Re: [O] Efficiency of Org v. LaTeX v. Word

[Nick Dokos]

Ken Mankoff <address@hidden> writes:

> People here might be interested in a publication from [2014-12-19 Fri]
> available at http://dx.doi.org/10.1371/journal.pone.0115069
>
> Title: An Efficiency Comparison of Document Preparation Systems Used
> in Academic Research and Development
>
> Summary: Word users are more efficient and have less errors than even
> experienced LaTeX users.
>
> Someone here should repeat experiment and add Org into the mix, perhaps
> Org -> ODT and/or Org -> LaTeX and see if it helps or hurts. I assume
> Org would trump LaTeX, but would Org -> ODT or Org -> X -> DOCX (via
> pandoc) beat straight Word?
>

The "study" is deeply flawed: Word users typed more text in 30 minutes
than LaTeX users? 20% more? For straight text? I don't believe it: I
think it's much more likely that the Word users were better typists on
average and I didn't see any mention of normalizing the results by
taking that into account. And for LaTeX, the editing environment is
of paramount importance: did they mostly use vi, emacs, emacs+auctex?
Other than a couple of vague sentences in the "Discussion" section,
there is no mention of how this variable was (or was not) controlled.
And did the LaTeX users have to type the preamble or were they allowed
to use a template? It just seems unbelievable that there is such
a big difference for straight text.

I can believe perhaps that typing a table by hand into LaTeX is more
error prone than typing it into Word (although to be honest, I have
never done the latter, so I don't really know). With org and radio
tables, this would be a non-issue on the LaTeX side.

As for equations, even the authors admit that LaTeX is better, although
they tend to minimize the differences as statistically insignificant (at
least between the expert classes), which strikes me as somewhat suspect
as well: there seems to be a 10% difference between the expert user
averages and a bigger one for novices, although the error bars might
overlap in the first case (although they don't look it). I don't think
that even that would make the difference insignificant, but we'd have to
analyze their raw results to make sure (which they do provide and which
I took a look at, but afaict they don't provide answers to the questions
I raised above; maybe we should suggest that the authors use org and
reproducible results methods).

Anyway, color me deeply suspicious of the "study". 

-- 
Nick