Which TTEST to use

[Rexkid]

Hi everyone,

This is my first post on this forum.

I'm on a Sports science course and during a practical we had each participant measure their bodyfat using different methods. We measured BMI, Bodyfat % (BF%) using skinfold measurements, and BF% using a biometric impedance analysis (BIA) (the scales with the current sent through the body). We also measured everyone in a BODPOD (http://www.bodpod.com/en/products/bo...pod-gs/bod-pod) which is a gold standard for measuring body compositon.

This a table of the correlation between the different methods and the BODPOD.

Males Females Overall
BMI 0.856 0.746 0.503
BF % (skinfold) 0.934 0.505 0.731
BF % (BIA) 0.900 0.785 0.868

I am now struggling as I need to run a Ttest for each value of this table but cannot figure out which one to use. Your help would be much appreciated.


Rexkid

[MrShorty]

I'm not sure what your question is.

If you are asking about the difference between the TTEST() function and the T.TEST() function, you should not see any real difference between the functions. TTEST() is the older function from older versions of Excel and the T.TEST() function was introduced in 2010. If you don't need backwards compatibility, then Microsoft wants you to use the newer function. https://support.office.com/en-us/art...3-a0eaad83c7ae

If your question is more about how to run a student's t test, I might suggest a more statistically oriented tutorial.

This site describe many basic statistical concepts with specific application in Excel: http://www.real-statistics.com/stude...asic-concepts/
Wikipedia, of course, has a page: http://en.wikipedia.org/wiki/Student%27s_t_test
Stattrek has several tutorials geared towards US high school students learning statistics: http://stattrek.com/probability-dist...px?Tutorial=AP

Since you have more than two different statistics, I also wonder if you are looking for an ANOVA type analysis (basically a formalized way of comparing more than 2 to determine if any one of the statistics is different from the others). Depending on the exact nature of the analysis you are performing, you may want to look into ANOVA.