UMP Tests and UMPU Tests in Statistical Inference, Study notes of Mathematical Statistics

Ump (uniform most power) tests and umpu (uniformly most power unbiased) tests in statistical inference. It includes examples of ump tests for uniform distributions and umpu tests for known cumulative distribution functions and negative binomial distributions. The document also explains how to show the non-existence of a ump test and how to obtain a umpu test of a given size.

Typology: Study notes

Pre 2010

Uploaded on 09/02/2009

koofers-user-ajn-1
koofers-user-ajn-1 🇺🇸

10 documents

1 / 1

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
TA: Yuan Jiang Email: [email protected]
STAT 710: Discussion #17
April 1, 2008
1 UMP Tests
Example 1. Let (X1, ..., Xn) be a random sample from the uniform distri-
bution on (θ, θ + 1), θ R. Suppose that n2.
(i) Show that a UMP test of size α(0,1) for testing H0:θ0 versus
H1:θ > 0 is of the form
T(X(1), X(n)) = 0X(1) <1α1/n, X(n)<1
1 otherwise,
where X(j)is the jth order statistic.
(ii) Does the family of all densities of (X(1), X(n)) have monotone likelihood
ratio?
2 UMPU Tests
Example 2. Let Fand Gbe two known cumulative distribution functions
on Rand Xbe a single observation from the cumulative distribution function
θF (x) + (1 θ)G(x), where θ[0,1] is unknown. Consider H0:θ[θ1, θ2]
versus H1:θ6∈ [θ1, θ2], where 0 < θ1θ2<1 are constants.
(i) Show that a UMP test does not exist.
(ii) Obtain a UMPU (uniformly most power unbiased) test of size α.
Example 3. Let X1and X2be independently distributed as the negative
binomial distributions with sizes n1and n2and probabilities p1and p2, re-
spectively, where ni’s are known and pi’s are unknown.
(i) Show that there exists a UMPU test of size αfor testing H0:p1p2
versus H1:p1> p2.
(ii) Let Y=X1and U=X1+X2. Determine the conditional distribution
of Ygiven Uwhen n1=n2= 1.
Office: 1275A MSC 1 Phone: 262-1577

Partial preview of the text

Download UMP Tests and UMPU Tests in Statistical Inference and more Study notes Mathematical Statistics in PDF only on Docsity!

TA: Yuan Jiang Email: [email protected]

STAT 710: Discussion

April 1, 2008

1 UMP Tests

Example 1. Let (X 1 , ..., Xn) be a random sample from the uniform distri- bution on (θ, θ + 1), θ ∈ R. Suppose that n ≥ 2. (i) Show that a UMP test of size α ∈ (0, 1) for testing H 0 : θ ≤ 0 versus H 1 : θ > 0 is of the form

T∗(X(1), X(n)) =

0 X(1) < 1 − α^1 /n, X(n) < 1 1 otherwise,

where X(j) is the jth order statistic. (ii) Does the family of all densities of (X(1), X(n)) have monotone likelihood ratio?

2 UMPU Tests

Example 2. Let F and G be two known cumulative distribution functions on R and X be a single observation from the cumulative distribution function θF (x) + (1 − θ)G(x), where θ ∈ [0, 1] is unknown. Consider H 0 : θ ∈ [θ 1 , θ 2 ] versus H 1 : θ 6 ∈ [θ 1 , θ 2 ], where 0 < θ 1 ≤ θ 2 < 1 are constants. (i) Show that a UMP test does not exist. (ii) Obtain a UMPU (uniformly most power unbiased) test of size α.

Example 3. Let X 1 and X 2 be independently distributed as the negative binomial distributions with sizes n 1 and n 2 and probabilities p 1 and p 2 , re- spectively, where ni’s are known and pi’s are unknown. (i) Show that there exists a UMPU test of size α for testing H 0 : p 1 ≤ p 2 versus H 1 : p 1 > p 2. (ii) Let Y = X 1 and U = X 1 + X 2. Determine the conditional distribution of Y given U when n 1 = n 2 = 1.

Office: 1275A MSC 1 Phone: 262-