Z Score, Z Statistics and P value
What is a Z score or Z statistics?
Firstly Z score and Z statistics is the same. It is a measure of the number of standard deviations a data point is away from the sample mean. Mathematically this is written as:
`z = (x - bar x) / sigma`
`z` = Z score or Z statistics
`x` = data point
`bar x` = sample mean
`sigma` = sample standard deviation
The mathematical equation of standard deviation is:
`sigma = sqrt((x - bar x)/N)`
What is P value?
The P value is the probability that a particular value assuming the null hypothesis is true.
Conventionally a P value of less than 0.05 is grounds for rejecting the null hypothesis.
For the lower-tailed test or left-tailed test, the P value is expressed as:
`P = Pr(TS <=ts | H_0) = cdf(ts)`
And for the upper-tailed test or right-tailed test, the P value is expressed as:
`P=Pr(TS>=ts |H_0) = 1-cdf(ts)`
And the P value for a two-tailed test is
`P = 2 xx min{Pr(TS<=ts |H_0), Pr(TS>=ts |H_0}`
`P` = P value of an observation
`TS` = test statistics
`ts` = observed value of test statistics from your sample
`cdf()` = Cumulative distribution function for the test statistic
`Pr(condition|H_0`) = probability of the observation if the null hypothesis `H_0` is true.
`min` function to select the minimum value
How to get P value from Z score?
Calculating Z score depends on the problem analysis.
Fortunately calculating P value from Z score is easy in Excel. Assuming the Z score has been calculated z then
p_value = NORM.S.DIST(z, TRUE)
What about T score / T statistics?
Likewise calculating T score or T statistics depends on the problem analysis. But it is related the t distribution.
Then in Excel the
p_value = T.DIST(t_score, degrees_of_freedom, TRUE)
I thought I write about this before launching into writing hypothesis testing formulae with LAMBDA where you will see z scores and t scores popping up. In the meantime, you might be interested in brushing up on your statistics in your DC-DEN!
Comments
Post a Comment