or equal to x. CDF(x) =
-

x2

f (x) .

Useful Excel Functions for Normal Distributions There are four Excel functions related to the normal distribution (1) NORM.DIST (2) NORM.INV (3) NORM.S.DIST (4) NORM.S.INV (2008 and earlier versions NORMDIST) (2008 and earlier versions NORMINV) (2008 and earlier versions NORMINV) (2008 and earlier versions NORMINV)

(1) NORM.DIST(x,mean,standard_deviation, cumulative) (a) When cumulative is set to False NORM.DIST returns f (x) (b) When cumulative is set to True, NORM.DIST returns the cumulative distribution function of x,
-

f (x) .
2

For example (to three decimal places), =NORM.DIST(10,4,6,FALSE) returns 0.040 =NORM.DIST(10, 4,6,TRUE) returns 0.841 (2) NORM.INV(probability, mean, standard_dev) NORM.INV inputs a cumulative probability [0,1], and returns, for a given mean and variance, the associated value of x. For example, =NORM.INV(0.75,4,6) returns 8.047 (3) NORM.S.DIST(z) NORM.S.DIST(z) takes as input a z-score and returns its associated cumulative probability. For example, =NORM.S.DIST(-2) returns 0.023 =NORM.S.DIST(2) returns 0.977 (4) NORM.S.INV(probability) takes as input a cumulative probability value [0,1] and returns its associated z-score. For example, =NORM.S.INV(0.3) returns -0.524 =NORM.S.INV(0.7) returns 0.524 END

Anda mungkin juga menyukai