Statistics Question: Calculating the Probability of a Type II Error

[eddiev10025]

I need help calculating the probability of a Type II Error (failing to reject the null hypothesis when it should be rejected).

I need help with the following tests:
1) z-tests (left-tailed, two-tailed and right-tailed)
2) t-tests (left-tailed, two-tailed and right-tailed)
3) F-tests (Equality of Variance and One-Way ANOVA
4) Chi-Square Goodness-of-Fit tests

eddiev10025

 

[Hany ali]

without the file ,we can't to help You ?

 

[eddiev10025]

Of course. See the attached file

 

[Chihiro]

1) https://support.office.com/en-us/article/z-test-function-d633d5a3-2031-4614-a016-92180ad82bee
Or using add-in (Data analysis add-in)
https://www.dummies.com/software/microsoft-office/excel/how-to-perform-z-test-calculations-in-excel/

2) https://support.office.com/en-us/article/t-test-function-d4e08ec3-c545-485f-962e-276f7cbed055
Or using add-in
https://www.rwu.edu/sites/default/files/downloads/fcas/mns/running_a_t-test_in_excel.pdf

3) https://support.office.com/en-us/article/f-test-function-100a59e7-4108-46f8-8443-78ffacb6c0a7
Or using add-in
http://www.datasciencemadesimple.com/f-test-in-excel/

4) http://www.real-statistics.com/chi-square-and-f-distributions/goodness-of-fit/

 

[eddiev10025]

Chihiro, Thanks!
I have been to all of these sites. The first three do not cover how to calculate Type II errors. Real-Statistics has a plug-in that is supposed to calculate Type IIs, butI do not want to have my students use a plug-in.

 

[Chihiro]

As I understand it, probability of Type II error can't generally be computed, since it's dependent on unknown (i.e. Population mean not the sample mean). It can only be calculated at given values.
But my understanding of statistics is bit sketchy

Do you mean hypothesis testing using statistical power?

1. http://www.real-statistics.com/hypothesis-testing/null-hypothesis/
Explains that β is acceptable level of Type II error.

2. http://www.real-statistics.com/sampling-distributions/statistical-power-sample/
Using Excel to estimate the power of test.

Or below, same concept, but bit more detailed.
https://amstat.tandfonline.com/doi/full/10.1080/10691898.2009.11889507

 

[eddiev10025]

Thanks for sending the link to the Journal of Stat Education article.

One should be able to calculate the P(Type II), or Power, 1-P(Type II), using Excel. Conrad Carlberg does it for t-tests and ANOVA tests in Statistical Analysis Microsoft Excel 2016. I am not confident with my attempts at two-tailed z and t tests and I am at a complete loss with Chi-Square. I have always used G*Power to calculate the sample size to get at least 80% power, or 20% probability of a Type II error. There must be a way of doing this in Excel without resorting to plug-ins. SigmaXL, by the way, does calculate power for some t-tests, proportions tests, and One-Way ANOVA. For some reason it does not due Chi-Square or z-tests for means. I have not found a POWER function in the REAL-STATs plug-in.

eddiev10025

 

[GraH - Guido]

The plug-in "Analysis Toolpak" is a standard one available in all Excel versions, but it needs to be checked manually via File -> Options.
Then it appears as a separate section on the ribbon.
I personally don't have any clue how to use them.