Welcome to the Q&A Forum

That's a really tough one. For paginated search results, I would normally go with @128K's suggestion - let Google follow the search results, but don't index them. Unfortunately, these are both search results AND your content, which is a bit unusual.

I would still consider the META robots tag. Here's my argument - while all of these results are unique, landing on question #21 isn't very useful to search visitors, and they might bounce. It probably makes more sense to land them all on page #1. While you'll lose some indexable content, I suspect the higher conversion and loss of duplicates would offset that.

Rand had a good post about pagination, but again, it assumes a more typical search-results scenario (like a list of product snippets and titles):

http://www.seomoz.org/blog/pagination-best-practices-for-seo-user-experience

Likewise, the pages aren't duplicates, so I'd avoid the canonical tag, in the usual sense. There are viable AJAX approaches, but some will make the full content invisible to spiders.

There are SEOs who advocate using the canonical tag, but canonicalizing to a version of the page that displays ALL of the results on one page. That way, visitors still go through the list, but Google would see the full page of answers. You could also default Google to a different count of results/page (like 100) but then default visitors to 10, etc. It borders on cloaking, but it's off-white at worst, in my opinion.

I lean toward this one, too, but Google has no particular bias against dynamic pages. SEOs use to warn against using "dynamic" URLs, because they tended to include a lot of parameters and create duplicate content, but there's plenty of evidence that ".php" pages are perfectly fine.

To me, it boils down to:

(1) As TraxNYC said, all else being equal, shorter is better

(2) It looks a bit friendlier to search visitors (and easier to read)

(3) It doesn't instantly reveal your underlying platform (and pose a security risk)