# Why are short answers considered to be poor quality for this site?

I have written some answers that are short, concise and accurate. But I get a message that indicates that my answer is of poor quality. I would appreciate hearing some thoughts on this.

• Please visit stats.stackexchange.com/help/how-to-answer and read it carefully, because it fully addresses the reasons for those messages. – whuber Feb 12 '17 at 18:22
• That was not my point. I have enough experience on this site to know how to write a good answer. Please don't insult my intelligence. – Michael R. Chernick Feb 12 '17 at 19:22
• I don't know which answers you mean, but just for the sake of example here is your currently second one: stats.stackexchange.com/a/261378. All it says is: "The result always holds when X and Y form a bivariate normal distribution. You will find this in most multivariate analysis texts as well as on some threads on this site." I find it exemplary unhelpful. You are stating that something always holds without explaining why and you claim that it's a duplicate without providing a link. You might have "enough experience" as you told @whuber but this is not a good answer. Just an example. – amoeba Feb 12 '17 at 20:11
• Michael, I apologize, because I did not intend any insult. You underestimate me: I can do a much better job of insulting people than that! I was only assuming that somehow you have overlooked this information and that you would receive a suggestion about where to find it as being a helpful and friendly contribution. Since you do know how to write a good answer--to date, you have posted 35 of them with 10+ net upvotes, out of 1246 answers--the real question on the table is: why don't you?? – whuber Feb 12 '17 at 20:33
• There's no sarcasm there, Michael. My question "why don't you" is an honest one. It has been puzzling users of this site--many of whom have not defended you as strongly as I have. Why do you persist in flaunting site norms by posting so many comments as answers? How do we avoid the perception that the primary purpose of your activities here is only to draw attention to your previous publications? What can we say that will change your behavior so that your contributions take a form that is consistently seen as valuable and constructive?? – whuber Feb 12 '17 at 21:04
• I should explain that if I seem to exhibit some frustration, it is not personal: it is because as a moderator, I have had to field complaints from long-time users. They note your credentials, they see the value and insight in some of the answers you post, but universally they bemoan the missed opportunities and they ask why you post so many poor, abbreviated, incomplete, and unexplained answers. It puzzles them and now it puzzles me, because my stock response--he doesn't yet know the site--just won't fly anymore. – whuber Feb 12 '17 at 21:13
• Short answers are often considered poor quality, because they tend to lack sufficient attention to detail, i.e. leaving assumptions unstated / overstated, not calling attention to edge cases, not fully answering the question. It's sometimes possible to give a terse, correct, and complete answer, but the opportunities to do so are rare. – Matthew Drury Feb 13 '17 at 2:03
• @whuber, with all due respect, I find this censure of Michael Chernick to be inappropriately harsh. I often read Michael's answers and I don't disagree that the many are incomplete, relegate key material to references, or have errors. But, I also think Michael is held to a different standard than others, perhaps due to his sometimes brash tone in comments. At least one very high rep user (who happens to be a moderator) routinely gives vague, incomplete, or duplicate/repetitive answers. He also has a similar rate of 10+ vote answers (~3%). I never see this kind of contempt directed at him. – gammer Feb 13 '17 at 4:12
• This mixes two quite different questions. The general question is easy to answer as the grounds are already documented. The specific question on why Michael gets this feedback sometimes is trickier and immensely more difficult to discuss without people getting upset or angry. I don't think it helps at all to infer or impute personal animus or drop dark hints about third parties. If anyone writes poor answers, downvoting, flagging, comments encouraging developing an answer and just posting a better answer are ways to respond in specific threads. – Nick Cox Feb 13 '17 at 9:13
• @gammer: No-one's expressed "disgust" or "contempt" for anyone, or their behaviour - I don't know where you're getting that from. Michael Chernick isn't being "held to a different standard": he's answered 11 questions over the last two days (including several now deleted or converted to comments), writing an average of about 31 words for each (initially - one he later edited to a fuller answer); no other users approach this combination of rate & brevity - & he's the one asking this question. – Scortchi - Reinstate Monica Feb 13 '17 at 12:11
• Regarding "I don't know where you're getting that". Yes, whuber was professional, but the underlying message -- "Look, your answers are terrible. I know you're smart so why are your answers so terrible? Is it just because you're a jerk ("flout site norms.."), or is your main purpose here just to brag about your prior academic successes? Everyone else here agrees with me. Trust me, I know. They complain to me all the time and ask why I, as a moderator, put up with it" -- could certainly be described in the terms I gave. Of course, it's subjective, but I don't think my reading was ridiculous. – gammer Feb 13 '17 at 13:07
• Michael: The reply to that is exactly the same here on Meta as on the main site. If your question is being misunderstood, do edit it to make it clear. It's three sentences long as it stands. – Nick Cox Feb 13 '17 at 13:16
• I hate to lengthen a long comment thread, but I simply cannot go without responding to the scurrilous comments posted by @gammer. What is most offensive about them is the mis-characterization beginning with "Look, your answers are terrible... " Not only does that stretch what has been written far beyond anything real or intended, it appears designed to hurt Michael. From his first appearance almost five years ago I have been grateful for Michael's participation and have worked hard to make the most of what he so graciously offers us. – whuber Feb 13 '17 at 14:19
• @gammer: It's clearly ridiculous: you've had to resort to re-wording & distorting what whuber wrote in order for it to evince characteristics you claim were originally present. If I were to allow myself the same latitude in interpreting your comments as you allow yourself in interpreting those of others, I'd conclude you were positively trying to sow discord. For my part I'm glad to see Michael back on CV, appreciate his contributions, & am sure he understands that differences of opinion on the best way to contribute are only that. – Scortchi - Reinstate Monica Feb 13 '17 at 17:30
• Thank you Scortchi. I very much appreciate you comments. – Michael R. Chernick Feb 13 '17 at 17:51

As someone who knows little about statistics I try to contribute here on Meta to repay what I learn on the main site.

Looking only on your newest answer: I did not learn much from it. It is just pointing towards external references, failing to be a self-contained answer. You mention a paper of yours, but do not provide a reference; how should anyone know which paper you are talking about?

I have written some answers that are short, concise and accurate.

To me that answer is not concise, because it leaves a gap between the question and what the answer might be. Therefore, some users will understand that type of answer as 'very-low-quality' and will flag it for peer-review.

I don't know how the process of hanging notices work, but I guess if the answer is peer reviewed as low quality, it is that (notice) or deletion.

Now, on a more subjective ground; because you are a high reputation user, and have shown yourself to know statistics, the expectations on your participation are high as a matter of quality and didactics, to me, at least.

• Thank you very much Andre. I always appreciate your help. You must have been an English major. – Michael R. Chernick Feb 13 '17 at 12:46
• +1 but I am pretty sure that the message We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed. that is currently displayed on a yellowish plaque below the answer is automatic, i.e. is triggered by some background checks, and is not due to flagging. – amoeba Feb 13 '17 at 13:52
• @amoeba: No, the post notice is placed there by moderators, who decide if the flag is justified. – Scortchi - Reinstate Monica Feb 13 '17 at 13:58
• @Scortchi Oh! Thanks. I did not realize that. – amoeba Feb 13 '17 at 14:00
• For my part, I have long recognized your efforts here on meta.CV, & I appreciate them. – gung - Reinstate Monica Feb 13 '17 at 17:02
• (+1) Looking at the linked answer, as an experiment I tried to quickly just add links to the implied citations (@MichaelChernick I believe these are the correct ones?). I am curious: Do people feel that this relatively simple and objective change is a significant improvement? – GeoMatt22 Feb 13 '17 at 18:04
• @GeoMatt22, thanks for the update. My opinion is that such edit improves the answer, but not to the point to make it self-contained. – Andre Silva Feb 13 '17 at 18:21
• @GeoMatt22, "significant improvement" is ambiguous to me. I certainly think it is an improvement. However, I would still say the post is a comment, not an answer. To paraphrase, the post says, 'you can find the answer to your question here &/or here'. The post doesn't quite contain the answer to the OP's question. It should still be a comment. – gung - Reinstate Monica Feb 13 '17 at 19:37
• +1. "I did not learn much from it" is an excellent test to apply to any putative answer. "Learn" should encompass more than "knowing where to go to learn": that is, a mere reference teaches us little. In addition, I would like to point out that "short, concise, and accurate" (SCA) do not suffice: one must add relevant to that list. I have downvoted SCA answers because they were not relevant and, when I pointed that out to their poster, those answers were never modified to make them relevant. Correct and irrelevant risks misinterpretation and so is little better than wrong. – whuber Feb 15 '17 at 14:21

Note that answers should explain why they're correct; they should give the arguments in any references not simply offer a link or reference (even when a question is a request for a reference, it should explain the reason for the recommendation).

Very short answers often fail to explain anything and quite a few amount to little more than pointing elsewhere.

Quite brief answers are on occasion acceptable, but if they're very short they need to be as complete as possible (that is, there's really nothing more that could be said). Since often very short answers are not suitable answers by Stack Exchange's criteria, they are automatically flagged by the system.

• I agree. The very fact that the SE system is programmed to automatically flag very short answers as likely to be low quality speaks volumes in itself. – gung - Reinstate Monica Feb 14 '17 at 16:25
• @gung I disagree. – Michael R. Chernick Mar 5 '17 at 5:34
• @Michael Which part of what gung said do you disagree with, and on what basis? Simply stating you disagree isn't of use (it won't serve to convince anyone if you can offer no basis for disagreement). – Glen_b Mar 5 '17 at 6:21