P Value

[liorabeth]

Just a quick question -

when P is set at <.05, does that mean a result of .05 is not statistically significant? I know that "<" means less than, however, from what people have told me, in this case it actually means less than or equal to .05 (or whatever number is set).

Can someone please clarify?

Thanks!

[Damoz.G]

I haven't come across any questions that have the number exactly as 0.05

Usually they pick a number other than that to avoid any confusion. However, if the inevitability was to occur, I'd say that the results would be insignificant. Thats just the Methods student in me speaking though. HAHA!

As long as you clearly explain your reason, I'm sure you will be fine.

[liorabeth]

Thanks, that what I (or the Further student in me) thought as well, but just wanted to make sure.

Good luck for Friday!

[TrueTears]

The convention is (at least in statistics and econometrics), that a test statistic must be greater than the critical value in order to reject the null. Hence the corresponding p value must satisfy strict inequality.

[Aelru]

The convention is (at least in statistics and econometrics), that a test statistic must be greater than the critical value in order to reject the null. Hence the corresponding p value must satisfy strict inequality.

I just love how scientific this answer sounds.

[alondouek]

If you have a significance level of 0.05 (5%), you can only reject the null hypothesis in favour of the alternative if your p-value of less than 0.05. If your p-value is equal to or greater than your significance level, you fail to reject the null hypothesis.

[IvanJames]

I hate to go against the grain here, but the Macmillan Grivas text book refers to the usual P value as ≤ 0.05, meaning a P value that is 0.05 or less is seen as statistically significant.

Is this wrong?

[alondouek]

I hate to go against the grain here, but the Macmillan Grivas text book refers to the usual P value as ≤ 0.05, meaning a P value that is 0.05 or less is seen as statistically significant.

Is this wrong?

It's not wrong, but it goes against every statistics unit I've taken at university (all two if them lol). As someone mentioned above, VCAA won't give you something ambiguous.

[TrueTears]

there's no right or wrong here, simply a matter of convention.