math and science people I need your help

[MR Pink]

In week 1, 3 and 6 the increase in % 1 testgroup was 100, 101.5 and 102.2

In testgroup 2 the results were 100, 100.7 and 101

How do I calculate with the t-test if these results are significant or not?

K for everyone that helps me out

 

[squatpuke]

It's been 20 years for me since I took Quantitative Methods....but I remembered the T-Test...

At least I remembered it was from my Stats class...

Sorry, all I could do was google...

http://www.google.com/search?q=t-test+statistics

 

[MR Pink]

I've already checked google to but I suck at math so couldn't make much out of it

 

[WODIN]

Mathworld my man.

http://mathworld.wolfram.com/Pairedt-Test.html

 

[MR Pink]

Mathworld my man.

http://mathworld.wolfram.com/Pairedt-Test.html

Yep that's it but can someone fill in that formula for me? I really don't know shit about math ... is the information in my first post enough to fill in this formula btw?

 

[MR Pink]

Yep that's it but can someone fill in that formula for me? I really don't know shit about math ... is the information in my first post enough to fill in this formula btw?

Bump for a wizzkid

 

[redguru]

In week 1, 3 and 6 the increase in % 1 testgroup was 100, 101.5 and 102.2

In testgroup 2 the results were 100, 100.7 and 101

How do I calculate with the t-test if these results are significant or not?

K for everyone that helps me out

Is that the entire question?

 

[WODIN]

LOL

http://www.physics.csbsju.edu/stats/Paired_t-test_NROW_form.html

Here's a calculator.

* You have 3 observations per group.

 

[Smurfy]

Stats was a long ass time ago (1990)

 

[redguru]

The reason I asked is if that 100 is 100 or 100.0

 

[MR Pink]

The reason I asked is if that 100 is 100 or 100.0

I guess 100 ?? whats the difference? ...the 3 numbers 100, 101.5 and 102.2 are a rise in % 100 being the 0 or starting point. The same goes for the 100, 100.7 and 101 ...don't know if that makes any sense but my brain is hurting

 

[redguru]

Not sure if this is correct, but maybe you can follow it, I haven't done t-tests since the 80's, Anyone else feel free to critique.

 

[samoth]

Umm... why are you doing this? You don't have any knowledge of stats or quants it seems, as 100 and 100.0 are different numbers.

Do you know how to calculate mu and sigma? Whether or not the numbers are significant depend on the level of confidence you are looking for; if you only need to be 2.6% confident, they sure, they're significant. If you're looking for 90% or more, then that's where the formulas come into play.

I can't verify RG's eqns, but the work follows through. If you can't follow what he did -- it's thorough and terms defined -- then you might need more help than we can easily give you.

 

[redguru]

I was looking for the post where I explained sigfigs to him, and I must have closed the window instead of posting it! Thanks Samoth, Now I don't have to retype it!