Tuesday September 26, 2006

The funniest thing I’ve seen so far today: In a recent online poll, 81% of the respondents thought online polls should continue. (Rick thinks I’m a spoilsport to suggest that online polls are a waste of time, even “just for fun.”) Update: Everyone agrees that I’m an ass for “making a big deal” out of this. See Fanless and the entire argument in the comments. BTW, I like to “click on buttons and look at graphs” as much as the next guy, and in that spirit I offer a link to the implicit association test at Harvard.

Tags: ,

 

comments powered by Disqus
  1. Rick    Tue Sep 26, 07:46 AM #  

    Two clarifications/corrections, Alesh:

    1. The “recent online poll” that you link to is being run by SotP and actually asks whether SotP online polling should continue. It’s a subtle difference, but one that probably doesn’t lessen the humor you find in it.

    2. You went a bit further than saying that these polls are “a waste of time” when you claimed that your training in psychology, specifically “behaviorism and cognative theory,” leads you the conclusion that SotP polls and those like them are actually harmful because “people inevitably (and obviously) do try to extricate meaning out of them,” something I requested you explain but have yet to receive a response to.



  2. mkh    Tue Sep 26, 08:22 AM #  

    People also try to extricate meaning from sheep entrails, the arrangement of the stars, and Fox News, but no one’s going to learn anything from any of those. However, I won’t tell people how to waste their grey matter.

    On-line polls are completely devoid of meaning, but they can still be fun.



  3. alesh    Tue Sep 26, 08:57 AM #  

    You’re right, Rick: I never responded. Nick did a pretty good job, but I suppose as to why they’re “harmful” there was a gap. Every time I started to think about this, I started to laugh, the milk i was drinking came out of my nose, and I got all distracted with cleaning it up. Plus, comments are so difficult to keep track of on SotP, what with that little window and all.

    I guess I’ll just direct your attention to the alarmingly high percentage of Americans that believe that Saddam Hussein had weapons of mass destruction and caused 9-11: we live in a society with lots and lots of stupid people who don’t think particularly deeply about stuff. So to the extent that online poll results are taken even remotely seriously (and really, what else would account for their popularity? that they’re meaningless fun? be real), they are harmful.



  4. gansibele    Tue Sep 26, 10:45 AM #  

    Alesh, all kinds of polls can be misguiding, manipulated or flat wrong if done incorrectly or maliciously. Those high percentages of Saddam WMD believers don’t come from online polls only – similar numbers also come from Gallup. So? As long as Rick does not claim they are scientific in nature, what’s the harm?



  5. alesh    Tue Sep 26, 12:04 PM #  

    Yes… My Saddam example was meant to illustrate people’s stupidity and gullibility, not the inaccuracy of polls. Said stupidity will make people take online polls seriously when they never ever should. Sorry about that bit of confusion.

    Polls by their very nature are imperfect. However, professional polling organizations are aware of those imperfections and take steps to minimize them. They also provide something called a “sampling error margin,” which can be helpful in interpreting the results, and gives the whole enterprise an air of legitimacy.

    Now, consider that 100% of the participants in an online poll are self-selecting (probably the single biggest flaw with these surveys), and you arrive at the unsettling notion that the sampling error margin for any online survey is 100%!

    I have the Associated Press style book in front of me opened to the “polls and surveys” section, and it gives a numbered list of eight general areas of concern w/r/t polls: “Information that should be in every story based on a poll includes the answer to these questions:”.

    Well, of course 99% folks who “administer” online polls are blissfully ignorant of these issues. In other words, the accuracy of online polls is precisely equal to the accuracy of reading sheep entrails. The problem is that since poll results involve numbers, their absurdity is non-obvious even to many people who DO recognize the absurdity of reading the sheep entrails.



  6. sheep entrails    Tue Sep 26, 12:17 PM #  

    You talkin’ to me? Hey!! YOU talkin’ to ME?



  7. gansibele    Tue Sep 26, 12:30 PM #  

    Then maybe your issue is with the stupidity of people, which is a very valid concern if a bit quixotic, but rather than an indictment of online polls it would be an endorsement – as in “people are so stupid they believe this poll reflects reality, when it’s meant as entertainment”. So Rick should not have polls because they make stupid people more stupid? I don’t follow the logic.



  8. Steve    Tue Sep 26, 12:38 PM #  

    Actually, the single most telling and humorous aspect of Alesh’s response is his presentation of credentials: i.e., training in psychology. Psychology! Study of the Mind (no matter) not Matter (never mind!). Open a window and gimme some air!

    Given the choice of sheep entrails or behavioral science, I go with the guts. Psychology isn’t science, art, or even sport: it’s a detritus-laden toad-load of unsubstantiated inconsequential generalizations deduced by undereducated quasi-intellectuals who unlike me like big words. Compared to the typical psychological text, internet polls are the Oracle of Delphi.



  9. alesh    Tue Sep 26, 12:50 PM #  

    gansibele~ I think Rick should stop doing the polls because they’re stupid. I think they’re stupid because they’re entertainment masquerading as information. I guess that’s about as succinctly as I can put it.

    Steve~ That’s a fairly typical reaction to psychology for an old guy. I estimate that the bulk of your knowledge of psychology dates back to the 1960’s, when Psychology’s standing as a science was rightly held in question.

    Suprise: a lot has happened in the field over the last 40 years! The developments include what may be a overreaction to some of the (not inaccurate) accusations of unscientificness—an emphasis on rigorous scientific procedure, with particular attention paid to the correct design of studies and experiments. (The field does still have its Freuidans and Jungians, but they’re in an ever-decreasing minority. I think the positive contributions of the field to society are fairly obvious to anyone under the age of 45 (and even most over 45).)



  10. mkh    Tue Sep 26, 01:55 PM #  

    I would feel better about psychology if we could have kept it a secret from the marketers.

    That said, I don’t have a problem with psychology, per se. I have had problems with quite a few psychologists, mind you, but not really with psychology.



  11. Rick    Tue Sep 26, 02:32 PM #  

    Gansibele: The harm is that these polls bug, Alesh. And as long as they bug Alesh, they should bug everyone.

    “The comments sections suck. You’re using Blogger. Blogger sucks. Your taste in music blows. You need your own domain name. Your polls are useless.”

    Alesh has been a wealth of information for everything that is wrong with SotP. I guess I need that to some degree, but after a while it just gets old. I appreciate the support, Alesh.

    Oh, and wearing a bib while eating usually helps with the milk clean up, partner.



  12. gansibele    Tue Sep 26, 03:38 PM #  

    “I think they’re stupid because they’re entertainment masquerading as information”

    You could say the same thing about Wikipedia. Or blogging in general for that matter.

    In fact you could say is more harmful, because it’s giving people with (for the most part) no formal training, no experience, no research capability or inclination, no set of standards and practices and no oversight (either institutional or from the marketplace) a tribune to comment on current events, report “news” and increasingly significantly shape public opinion.

    But I’d hope that smart people can see the difference, and the same goes for the polls.



  13. gansibele    Tue Sep 26, 03:40 PM #  

    And I hate that Textile converts dashes into crossed out words.



  14. Tom Cruise    Tue Sep 26, 03:51 PM #  

    Only Xenu sees all!



  15. alesh    Tue Sep 26, 03:55 PM #  

    Rick~ That’s about right. And bonus points for memory. Also, whenever I try to figure out what day one of your posts was posted, it takes a unbearably long time of careful scrolling and looking for the time stamp mixed in with the posts. And the last time I clicked on a link someone had painstakingly put into a comment, it opened up in that tinny little window and I almost had a cow.

    I’m only taking the time to tell you this because I really enjoy your blogging, and I hate to see all that work to into such a crummy container.

    Ah well. But the “should we keep having these polls” poll tops everything. It makes it all worth it.

    gansibele~ You could say the same thing about Wikipedia?! That it’s masquerading as information?!? Where on earth did you get THAT idea??

    Hmm… oh, I see your argument. I guess it’s difficult to say if you’re serious about it. The informational value of blogging is equally suspect as online polls. I guess if you honestly believe that then you’ve totally missed the gist of my argument. And I suspect you DO believe that the polls hold information and value. Or maybe you’re just pulling my leg. It’s actually difficult to tell.

    Yes, textile can be annoying. But I swear . . . there’s a mandatory ‘preview’ step built into the system. Why not use it, and at least, you know, take a glance?

    The effect you’re trying to achieve—an em dash—can be done by doing two hyphens between words, with spaces.



  16. Steve    Tue Sep 26, 04:13 PM #  

    mkh: I’m still laughing over the sheep entrails. Also, I can relate to your remark about psychologists…..I don’t mind assasination per se, but yeah, I’ve had a problem or two with specific assasins.

    The polls are fun, and Alesh is just wrong (again. But lovable. Especially with milk dribbling on his chin). And of course they’re bullshit. B.F.D. Amazing how somebody trained in real bullshit (psychology) doesn’t like Bullshit Lite. Alesh—how do you feel about not liking Bullshit Lite? Would you like to come on the Orka Winfrey Show and discuss your Bullshit Lite feelings? That’s what we psychologismitists call “therapy.” You can read my book and learn. But buy it first. Buy-buy.



  17. gansibele    Tue Sep 26, 04:53 PM #  

    I guess I did miss the gist of your argument. It seemed to me you were saying polls—and later generalized to “information”—even if they are obviously for entertainment purposes, should not be done by amateurs lest they be harmful because stupid people believe them. I’m merely pointing out to you other examples. Or maybe you don’t believe people eat whatever Malkin spews like manna from heaven. Maybe the blogging hit too close to home.

    I don’t glance at previews—same reason I don’t use spell check—I have enormous confidence that I was right the first time. And the dork who decided the dash key was a good shorcut for the crossing out code needs to learn about interface design. But how do you like those dashes now?



  18. Rick    Tue Sep 26, 06:34 PM #  

    Make you a deal, Alesh. You quit busting my chops about my blog under the pretext of concern for my container, or whateverthehell you said, and I’ll continue to say nothing about your continued use of that GD Textile that no one in their right mind enjoys using, the effing ads, or the news bytes that don’t show the comment count on them. Amongst other things.

    Because, you know, it’s your blog.



  19. alesh    Tue Sep 26, 06:48 PM #  

    gansibele~ hey—you got it (the dashes, that is). w/r/t the poll/blog thing, I got it. You’re pulling my leg. Thank god.

    Rick~ No deal. I understand your frustration with Textile, and I understand why my defense of that particular system falls on deaf ears, especially to those that are properly proud of having mastered the A HREF= business. All I can really say is that if Textile was more widely implemented, the world would be a happy place. Since it’s only here and there, it causes more confusion and frustration then satisfaction. It’s something I have to live with it.

    You’re right, the news bytes (i like that term) should show the comment count. I’ll work on that.

    I might take a break from busting your chops, but not just yet: a recap of your skirmish with Val is up soon!



  20. mkh    Tue Sep 26, 06:59 PM #  

    I’ve learned to know—and love!—Textile’s markup system. Except for links. Anyone who manages to replace HTML link syntax with something worse in the name of “simplification” deserves a few spins on Satan’s rotisserie, that’s for damned sure!



  21. Rick    Tue Sep 26, 07:03 PM #  

    I know you can’t help yourself. But aren’t there enough other things going on in Miami? More obscure bands to cover? Air conditioners being installed? Unknown holes being jack hammered into I-95? Those are the things people come to Critical Miami expecting to read about.

    Really.



  22. alesh    Tue Sep 26, 07:04 PM #  

    A couple of Wikipedia articles for your reading pleasure:

    Margin of error
    Nonprobability sampling

    In the former, the margin of error is calculated ONLY IF you’re sure you have probability sampling (that is, a true random sample) (if you’d like to pile absurdity upon absurdity, you can calculate the margin of error for rick’s 24-responder poll as though it used a probability sampling. You get something outrageous like 26%.)

    In the latter article, there is a list of techniques used in various sorts of informal nonprobability sampling. Of course an online poll doesn’t appear on the list. That’s because an online poll doesn’t have a self-selecting error—it has a SELF-SELECTING DESIGN.



  23. Rick    Tue Sep 26, 07:10 PM #  

    Dude…no one is making the argument that these polls are anything close to being scientific or valid…except you.



  24. alesh    Tue Sep 26, 07:19 PM #  

    Riiiiiiiiiiiiiiiiiiiight. It’s kind of like having a bakesale and not making any argument that your cupcakes are edible. Good play, partner.



  25. alesh    Tue Sep 26, 07:20 PM #  

    mkh~ I love linking with TXP. The only trick is knowing which unctuation marks can be appended to the end of the URI, and which will kill it. In my experience,

    . , : and )

    work well, while

    ? and !

    cause problems, and force me to resort back to HTML.



  26. gansibele    Tue Sep 26, 08:55 PM #  

    No Alesh, it’s more like having a bakesale and saying your cupcakes are edible and then having some dork with a diploma in Advanced Food Sciences hound you because they are not scientifically nutritive.



  27. alesh    Tue Sep 26, 09:44 PM #  

    No, it’s not like that, gansibele. It’s pretending to collect data and collecting garbage instead. If you’re serious, then you really just don’t understand the issues, and I’m afraid I’m not the one to try to teach them.

    We live in a world where politicians and others abuse numbers to make evil points all the time. In this case, I understand that there is no motivation to mislead, but believe me that the distortion is MUCH MUCH greater. I believe that these polls are doing actual harm to middlebrow people’s understanding of statistics by presenting themselves are “real information.”

    And whatever claims you make to being “just for fun,” the sad fact is that these polls WILL be taken for real information by a great many people.



  28. Steve    Tue Sep 26, 10:17 PM #  

    I believe that these polls are doing actual harm to middlebrow people’s understanding of statistics by presenting themselves are “real information.”

    Yes, Alesh, we understand you believe that. And your evidence for that is…..what again? Got data? “Dude, where’s my proof?”

    Maybe you took an internet poll and discovered that a certain % of SOTP respondents are, oh, no, wait. That won’t work, will it. No, this is just elitist blather, a college educated naif worried that lesser mental mortals might be fooled into thinking that a silly fun little poll among like-minded souls could be mistaken for capital-G Gospel.

    After all, a sensible lad like yourself actually got sucked in for 4 years majoring in a subject one step above sheep-entrail alchemy, and came out a True Believer.

    As for Textile, isn’t that a college in Philadelphia?



  29. gansibele    Tue Sep 26, 10:17 PM #  

    Dude, get off your high horse. I work in advertising. I analyze research all the time. I attend focus groups. I know how to read a poll. I don’t need any “teaching”. Those evil politicians you talk about don’t need online polls. They manipulate “scientific” polls all the time – Gallup, Zogby, Luntz, Forrester, Harris, you name it. Remember Kerry’s exit polls? So, yes, it is exactly like that. Believe me I understand perfectly why these silly polls are flawed. What you don’t understand is that you are grandstanding and making a mountain out of a molehill.



  30. alesh    Wed Sep 27, 06:32 AM #  

    Go back and read my post and tell me again that I’m making a mountain out of a molehill. I expressed amuseuemnt. That’s it.

    Then gansibele came back with a response which amounts to: “online polls are no different then any other sorts of polls.”

    Since then, I’ve explained a million different ways that they are different, and it’s fallen on deaf ears. The whole thing is silly, if amusing.



  31. mkh    Wed Sep 27, 07:31 AM #  

    Alesh, typo or not, I think “unctuation mark” is my new favorite word. It’s like the Scarlet Letter for slimy kiss-asses.

    As for the rest of this discussion, I think it is high time we modify the old aphorism to “lies, damned lies, and opinion polls.”



  32. gansibele    Wed Sep 27, 09:11 AM #  

    You brought it up twice already in two different comment threads, complete with big concerns about number abuse, evil politicians, Wikipedia links and “actual harm to middlebrow people’s understanding of statistics”. All for an amusing silly online poll? Ok, not a mountain. Let’s say the issue doesn’t amount to a hill of beans.



  33. Rick    Wed Sep 27, 09:39 AM #  

    Alesh: It would have helped some had you just explained yourself after making the statement that the polls are psychologically harmful. You didn’t, and still refuse to despite requests by Steve and myself.

    And, yeah, gansibele is right, you took it a lot higher than “amusement,” and to say otherwise is a cop out.

    .



  34. alesh    Wed Sep 27, 01:30 PM #  

    Rick~ Huh? “psychologically harmful”? “higher then amusement”?? I’m afraid I just have no idea what you’re talking about now.

    gansibele~ Yes. I know now that you work with numbers, statistics, and polls all day. I think it’s safe to assume that you have a better understanding of why these polls are BS, from an informational standpoint, then the average person. Not sure why this is such a troubling concept?



  35. Rick    Wed Sep 27, 02:06 PM #  

    Here you go, Alesh. This is your idea of “expressing amusement”.......

    “I was trained in psychology. OK, it’s not the hardest of sciences. Actually, I mostly studied behaviorism and cognative theory. Both are pretty hard sciences, and they take a particular interest in the design of studies. So I know whereof I speak.

    Online polls are often defended as being “just for fun.” Unfortounately they don’t get nearly enough disclaimers. There are SO MANY THINGS wrong with these type of surveys as a way to gather meaningful information as to make them do much more harm then good, because people inevitably (and obviously) do try to extricate meaning out of them.

    ......

    “I guess I’ll just direct your attention to the alarmingly high percentage of Americans that believe that Saddam Hussein had weapons of mass destruction and caused 9-11: we live in a society with lots and lots of stupid people who don’t think particularly deeply about stuff. So to the extent that online poll results are taken even remotely seriously (and really, what else would account for their popularity? that they’re meaningless fun? be real), they are harmful.”

    ......

    “Polls by their very nature are imperfect. However, professional polling organizations are aware of those imperfections and take steps to minimize them. They also provide something called a “sampling error margin,” which can be helpful in interpreting the results, and gives the whole enterprise an air of legitimacy.

    Now, consider that 100% of the participants in an online poll are self-selecting (probably the single biggest flaw with these surveys), and you arrive at the unsettling notion that the sampling error margin for any online survey is 100%!

    I have the Associated Press style book in front of me opened to the “polls and surveys” section, and it gives a numbered list of eight general areas of concern w/r/t polls: “Information that should be in every story based on a poll includes the answer to these questions:”.”

    ......

    “We live in a world where politicians and others abuse numbers to make evil points all the time. In this case, I understand that there is no motivation to mislead, but believe me that the distortion is MUCH MUCH greater. I believe that these polls are doing actual harm to middlebrow people’s understanding of statistics by presenting themselves are “real information.”

    And whatever claims you make to being “just for fun,” the sad fact is that these polls WILL be taken for real information by a great many people.”

    ......

    My job is done here.



  36. alesh    Wed Sep 27, 02:26 PM #  

    Good job, Rick. Now go read the initial post. That’s the amusement. Everything you posted just now was in response to your and gainsibele’s insisting that your polls are no less accurate then any other polls.

    A claim, by the way, that remains without a shred of evidence. I’m glad to hear you’re finally dropping it, though.



  37. NicFitKid    Wed Sep 27, 03:22 PM #  

    Damn, I lose my internet access for a week and look at all the fun I miss.

    I have to agree with Alesh, online polls are useless and misleading, but I do get a kick out of seeing Rick running an online poll to determine whether he should continue his online polls. Gosh, I wonder how all the self selecting participants will vote on whether they want to keep selecting themselves? But hey, we can’t all afford our own professional polling firms, so Rick might as well keep going with it. It’s not like he’s trying to influence public opinion for the purpose of world domination (although that might be kind of cool).

    We already live in the world of reification, where every nebulous concept can be converted into numbers and data, shrink wrapped and shipped out to the trenches of our political wars. Rick’s polls are a tiny part of that world, but hardly malevolent in their intent.



  38. gansibele    Wed Sep 27, 03:51 PM #  

    For heaven’s sake, I never said the online polls were no less accurate than other polls. I just pointed out that any poll can be misleading or manipulated and only in response to your examples. All I said about the online polls was that I don’t see the harm done (of which you don’t have a shred of evidence either, as Steve pointed out).

    Now if you would drop it too Alesh, and not continue to bring it up, that’ll be a Kodak moment for all.



  39. Steve    Wed Sep 27, 04:16 PM #  

    Let’s take a poll! How many people want this “debate” to (a) continue as is, (b) escalate, (c) fade out and fuck off, (d) Die, Die!, (e) I Hate Polls, (f) Vamos a Cuba. Vote for as many choices as you like and as often as you want. Results will be reported as gospel in next month’s Psycholo-gee-whiz Today (look for Orka Winfrey on the cover!).



  40. gansibele    Wed Sep 27, 05:00 PM #  

    Vamos a Cuba! Then we can continue mojitos in hand while polling Party cadres whether they like Fidel or Raul.

    On the alternative, I’m fading out and fucking off…



  41. Manola Blablablanik    Wed Sep 27, 07:34 PM #  

    I’m simply astounded that all of you overlooked the value of tea leaves, coffee grounds, birthmarks, fresh placenta and palm wrinkles! Hmpf!



  42. alesh    Thu Sep 28, 08:21 AM #  

    gansibele~ If we’re talking about online polls, and you say “all kinds of polls can be misguiding, manipulated or flat wrong if done incorrectly or maliciously”, it comes across as meaning either (1) that online polls are pretty much the same thing as other types of polls or (2) that there is a “correct” way to do an online poll. Note: there is no such thing as a distinction between a “correctly done” online poll and an “incorrectly done” online poll.

    BTW, it’s officially time to drop this

    I’m going to admit that I’ve taken it way too far, though I still insist that it didn’t start out that way.

    BTW, as an addendum to Christian’s linkfight, SotP has links from 95 blogs, CM 89.