Newsgroups: comp.robotics
Path: brunix!uunet!utcsri!torn!watserv2.uwaterloo.ca!mach1!veen2976
From: veen2976@mach1.wlu.ca (veenstra christopher)
Subject: Scientific Paper Survey, results plus repost
Message-ID: <C0nFw9.3z9@mach1.wlu.ca>
Organization: Wilfrid Laurier University
Date: Sun, 10 Jan 1993 17:43:20 GMT
Lines: 121

			     Scientific Paper Survey

  Well, I only got 6 replies (there's still time if you haven't sent me one)
and the general opinion seems to be that most, or at least many, scientific
papers are poorly written.  I realize that 6 is not a sound basis for
analysis but it's all I have to go on, so I'll just have to mention that
when I'm presenting the data.

  Conclusions up to now:  4 of the 6 people stated that of the papers that they
read, probably about 80% were poorly written, while one said that that 95%
are well written, though that doesn't mean it's anything but "well-written
garbage".  The 6th person stated that there were some problems but
attributed it to the writers' not having English as their native language.
Only one person commented on the technical content of the papers, and
he stated that, though not nearly as bad as the general opinions of the written
aspect of the papers, the technical content was poor in a number of papers.
He basically implied that there are some people writing papers, and not
presenting their data in a concise and understandable (reproducable) manner.
This was something that I wasn't expecting to see.  I've always thought that
people in the sciences have been known for stating just the facts, and
explanations generally tend to be either mechanical, or very short winded.
That writers would be presenting their data in a poor fashion didn't even
occur to me.  Still, only one person commented on this so I believe that
poor technical content is probably not too common.

  Also, I found that 4 of the people who responded were basically accustomed
to the written style of the papers that they read.  At least, that's my
interpretation.  2 people didn't want to waste their time trying to figure
out what poorly written papers were trying to say, one going so far as to
reject papers given to him due to their unreadability.  The others seemed to
say that the papers were not generally well written, but didn't go so far as
saying that they would refuse to read them, which makes sense because there
could be some useful information buried within them that might make it
worthwhile going after.  

  So far as the survey and the results go, I believe that this has been very
beneficial study.  The information that I have received is going to allow me
to make a very interesting, and thought provoking presentation.  If you have
any further comments/questions feel free to email me at the address at the
bottom of this message.

  Once again, let me thank you all for your time and trouble.

Chris Veenstra
veen2976@mach1.wlu.ca

------------------------

Comments:
  These are the more quotable comments that people wrote to me.  I've tried
to pair them up, pro and con, to give an equal representation of what people
said.  I've still got until Thursday to gather data for my presentation so
if you wish to send any more comments then feel free to, and I will be
appending a copy of my survey to the end of this article to try to expand my
data group further.

Written Style:
(pro)
Unless the paper is extremely flowery or excessively dense, I don't
worry to much about the style of the papers I read--if I can extract
the information I'm interested in, then the papers is good enough.
  
(con) 
Of the papers that I read or referee, a distressing fraction are well written.
I have recently rejected two papers because they were so poorly written that I
couldn't tell what the results were. Published papers tend to be better
written (at least in terms of the mechanics of the English) but frequently
have disappointingly little new in them.

Technical Content:
(con)
> 1)  Do you read/write/referee papers?  For each that apply to you, what do
>         you look for?
    Read and Write: For both a look for clarity, in explinations, (all
  too often important assumptions are left out :-(. I also look to
  see that if a model or example is given and if enough data is given
  so that the readed could recreate the example. (Again, often data
  is omitted or vague, which makes me question the validity of the
  paper. which makes me wonder what they are trying to hide.) So I
  guess I look for clarity and completeness. If one gives derivations
  the should be complete enough that the casual user should be able
  to follow them. If one gives examples, then give all pertinate
  data. (A good example of vague data is when some one says that
  the operating speed was w_n/2.5, where w_n is the natural frequency,
  but don't tell you what w_n is or how it was computed.)

  <Author of survey comment>  No one wrote about good technical content,
though only one commented about poor content so I believe that we can assume
that the others didn't see it as an issue.  This would imply to me that the
technical content of the papers is generally fairly good and probably well
(or adequately) presented.

============================================================================

			Scientific Paper Survey

   1)  Do you read/write/referee papers?  For each that apply to you, what do
	   you look for?

   2)  Of the papers that you read or referee, what percent do you feel are
	   well-written?

   3)  What is your position/company you work for?  (This is so I'll know
	   approximately what the technical level of the people replying is.)

============================================================================

  For those of you who missed the original survey, I'm trying to gather some
opinions on the content of scientific papers.  The data is being gathered for
a presentation that I am going to make on Thursday regarding the written
content of papers.  If you have a few moments to spare, I would greatly
appreciate your replying and telling me what you think.  There seems to be
some interest in the results, so I will repost the final results at the end
of the week.  Thank you for your time and trouble.


Chris Veenstra
veen2976@mach1.wlu.ca
4th Year, Physics and Computing
Wilfrid Laurier University
Waterloo, Ontario, Canada
