Hello,

I have 2 problems dealing with "is this significant" (at a 1% level, and a different problem with 5%).

Unfortunately the tables are on a PDF file so I cant post them on here. So if someone is willing to help me, I can send them the file.

It would be greatly appreciated if someone could help me


And if you scratch my back, I will scratch yours

Just have 1x significance problem left. I need to complete it by noon. So if anyone can help me by then, that would be great!

I will be checking back on here every 20 minutes.

The problem refers to a coeffecient of 13.5. The number below this is [6.63]* where * means that it is statistically significant at a 5% level.

Question is asking if it is significant at the 1% level and says "hint: calculate t-observed value and compare it to the t-critical value".

So I don't believe I can use the * to say that it is not significant at the 1% level (only the 5% level) because it doesn't follow the Hint that was given (and maybe it is the wrong way of doing it).

Let me know if you can help! I have the PDF file. Again, I will be checkin back on here every 20 minutes (till noon)


-Thanks for reading!

test. will be edited if i see that i can post here

cant edit, so double-post. sry for that.

a little late, but maybe it still helps a little.

the t-test helps you to make a decission between your H0 (there is no effect; m1 = m2, m1 >= m2, m1 <= m2) and H1 (the independent variable has an effect on the dependend one; m1 != m2, m1 < m2, m1 > m2).
Therefor the t-test estimates how likely the observed results (the data you've collected in your sample) are to occure if your H0 is assumed.
If the chance is less than alpha % (with alpha beeing 0,05 or 0,01 in your case) we say per definition, that the difference between m1 and m2 is significant so we dont think that it is very likely that there is no effect from the independent variable on the dependent one. In this case we refuse H0 and go with H1.

To do this the difference between m1 and m2 (or a fixed value in case of a one sample t-test) is valued by the standarderror of this difference. thats your t = formular.
This t-value, you've calculated lies somewhere in the t-distribution (thats the figure most likely printed on top of your .pdf table).
If you have a big positive t value it is located somewhere on the far right side of the Distribution, if you got a big one negative t-value its located on the far left side.

Your .pdf table lists some critical t-values for a given alpha and given degrees of freedom.
This critical t-value lies somewhere in the t-distribution as well (or to be more precise: it lies at the point where the area under the slope on the outer side of this t-value is alpha% [or alpha/2 percent in case of a non-directed H1]).
Its hard to describe (especially with my bad english) so i bring a picture with me



the pic showes the critical t value for alpha = 5 with an unspecified H1 (so there are 5/2 % on the far left side and 5/2 on the far right side).
To make it a little bit easier we will just look on the right side for now.
If your calculated t-value lies somewhere in the outer 2,5 % area it is significant. To calculate this area you need an integral which is not so easy to calculate by hand so back in the good old days without computers all around people thought it would be a great idea to have a table containing the t-values that stand for the common alpha-areas (which is the critical values-table). If i calculated a t-value bigger than the critical t-value listed in this list i knew that it lies somewhere right side of the crit-value and therefor p (the probability or in other words, the area under the slope) has to be smaler than alpha.

Since you have solved problem 1 (with alpha = 5) allready it is now very easy to solve problem 2 as well.
You simply take the t-value that you've calculated allready and look if it is bigger than the t-critical value in the list in column (or row) ,001 (instead of .05).

If you want to troll your teacher a little bit (which would be legitim since he/she didn't explain this stuff to you in a propper way) ask him where to find this critical-values-tables in spss, R or the little statistics-app on your smart-phone.

hope that helped at least a little bit