In: Statistics and Probability
1. This question takes you through the design-based approach to
sorting out sampling
distributions, their means and variances, and the estimation of
these quantities when
you sample with and without replacement.
Consider a population of 5 individuals. The variable is their
annual income in 1000s
of dollars per year, and you want to estimate the population mean
income based on
a random sample of size two. The values of income are as
follows:
Individual Income
A 80
B 18
C 24
D 52
E 24
(a) Make a list of all the samples of size 2 possible without
replacement. Find the
sample mean and sample variance of each of these samples using the
formula
yS = 1/n sigma yi
s^2 = 1/(n-1) sigma(yi - y(average))^2
(b) Find the sampling distributions of the sample mean and
sample standard devi-
ation based on (a).
(c) Find the expected values of the sampling distributions of by
using
E[W] = sigma wP[W = w]
(d) Find the variance of the sampling distribution of by
using
V [W] =sigma (w - E[W])2P[W = w]
(e) Compare the results from (c) and (d) with the population mean,
the population
variance as calculated from
S^2 = 1/(N-1) sigma (yi - yU)^2
and the variance of the sampling distribution of as calculated
from
V [y] = S2/n(1 -n/N)
(f) For sampling with replacement, nd the sampling distribution of
the sample
mean by listing all the samples that can occur, nding the sample
mean values
for each sample, and nding the probability that each value occurs.
Be careful:
in one kind of sampling, the order in which the individuals are
sampled can't
be ignored.