Sampling and Sampling
Distribution
Population Sample
The totality of the items, things, or people A representation of the population where
P S
under consideration. one hopes to draw valid conclusions from
Entire group that is under study or about the population.
investigation A subset of the population, either by
random or nonrandom sampling techniques.
Any measurable characteristics
of a population p s Any measurable characteristics
of a sample
parameter statistics
Reasons why a researcher would rather take a sample
than study the entire population:
It saves time and money.
It may be physically impossible to get the
entire population.
Some tests are destructive in nature.
The sample results are sufficient.
Population : parameter :: Sample : statistics
Size N n
Mean x
Standard Deviation s
Proportion P p̂
IMPORTANT TERMS:
Census a 100% survey or sampling
Random Sampling
selection of n elements derived from a population N, which is the subject of the
investigation or experiment, where each sample point has an equal chance of being
selected using the appropriate sampling technique.
• a random starting point is
• a selection of a subset of a selected, and then every kth
population where each element member of the population is Systematic
Simple
has an equal chance of being selected. Random
Random selected. • k - sample intervals chosen by
• Also called as Lottery sampling. the researcher Sampling
Sampling
• a selection of a simple random sample from each
of a given number of sub-populations, or strata. • selection of clusters from the available clusters in
Stratified the population. Each member of the selected
• a selection of a simple random sample from each Cluster
Random of a given number of sub-populations, or strata. clusters is then included in the sample.
• The sample is constructed by classifying the • Also called area sampling. Sampling
Sampling population into strata, based on some • It is applied on geographical basis.
characteristics of the population such as age, • Generally done by first sampling at the higher
gender or socio-economic status. The selection of levels before going down to the lower levels.
elements is then made separately from within each
stratum, usually by random or systematic sampling
methods.
Cluster- heterogeneous sub-groups
Strata - homogeneous sub-groups
• done using a combination of different
sampling techniques.
• When selecting respondents for a
national election survey, we can
Multi-stage make use of the SRS first for regions
and cities. Then we can utilize
Sampling stratified sampling to determine the
number of respondents from the
chosen areas and clusters.
Nonrandom Sampling
used when the sample is not a proportion of the population and when there is no
system in selecting a sample.
• only those whom the
researcher meets by chance • includes a specified number of
Accidental Quota
are included in the sample persons of certain types to be
Sampling when using this technique. taken as sample. Sampling
• most convenient and fastest
sampling technique that make
Convenience • used in a very small sample Purposive
use of the telephone, mobile
Sampling size. Sampling
phones, or the internet.
Important Terminologies
Sample Statistics
merely a number assigned to the outcome of that random circumstance.
a special cases of random variables
Sampling Distribution
*the probability distribution of possible values of the statistics for repeated samples of
the sample size taken from the same population.
*used to find the probability that a sample statistics will be within a specified distance
of the unknown population parameter.
Standard Deviation of a Sampling distribution
*measures the variation amount possible values of the sample statistics over
all possible random samples.
Standard Error of a Sampling Distribution
*standard deviation of the sampling distribution.
*describes the estimated value of the standard deviation of a statistic.
Sampling Distribution of Sample Means
Suppose that we take a sample size n=2, with
replacement, from the discrete uniform population with values
0,1,2,3.
{0,1,2,3}
n=2
discrete uniform population : means equal probability p(x) = 1/4
with replacement : there are a total of 42 =16 possible samples
Characteristics of Sampling Distribution of Sample
Means
Given a population with a finite
mean µ and a finite nonzero
variance σ2, the sampling
distribution of the mean
approaches a normal distribution
with a mean of µ and a variance
of σ2/N as N, the sample size,
increases.
Standard
Mean CLT
Deviation
Proving the Characteristics of Sampling Distribution of
Sample Means by Example
Sampling Distribution of Sample Proportions
One Sample Proportion Difference of Two Sample Proportions
Sampling Distribution of Proportion
Are in favor of the death penalty , or are you opposed to it? In surveys done
in the past, women are more likely to oppose the death penalty than men. In 2008
General Social Survey, the sample proportions opposing the death penalty were
for for women and for men. Let’s suppose that, in truth,
37% women and 27% men oppose death penalty.
