If there is an option to submit your paper in either Latex or Word, which format should be given preference? Does it affect the chances of acceptance? My paper relates to computer science & engineering, with a few mathematical formula, diagrams, and tables.
This is coming from a more engineering, physics and mathematics point of view. In many sub-area of these fields, LaTeX papers have a more distinct professional feel. This is mostly related to the formatting of the equations and the symbols. Furthermore, numerous authors in mathematics are openly LaTeX advocate. As such, I believe (but this is no hard fact) that you might get a slightly more positive bias using LaTeX than Word. Keep in mind that the most important part is to submit a clear well-written document and that the format is second to this former fact.
Let's start by acknowledging that there is a very good reason why LaTeX is the standard system for writing papers across all of math and large parts of computer science, physics, statistics, and other technical areas. No one has yet come up with a system that produces better-looking, more esthetically pleasing, well formatted documents that do as effective a job of communicating complex technical ideas, especially those involving lots of mathematical symbols and formulas. By contrast, documents generated by Microsoft Word are generally... meh (and if they contain math formulas... double meh).
With that said, I believe that the answer to your question is: yes, LaTeX papers and Word papers are far from equivalent in terms of their chances to be accepted to a computer science journal. I see several issues here, some direct and some indirect:
One effect is that, as I was opining above, a properly written LaTeX paper will simply do a better job of communicating the content of your ideas to the reader than the same paper written in Word. Thus, the cognitive load on the journal editor and referees that stands between them and understanding what you are trying to say and what the contribution of your paper is will be higher with a Word paper. This will translate to a difference in the chances of acceptance (better chances for the LaTeX version).
Another effect involves the fact that if you are writing a technical computer science paper in Word then it is extremely likely that you are a very inexperienced researcher, possibly from outside of academia. That by itself makes it a lot less likely that your paper is well-written (aside from which software it is written in) and that the ideas in the paper are good enough to make it worthy of accepting. And the editor and referees know that, and are likely to take the paper less seriously as a result (this is the bias alluded to in other answers, and lamented by some), subconsciously and/or consciously, even if you are in fact an experienced researcher. This would again translate to a reduced chance of acceptance for a Word paper. Perhaps this bias is unfair, but there it is.
Even without the bias I mentioned above, for the reason explained in item 2 above I still predict a clear statistical correlation between the Word/LaTeX typesetting information and the quality of the paper (and therefore the chances of acceptance). This correlation of course does not imply causation; without the bias, it would not be the typesetting system that was used that caused the paper to be rejected, but the overall low quality of the paper. Nonetheless, the fact remains that by writing your paper in Word, you would be choosing to join a group of authors whose papers have a statistically lower rate of acceptance, even if your situation is somehow different than most of them.
As a final caveat, if your area of computer science has papers with very few mathematical formulas, then what I wrote above is probably still correct, but to a lesser extent, i.e., the chances of acceptance would be closer to equal, but I still predict a small advantage for LaTeX.
The final look and feel of your paper must look like it belongs in that journal, regardless of which software you use to produce it. I will, as a reviewer, subconsciously bias towards a reject if I find that the paper looks amateurish. This means it is an uphill climb to change my recommendation to an accept.
If you are comfortable using Latex (i.e. it won't slow you down to use it), and it is well-supported by the journal (i.e. they provide a style file), I would recommend it, as it produces a paper that is much easier for the referees to understand. Reducing any barrier to referee understanding is great!
But just because a journal accepts both Latex and Word, doesn't mean there isn't an internal preference. Sometimes you have to dig a little bit to find this. Look at the publishing details: some journals will waive part of the publication fees if you give them a Latex file; others state they will take longer to move from acceptance to publishing with Latex than Word.
Personally, I think that the advantages of Latex in terms of appearance and bibliography-handling are large enough that I put up with its other annoyances, and use it whenever possible.
No. Following submission instructions found in "Instructions to Authors", whatever they may say, will pretty much insure that your paper gets reviewed, and not whatever the digital equivalent of "returned unread" is, but reviewers generally see a pdf, and couldn't care less how you wrote the paper. They will care if you gave the manuscript the attention it merits, regardless of the tools you use to prepare it.
As an aside, when you choose to work in Latex, you need to think about how that interacts with any collaborators who will need to edit and mark up your documents. It's certainly not undoable to work with Latex nonusers, but it does require thought and planning.
Note: this answer comes from someone who does not review computer science articles. There seem to be a strong belief in your field that using LaTeX is The Way to go. I would suggest to comply for your career's sake. My answer still applies to other fields.
I truly hope it doesn't. It would be extremely unprofessional for reviewers and editors to favor their pet typesetting tool when assessing the content of a paper. Especially knowing that in many cases, the paper will be re-written from scratch by the typesetter and thus it becomes irrelevant.
But possibly, writing papers with LaTeX might reduce your research output, and thus the quality or number of the papers you submit since it's been observed* that using LaTeX is less efficient than MS Word to write academic papers (with the exception of a small subset of fields).
So, unless your document is made mostly of equations, data suggests you will make fewer errors and spend less time writing your paper using MS Word than LaTeX.
The area is computer science & engineering with few mathematical formula, diagrams, and tables
You need to use Latex. Full stop.
Even when you don't know Latex, you are supposed to learn basic math formula, figures, tables in less than a day. (Tables are pieces of cake with this online tool http://www.tablesgenerator.com/)
I never heard of anybody in Computer Science that does not know Latex. Call me a jerk and down vote if you like, but such a person will become a topic of a joke.
If one can't prepare a document in Latex, how can (s)he do any programming? And if they can't do programming, I can't take their Computer Science research very serious.
Everybody is talking about how Latex produces beautiful math equations etc, and forget how it is efficient even on basic tasks. I was once a Word user. In fact, I wrote my undergraduate thesis using Word 2003, and it was a painful experience, even though I didn't have any math equation in my thesis.
- Creating a Reference section, and citing a paper in Word is painful, and time consuming.
- Changing reference format from John Smith to J. SMITH (ACM style) is also painful, and it can take you days. Changing the citation from (Smith 2003) to  is also painful.
- Referring to a figure is also painful.
- Formatting (pseud-do) code is also painful.
- Keeping the format Heading1, Heading2 etc is really painful.
The list does not end here. All of these are pieces of cakes in Latex, and the most important thing is it gives you the feeling of total control of your text.
But possibly, writing papers with LaTeX might reduce your research output, and thus the quality or number of the papers you submit since it's been observed* that using LaTeX is less efficient than MS Word to write academic papers (with the exception of a small subset of fields)
I don't care if anybody/paper says preparing documents in Word is more efficient. It's like saying typing with two fingers are faster than typing with 10 fingers. It is true for some people, but it will become a joke for many people.