Econometrics, Lecture 2, 2014-02-25

• Dr Bartłomiej Bartoszewicz
• Email:

[e-mail]

• Office hours: if necessary - after each lecture,

Tuesday, 9:30-10:00

• Final examination: Friday, 13-06-2014, 17:00,

room to be announced

• Final examination: 5-6 short problems to solve, 60

min., total 30 points, at least 16 points to pass

• Bring your calculators with you!

• Tables: print statistical tables and bring them to

every lecture and to examination!

Methods of statistical inference

• Point estimation -> Ordinary Least Squares, linear

regression

• Interval estimation -> coming soon
• Hypotheses testing -> example: testing statistical

significance of correlation coefficient

Econometrics, Lecture 2, 2014-02-25

Statistical Hypothesis Testing

• Elements of a significance test:

– Working hypothesis
– Null hypothesis H

0

and alternative hypothesis H

1

– Test statistic
– Probability distribution of the test statistic under H

0

(assumption that the null hypothesis is true)

– Rejection rule, rejection region, critical value, p-value
– Test conclusion

Econometrics, Lecture 2, 2014-02-25

Statistical Hypothesis Testing

Econometrics, Lecture 2, 2014-02-25

Statistical Hypothesis Testing

• α – significance level = size of the test
• 1 - β – power of the test

• Significance test is constructed (and the test statistic is

chosen) so that for a given (small) significance level α
the power of the test is large (as large as possible).

Econometrics, Lecture 2, 2014-02-25

Example

Significance test of correlation coefficient

Econometrics, Lecture 2, 2014-02-25