Why was this question about research on LaTeX considered too opinion-based?

Question

Recently my question about the validity of some research findings on LaTeX usage was placed on hold as "primarily opinion-based", however it is not clear why this question falls outside the general standards of the community, when other similar questions do not.

While there can certainly be multiple valid ways to answer the question any appropriate approach would need to be rooted in factual information about TeX and the specific findings of the study. The answers would not be pure speculation but the interpretation of specific information from the study in the light of specific information about TeX.

I can see no difference from this question and other open questions where answers could result from different approaches such as the following examples:

• How can I convert my TeX-illiterate coworkers to LaTeX?

• Advantage of LaTeX presentations?

• How can I explain the meaning of LaTeX to my grandma?

What about this question falls outside of the general community standards for questions that could have multiple valid answers?

Answer 1

We get a relatively small number of questions which may be viewed as off-topic or opinion based on the site. For these, there is a certain amount of discussion which can be had and as pointed out by percusee you can argue things are not entirely consistent. As we are talking about value judgements that is perhaps not surprising.

Looking at the question here, there are arguments for both 'off-topic' and 'opinion-based'. I will take the two arguments separately.

In terms of topic, the PLOS ONE article seeks to compare TeX (LaTeX) with a word processor (Word). Thus at least in part it's already about something we don't cover: word processors. However, the bigger issue here for me is that the issue your question is raises is

Based on this study can we conclude that Word works better than LaTeX?

That's not about TeX and it's not about Word: it's about measuring productivity, specifically from the study in an academic setting but presumably more generally. That's a technical area (I'm sure there are experts on this type of measurement) but it's not what we cover as a site.

On the 'opinion-based' side, this is about more than 'there might be more than one answer'. (While the StackExchange model is best for questions where a single answer can be 'more correct' than any others, there is certainly some room for more nuanced situations.) Asking for a critique of a paper comparing Word and LaTeX on a site about TeX is only going to go one way, but it's hardly going to be objective. (Something like the 'grandma' question is clearly a bit of fun, so people are likely to give them more leeway, whereas some of the promotion questions may be given more slack as they are not directly about criticising other approaches.) The entire thing seems much better suited to a threaded discussion format, which chat does better than a question but which is really better on a mailing list or on Usenet.

Over all, to me the two areas together suggest that this is not a good fit for the site: TeX-sx isn't the best place for everything about TeX. The same might be said of some other questions, and I'd be wary of saying 'As X is open, Y must be allowed too.'.

[Note: While my name is currently down as closing the question, this is because I was asked to 'tidy up' the close reason from 'off-topic' to 'opinion-based' as this seemed perhaps better. I did not vote to close initially.]

Answer 2

Well as I was involved a bit I can justify my take;

1. First and foremost, we are not consistent at about anything. Hence, why this is open but the other is closed is not going to solve anything. Half of the TikZ questions should have been closed but somehow didn't. The examples you give are also the same. So there is always a grappling aspect of any question. You might not like it (I also don't) but this is a social construct. Not a rule-based media. Moreover, as we discussed a lot of times earlier, the SO rules are self-conflicting so there is nothing to gain sticking to the rules that you might think appropriate.

2. Then the obvious follow-up is that why this one didn't fit the bill. I can't talk for others but in my point of view, the article is really weak to start with. Even without downloading the extra files and checking the numbers, you can tell that it is, in many places, biased and unforgivably misinformed about TeX. So opening a discussion in the question/answer format is just going to compile a lot of answers which are reinforcing the obvious stuff about TeX.

3. It has weird conclusions such as saving time and wasting money and that kind of nonsense that any academic can spot the logical error. This item is not about why question is closed but about an article however, in turn, I have to confess that it is being taken too seriously for such weak, and borderline unethical, conclusions.

4. Finally, even though the article is weak and erroneous, we can't argue anything against it because we don't have any counter evidence. Alan Munn already took the initiative and set up the chat room for experiment design. So that covers much more than what this question would cover had it stayed open. I would recommend that we push our efforts in that direction instead of typing long answers to the question and gaining nothing. So in that sense, the question is truly opinion-based and should not be mixed up with half-baked and in some cases fun questions you have sampled in your post.

Answer 3

I was also involved (as a reviewer). I'm neither statistician nor mathematician, so I can't comment on the stats or math involved. I am, however, a typesetter, and my research on editing systems for structured documents was cited.

What fails to come through properly in the article is that the real distinction is between the Word interface and the LaTeX interface[s]; and we know that most of the interfaces to LaTeX — even LyX — are not the same as Word, despite the evidence I found in my own research that the majority of computer users prefer a typographically synchronous interface (what people inaccurately call WYSIWYG).

In my experience, most authors would indeed probably work faster using Word (leaving aside mathematics, which the authors of the paper do acknowledge to be a special case). The reason seems to be that they just want to get the paper done, and they absolutely do not want to have to learn anything new to achieve that.

A few people, of course, do see the advantages of LaTeX, but even then, there is a strong tendency for them to use obsolete or unnecessary markup because they have been improperly instructed (or have resorted to guesswork), and they have not encountered the concepts of document structure and automation which lead to efficient use of LaTeX.

I have urged in a response on comp.text.tex that we (the [La]TeX) community should undertake some more research on this, and I am delighted to see Alan Munn's initiative.