Finding the Value from a Normal Distribution Corresponding to a Given Proportion
Section 7.3 Sampling Distributions and the Central Limit Theorem
The probability historgram for the sampling distributon of `bar x`
The Central Limit Theorem
If we start with a population with mean `mu` and standard deviation `sigma`,
and we draw random samples each of size `n` where `n > 30` and we compute the sample mean, `bar x`,
for each of sample, then the distribution of `bar x` will be approximately normal with the mean of all the `bar x`'s approaching `mu`
(i.e., `mu_(bar x) = mu`) and the standard deviation of the `bar x`'s will
approach `sigma/sqrt(n)` (i.e., `sigma_(bar x) = sigma/sqrt(n)`).
The Central Limit Theorem for Proportions:
If we start with a population that has a characteristic that appears
in `p` proportion, and we draw random samples of size `n` where we know that `n*p >= 10` and `n*(1-p) >= 10`,
and we look at compute `hat p` for each sample, then the distribution of `hat p`
will be approximately normal, with the mean of the `hat p`'s equal to `p` (i.e., `mu_(hat p) = p`),
and the standard deviation of the `hat p`'s approximately equal to `sqrt( ((p*(1-p))/n )` (i.e.,
`sigma_(hat p) =sqrt( ((p*(1-p))/n )` )
Section 7.5 Normal Approximation to the Binomial Distribution
Review:
The mean of a binomial random variable `X` with `n` trials and
probability of success equal to `p` is given as `mu_X = n*p`
The variance of a binomial random variable `X` with `n` trials and
probability of success equal to `p` is given as `sigma_X^2 = n*p*(1-p)`
The standard deviation of a binomial random variable `X` with `n` trials and
probability of success equal to `p` is given as `sigma_X = sqrt(sigma_X^2) = sqrt( n*p*(1-p) )`
The Normal Approximation to the Binomial:
If we start with a binomial distribution that has a success probability of `p`,
and we look at a random variable `X` based on `n` trials,
then, in those cases where `n*p >= 10` and `n*(1-p) >= 10`, we will find that `X`
is approximately normal with mean `mu_X = n*p` and standard devaition `sigma_X = sqrt( (n*p*(1-p) )`.
The continuity correction
Before looking at any explanation of the differences between
proportions and the binomial distribution, we want to
look at the changes in mean and standard deviation
related to transformations in values.
Section 7.6 Assessing Normality
What does it take to say that something is not approximately normally distributed.
The existence of one or more outliers.
Significant (whatever that means) skewness (we want the mean and median to be close)
More than one mode (but that depends on how hard we
look and how far apart the two supposed modes are)