CHAPTER 6: DISCRETE PROBABILITY DISTRIBUTIONS
Download
Report
Transcript CHAPTER 6: DISCRETE PROBABILITY DISTRIBUTIONS
A
Characteristic is a measurable description
of an individual such as height, weight or a
count meeting a certain requirement.
A Parameter is a numerical summary of a
characteristic from an entire population such
as mean, proportion or standard deviation.
A Statistic is a numerical summary of a
characteristic from a sample taken from an
entire population such as mean, proportion
or standard deviation.
A
Point Estimate is the Statistic from
experimental data that estimates the
Parameter.
The confidence Interval is an interval that
contains the Parameter (with some level of
confidence) and is based on the Point
Estimate.
Confidence Level is the expected proportion of
intervals that will contain the Parameter.
Margin of Error is distance from the Point
Estimate to the ends of the Confidence Interval
(the possible error of the Point Estimate.
To
Find the Confidence Interval of a
Proportion, the Z-distribution can be used to
find the Critical Value IF the requirements for
using the Normal Approximation of a Binomial
are met (Section 8.2).
For a Proportion the experiment consists of a
sample size (n) and the count (x) from the
sample that meets the criteria producing the
point estimate of p ( p ) and a confidence
Level and p x . If given p , then x p * n .
n
The
Margin of Error is the
(Standard Deviation) *(Critical Value).
Z
Find the Critical Value using the Z1
2 ).
Distribution as in Chapter 7: InvNorm(
From 8.2, recall that the standard deviation
p*q
of p is given by n . So the Margin of Error
is calculated by Z * pn* q .
p*q
pZ *
So the Confidence Interval is
n .
Given x, n, and a Confidence Level, find p
Confidence Intervals. Also do for n and .
Use 1-PropZInt.
2
2
2
If
a certain Margin of Error (E) is required, a
sample size must be calculated.
If
a p is available, then the sample size
needed to get a value of E or less is
Z
2
p * q
E
If
p is not known, then the sample size
needed to get a value of E or less is
Z
2
0.25
E
If the Confidence Interval is given, the Margin of
Error and the Point Estimate can be found from
the interval maximum and minimum values:
Maximum Minimum
Margin of Error =
2
between the two values
Point Estimate =
interval.
Maximum Minimum
2
Half the distance
The middle of the
To
find the confidence Interval for a Mean,
find the Margin of Error (E) and add it to and
subtract it from the point estimate x :
x E, x E
The
Margin of Error (E) is a factor (Critical
Value) multiplied by the Standard Deviation:
E CV
. .* or E CV
. .* s
The Critical Value depends on the
Confidence Level desired.
Find
the Critical Value Z 2 from the
Confidence Level Required using the
InvNorm( 1 2 ).
Create the table of Confidence Levels and
Critical Values for various Levels.
Z *
Then the Margin of Error is 2
.
And the Confidence Interval is x Z 2 * .
Given an x , a and a required Confidence
Level find the Critical Value, the margin of
Error and the Confidence Interval.
If
the population standard deviation is unknown,
the sample standard deviation must be used.
But for the sample standard deviation we can not
use the Z-table to find the Critical Value. The
sample standard deviation is probably not exact,
so a wider different distribution is needed.
The larger the sample size the better estimate
the sample standard deviation is of the
population standard deviation.
Therefore the Critical Value will depend on the
Confidence Level AND THE SAMPLE SIZE.
The
distribution used for the Mean with
an unknown Standard Deviation is the
Student or t Distribution and will be found
in the t-table.
The columns of the table are the 2 ' s from
the Confidence Level. The rows are the
Degrees of Freedom and are one less than
the sample size n 1.
t
Find 2 for several sample sizes and
Confidence Levels.
Also use InvT( 1 2 , n-1).
Note
that the t-distribution is symmetric like
the Z distribution but wider.
Note that the t-distribution is not one
distribution but many – one for each sample
size. The smaller the sample size the wider the
distribution.
As the sample size approaches infinity the t
distribution approaches the Z distribution.
After
finding the Critical Value, the margin of
Error and Confidence Interval can now be
calculated as was done with the situation with
a known standard deviation.
Then the Margin of Error is t * s .
2
And the Confidence Interval is x t * s .
Given an x , an s and a required Confidence
Level find the Critical Value, the margin of
Error and the Confidence Interval.
Also use Tinterval.
2
Finding
the Confidence Interval for a
Standard Deviation (s) or Variance ( s 2) is
different than the others have been.
There is no Error to be calculated.
The Critical Values comes from a different
distribution that is not normal.
2
The (called Chi-squared) Distribution is
skewed right, starts at 0 on the left and goes
to
on the right.
There are two Critical Values that are both
positive.
The
columns are split into two halves. The
half on the right are for R2 - Chi-squared Right
and use the value of 2 . The half on the left
are for L2 - Chi-squared Left and use the
value of 1 2 .
The
rows are again the Degrees of Freedom
(n - 1).
To
find the Confidence Interval of a Variance
n 1 s 2 n 1 s 2
use:
,
R2
L2
To find the Confidence Interval of a Standard
Deviation use:
2
2
n
1
s
n
1
s
,
2
2
R
L
There are no calculator functions
Must be given s, n and a Confidence Level.