p-values using fitglm and standard t-tests

Using fitglm in Matlab R2014a on Windows 7, I am getting the following error when trying to call column vectors from elements of a cell (tr) into the function: mdl = fitglm(tr{1}(:,1),tr{1}(:,2)) Undefined function 'statgetargs' for input arguments of type 'cell'.

Error in glmfit (line 129) [errid,errmsg,link,estdisp,offset,pwts,const] = ...

Error in GeneralizedLinearModel/fitter (line 919) [model.Coefs,model.Deviance,stats] = ...

Error in classreg.regr.FitObject/doFit (line 220) model = fitter(model);

Error in GeneralizedLinearModel.fit (line 1245) model = doFit(model);

Error in fitglm (line 133) model = GeneralizedLinearModel.fit(X,varargin{:});

This error will persist every time I use the function until I reboot Matlab. It is not fun.

Are there any users that have seen this error?

Also, I have a routine I wrote years ago that calculates the t-statistic of a linear model, and I should be able to calculate the p-value easily with basically a table (like in my old stats book) or in a pdf. Does anyone know how to calculate p-values to test the null hypothesis of a linear model having a slope of 0 without using such overly complex functions as fitglm????!!! It should be as easy as a table look up from my old function! Then I wouldn't need to find the cause of this annoying bug in fitglm.

Thank you, Brad