• Sampling the Slyme

    Random sampling is one of the most powerful epistemic tools ever developed by humankind. It provides us with the ability to gain statistical insight into an entire population without the massive expenditures  involved in conducting a full census.

    On a recent post, a fellow calling himself “Steersman” (sounds a bit pretentious to me, but hey, it’s an internet pseudonym) suggested that I was being unfair in the way that I have been sampling the Pit, picking out the worst bits and drawing attention to them. In the interests of statistics and fairness, then, let us sample the pit randomly.

    Lsuoma’s version of the Pit has been running since 7/3/2012 at 8:58pm which is exactly 460 days ago as of the time of this posting. This new Pit has been operational for 11,040 hours, or 662,400 minutes. If only there were some way that we could randomly slice up that time frame using a truly random web-based process.

    What I’ve done here below is convert the list of numbers generated by Chibbi into their corresponding posts in “the endless thread” which is the central axis of slymy discussion:

    Jul 18, 2012 4:56 pm  
    Jul 23, 2012 12:53 pm
    Aug 05, 2012 1:42 am
    Aug 23, 2012 1:35 am
    Sep 02, 2012 5:10 pm
    Sep 19, 2012 8:23 pm
    Sep 24, 2012 12:14 pm
    Oct 21, 2012 9:56 pm
    Nov 04, 2012 3:00 pm
    Nov 20, 2012 6:33 pm
    Nov 23, 2012 6:52 am
    Dec 01, 2012 7:00 pm
    Dec 21, 2012 9:54 am
    Jan 21, 2013 4:58 am
    Mar 10, 2013 3:54 am
    Mar 23, 2013 1:12 pm
    Mar 23, 2013 7:22 pm
    Apr 16, 2013 3:45 pm
    May 04, 2013 8:43 am
    May 07, 2013 2:15 am
    May 27, 2013 6:06 am
    Jun 06, 2013 11:00 pm
    Jun 23, 2013 7:11 am
    Jul 06, 2013 10:36 am
    Aug 14, 2013 7:49 am
    Aug 16, 2013 1:06 am
    Aug 30, 2013 2:28 am
    Sep 12, 2013 2:59 pm
    Sep 24, 2013 12:46 am
    Sep 29, 2013 4:28 am

     

    The process was very simple, each of Chibbi’s numbers represents the number of minutes since the founding of the forum, and I simply linked to the next post after that time had elapsed.

    Whereas fans of the Pit will tend to judge it by its most brilliant posts and posters, and critics (such as myself) will tend to judge it by its worst posts and posters, the process of random sampling eliminates such biases and provides us with an unbiased representative sample of the endless thread from its foundation until today.

    I’m not going to provide my own subjective evaluation of the content of these links right here and now. Instead, I’m going to leave the links up for those who want a fair sample to peruse at their own leisure, or to share with critics or supporters of the forum. If you want to avoid positive-spin cherry-picking or negative-spin nightshade-gathering in favour of a truly unbiased and randomized approach, this is the probably your best bet.

    Should the content from list of links change your own impression of the Pit, whether for good or ill, please feel free to leave your thoughts in the comments.

    Category: Damned Lies and StatisticsSlymePit

    Article by: Damion Reinhardt

    Former fundie finds freethought fairly fab.

    One Pingback/Trackback

    • whatever

      http://vimeo.com/44941848

      Get a life you fucking stalker.

    • Axel Blaster

      The Random sampling solves any selection bias that Steersman could have objected to. But, the biggest interpretation problem remains. Those post are context dependent. Therefore the gaps will be filled with preconceived notions or random interpretations. You are applying math to human variables that can’t even be predicted, IMHO.

      Consider these two questions:

      1. How is this different than opening the bible at random, to know what God wants to tell you?

      2. Since this is mostly a worldview issue, do you believe that the impressions about the merits or flaws of a forum can be settled satisfactorily, when folks have been fighting for more than a century about the words and attitudes in The Adventures of Huckleberry Finn.

      Worldview discussions rarely end with reasoned arguments. People die first. These discussions reflect personal values and arguments are only used to self-deluded oneself that there is any reason behind. Nobody wants to admit that their argument is not backed by hard logic and flawless reasoning.

      I’m beginning to think that Hugo Mercier’s Theory could be the right one:

      http://www.sjsu.edu/people/anand.vaidya/courses/c5/s2/Why%20Do%20Humans%20Reason%20Sperber.pdf

      Brings to mind TF00t, recently.

      • How is this different than opening the bible at random, to know what God wants to tell you?

        Well, it’s an exercise in critical evaluation rather than blind obedience, so there is that.

        I read through the entire Bible back in 2011 so I have a pretty good idea what would happen it you randomly sampled 30 chapters out of the entire book. You would get mostly boring stories rather than the outstanding aphorisms of the Jefferson Bible or the really nasty verses that atheists usually pass around as memes.

      • Axel Blaster

        exactly.

      • Since this is mostly a worldview issue, do you believe that the impressions about the merits or flaws of a forum can be settled satisfactorily, when folks have been fighting for more than a century about the words and attitudes in The Adventures of Huckleberry Finn.

        No, I do not believe a fully satisfactory settlement can be reached.

        I do believe, however, that reading through a random sample of any given forum provides a much better idea of what usually goes on there than reading through a collection of the very best (or the very worst) which is precisely how this issue is usually pursued. I believe that we can add just a little bit of fairness into our subjective approach.

    • Pogsurf

      You’ve chosen a cake, sliced it into beautifully even slices and told us you don’t like this particular cake. Haven’t you forgotten to tell us why you don’t like this cake?

      Can’t see this thing getting off the ground unless someone supplies an actual proposition to test.

      • kiiski

        The cake is a lie!

      • ThePrussian

        Dunno if this should get applause of banning

      • Forgotten? You need to read more closely my friend.

        I’m not going to provide my own subjective evaluation of the content of these links right here and now. Instead, I’m going to leave the links up for those who want a fair sample to peruse at their own leisure, or to share with critics or supporters of the forum. If you want to avoid positive-spin cherry-picking or negative-spin nightshade-gathering in favour of a truly unbiased and randomized approach, this is the probably your best bet.

        Of course I have to agree that specific propositions should be tested. Can you think of any worth testing?

      • Pogsurf

        I’d like to test the criticism of the Pit which is in your head, but if you won’t make it explicit (and testable) then that is going to be difficult.

      • Firstly, I want to make it clear why I would choose not to include any of my own subjective criticisms in a post dedicated to the idea of giving the Pit a fair shake of the dice. In the future, I will refer back to this post when asking people to test specific propositions about the Pit, such as the oft-repeated claim that the Pit is a hate site.

      • Pogsurf

        “Whereas fans of the Pit will tend to judge it by its most brilliant posts and posters, and critics (such as myself) will tend to judge it by its worst posts and posters, the process of random sampling eliminates such biases and provides us with an unbiased representative sample of the endless thread from its foundation until today.”

        You admit here that you are critical of the pit, and that your bias towards criticism skews your overall view. You also believe that fans do the opposite thing (just take the best bits and judge it on that), but you don’t show your workings, so we cannot tell how this belief was arrived at.

        You have also omitted to include the possibility that there may be people who try to take a rounded view, where they consider both the best bits and the worst bits and try to consider the Pit on balance.

        Without being willing to properly explore your own biases I don’t believe you can be quite as good an honest broker as you are trying to be.

      • This is all true, but it doesn’t make a case for exploring those biases here and now when the project is trying to create a fair sample.

    • John Greg

      As has been pointed out, your random sample fails completely because it leaves out all context. You simply cannot generalize a community via its commentary if you have no context for each individual comment that you have linked to. As Axel quite accurately pointed out:

      “Those post are context dependent. Therefore the gaps will be filled with
      preconceived notions or random interpretations. You are applying math
      to human variables that can’t even be predicted.”

      • Did you even look at the links to see how many of them require context (and whether that context is easy to find upthread) as opposed to how many are self explanatory on their face?

      • John Greg

        I’ve checked out eight or nine, or so. Of that grouping, I felt that only one could, with a squint and a wink, stand on its own without having to find the preceding context.

        However, that is irrelevant. You state that random sampling (clearly, you intend to explicitly include yours as well) “… provides us with the ability to gain statistical insight into an entire population….” But in this instance it does no such thing. In dealing with this form of content — a long forum with several hundred participants covering some 130,000 comments — you MUST have context, detailed and specific context.

        Your basic argument is that your random sample provides enough data to make a generalised determination of the character of ALL pitters. And that’s just nonsense. Axel put it better than I, but you seem to have ignored most of his comment to focus on his least important statement.

        As to your comment following this, if you really think that those first two simplistic juvenile diminutives warrant deletion of the entire post, well, I am gobsmacked. Such childish intolerance; such Orwellian language policing. And if you are one of those people who think a simple diminutive liike the third one is insulting and warrants deletion, well, words fail me.

        “… at Skeptic Ink we discuss ideas rather than attacking personalities.”

        Poppycock. Pure poppycock.

      • Your basic argument is that your random sample provides enough data to make a generalised determination of the character of ALL pitters

        Where are you getting that from, exactly?

        Such childish intolerance; such Orwellian language policing.

        If you don’t like it, you can go back to the Pit and say whatever you like. I’m looking for a certain quality of commentary here, and if you cannot refrain from pointless childish insults, you are not what I am looking for.

      • John,

        I deleted your other recent post.

        Insulting diminutives such as “Peezus, InZvanity, and Ophie” are not welcome here. I have rules, and you should read them before commenting again. This is not the Pit, at Skeptic Ink we discuss ideas rather than attacking personalities. Comments which insult people, especially people I don’t happen to admire, will be deleted.

    • kiiski

      In the earlier discussion on PZ Myers and hypocrisy, you commendably restricted the admissible evidence to things PZ himself has said, rather than what people he associates with have said. Why, then, should another site be evaluated as a collective?

      • That is an excellent question. Individuals should be judged by their own words, of course, but websites which generate content collectively should be judged by the quality of that content. There is a world of difference between evaluating http://www.slymepit.com and judging any of the individuals who have ever posted there.

      • I agree. I’ve tried to be specific in my criticisms re the pit. I also have used the URL slymepit.com, and referred to it as a message forum, rather than as a vague collection of people. I also try to keep my critiques about *behaviour* and words/ideas. That would be the most productive thing to focus on here, IMO. Quantifiable, observable things. Avoiding ‘mind reading’ and other such psychological fallacies.

        Specifically, if you are trying to make a case that something going on at the pit is literally *harmful*, I think it’s your responsibility to specify exactly what that harm is, and to present some usable criterion for measuring this effect. That is, if you’re trying to make a factual claim, rather than just stating your own opinions.

        On the other hand, I don’t think it’s absolutely *necessary* to make some statement about ‘harm’ before presenting a critique; I’m just not sure at what level you are trying to make some case for this that or the other. Again, it would be really helpful if you got down to concrete specifics and tangibles. IMO.

    • This is the right approach in the long-run, i.e. to do a proper statistical analysis, rather than to simply ‘frame’ things in one’s own preconceived biases. And I do commend you from refraining to draw conclusions in this post.

      There are some issues I would like to raise, however. The most important, I think, is that you should specify what hypotheses you’re testing *before* you sample. The above sampling is fine, but you’ve already ‘spoiled it’ by sampling before making your predictions. I think it would be a mistake to ‘refer back to this post’ in the future to test specific hypotheses. Instead, each test should perform a new sampling (which you’ve nicely shown is easy enough to do, so there shouldn’t be any problem with that). Otherwise, you risk over-fitting the hypotheses to the pre-selected data and hence biasing them simply by treating a single random sample as some sort of eternal archetypal set. However, this is not a big problem to overcome; just do a fresh sample each time you want to test some proposition.

      As for some people’s critique of ‘requiring context’, though this is true in general, it wouldn’t stop one from testing *some* hypotheses. It just depends on the hypotheses, which is another reason the hyps should be selected for testing before the sampling is done.

      I think it would be best to start with very simple, basic hypotheses, see how well this method works to identify issues, and build from there, if it is successful, with further refined hyps in the future. Don’t have any example hyps off the top of my head, but as Axel Blaster mentioned, if you make too-complex hypotheses too early, there will probably be too much ambiguity in interpretation to make a convincing point one way or the other about anything.

      Also, while I do commend you not jumping to conclusions in this post, I do also agree with pogsurf’s point that you *really should* explicitly state your hypotheses in concrete words and stop with the innuendo and smear-ish tactics. So, I think if you’re interested in pursuing this investigation in an intellectually honest, reasonable way, you should probably do a post or three developing/making-concrete your specific hypotheses and what predictions your hypotheses would make. Basically, spell out your beef(s), whatever they happen to be.

      I, too, have a few beefs with some folks’ behaviour at the SP, and specifically with one particular event involving Lsuoma’s moderation behaviour, which I spelled out in a post there; but I try not to generalize to ‘the Slymepit’ or ‘Slymepitters’ or whatnot. On the other hand, you may have legitimate points, so I’m not against you making such complaints in principle. But if you do, I think it’s in everyone’s interests, yourself included as a skeptic/rational person, to be as up-front and intellectually honest about it as you can, with as much transparency as possible. Hence why spelling out your specific points/hypotheses as concretely as possible would be a big benefit, IMO.

      (Off to check out your current sampling….)

      • There are already a number of interesting hypotheses floating around, but to be honest most of them really suck. One popular hypothesis (often treated as a self-evident truth) over at FtB is that the SlymePit may be aptly characterized as a hate site but I would guess that any unbiased sampling of the Pit would lead to this hypothesis being firmly rejected. In fact, I’m guessing that the sample is going to turn out mostly ho-hum with just a handful of hateful remarks. Of course, that raises the question of just how much shit one should tolerate in the swimming pool, but we’ll come back to that later.

        Once we come to agree on a fair sampling methodology, and a fair survey methodology, then we can start testing out specific hypotheses. I don’t think any of those three steps will be easy, but it would be a huge improvement over the current method, which is usually to highlight either the most brilliant satire or the most hateful outbursts and try to hand wave the rest away.

        Basically, spell out your beef(s), whatever they happen to be.

        I’m not sure why you believe that we would expect my “beefs” to show up in a random sample. One of my problems with the Pit is that the most hateful language generally gets a pass from otherwise ethical people, but that is a reference to the extreme tail of the posting distribution.

      • John Greg

        Damion said:

        “One popular hypothesis (often treated as a self-evident truth) over at FtB is that the SlymePit may be aptly characterized as a hate site but I would guess that any unbiased sampling of the Pit would lead to this hypothesis being firmly rejected.”

        Well, I am surprised. Thanks for that.

        “Once we come to agree on a fair sampling methodology, and a fair survey methodology, then we can start testing out specific hypotheses.”

        My argument, theory, belief, whatever, is that there can be no legitimately effective or accurate sample of the Pit, for the purposes of determining its overall character, without being a fulltime participant and reading almost every comment therein posted. Because, because, because, there is such a vast and wide ranging opinion / ideology / socio-political base to the Pit, that it requires that much data to really understand it.

        The Pit is not homogenous or uniform in any way. There are some, for lack of a better term, ideas that are held in almost, but not quite, universal accord, but please note the use of the word “almost”. Nothing, so far as I can determine, is universal at the Pit. There are even people, regular Pitters, who frequently defend certain actions and statements from FfTB and Skepchick. And they’re not censored or hounded out; they’re just debated and disagreed with.

        “… the most hateful language generally gets a pass from otherwise ethical people….”

        That is because of the insistence, again, almost but not quite universal, on freedom of expression; freedom of speech. You cannot insist on full freedom of speech, and then police it just because some figure (n) is hurt / uncomfortable / ethically opposed / religiously opposed to certain terms.

        It also reflects and demands acknowledgement that there is no universal agreement, either here, or at the Pit, or at FfTB, or Skepchick, or even A+, upon what “the most hateful language” actually is.

      • You cannot insist on full freedom of speech, and then police it just because some figure (n) is hurt / uncomfortable / ethically opposed / religiously opposed to certain terms.

        Freedom of speech does not mean that all speech goes by without any opposing speech.

        I can insist upon your right to march openly with the Nazis in Skokie and still vehemently argue that it is a terrible ideology, both factually and morally wrong.

        I can insist upon your right to call someone a “cobweb cunt” or a “talking prune” and still argue that you are wrong to do so.

      • John Greg

        Agreed. Fully.

        But I do not consider that “policing” it; I consider that to be you using your freedom of speech to disagree with how “I” use my freedom of speech. We both win.

    • Here’s a brief description of each comment in your sample. The short of it: pretty ho hum.
      (no preview button, so not sure if it will come out looking a mess)

      1 2012-07-18 16:56 Gumby making a point about mockery vs. bullying
      2 2012-07-23 12:53 A personal account/observation about the degradation of quality of commentary at Pharyngula
      3 2012-08-05 1:42 A continuation of joking about RW’s lifestyle, referring to past rock star deaths via choking on vomit, and making reference to This Is Spinal Tap’s parody of a drummer who chokes on someone else’s vomit
      4 2012-08-23 1:35 Regarding sexual selection of genitalia, specifically duck penises, and characterizing gender feminism as woo
      5 2012-09-02 17:10 Comment about a Japanese-seeming name suffix meaning ‘dunce’
      6 2012-09-19 20:23 Link to an article on the religious right
      7 2012-09-24 12:14 Questioning RW’s standards of behaviour/discourse
      8 2012-10-21 21:56 Expression of incredulity
      9 2012-11-04 15:00 Mykeru posting one of his videos
      10 2012-11-20 18:33 Discussion of unskeptical views re Hamas and terrorism
      11 2012-11-23 6:52 Scented Nectar posting a parody video about Surly Amy’s crying at TAM
      12 2012-12-01 19:00 Joke photo of a T-Shirt re DJ Grothe meeting RW
      13 2012-12-21 9:54 Re the Newtown massacre and Svan’s/FTBers’ interpretations, re masculinity/race
      14 2013-01-21 4:58 Continuing discussion of gendered words such as ‘bitch’ as an insult vs inherently sexist
      15 2013-03-10 3:54 Re feminist-inspired woo and cultural relativism
      16 2013-03-23 13:12 Satirical/parody shop by Jan Steen re Pharyngula as an echo chamber
      17 2013-03-23 19:22 Reply re a video link
      18 2013-04-16 15:45 Re ‘dramatic readings’ and/or gumby/comic-sans-font as ‘harassment’
      19 2013-05-04 8:43 Personal account re zen and hipsters
      20 2013-05-07 2:15 Question about ‘patriarchy’
      21 2013-05-27 6:06 Banter about wearing ‘Surlypastry’ parody at a conference
      22 2013-06-06 23:00 “I am disappoint” shop re a Michael Nugent post
      23 2013-06-23 7:11 One-liner re a video
      24 2013-07-06 10:36 Comment about Anita Sarkeesian’s video series
      25 2013-08-14 7:49 Comment expressing hope re decline of A+
      26 2013-08-16 1:06 Screencap and critique of PZ re comparison of ElevatorGate with Dennis Markuze/David Mabus
      27 2013-08-30 2:28 Welcome message to new pitter; commentary re Shermer libel suit
      28 2013-09-12 14:59 Joke about screwdriver vs. hammer (an in-joke at the pit re a wager between some pitters)
      29 2013-09-24 0:46 Comment about fickleness; critique of SomeGreyBloke
      30 2013-09-29 4:28 Satirical shop from Ape+lust featuring Oolon and Big Red

      • Would you characterize any of these posts as hateful? Perhaps unnecessarily degrading or dehumanizing? This is what I would call the FtB view of the Pit, and it is worth testing.

        What fraction would you characterize as lighthearted satire no more harsh than the typical editorial or editorial cartoon? This is how many Pitters tend to see themselves, in my experience.

        As to subject matter, what fraction of these posts would you say are directed against outspoken atheist feminists?

      • As I mentioned, I try to avoid publicly ‘mind reading’, and even if I were to speculate, there are inherent problems with characterizing something as ‘hateful’, i.e. ‘hate speech’. If we were to go by some of the legal definitions available, however, I think you could clearly conclude that, no, none of the comments fall under legal definitions of ‘hate speech’ that I’m aware of. Some background: http://en.wikipedia.org/wiki/Hate_speech

        As for comparing it to editorial satire, you again run into the problem that the pit is not a collective. There is no organization (such as at a newspaper or magazine) for which there is an ‘editorial’ perspective. Each pitter is their own editor (well, aside from that incident with Lsuoma I mentioned, which is why I am not posting on the pit until that issue is resolved). So, where one pitter might see their satire as ‘lighthearted’, even if expressed ‘harshly’, another might see that particular piece as ‘crossing a line’ and might or might not bother to say so. There were several instances when I was active there where I considered something crossing a line but didn’t bother to say so because either a) there are only so many hours in a day, and it wasn’t significant enough to devote the time to argue over it, or b) there was enough uncertainty in my own interpretation that I couldn’t in good conscience declare that I *knew* that my interpretation was ‘what they really meant’. On the other hand, there were also several occasions where I did object openly and spent the time to argue over it.

        I think the hypotheses of ‘hate speech’ and/or ‘satire’ are examples of the ‘too complex’ hypotheses that would be premature to try to explore at this stage. Something simpler — such as a mere word-frequency count, or a count of the number of topics/subjects (e.g. how often they bring up a particular person), without any subjective interpretation of what that means, just a statistical measure of it — would be more likely to be useful at this point. IMHO, you need to start with a common ground of evidence that anyone and everyone could potentially agree on, and then venture out from there in various directions.

        Here’s an example hypothesis based on word-frequency counting that could be of interest: I hypothesize that if one were to sample the frequencies of words used at the SP and compare them to those used at FTB (say, Pharyngula, to be concrete), then you would find that there is a *greater* variety (variation) of words used at the pit than at FTB. For me, this would signify a greater intellectual diversity, however, that would be my subjective interpretation, and I’d leave that out of the ‘study’ as a conclusion. It would still be an interesting finding if it were true, nonetheless. Once this is established as a basic fact, one way or the other, then it would be more understandable to most onlookers that you’d want to go further to try to understand this fact by extending the study to test specific hypotheses which explain the finding (e.g. my hyp about intellectual diversity). If you jump right in to a fairly complex hyp from the get go, then onlookers (especially the many who are already biased one way or the other) are more likely to reject the whole thing altogether, to the point of even ignoring the interesting basic facts (assuming it were the case that one or the other site had a greater variety of word-frequencies).

        As for ‘what percent directed at outspoken atheist feminists’, how do we define ‘feminist’? Huge can of worms there. Enormous, I might say. Many pitters themselves would fit many contemporary definitions of ‘feminist’. Again, if instead we were to limit the study to something more objective and concrete: What percentage of posts directed at a particular person? Then you could side-step the worms and just look at the basic facts first.

        As for future questions, I think you should really take into consideration *variation* and not just measures of central tendency (mean, mode, median). Are you trying to make a case that the pit is single-minded or conformist in some way? Then a potentially good way to test that would be to look at how much variation there is in posts (especially posts by different people). Low variation (compared to some background ‘null hypothesis’ population) would tend to indicate more single-mindedness, high variation less. Are there clusters of different ‘types’ of posters in the population at the pit? Is there some way to measure the large population of lurkers who rarely or never post? Is there some way to measure whether what is discussed at the pit has some sort of influence outside the pit? Or even within the pit itself? (I.e. are there de facto opinion leaders who end up ‘shaping the narrative’?)

        All of those things could be measured without first slapping on an interpretation to them. See what’s actually there first, and then extend those observations later with hyps to explain the facts.

      • As for ‘what percent directed at outspoken atheist feminists’, how do we define ‘feminist’? Huge can of worms there. Enormous, I might say. Many pitters themselves would fit many contemporary definitions of ‘feminist’.

        I think you may be playing a bit dumb here. You should know, from your time in the Pit, that there is a particular social group of atheist feminists that get much (perhaps even most) of the attention. They tend to blog and comment at FtB or Skepchick, although some of them do other things like write for Slate.

      • John Greg

        Damion said:

        “… there is a particular social group of atheist feminists that get much (perhaps even most) of the attention.”

        Yes, this is generally true. But as I said above [edit: above > elsewhere (due to the hierarchy of the blog)], the reason behind that focus is because of the perceived dishonesty, hypocrisy, etc., of those so-called atheist feminists, not because of their atheism/feminism; not because of where they tend to congregate. And, once again, this also points to the facts behind the origins and creation of the Pit in the first place.

      • “I think you may be playing a bit dumb here.”

        Pfft. Cautious. Intellectually honest. Not dumb. If you think feminism is easy to define, try defining it. You’ll invariably miss some people who call themselves feminists, and include some who vehemently reject the term.

        ” there is a particular social group of atheist feminists that get much (perhaps even most) of the attention.”

        Well, of course! But you didn’t originally ask me about *that* group. You asked about ‘outspoken feminist atheists’. That is why *you* need to get more concrete and specific instead of waving around ambiguous labels. That is why I’m recommending to go for the simpler hypotheses, the ones that don’t require thorny and contentious definitions.

      • I don’t have a convenient non-insulting shorthand for the particular group of people I’m thinking about, but they all self-identify as advocates for feminism, atheism, and social justice and most of them get paid to fly out to Skepticon.

      • Why don’t you, as I suggested earlier, simply study how frequently *individuals* get mentioned/criticized? That way you don’t have to define any groupings ahead of time. If you then see within *that* result an apparent grouping, then you can start to refine the hypothesis to be about some particular ‘defined’ group.

        I’m reminded of Daniel Kahneman’s usage of “System I” and “System II” to avoid the ambiguities and controversies over labeling modes of thinking as ‘intuitive’ or ‘rational’.

        Key here, I think will be to provide a reasonable null hypothesis, perhaps by using a ‘control’ group such as comments from some other kind(s) of topical message forums.

      • Are you trying to make a case that the pit is single-minded or conformist in some way?

        I was not planning on it, that sounds like a vastly more complex and difficult hypothesis to test than something like “The forum is generally more lighthearted and satirical than it is hateful and ranty.”

      • You could also perhaps do some sort of opinion survey, using random quotes from the pit and asking *random* people to rate them on scales of ‘light-hearted’, ‘satirical’, ‘snarky’, ‘hateful’, ‘ranty’, etc. This would at least give a qualitative representation of what a random onlooker might expect to see with a quick, cursory glance at the pit. Properly done surveys tend to be quite difficult to pull off, though, without significant time and effort. If you have a possibility of getting some modest research funding, though, it would be a potentially novel research project.

      • I like where you are going with this.

      • John Greg

        TT said:

        You could also perhaps do some sort of opinion survey, using random quotes from the pit and asking *random* people to rate them on scales of ‘light-hearted’, ‘satirical’, ‘snarky’, ‘hateful’, ‘ranty’, etc.

        On the face of it that sounds like a good idea but, yet again, you will run into the problem of context. Any kind of random quote taken out of context is, ultimately, meaningless (anywhere, not just the Pit). And with a lot of the stuff at the Pit, the context goes back a couple of years and, for the uninitiated, is not at all easy to find.

      • Yes, of course. It would be a very limited hypothesis (essentially only measuring how ‘random’ people would react to a random pit post out of context). Still, it’s something that could in practice be studied, so long as the limitations are admitted up-front.

        I’m treating this blog post/comments as ‘brainstorming’, just tossing out whatever comes to mind. There will be weak ideas in the mix, for sure. But for the purposes of brainstorming, it’s better to present a variety of ideas without too much self-censorship, with the understanding that there will be a necessary ‘culling’ of options down to the best ones, at a later stage. If this happens to be one of the weaker ideas on second thought, so be it.

      • John Greg

        Damion said:

        “As to subject matter, what fraction of these posts would you say are directed against outspoken atheist feminists?”

        What does it matter whether or not any given percentage of posts is so directed? It is the reason behind the posts that is relevant, and that reason tends to be to focus upon the perceived hypocrisy, the apparent deceit, the consistent misrepresentation of others, the intentionally fuzzy expression and ambiguity, and so on and so forth, delivered by the ‘target’ posters, and the damage they do to the base of the supposed community of their ideology.

        The principal reason that most of the individuals so directed might be so-called outspoken feminists points to facts of the origin of the Pit, not the fact that they are feminsists; not the fact they are atheists; not the fact that they are outspoken.

      • John Greg

        Damion: I did not make the post above, the one that this post is replying to. Is the software here a bit borked? Or did you accidentally “edit”, rather than “reply” to an earlier post of mine?

        Also, it looks to me like the post that the above one seems to be replying to is now lost.

      • Shit, I have no idea. Checked the pending queue, there is nothing in there.

        These threads can get confusing, if there is something you need me to reply to in particular that I missed, please let me know.

      • John Greg

        Well, for whatever it is worth, I was responding to your statement:

        As to subject matter, what fraction of these posts would you say are directed against outspoken atheist feminists?

        My argument was that it is irrelevant what fraction or percentage of the posts are directed against outspoken atheist feminists because the reason those posts are so-directed is not because the objects/recipients of the posts are outspoken atheist feminists, but because of what those indivduals say and how they say it, i.e., the self-contradictions, the hypocrisies, and the frequent deceit within their rhetoric.

        In my opinion, simply asking that question risks derailing, misfocussing, and misinterpreting the real and legitimate reason behind the posts.

        People Pit-side don’t point fingers and shoops at, for example, Zvan, Watson, and Myers because they are outspoken atheist feminists, they point fingers and shoops at those individuals because they are so frequently deceitful, two-faced, and hypocritical, and do not even follow their own ideological rules. And that is a very important distinction.

      • I think we would be hard pressed to find someone who is all of these things:

        1) Reasonably well-known as a skeptic/atheist
        2) Outspokenly advocating feminist reforms
        3) Not taking heat from the Pit for #2

        Perhaps I’m not thinking broadly enough, maybe we can come up with someone who fits the bill.

      • John Greg

        How’s about Christina Hoff Sommers; Camille Anna Paglia; Harriet Hall?

        Just for a start?

      • I had no idea that Sommers and Paglia are known as atheists or skeptics, nor that Hall is known for promoting feminism. I’ll need to read up on them all.

      • John Greg

        Whoops! Right you are. My bad.

        /blushes at dumbassity

        OK, OK, recovery time: Harriet Hall indeed is known for promoting feminism, and has been so known for decades. Have you missed the years-long discussions of her very successful battles with the military patriarchy, in the name of feminism?

        Oh, and by-the-by, Paglia identifies as an atheist. Whether or not she is known as one, I guess, depends on how well you know her — or of her.

      • (Sorry, meant to be addressed to Damion, but replied to JG by mistake.)

        The fact remains that there are very many *actual* feminists who either are not involved in drama, or who reject the kind of ‘feminist reform’ being put forth but ‘the usual suspects’.

        Want to know one of them? I seem to recall a young feminist by the name of Stef McGraw. Yes, I believe that was her name.

        This is the major point I’m trying to make with my ‘dumb’ distinctions over the wooly term ‘feminist’. Stef McGraw was an outspoken atheist feminist. She was nearly universally supported by pitters (and many non-pitters).

        Will your criterion of measuring the pit’s ‘antipathy towards outspoken atheist feminists’ take this crucial fact into account, or will it ignore it? That is key. Because if you ignore it, then you will be ignoring the raison d’etre of the pit itself. The pit began *literally* to *defend* an outspoken atheist feminist from the social-bullying modus operandi of FTB/Skepchickers/whatever-you-want-to-call-them. See http://scienceblogs.com/erv/2011/07/01/bad-form-rebecca-watson/ for one of the ‘genesis’ posts.

        Here are some other names of outspoken atheist feminists who have been supported by pitters: Maria Maltseva, Sara Mayhew, YouTuber stclairose, Miranda Celeste Hale, Paula Kirby, EllenBeth Wachs, et many al. There are promising new faces as well, such as Jen August, who has joined the fray against the A+ BlockBot. See her fabulous post here: http://jenaugust.wordpress.com/2013/08/30/feminism-the-hard-way/

        If you’re not intellectually honest about this (and I’m not saying you aren’t, just hypothetically speaking), then whatever you present as research will lack credibility and integrity.

        And, I repeat, there are people at the pit who still identify as feminists of the ‘equity’ kind, in distinction to the ‘gender’ kind. And they are outspoken, and atheists. If your study fails to recognize this *basic fact*, then it will be rightly criticized on that basis at the very least.

      • Seems to me that we need to differentiate between the sort of feminists who are advocating for feminist policy reforms and boycotts within the atheist movement and the sort of feminists who believe in some form of equality for the sexes, which is pretty much everyone.

        Perhaps I am much mistaken, but I cannot recall Maltseva, Mayhew, Hale, Kirby, or Wachs advocating for the kind of reforms that we’ve seen promoted at Skepchick and FtB.

      • Maltseva is (as far as I can recall, could be mistaken slightly) pro-harassment policies, but not the overly restrictive kind pushed by the A+ers. Wachs most certainly fell on the A+ side on many issues, and still probably holds many of her previous views (just not on the evilness of the opposition anymore, nor on the benevolence of A+ers anymore).

        I don’t think it’s as simple as ‘advocating for feminist policy reforms’. If you recall, Richard Dawkins himself (and his foundation, including both Paula Kirby and Elizabeth Cornwell among many others) instituted funding for child care at conferences to encourage more family-oriented adults (majority women) to be able to attend conferences while maintaining family commitments. Did the pitters oppose that? No. Because *it’s a good policy*!

        It’s not ‘policy’ vs ‘no policy’, it’s ‘smart policy’ vs ‘dumb policy’. And it’s ‘allow debate over policy before shoving it down peoples’ throats’ vs ‘disallow any debate over policy and label any dissenters rape apologists, misogynists, trolls, and worse’.

        You should know this, having spent some significant portion of time at the pit yourself. Frankly, I’m surprised you don’t, or maybe you’ve forgotten.

      • I don’t think it’s as simple as ‘advocating for feminist policy reforms’.

        Of course it is not that simple. Sometimes the equity feminists and gender feminists can agree on the utility of certain reforms, such as child-care at conferences.

        The more I think about it, the more I’ve come to think that it’s not so much about the reforms themselves as it is about kowtowing to a certain high-profile social group. The policy proposals are just one way of asserting power and privilege. Blacklisting selected speakers is another way, whitelisting others is another way. Mostly it is about control of the commanding heights of the conference stages. Well, that’s my going theory at the moment, anyhow.

      • Harriet Hall’s position on feminism is that she doesn’t write about it too terribly much. When she does, though, she is on the defensive, under attack by the outspoken SJW/Tumblr wave feminists for not being deferential to their values and norms.

        I am a feminist too, even though my brand of feminism may not meet your expectations of how a feminist should act. There are different roads to the same destination. Don’t disparage mine.

        If you cannot recall, you can probably guess how the social justice wing of the movement reacted to this plea for peaceful coexistence.

      • Not *writing* about feminism and not *being* a feminist are two entirely different things (aside from the common feature of being *about* feminism). Harriet Hall is a) outspoken, b) an atheist, c) a feminist. All at once. This is why you need to be more concrete and specific in your labels/groupings.

        May I suggest getting feedback from some of the feminists who’ve been mentioned? I would suggest at least talking to Maria Maltseva and EllenBeth Wachs, both of whom I believe we mutually know through FB. I’m certain they would have a valuable perspective on if/how you could/should group/not-group the people involved.

      • When I say “feminist” I mean to pick out a specific group of people, with a distinct set of goals.

        If by “feminist” you mean merely those who believe in equality of the sexes, you’ve failed to pick out a distinct group of people within the freethought community. I’ve never once met a single freethinker who does not fit that description.

      • John Greg

        So, Damion, are you just going to continue to dismiss any and all atheist feminsts that TT and I bring up because they fail to meet your narrowing and changing definition of applicable atheist feminists; a definition that changes and narrows each time you post?

        I would certainly call that a textbook example of shifting the goal poats and intellectual dishonesty.

      • If by feminist you mean merely people who believe in equality of the sexes, you’ve failed to pick out a distinct group of people within the freethought community. I’ve never once met a single freethinker who does not fit that description.

    • Another thing that just occurred to me: I’m pretty certain your method of sampling has potential flaws in terms of assigning greater probability for sampling to posts which occur during ‘slow’ periods at the pit. If you’re assigning uniform probability to each minute since the founding, and then ’rounding up’ to the next post after that minute, then you are necessarily treating posts with longer delays between them as disproportionately more probable than posts which occur in rapid succession. For example, during a heated debate/discussion, posts might be flying by at one every two or three minutes, sometimes faster. Whereas at other times, maybe only one or two in an hour. If there were no connection between rate of posting and subject matter/style, then there’d be no problem, but I’m pretty sure there *is* a connection/correlation between these things. Probably the better thing to do would be to give each post uniform probability.

      • One method of sampling provides a sample of the inside view (for people that follow the whole thread) and another method of sampling provides a sample of the outside view (people who drop in at random to see what’s up).

        I considered doing it both ways, but my goal was to get a statistical sample of what a curious onlooker would be likely see if they dropped in at some random time. Posts that stay up longer on slow days are more likely to be viewed by the average lurker from outside of the inner circle, because they don’t scroll off the screen as quickly.

      • That’s a somewhat valid response. I was thinking something similar, but didn’t put much weight on the ‘curious onlooker looking at a single latest post’ model. Honestly, I don’t think that’s how any significant proportion of people would experience the pit. I would expect that a curious onlooker would at least read *more than one* post, likely in succession, in order to gather some context of the ongoing conversations. Certainly that’s what the large number of hidden lurkers do.

        However, perhaps it’s possible to do both types of sampling and compare the results to see if there’s any major differences. You could even reasonably model the ‘dip in for a bit of context’ scenario by randomly choosing a time and then randomly choosing a certain number of posts to add to the sample in succession (exponential distribution should do the trick for choosing this number). Then you could choose n, the number of time-samples to make, and then k1, k2, …, kn for the number of context-samples to take, and put them all together into N=sum(ki) total samples. Or something like that. Just an idea.

        In any case, it’s not a fatal limitation, and can easily be dealt with. As long as you’re up front about it and acknowledge the effects of it, I don’t see it as a stumbling block.

      • There is nothing in the sampling methodology that necessarily limits us to just one post, unless we are doing something quantitative like word counts, and even then we are free to scoop in a fixed number of nearest neighbors.

    • I’m not a scientist so my understanding may be flawed here, but I feel that in this exercise Damion has already reached a conclusion and he is seeking a justification for his beliefs.

      This post is the beginning of an excercise in backing away from our emotions and assumptions. If my lack of open partisanship makes you uncomfortable, there is little I can do to help.

      I’m open to testing any proposition using sampling, whether it fits my preconceptions or not. Feel free to throw some out, preferably before looking at Thaumas’ summary comment.

    • Allison Kirkpatrick

      I love the Slymepit and have followed it from its beginning. It’s one of the healthiest online communities I’ve ever frequented, thanks to it’s light-handed moderation policy.

      The content is often hilarious, sometimes boring or uninteresting (e.g. computer games, television), and occasionally disappointing, but I’d take the SlymePit over the dysfunctional messes that are FreeThoughtBlogs, Atheism Plus, etc., any day. The silly namecalling (e.g. “Rebecca Twatson”, “Stefunny Svan”, “PedOphelia Benson”, etc) is rather juvenile, but it does not bother me, since those people have done far worse for the last couple of years.

      • Do you post there?

      • Allison Kirkpatrick

        Yes, but very infrequently.

    • Pingback: Sampling the A-Plus Forum | Background Probability()