It has been my experience that different academic fields have converged to different statistical software apps as their "standard." E.g., in Economics Stata is very common but virtually unheard of in Social Science, where SPSS is ubiquitous.

I sometimes wonder if academic papers sometimes get a hard time from peer reviewers simply because the statistical outputs are formatted in a way that they're not used to (because the paper used a different statistical software app than the dominant one in the field).

This becomes a real issue because different software apps sometimes refer to the same concept with different names. E.g., what may be termed "standard error of the regression" in one app may be termed something slightly different in another.

I want to use R, but at the same time appeal to peer reviewers who may be used to the dominant software in the field. There is a real need for something in R that will allow the output to mimic the output from some other software. Is there already some way to do this?

  • 5
    $\begingroup$ I don't think I have ever seen a peer-reviewed paper that just dumped the output of a stats package (except software reviews and comparisons). There's also a fundamental problem with an implicit assumption of your question: that R has some kind of standard "output". What it prints is always the result of some specialized function that formats the values. R has no standards for such outputting, which is left up to those who have written those functions--which includes tens of thousands of contributors. In this context it would appear you are asking for something that could never exist. $\endgroup$
    – whuber
    Commented Jul 30, 2015 at 16:00
  • 2
    $\begingroup$ Look into packages like knitr, pander, xtable, etc. that let you render professional looking output via LaTex. $\endgroup$
    – A. Webb
    Commented Jul 30, 2015 at 16:38
  • 1
    $\begingroup$ If for some reason you needed to present output rather than some brief summary information, you could simply edit the formatting; as long as you're not changing the information, I don't see any issue. But R is usually pretty easy to make give output any way you like - many packages return an object that is printed by either a print or summary function written for that object. You could quite readily write your own function to turn the object into output. In many cases the existing summary or print function is written directly in R, and you can readily use it as a basis for your own. $\endgroup$
    – Glen_b
    Commented Jul 30, 2015 at 21:31


Browse other questions tagged or ask your own question.