{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Probabilities of 005 or 001 therefore when testing

Info iconThis preview shows pages 2–5. Sign up to view the full content.

View Full Document Right Arrow Icon
probabilities of 0.05 or 0.01 . Therefore, when testing against a two- sided alternative, the information in the table offers a choice of the significance level as either: 0.10 = α (a 10 % level) or 0.02 = α (a 2 % level). In applied work, computer software is used for the calculation of descriptive statistics, the test statistic and an accompanying p-value. A p-value for the two-tailed test is calculated as: p-value > = - - 2 y 2 x ) 1 n , 1 n ( s s F P 2 y x For a chosen significance level α , the decision rule is to reject the null hypothesis if: p-value < α
Background image of page 2

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

View Full Document Right Arrow Icon
Econ 325 – Chapter 10.4 5 A one-sided alternative can be used as follows. Test the null hypothesis: 2 Y 2 X 0 : H σ = σ or 2 Y 2 X 0 : H σ σ against the alternative: 2 Y 2 X 1 : H σ > σ If a comparison of the calculated sample variances shows 2 y 2 x s s < then there is no evidence to reject the null hypothesis. On the other hand, if 2 y 2 x s s > then the null hypothesis may or may not be rejected. To make a decision, calculate a test statistic as the variance ratio: 2 y 2 x s s For a chosen significance level α , the decision rule is to reject the null hypothesis if: c 2 y 2 x F s s > where c F is the F-distribution critical value that satisfies: α = > - - ) F F ( P c ) 1 n , 1 n ( y x Econ 325 – Chapter 10.4 6 Example Annual total sales of a company are reported for two sample periods: Period 1: active price competition in the industry (4 years) Period 2: price collusion in the industry (7 years) It is hypothesized that total sales should vary more in an industry with active price competition compared to a market with price collusion. Denote 2 X σ and 2 Y σ as the population variances in the two sample periods. Test the null hypothesis: 2 Y 2 X 0 : H σ = σ against the alternative: 2 Y 2 X 1 : H σ > σ From the data set, the sample statistics are: Period 1: 4 = x n 114.09 = 2 x s 10.68 = x s Period 2: 7 = y n 16.08 = 2 y s 4.01 = y s
Background image of page 3
Econ 325 – Chapter 10.4 7 The test statistic is the variance ratio: 7.095 16.08 114.09 = = 2 y 2 x s s This can be compared with the critical values reported in the F-distribution Appendix Table.
Background image of page 4

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

View Full Document Right Arrow Icon
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}