Download 1 Moment Statistics - University of Houston

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Probability wikipedia , lookup

Foundations of statistics wikipedia , lookup

Statistics wikipedia , lookup

History of statistics wikipedia , lookup

Transcript
Notes - Econ 2370 - Statistics and Probability
1 Moment Statistics
Advantages
Mathematically interrelated and related to other moments.
All have the same assumptions.
They provide the only measures of skewness and kurtosis.
They provide suÆcient information to reconstruct a frequency distribution function.
Assumption Interval or ratio data, univariate system
General Moment Equation
N
X
1
k
k =
i
N
i=1
= any variable having its origin at some point .
= real or arbitrary origin (operationally dened)
k = the kth moment about .
Moments about the origin = 0, Xi = raw score
First Moment (M1 ) - Mean = X
N
N
X
X
1
1
M1 =
Xi =
(freq X )
N
N
i=1
Second Moment (M2 )
M2 =
Third Moment (M3 )
1
N
XX
N
i=1
i
XN
= N1 (freq X 2 )
i=1
i=1
N
N
X
X
1
1
4
M4 =
Xi =
(
freq X 4 )
N
N
i=1
Spring, 2000
2
N
N
X
X
1
1
3
M3 =
Xi =
(
freq X 3 )
N
N
i=1
Fourth Moment (M4 )
i=1
1
i=1
Notes - Econ 2370 - Statistics and Probability
(x1 = Xi
Moments about the mean = X,
First Moment (m1 )
X ) = raw score
N
X
1
m1 =
xi = 0
N
i=1
Second Moment (m2 ) - Variance
N
X
1
x2i = M2 M12
m2 =
N
Standard Deviation (s) = pm2 .
Third Moment (m3 )
i=1
N
X
1
m3 =
x3i = M3 3M1 M2 + 2M13
N
i=1
For a symmetrical distribution, m3 =0. This function(m3) is related to skewness,
but it is inuenced by the size of the unit of measure.
Fourth Moment (m4 )
N
X
1
m4 =
x4i = M4 4M1 M3 + 6M12 M2 3M14
N
i=1
m4 is directly related to kurtosis, but it inuenced by the size of the metric unit.
Spring, 2000
2
Notes - Econ 2370 - Statistics and Probability
(z1 = xsi ) = standardized score
Standardized Moments = X,
First Moment (a1 )
N
X
1
a1 =
zi = 0
N
i=1
Second Moment (a2 )
N
X
1
zi2 = 1
a2 =
N
i=1
Third Moment (a3 ) - Skewness
a3 =
1
N
Xz
N
i=1
3
i
3
= M3 p3M1 M2 +2 23M1 = ms33
( M2 M1 )
Fourth Moment (a4 ) - Kurtosis = (a4 -3.0)
a4 =
1
N
Xz
N
i=1
4
i
2
3M14 = m4
3 + 6M1 M2
= M4 4M1(M
M2 M12 )2
V ariance2
Interpretation of Moment Statistics
Mean (M1 ) - 1st moment about the origin - central tendency measure.
Variance (m2 ) - 2nd moment about the mean - dispersion measure.
Skewness (a3) - 3rd standardize moment - skewness measure.
a3 = 0 ! symmetrical
a3 > 0 ! positively skewed
a3 < 0 ! negatively skewed
for a3 between 0.2, the distribution can be assumed to be normal with respect to
skewness.
Kurtosis (a4 -3.0) - (4th standardized moment - 3) - kurtosis measure.
a4 3 = 0 ! same peakedness as normal curve.
a4 3 > 0 ! more peakedness than normal curve.
Spring, 2000
3
Notes - Econ 2370 - Statistics and Probability
3 < 0 ! attter than normal curve.
for a4 3 between 0.5, the curve can be considered normal with respect to kurtosis.
a4
Example - # of years attending University of Houston
Y ears Students
(X )
(F ) F X F X 2 F X 3 F X 4
1
2
3
4
5
6
N=21
5
4
3
7
1
1
P X = 61
P X = 221
P X = 907
P X = 4025
2
3
4
M1
M2
M3
M4
5
8
9
28
5
6
5
16
27
112
25
36
= 2.90476
= 10.52381
= 43.19048
= 191.66667
5
32
81
448
125
216
m1 = 0.00000
m2 = 2.08617
m3 = 0.50167
m4 = 9.02988
5
64
243
1792
625
1296
a1 = 0.00000
a2 = 1.00000
a3 = 0.16649
a4 = 2.07483
Mean = M1 = 2.90476
Variance = m2 = 2.08617
Standard Deviation = pm2 = 1.4443572
Skewness = a3 = 0.16649 ! assumed normal
Kurtosis = (a4 3) = -0.92517 ! atter than a normal distribution curve.
Spring, 2000
4
Notes - Econ 2370 - Statistics and Probability
2 Single Sample Tests
2.1
Single sample test of the mean (
^) where
is known
Make Assumption
Level of measurement - interval
Model - random sampling, population nornally distributed, = (some known value)
HO: = (some number)
Obtain a Sampling Distribution
Z distribution
= (some number)
X = pN
Choose a Signicance Level and Critical Region
= 0.05, 0.01 or 0.001
One or two tail test
H1: 6= (some number)
Compute a Test Statistics
Z=
X X
Make a decision
Formula for computing the condence interval
C ((X
Spring, 2000
Z 2 X ) (X + Z 2 X )) = 1 5