R help - Feb 2007 - formula for test statistics of arima coefficients

I have an object, ARestopt, that is the result of a call to the arima
function. I want to calculate test statistics ( null is coeffs are zero
) for each of the estimated coefficients.
My coefficents are in a vector called ARestopt$coef and my covariance
matrix is  ARestopt$var.coef. 
 
I thought my formula for calculating them should be  
 
ARestopt$coef/(sqrt(diag(ARestopt$var.coef)  
 
but then I started thinking about the square root of n in the variance
for estimates. do the variances in the covariance matrix already
incorporate square root of n or should i be putting a sqrt(n) in also ?
I've seen a similar formula in an R book somewhere but I can't recall
where.  If someone
could tell me if above is correct ( disregard the issue of the
covariance matrix maybe not being a consistent estimate of the
covariance ) or the book or doc where I have seen the formula , 
I  would really appreciate it.
--------------------------------------------------------

This is not an offer (or solicitation of an offer) to buy/se...{{dropped}}