chapter 10 statistical inference for two samples more than one but less than three! chapter 10b <...

40
Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X <

Upload: warren-porter

Post on 19-Jan-2016

227 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Chapter 10 Statistical Inference for Two Samples

More than one but less than three!

Chapter 10B< X <

Page 2: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Things Statistical

Page 3: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

• A special case of the two-sample t-tests occurs when the observations on the two populations of interest are collected in pairs. They are not independent observations.

• Each pair of observations, say (X1j , X2j ), is taken under homogeneous conditions.

•Each observation in one sample is paired naturally or by design with an observation in the other sample.

•The test procedure consists of analyzing the differences between each the pairs.

10-4 Paired t -Test

Page 4: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-4 Paired t-Test

1 2

1 2 ; 1,...,i i i

D X X

d x x i n

Page 5: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-9

Page 6: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-9

Page 7: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-9

Page 8: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Paired Versus Unpaired Comparisons

Page 9: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

A Confidence Interval for D

Definition

Page 10: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-10

Page 11: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-10

1.21

12.68 / 14d

d

s

Page 12: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-10

Page 13: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.1 The F Distribution

10-5 Inferences on the Variances of Two Normal Populations

We wish to test the hypotheses:

The development of a test procedure for these hypotheses requires the F distribution.

Page 14: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.1 The F Distribution

Page 15: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

More F Distribution

Page 16: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Look at this…

The lower-tail percentage points f-1,u, can be found as follows.

Page 17: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.2 Hypothesis Tests on the Ratio of Two Variances

Page 18: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.2 Hypothesis Tests on the Ratio of Two Variances

2 21 12 22 2

/

/

SF

S

Page 19: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-11

Page 20: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-11

Page 21: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-11

Page 22: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.3 Type II Error and Choice of Sample Size

Page 23: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-12

Page 24: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-5.4 Confidence Interval on the Ratio of Two Variances

1,1,2/22

21

22

21

1,1,2/122

21

1,1,2/1,1,2/1

21

21

22

22

1212

1212

s

s

s

s

interval, confidence get the wealgebra little aWith

.1)(

on,distributi Fan has /

/Since

nnnn

nnnn

ff

fFfP

S

SF

Page 25: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-13

Page 26: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-13

Page 27: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-6 Large-Sample Test on the Difference in Population Proportions

0 1 2

1 20

1 2

Null hypothesis: :

ˆ ˆTest Statistic:

1 1ˆ ˆ(1 )

H p p

P PZ

P Pn n

zzppH

zzppH

zzorzzppH

0211

0211

2/02/0211

:

:

:

Alternate Hypothesis

Rejection Criterion

1 2

1 2

ˆ X XP

n n

Common punder H0

Page 28: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-14

A depressed stat studentwho was unable to reject his null hypothesis. Very disturbed.

Page 29: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-14

Page 30: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-14

Page 31: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Minitab Output for Example 10-14

Page 32: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Inferences on Two Population Proportions- beta errors

Unless p1=p2, you cannot use the estimate of the standard deviation as in the hypothesis tests. Instead we use,

21

2211

21

2211

ˆˆ

21212/

ˆˆ

21212/

2

22

1

11ˆˆ

)1()1( and with

)()/1/1(

)()/1/1(

:iserror beta thee,alternativ sided- twoFor the

)1()1(

21

21

21

nn

pnpnq

nn

pnpnp

ppnnqpzΦ

ppnnqpzΦβ

n

pp

n

pp

PP

PP

PP

Page 33: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-6.2 Type II Error and Choice of Sample Size

Page 34: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

10-6.2 Type II Error and Choice of Sample Size

Page 35: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Confidence Interval on the Difference in the Population Proportions

Page 36: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-15

8-7

Page 37: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Example 10-15

Page 38: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Summary Table and Road Map for Inference Procedures for Two Samples

Page 39: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

Summary Table and Road Map for Inference Procedures for Two Samples

Page 40: Chapter 10 Statistical Inference for Two Samples More than one but less than three! Chapter 10B < X

We have concluded our discussion on two samples

Should we not go to three or more sample tests?

A statistics student countingthe number of samples…