4

Before using the Psychology and Neuroscience stackexchange I was more familiar with the MathOverflow which is a more mature member of the stackexchange community. This platform is used by professional mathematicians that practice both pure and applied mathematics. Furthermore, on the MathOverflow it isn't uncommon and it isn't considered unproductive to ask questions which involve a big list.

Consider the following examples:

  1. Big list of repositories of mathematical preprints and postprints
  2. Proofs without words

Now, I recently asked a specific question on the existence of good neuroscience research blogs. I gave concrete examples in the case of AI to clarify what I meant, distill being an example. After a couple days without an answer I decided to share the question with a neuroscientist I have communicated with in the past, Konrad Kording.

Dr. Kording, runs a well-known neuroscience lab and isn't merely providing a list of opinions. If you check each of the blogs in his answer they are all neuroscience research blogs. I must add that the first link to https://bitsandbrains.io/ fits the question description in particular.

Consider these posts taken from bitsandbrains:

  1. Some thoughts on building interdisciplinary biological/data science teams
  2. Bilateral Homology
  3. Connectome Coding: Discovering the Missing Link between Genotypes and Phenotypes

This leads me to the following questions:

  1. Before downvoting the question, was any detailed consideration given to the answer of Dr. Kording?
  2. Is my question, which I discussed with other neuroscientists, imprecise?
9
  • 3
    Thanks. You have to realize our site is haunted by poor Qs and broad questions are one of the culprits. The Q hasn't been close-voted though so I guess there's no big deal here anyway. Qs are downvoted en mass across the SE network, including mine when I enter a new site for the first time. Every site is different in many ways. Having said that, your Q seems well researched and I would reckon the Q to be valid (I +1d, for what it's worth).
    – AliceD Mod
    Commented Aug 19, 2019 at 18:39
  • 2
    @AliceD I am trying to encourage some of my neuroscientist friends to join. psychology.stackexchange.com/questions/13232/… was answered by a colleague, Christian Jarvers, after I shared it with him and a few other neuroscientists.
    – user3503
    Commented Aug 19, 2019 at 18:42
  • 1
    I'd say that I'm more concerned that neuroscientists I share the site with end up being discouraged and abandon this SE site.
    – user3503
    Commented Aug 19, 2019 at 18:43
  • Agreed with @AliceD. An example on this site is: psychology.stackexchange.com/q/9149/7001. I would generally recommend avoiding subjective criteria such as "good", "interesting", or "important" (though the example I gave uses "interesting"); specify more objective criteria in the question content instead.
    – Arnon Weinberg Mod
    Commented Aug 20, 2019 at 6:09
  • 1
    I think it is essential to keep 'list' questions "Not overly broad". You do exactly that by being quite detailed in the type of blogs you are interested in. Good question! Welcome to this stack!
    – Steven Jeuris Mod
    Commented Aug 20, 2019 at 8:50
  • Thank you Steven. I appreciate the efficient process for addressing this issue.
    – user3503
    Commented Aug 20, 2019 at 9:00
  • @StevenJeuris I like your answer on the other meta question, but the one answer on the question in question does not explain why/how, is short, and doesn't really share experiences and I find it hard to believe that the question will attract good answers by the standards you laid out.
    – StrongBad
    Commented Aug 20, 2019 at 15:52
  • 1
    @AidanRocke I think a better way to attract people and not have them be discouraged is to ask "research level" questions that the people you are talking to would be excited to answer instead of showing them the weakest side of the SE system.
    – StrongBad
    Commented Aug 20, 2019 at 15:53
  • @AliceD I have added an addendum to my answer in order to address the points raised by StrongBad, Bryan Krause and yourself.
    – user3503
    Commented Aug 21, 2019 at 0:14

3 Answers 3

4

Like StrongBad mentioned in a comment on the original question, questions where any answer is correct are not a good fit for SE.

In fact, in the help under "Don't Ask" the very first item is:

To prevent your question from being flagged and possibly removed, avoid asking subjective questions where …

every answer is equally valid: “What’s your favorite ______?”

Different sites have different levels of permissiveness of certain questions that don't follow this rule, but that's probably where the downvotes come from.

2

As I said in my comment explaining my down vote and VTC, while the question is clear and limited in scope, I think every answer will be equally correct. The example I gave was consider the following two answers

Alice's blog cover neurotransmitters in depth. I really like her post on GABA and anxiety

and

Bob's blog covers deep learning at an introductory level. I really like his post on using PyTorch for object identification.

I think both would be valid answers, but how do we judge which is "correct". This is a clear cut example of a don't ask question.

Now of course we can make exceptions, but I am always a little hesitant about requests for exceptions from relatively new users. I also think justification based on practices at MO are not good since they are historically different from the rest of the SE system. Of course other sites in the SE system allow big list questions. In fact, I recently asked about creating a big list question on a site I moderate.

In summary. I think big list type questions should be the exception and not the rule. We should make a case-by-case determination for each request and we should attempt to come to a consensus regarding what the answers will look like prior to asking the question. If we choose to make a big list question, I think it should be made community wiki and the first comment shold link back to a meta discussion.

1
0

To be fair, I understand where users like Bryan Krause and StrongBad are coming from. Unlike the MathOverflow or the Theoretical Computer Science stackexchange the average quality of most questions on this site isn't research-level. I am hopeful this will change and I am actively trying to convince friends of mine that are neuroscientists and cognitive scientists to contribute here. However, I think we should avoid categorically downvoting 'big-list' questions because they can be useful if worded carefully.

I worded my question carefully. The term 'best' and 'favorite' were consciously avoided and I laid out a specific set of criteria to qualify 'good research blog'. It has to cover:

  • methods for experimental, computational and theoretical neuroscience
  • comprehensive reviews of basic neuroscience research

To provide a lower-bound on the quality I expect I gave concrete examples. Distill being the ideal I had in mind.

Finally, I must note that Memming Park who runs a lab at Stony Brook has previously made a case for the value of such questions. (He uses an example from the MathOverflow to back his point.)

Addendum:

From what I can gather from AliceD, Bryan Krause and StrongBad I understand that there is a fear that on this forum a big list type question would invite spammy answers. I must clarify that from the beginning I wanted to know whether an online journal like distill existed for neuroscience research. I passed this question to both Konrad Kording and Adam Marblestone who both have a very broad understanding of neuroscience research.

Now, to avoid the 'big list' conundrum I decided to add the reference request tag. I am indeed looking for a particular kind of reference and what you can infer from Konrad's answer is that an online journal like distill doesn't exist for neuroscience research.

I think this should allow me to accept Konrad's answer and also avoids the risk of spammy answers.

4
  • 2
    If you are looking for a journal like Distill, that doesn't seem like a big list question, because I just don't think there are going to be that many. That is a very different question, however, from what you have asked and if you want to ask that, you should point out the features of Distill you want to see.
    – StrongBad
    Commented Aug 21, 2019 at 13:40
  • 2
    Also note, I am not trying to attack you or discourage you or other to participate here. If the community decides that your question, or big lists questions in general, are on topic, great. Regardless of the outcome, I hope you will continue to participate in the community and reach out to colleagues to try and grow the community.
    – StrongBad
    Commented Aug 21, 2019 at 13:42
  • "Now, to avoid the 'big list' conundrum I decided to add the reference request tag." I feel this is a misuse of the reference-request. From the description: "It should NOT be applied merely because answers with references are desired. "
    – Steven Jeuris Mod
    Commented Aug 23, 2019 at 11:25
  • Ok. In this case I will remove it.
    – user3503
    Commented Aug 23, 2019 at 11:34