Question quality - thoughts?

[nickloman]

I might be pre-empting Zigster on this but I am increasing seeing a large number of ill-thought-out, low-quality questions on the board, particularly in the Bioinformatics category.

I realise this is par for the course on the Internet but can we as a community think of a way of improving matters. We need to discourage questions being asked which can be rapidly solved by methods such as a) a simple Google search b) reading back through old threads c) reading the software documentation or perhaps even contacting the author of the software in question.

Also, I hesitate to bring this up but a little bit of thanks from OPs for answering their questions wouldn't go amiss. I realise we have all sorts of languages and cultures represented here but regular contributors give up their own time to answer questions with little reward. A simple thank you goes a long way!

This is in no way a criticism of ECO or SeqAnswers as a forum (I love you both), more an attempt to kick off discussion about how the board can continue to improve without the signal:noise ratio deteriorating!

Do we need better policies? More moderation? What?

Cheers

Nick

[nilshomer]

I might be pre-empting Zigster on this but I am increasing seeing a large number of ill-thought-out, low-quality questions on the board, particularly in the Bioinformatics category.

I realise this is par for the course on the Internet but can we as a community think of a way of improving matters. We need to discourage questions being asked which can be rapidly solved by methods such as a) a simple Google search b) reading back through old threads c) reading the software documentation or perhaps even contacting the author of the software in question.

Also, I hesitate to bring this up but a little bit of thanks from OPs for answering their questions wouldn't go amiss. I realise we have all sorts of languages and cultures represented here but regular contributors give up their own time to answer questions with little reward. A simple thank you goes a long way!

This is in no way a criticism of ECO or SeqAnswers as a forum (I love you both), more an attempt to kick off discussion about how the board can continue to improve without the signal:noise ratio deteriorating!

Do we need better policies? More moderation? What?

Cheers

Nick

I agree with that assessment, but it is very difficult for someone to break into NGS analysis without having a seasoned NGS expert a few lab-benches/offices/seats away. Also, I would liken NGS to the GNU/Linux community, which has had no success with keeping users by saying RTFM. That being said, maybe the forum could be reorganized to have beginner/expert/power-user sections? Beginner forum(s) would have questions like "how do I get cufflinks to work?" or "what is a paired end?", while the power-user forum would ask more advanced questions.

This is indeed a useful discussion, one which should not be ignored.

[lh3]

Maybe we need a sticky FAQ (maybe wiki is better) on general questions like: "what is a paired end read?" or "what is fastq/quality?", although this may not help software-specific questions. As to c), I think it would be good to post some software related questions before asking the developer. In this way, this makes other users to see the answers and frees the developer from many similar questions because more people are looking at the forum than in each individual mailing list.

[ECO]

I agree guys. I'll post more thoughts later. But I also don't want to overtly cater to people who can't use "The Google".

I'm not sure we should sink to a FAQ for "what is a paired read" or "what is FASTQ", I do like the beginners section, where any level of question goes. But I agree, it seems like a lot of people are using the Bioinf. forum as their own personal research service...we could just move the beginner threads to the newbie forum...

...Definitely worth more discussion.

[spenthil]

Some things I have found to work in other communities to various degrees of success:
1. A *short* set of instructions on message board etiquette.
2. A way of marking question oriented posts as answered.
3. The biggest one - better search. For example, use the vBulletin tools and search "paired end reads" - good luck. On the flip side, google and you get the answer right away. Integrating it with vBulletin seems possible.

I have personally found FAQs annoying - maintaining them is a pain, and I can only imagine such a highly technical and evolving subject would make this worse.

As a newbie myself, I can't imagine myself figuring out the technical complexity of a lot of questions. This puts the burden back on board members (specifically admins) to reorganize improperly categorized questions. It also encourages duplication between the levels, not really solving the problem.

[ECO]

Google...."what is a paired read site:seqanswers.com", works pretty well...but I agree.

[spenthil]

Google...."what is a paired read site:seqanswers.com", works pretty well...but I agree.

Click on my "google" link above, it does exactly that (don't challenge my google'fu) :P. My point is (which I did not make clear at all), I bet a lot of members aren't aware of the "site:" keyword - I feel integrating that/educating them would go a long way.

[ECO]

Agree agree. It's straightforward to enable Google search directly on the forums, will put that on the list.

[Joann]

Basics

Why not have a separate "Basics in Bioinformatics" forum to log the "paired end" type questions specially open to beginners and newbies who might then also want a chance to try providing answers among themselves, (thus raising their posting counts). This may lead to a collection of

1) frequently answered questions helping to shape discussion in introductory course materials

2) a set of interesting, useful, active and recommended links to related material of all types that might be usefully cited in course work papers by non-major undergrads, MBA candidates and others interested in NGS technology

even

3) technical clarifications as in "after having read the manual..."?

[krobison]

What I find more depressing are the cases where someone is already sitting on data & now asking how they should analyze it without stating what they are looking for. Sequencing is seen as so powerful that it apparently obviates the need to design your experiments or have hypotheses to test.

[ECO]

Any volunteers for more moderators? If we went with a Basics in Bioinformatics/Seq Analysis forum...I'd need to rely on more than just myself to report and/or move threads there.

[flxlex]

@eco: what is a moderator's job description, and do you have a workload estimate?

[ECO]

@eco: what is a moderator's job description, and do you have a workload estimate?

Moderator duties include keeping discussions organized in their proper forum, closes redundant threads and reminds folks to search (which I admittedly need to do more often), deletes spam, bans spammers...basically improving the signal-to-noise of the forums wherever possible.

The workload is definitely whatever you can handle. It would be great if you visited at least a couple times a week.

If you (or anyone reading) can't commit to that, it's almost the same amount of work to simply use the "Report Post" function which will send the thread to all the moderators for direct and decisive action ( ).

Thanks for asking.

[Zigster]

I sense the anonymity of usernames is driving some of the careless posting, although it might encourage beginners to ask questions without fear of embarrassment..

Not sure what the capabilities of vbulletin are but perhaps you should be required to use your real name to start new threads?

there is some precedence for this:

http://www.vbulletin.com/forum/archive/index.php/t-195570.html

[GW_OK]

Real names will work until someone types in a fake name.

How about a 1 or 2 day account-creation to able-to-post (or create-new-post) waiting period.

(Or a you must have replied to x posts to create your own post threshold. Be able to join conversations before starting a new one)