How to calculate 2-tailed P value using t value and degree of freedom?

Hi all, I have gotten the t value and degree of freedom of paired data in T-Test .I need to calculate the 2-tailed P value using t value and degree of freedom.I know that the function 'tcdf' can calculate 1-tailed P value.But I still don't know how to calculate 2-tailed P value. I will be very appreciate your answers. Thank you!

Answer 1

[Star Strider]

You can simply cut to the chase and calculate it directly:

tdist2T = @(t,v) (1-betainc(v/(v+t^2),v/2,0.5));    % 2-tailed t-distribution

where ‘t’ is the t-statistic and ‘v’ the degrees-of-freedom.

[William]

Thank you! I got a more concise expression : 2tailedP=2*tcdf(-abs(t),v) .where ‘t’ is the t-statistic and ‘v’ the degrees-of-freedom.And tcdf is matlab function.I would like to let you help me to see if this expression is right.

[Star Strider]

My pleasure!

I know my anonymous function is correct (I created it to Answer a Question for someone who didn’t have the Statistics Toolbox, although its purpose there was to calculate confidence intervals), so comparing them, they are equal:

t =  4;
v = 10;
tail2P = 2*tcdf(-abs(t),v);
tdist2T = @(t,v) (1-betainc(v/(v+t^2),v/2,0.5));    % 2-tailed t-distribution
T2 = [1-tdist2T(t,v)  tail2P]

produces:

T2 =

       2.5183e-003     2.5183e-003

By the way, you can’t start a variable or function name with a number in MATLAB, so I changed your variable name slightly.

[William]

Thank you very much.I am very appreciate your answers!

[Star Strider]

My pleasure!