7
$\begingroup$

Recently I wanted to mark a duplicate of a question about defining maximum likelihood, but was unable to find the "main" question since it used "MLE" in title rather then "maximum likelihood" (Maximum Likelihood Estimation (MLE) in layman terms). Finally someone merged the initial question, the one that I suggested as a duplicate, and the one that I refer to in the link.

If there exist (and they do) questions that are prone to be duplicated, like the one I refer to, wouldn't it be a good idea to slightly edit their titles to be more informative and easier to find via search? In this case I would edit "MLE" and changed it to "maximum likelihood (MLE)" however there were discussions on if it is good to make such edits or not.

What do you think? Of course, I understand that "minor edit" could be easily misunderstood and abused.

EDIT

OK, there doesn't seem to be disagreement in here so I edited this title.

| |
$\endgroup$
  • 3
    $\begingroup$ There is a typo in your title (or is it a pun/joke I am too dense to get....) $\endgroup$ – Andy W Feb 14 '15 at 13:06
  • 3
    $\begingroup$ Where an abbreviation is ambiguous (which is surprisingly often), it's pretty much an essential edit -- but it often makes sense to keep the abbreviation as well, or at least make sure it's present in the body text. $\endgroup$ – Glen_b Feb 16 '15 at 0:09
9
$\begingroup$

Although I also prefer shorter titles, I am less worried about their length than @amoeba. At any rate, I agree that you should go ahead and edit the title. The title and the tags are often what people have access to when searching / perusing the site. Getting them right has an enormous influence on the organization and usability of the site. Your own case is a great example of this; you couldn't find the thread you wanted—and knew was there—because the title wasn't quite right. In cases like that, I would always strongly suggest editing the title so that it is easier to find and accurately reflects what the thread is about.

| |
$\endgroup$
  • 1
    $\begingroup$ +1, I don't think there is any disagreement between us here. My point was that some abbreviations are so common (PCA? ANOVA?) and some titles are so long, that they would be better of without explicit spelling out of these abbreviations. $\endgroup$ – amoeba Feb 14 '15 at 16:09
  • 1
    $\begingroup$ Yeah, I wouldn't spell out ANOVA either, @amoeba. $\endgroup$ – gung - Reinstate Monica Feb 14 '15 at 16:15
7
$\begingroup$

I think in this particular case it would be a good edit, but mainly especially because the title "MLE in layman terms" is otherwise very short; there is some room to expand it. Go ahead and replace it with "Maximum likelihood estimation (MLE) in layman terms", why not!

But often Sometimes titles with abbreviations are quite long already, and then I would be reluctant to replace e.g. "PCA" by "principal component analysis (PCA)" -- and even occasionally make an opposite edit, for the sake of brevity. I do always try to make sure, however, that the abbreviation is spelled out in the question body (and occasionally edit solely for this purpose).

In this case it was ("MLE" was fully spelled out); so I am not sure I understand why you failed to find it via search. What did you search for?

| |
$\endgroup$
  • 4
    $\begingroup$ Specifically in this case, "ML "is often used here to mean "machine learning". So, to spell out the obvious, ambiguous abbreviations should certainly be expanded. One of my pet dislikes is "IV", which despite commonly meaning independent variable often also means instrumental variable (particularly to economists). $\endgroup$ – Nick Cox Feb 15 '15 at 0:53
  • 1
    $\begingroup$ Besides Principal Component(s) Analysis, I've seen PCA crop up in a statistical context used for each of the following: Partial Correlation Analysis, Percentage of Calls Answered, Prostate Cancer Antigen, Patient Controlled Analgesic, Prescribed Concentration of Alcohol (and likely others I don't recall at the moment). On the other hand, I think I've only seen "ANOVA" used to mean one thing, so ANOVA is pretty safe. Three letter acronyms (like PCA or GLM) may often be ambiguous, especially for some areas (a reader from the medical field for example, or from computing, may misconstrue PCA). $\endgroup$ – Glen_b Feb 16 '15 at 0:32
  • 1
    $\begingroup$ @Nick Indeed I just now came from expanding "ML" to "machine learning" in a title $\endgroup$ – Glen_b Feb 16 '15 at 0:33
  • 1
    $\begingroup$ GLM can mean general linear model or generalized linear model. Often the poster has in mind some syntax so named, but it's ambiguous without detail. $\endgroup$ – Nick Cox Feb 16 '15 at 0:37

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .