# Should references always be used in answers?

The best answers I've seen, on this site and on programmers, often include references to experts on refactoring, design patterns, and other computer science fields. I have learned a good bit about refactoring and design patterns and implement them myself, but I don't know many of the names for them. The best reference I could come up with if I do happen to know the name is Wikipedia, which I think you'll agree is not as strong as say, Knuth.

I often look at a question, think of an answer, then see the posted answer is the same answer as mine, but with references and experts to back it up. Should I not be as active until I have a better background in this area, or should I still try to answer questions?

## 2 Answers

Two things:

1. Answering questions is how you get "a better background in this area".
2. you can learn much by answering.

I find the process of writing an answer guides my thoughts to really take a problem apart - in the process I often think of things I should be googling.

Also you can always go back and edit your answer to refer to the "experts".

So in short: answer away.

• +1 Often what I do when I submit an answer. – Mark Loeser Jan 23 '11 at 2:32

Personally, I would make the decision based on the answer itself. Does the answer make sense without the references (Common-Sense wise at least)? If so, go ahead and post, and you can always back it up with references later if you would like. If it does not make sense without the references (it seems counter-intuitive, goes against popular thought or is just plain strange) then I would back it up by references from the get-go.

A pair of (simple) examples:

Straight Forward:

Q: is this code good: Some code block using eval and goto?

A: No, typically code that uses either eval or goto is considered bad practice because it makes testing harder, and is much harder to read, among other issues they cause.

Counter-Intuitive:

Q: If rand() is good, is rand() * rand() better?

A: No, it actually makes the output much more predictable (from a statistical standpoint at least).

While it answers the question, unless you have a strong background in math you likely won't understand it (or it'll seem wrong since there's "twice as much random" in the formula).