View
216
Download
0
Embed Size (px)
Citation preview
Ekstrom Math 115b
Mathematics for Business Decisions, part II
Variance
Math 115b
Variance gives a measure of the dispersion of data
Larger variance means the data is more spread out
Several formulas for different types of variables
Variance
Finite R.V.
Easiest if a table is set up
Variance: Finite R.V.
x
XX xfxXV all
2
x
Given
Given Compute Compute Compute
xf X Xx 2Xx xfx XX 2
Variance: Finite R.V.
Ex. Find the variance for the following finite R.V., X
x fX (x)
1 0.30
3 0.15
5 0.10
7 0.15
9 0.30
First find the mean
Variance: Finite R.V.
x fX (x)
1 0.30
3 0.15
5 0.10
7 0.15
9 0.30
5
30.0915.0710.0515.0330.01
X
Calculate x-value minus the mean
Variance: Finite R.V.
-4
-2
0
2
4
x
1 0.30
3 0.15
5 0.10
7 0.15
9 0.30
xf X Xx
Calculate the square of x-value minus the mean
Variance: Finite R.V.
16
4
0
4
16
x
1 0.30 -4
3 0.15 -2
5 0.10 0
7 0.15 2
9 0.30 4
xf X 2
Xx Xx
Multiply the squared values by their respective p.d.f. values
Variance: Finite R.V.
4.80
0.60
0.00
0.60
4.80
x
1 0.30 -4 16
3 0.15 -2 4
5 0.10 0 0
7 0.15 2 4
9 0.30 4 16
xfx XX 2 2Xx Xx xf X
Add the final column to find the variance
Variance: Finite R.V.x
1 0.30 -4 16 4.80
3 0.15 -2 4 0.60
5 0.10 0 0 0.00
7 0.15 2 4 0.60
9 0.30 4 16 4.80
2Xx Xx xfx XX 2 xf X
Variance = 10.8
Ex. Find the variance for the following finite R.V., X
Variance: Finite R.V.
x
1 0.10
3 0.25
5 0.30
7 0.25
9 0.10
xf X
First find the mean
Variance: Finite R.V.
x fX (x)
1 0.10
3 0.25
5 0.30
7 0.25
9 0.10
5
10.0925.0730.0525.0310.01
X
Calculate x-value minus the mean
Variance: Finite R.V.
-4
-2
0
2
4
x
1 0.10
3 0.25
5 0.30
7 0.25
9 0.10
xf X Xx
Calculate the square of x-value minus the mean
Variance: Finite R.V.
16
4
0
4
16
x
1 0.10 -4
3 0.25 -2
5 0.30 0
7 0.25 2
9 0.10 4
xf X 2
Xx Xx
Multiply the squared values by their respective p.d.f. values
Variance: Finite R.V.
1.60
1.00
0.00
1.00
1.60
x
1 0.10 -4 16
3 0.25 -2 4
5 0.30 0 0
7 0.25 2 4
9 0.10 4 16
xfx XX 2 2Xx Xx xf X
Add the final column to find the variance
Variance: Finite R.V.x
1 0.10 -4 16 1.60
3 0.25 -2 4 1.00
5 0.30 0 0 0.00
7 0.25 2 4 1.00
9 0.10 4 16 1.60
2Xx Xx xfx XX 2 xf X
Variance = 5.2
Variance is used to find standard deviation
Standard deviation is ALWAYS the square root of variance
Standard deviation is represented by sigma
or
Standard deviation is the “typical amount” of variation from the mean (approx. 2/3 of all data lies within 1 standard deviation of mean)
Variance
XVX 2XXV
Shortcut for binomial R.V.
Variance: Binomial R.V.
ppn
XVX
1
ppnXV 1
Ex. Suppose X represents to total number of students that pass a particular class in a given semester. If there are 34 students in the class and historically 83% of the students pass, find the standard deviation of X.
Soln:
Variance: Binomial R.V.
1903.2
7974.4
83.0183.034
1
ppnX
A solution could also be attempted using the BINOMDIST function
Recall binomial R.V.’s are finite R.V.’s
Complete table as done in previous examples
Variance: Binomial R.V.
Similar formula for continuous R.V.
Value is found using Integrating.xlsm
Recall,
Variance: Continuous R.V.
x
XX dxxfxXV all
2
XVX
Ex. Suppose is a p.d.f. on the interval [0, 1]. Find the mean of X, the variance of X, and the standard deviation of X.
Soln:
Variance: Continuous R.V. 5.033 2 xxxf
5.0
5.0331
0
2
1
0
dxxxx
dxxfx XX
Soln:
Variance: Continuous R.V.
2582.0
0667.0
XVX
0667.0
5.0335.01
0
22
1
0
2
dxxxx
dxxfxXV XX
Ex. Let T be an exponential random variable with parameter . Find and .
Soln:
Variance: Continuous R.V.
8 TV T
64
80
8/812
0
2
dtet
dttftTV
t
TT
8
64
TVT
Note that for an exponential random variable, the variance is equal to and the standard deviation is equal to .
Ex. Let W be a uniform random variable on the interval [0, 30]. Find and .
Variance: Continuous R.V.
WV W
2
Estrom Math 115b
Soln.:
Variance: Continuous R.V.
75
1530
0 3012
30
0
2
dww
dwwfwWV WW
6603.8
75
WVW
Note that for a uniform random variable, the variance is equal to .
Variance: Continuous R.V.
12
2ab
Different types of variables can have similar parameters (mean & std. dev.)
We can transform the variables
New variable S defined as
Variance: Standardization
X
XXS
We say S is the standardization of X.
The mean of S will be 0
The standard deviation of S will be 1.
Variance: Standardization
Determining probabilities using standardized values
Ex. Suppose X is an exponential random variable with parameter . Determine where S is the standardization of X.
Soln: Recall
Variance: Standardization
5 2SP
X
XXS
So,
Then,
15
105
25
5
22
XP
XP
XP
XPSP
X
X
9502.0
1515
0
5/51
dxeXP x
Variance: Standardization
Formula:
A sample is a collection of data from some random variable (finite or continuous)
Variance: Sample
n
ii xx
ns
1
22
1
1
Standard deviation of a sample is found by taking the square root of variance
Formula:
Ex. Find the mean, variance, and standard deviation of the sample 14, 16, 17, 21, 22.
Variance: Sample
n
ii xx
nss
1
22
1
1
Soln:
Mean:
AVERAGE function in Excel
Variance: Sample
185
2221171614
x
Variance:
VAR function in Excel
Variance: Sample
5.11
16914164
1
1822182118171816181415
1
1
1
22222
1
22
n
ii xx
ns
Standard Deviation:
STDEV function in Excel
Variance: Sample
3912.3
5.11
2
ss
Why are sample values important?
Sometimes unreasonable/impossible to achieve all values for a random variable
We can assume and
Samples help predict values for the random variable
Variance: Sample
Xx XVs 2
Formula:
Compares a group of sample means
Variance: Sample Mean
n
XVxV
nX
x
Ex. Find the variance and standard deviation of the sample mean for the following data set:
14, 16, 17, 21, 22
Soln:
Variance: Sample Mean
3.25
5.11
n
XVxV
5166.15
5.11
nX
x
How do you interpret ?
If there were samples of size 5, the sample mean would be about 18 (from previous calculation) and, on average, the sample mean would be within 1.5166 units about 2/3 of the time.
Variance: Sample Mean5166.1x
What to do:
1. Calculate standard deviation of errors (use STDEV function)
2. My standard deviation is about 13.53
3. We assume mean is 0 even though we calculated a value that was different
Variance: Project