Significance level

[wittmannv]

Hello,

in the presettings a significance level of 0,05 can be found. Does this mean, the generated significance levels after bootstrapping are restricted to the 5%-Level?
Is it posible to accept hypotheses at significance levels of 10%, 5%, 1% and 0,1%?

Thank you very much!

[cringle]

No, in fact you get t-values and p-values. The level is important to build the confidence intervals.

Kind regards,

[wittmannv]

Hello,

thanks for the answer. But does the presetting of 0,05 mean that I can accept my hypotheses only until the 5%-level? I want to accept them until the 10% significance level, so do I have to adapt the number in the presettings to 0,1?

Thank you very much!

[cringle]

Sure

[wittmannv]

Sorry for asking again but as you realize I am a bit confused...:
I suppose this is also the case for the two-sided test as 0,1 is directly about the significance level?
Or does a presetting of 0,05 in a two-tailed test mean that it could accept hypotheses until 10%?

Thank you very much!

[Hengkov]

Hi,

To choose one-tailed or two-tailed in the run bootstrap depending on the direction of your hypothesis. If you have hypothesized direction, then select the one-tailed, and if not then select the two-tailed. One-tailed tests will always be smaller than the two-tailed if significance level is the same. For example, for a significance level of 2.5% obtained for the one-tailed probability is 1.96. And for the 5% significance level obtained for the two-tailed probability is 1.96. Thus, one-tailed test is two times two-tailed test (2 x 0.025= 0.05). To further understand this, see My book "Applied Statistical Data Analysis for Social Science Research with IBM SPSS (2015), + 430 pages"

http://cvalfabeta.com/0538-detail-apl._ ... _spss.html

Greetings