ECON206_1011_01_Handout_08

# ECON206_1011_01_Handout_08 - ECON 206 METU Department of...

This preview shows pages 1–3. Sign up to view the full content.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ECON 206 December 13, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 1 LECTURE 08 HYPOTHESIS TESTING - II Outline of today’s lecture: I. Comments about Hypothesis Testing ................................................................................. 1 II. Small Sample Hypothesis Testing..................................................................................... 2 A. Small Sample Tests for Mean ....................................................................................... 3 B. Small Sample Tests for Comparing Two Population Means ........................................ 7 III. Testing Hypotheses Concerning Variances ..................................................................... 9 A. Tests for a Population Variance .................................................................................... 9 B. Tests for Comparing Two Population Variances ........................................................ 13 I. Comments about Hypothesis Testing Attention 1 Since there is always a room for Type II error, β when we carry out hypothesis testing: o You should say “Fail to Reject” the null hypothesis, H Æ You should not say “accept” null hypothesis, H Attention 2 Use the equality sign in null hypothesis when you do a one tailed-test. o In fact it is a controversial issue in statistical science. Some statisticians prefer to use ≤ or ≥ signs in null hypothesis. However, when we use the null hypothesis such as : . 8 H μ = , calculating the level of α is simple: we just found (value of test statistic is in RR when : 0.8 is true) P H μ = . If we used : . 8 H μ ≥ as the null hypothesis, our previous definition ECON 206 December 13, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 2 of α [ (value of test statistic is in RR when : 0.8 is true) P H μ = ] would be inadequate because the value of (value of test statistic is in RR) P will change for different possible values of μ since in this case any value for μ is possible within the interval 0.8 μ ≥ . In cases like this, α is defined to be the maximum (over all values of 0.8 μ ≥ ) value of (value of test statistic is in RR) P . This maximum value happens when 0.8 μ = , the “boundary value” of μ in : . 8 H μ ≥ . Thus, using : . 8 H μ = instead of : . 8 H μ ≥ leads to the correct testing procedure and the correct calculation of α without needlessly raising additional considerations. II. Small Sample Hypothesis Testing In order to apply large-sample hypothesis-testing procedures, the sample size must be large enough that ( ) ˆ ˆ / Z θ θ θ σ = − has approximately a standard normal distribution....
View Full Document

{[ snackBarMessage ]}

### Page1 / 20

ECON206_1011_01_Handout_08 - ECON 206 METU Department of...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online