I ran an linear mixed effects model in r. The summary statistics for my fixed effects has estimates of zero but gives me a t-value and p-value (see variable Buffer_400 in image below). enter image description here

How do I interpret this output? My understanding is t-value = Value / Std. Error. If I am correct, how can I get a t-value =-2.297 when the std. error is zero? Is it because the standard errors are less than 0.0000000? Thank you.

  • $\begingroup$ Note that, unless the inputs are standardized, regression coefficients are expressed as the change in Y for a unit change in X. Given that, you are correct in noting that the coefficient and standard error for Buffer_400 are smaller than the smallest magnitude possible in the printed output. $\endgroup$ – Mike Hunter May 18 '16 at 23:02

Change the units of buffer_400. This is almost certainly a problem of the units being too large - like measuring human height in miles. It isn't dividing 0 by 0, it's just that the values are all in the 8th or later decimal places.

So, make a new variable buff_400_100000 or something that is buffersize divided by 1,000,000.

  • 1
    $\begingroup$ Perfect. Thank you very much. I tried it and it worked. $\endgroup$ – Guphadi May 18 '16 at 23:15
  • $\begingroup$ I'm glad it worked and welcome to the site. It is usual, here, to "accept" an answer that solves the problem. This lets others know it has been solved. You can check the check mark near the question. Thanks $\endgroup$ – Peter Flom May 18 '16 at 23:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.