Page 277 of 628 FirstFirst ... 272 273 274 275 276 277 278 279 280 281 282 ... LastLast
Results 6,901 to 6,925 of 15681
  1. #6901
    Join Date
    Aug 2006
    Posts
    6,799
    Year?!

  2. #6902
    Join Date
    Aug 2019
    Posts
    1,083
    Quote Originally Posted by rod9301 View Post
    I'm not writing of statistics, studied them for my nuclear physics degree, and actually enjoy them.

    I'm just objecting to a meta study that uses a number of poorly constructed studies to see if masks work or not

    If you use garbage data, the meta study will be worthless.

    On the other hand, physics will tell you how much benefit a mask will give either the wearer or the people around him.

    It seems that the anti mask people are using it as religion, and want to prove that masks don't work.

    Not clear why, masks are not that hard to use.

    Sent from my Redmi Note 8 Pro using Tapatalk
    This is nonsense. You realize it's entirely possible a mask can block droplets by x% and still be ineffective at reducing transmission in the real world?

    You can nitpick the studies all you want, but the reality is it's extremely likely that masks have little effect if 14 RCT's find that to be the case. Or put another way, if masks have significant benefit, it's extremely unlikely 14 RCT's are going to find that they have no benefit. And by the way, the CDC thought this study was valid enough to have on their website - it seems most people on here have a high degree of faith in that organization.

    We have numerous studies showing no benefit from masks, and the data from COVID makes it clear as day.

  3. #6903
    Join Date
    Aug 2019
    Posts
    1,083
    Quote Originally Posted by The AD View Post
    Ron Johnson thinks he can demonstrate that Babe Ruth is the greatest baseball player of all time by only showing us Babe Ruth's stats.
    rideit level analogy right here.

  4. #6904
    Join Date
    Nov 2005
    Posts
    6,122
    Is there ever really a good moment to be reminded that the easiest lie to tell is to build a randomly controlled trial to study a rare event using a too-small sample and then (predictably and meaninglessly) find "no statistically significant" effect using said study?

    It's kind of basic but it feels like you don't necessarily have to be a ron johnson to miss that. For instance, if you just see the "Breaking news! X has no effect!" headlines it's often buried pretty deep that if you study 4 people for 2 weeks for skin cancer you should expect sun screen, sun, tanning beds and topically applied agent orange to all have no statistically significant difference from the control group. Touting such a result is two steps beyond a lie, of course. Putting together a meta-analysis of studies with no statistically significant results is next level, like big tobacco level.
    Last edited by jono; 07-23-2021 at 08:45 PM. Reason: the password is still yes, ron. but you knew that.

  5. #6905
    Join Date
    Nov 2005
    Posts
    6,122
    Quote Originally Posted by bodywhomper View Post
    Year?!
    Good point.
    A woman came up to me and said "I'd like to poison your mind
    with wrong ideas that appeal to you, though I am not unkind."

  6. #6906
    Join Date
    Aug 2019
    Posts
    1,083
    Quote Originally Posted by jono View Post
    Is there ever really a good moment to be reminded that the easiest lie to tell is to build a randomly controlled trial to study a rare event using a too-small sample and then (predictably and meaninglessly) find "no statistically significant" effect using said study?

    It's kind of basic but it feels like you don't necessarily have to be a ron johnson to miss that. For instance, if you just see the "Breaking news! X has no effect!" headlines it's often buried pretty deep that if you study 4 people for 2 weeks for skin cancer you should expect sun screen, sun, tanning beds and topically applied agent orange to all have no statistically significant difference from the control group. Touting such a result is two steps beyond a lie, of course. Putting together a meta-analysis of studies with no statistically significant results is next level, like big tobacco level.
    14 RCT's is not comparable to a study of 4 people for 2 weeks. Where are you finding that this study has no statistically significant results? You are just desperate to dismiss it, and still haven't posted anything you think is better.

    At this point it seems you guys must have such soft minds that it's just impossible for you to accept that masks don't work. I guess you view it as a loss for your team or something. Or maybe it's just because Trump wasn't into them so you must support them.

  7. #6907
    Join Date
    Nov 2005
    Location
    Down In A Hole, Up in the Sky
    Posts
    31,053
    Aren’t YOU the head cheerleader for Team No Mask?
    Oh, the irony…you sure are a glutton for well deserved DK abuse!
    Thanks for giving us a chuckle again, every forum has a village idiot.
    Forum Cross Pollinator

  8. #6908
    Join Date
    Dec 2005
    Location
    Watching over the valley
    Posts
    3,792
    How do we vaccinate threads against Ron Johnsons? I mean, how do we keep his stupid ass out of otherwise decent threads?

    sent from Utah.
    sigless.

  9. #6909
    Join Date
    Nov 2005
    Location
    Down In A Hole, Up in the Sky
    Posts
    31,053
    I don’t know, but I found a tweet thread from the kind of moron who might listen to a simpleton like rj.

    https://twitter.com/bethanyshondark/...74070442217478
    Forum Cross Pollinator

  10. #6910
    Join Date
    Aug 2006
    Posts
    6,799
    Quote Originally Posted by rideit View Post
    I don’t know, but I found a tweet thread from the kind of moron who might listen to a simpleton like rj.

    https://twitter.com/bethanyshondark/...74070442217478
    Fkna, you must be having a slow day….

  11. #6911
    Join Date
    Nov 2005
    Location
    Down In A Hole, Up in the Sky
    Posts
    31,053
    Nah, I just cross posted it. She is a good example of the deluded fools. Now she has turned to prayer.
    Forum Cross Pollinator

  12. #6912
    Join Date
    Jan 2008
    Location
    truckee
    Posts
    18,413
    Quote Originally Posted by jono View Post
    ^ Violent agreement, yes.



    I think I'm still smarting a little from the 6 microns = droplet fiasco. It's been a weird year.
    I still haven't gotten over the idiocy of the early days--sitting knee to knee at a banquet table with a bunch of strangers, not a mask in sight, but carefully bumping elbows instead of shaking hands. Or how paranoid I was hiking in the mountains. I would have preferred to see a mountain lion than another hiker coming at me on the trail. But then weren't we worried back then about getting covid from our dogs and cats? Why not mountain lions?

    Re type 2 (or is it type II) error--an underpowered study missing a true benefit. That's why if you're doing it right you call in the statisticians ahead of time to figure out how big a study you need to minimize type 2 error. Based on the expected magnitude of the benefit as suggested by preliminary studies.

  13. #6913
    Join Date
    Nov 2005
    Posts
    6,122
    Early days lasted much too long. On the upside, I discovered that hand sanitizer removes pine pitch, which is going to be useful for the rest of my life!

  14. #6914
    Join Date
    Nov 2005
    Posts
    6,122
    Quote Originally Posted by basinbeater View Post
    How do we vaccinate threads against Ron Johnsons? I mean, how do we keep his stupid ass out of otherwise decent threads?

    sent from Utah.
    I'm sure I've responded to him more than makes sense so I finally parked his answers in the troll thread. IDK if that's better. On some topics there are honest-thinking people with similar issues who keep their mouths shut rather than be thought a fool and he's just the guy to open it and remove all doubt. The disingenuousnous seems like a trump card, though.

  15. #6915
    Join Date
    Oct 2008
    Location
    valley of the heart's delight
    Posts
    1,379
    Trolling or honest questions? Part of the issue is average level of education, complicated by the specialist level of education of the various pandemic experts. I watched a few TWiV episodes where they discussed submitting an op-ed to NYT, and complaining about the editors stripping out all the science. NYT's justification is their publication needs to be understandable to people with fifth grade level education. I posit that NYT's readers are smarter than the national average. With a fifth grade education, statistics doesn't exist, they don't even have algebra. Critical thinking skills aren't well developed either. Even among college grads, in many fields, their knowledge may not exceed a fifth grade level - how many sociology grads can complete a square? Or understand compound interest?

    Then, for the trolls, there's plenty of dishonest argument styles, and free internet training for anyone who wants to improve their trolling. Not a useful line of work, but trolls don't are about being useful. Anyway, I say there's trolls in this thread. Prove me wrong.

    There's some honest questions too, and several of us err toward that interpretation. Probably the right way to err, though it sometimes feeds the trolls.

  16. #6916
    Join Date
    Aug 2018
    Location
    Frantically crawling out of the backseat
    Posts
    668
    I know it's not as fun as arguing about freedom and the tyranny of nose+mouth coverings; but again,
    Quote Originally Posted by Lone Star View Post
    88% of the folks who show up symptomatic will turn out to not have had the vaccine
    Quote Originally Posted by LongShortLong View Post
    95% trial efficacy...means for every 20 people who developed severe symptoms, 1 of them was vaxxed
    is not what vaccine trial efficacy means. It means how much the prevalence (<-- I used the wrong word here, should've been "incidence"; thanks oldgoat) of the disease was reduced in the vaccinated group versus the placebo group in a clinical trial.

    And for the record, the Pfizer and Moderna clinical trials measured confirmed COVID cases (1 symptom + PCR positive) as endpoints for vaccine efficacy, not severe COVID. Trials found both vaccines to be 100% effective at preventing severe COVID.

    Looking at the Pfizer trial: 36,523 people participated in the study, 18,325 in the placebo group, 18,198 in the vaccine group. 162 people in the placebo group (0.88%) developed symptoms compared to 8 people in the vaccine group (0.04%), for a difference of 0.84% (0.88 - 0.04%). 0.84%/0.88% = 95.03%. That's how the clinical statisticians came to the vaccine efficacy of 95.0%. Of the 170 people who developed symptoms, 8 of them were vaccinated. 8/170 = 95.3% (which is coincidentally similar to, but distinct from, 95.0%). Of the 18,198 people who were vaccinated, 8 developed symptoms; so we can expect ~1 out of every 2,000 vaccinated people to have breakthrough disease (with Wuhan Classic).

    Just for the sake of illustration, let's imagine that 100 people in the vaccine group developed the disease, and see how that changes the numbers. 162/18,325 = 0.88% of the placebo group developed symptoms (this value is unchanged from the Pfizer study). 100/18,198 = 0.55% of the vaccine group developed symptoms for a difference of 0.33%. 0.33%/0.88% = 37.8% vaccine efficacy. Of the 262 total symptomatic individuals (in this imaginary example), 100 of them were vaccinated. 100/262 = 61.8%, which is a very different value from the vaccine efficacy of 37.8%.

    It's inherent in the calculations that as the magnitude of the difference increases between the number of subjects in the placebo group that developed symptoms and the number of subjects in the vaccine group that developed symptoms, the closer the vaccine efficacy and the percentage of symptomatic individuals who received the vaccine will be.

    Vaccine efficacy means how much the risk of developing disease is reduced by receiving the vaccine. 95% efficacy is the same thing (statistically) as saying that out of every 20 people who would have developed symptomatic COVID had they remained unvaxed, 1 will develop symptomatic COVID even after vaccination.
    Last edited by CS2-6; 07-25-2021 at 05:18 AM. Reason: oldgoat pointed out I used the wrong word ("prevalence"); I corrected it to avoid possible confusion
    Quote Originally Posted by digitaldeath View Post
    Here’s the dumbest person on tgr
    "What are you trying to say? I'm crazy? When I went to your ski schools, I went on your church trips, I went to your alpine race-training facilities? So how can you say I'm crazy?!"

  17. #6917
    Join Date
    Oct 2008
    Location
    Colorado
    Posts
    5,073

    To Vaccinate or Not---The Rat Flu Odyssey Continues

    Quote Originally Posted by CS2-6 View Post
    I know it's not as fun as arguing about freedom and the tyranny of nose+mouth coverings; but again,


    is not what vaccine trial efficacy means. It means how much the prevalence of the disease was reduced in the vaccinated group versus the placebo group in a clinical trial.

    And for the record, the Pfizer and Moderna clinical trials measured confirmed COVID cases (1 symptom + PCR positive) as endpoints for vaccine efficacy, not severe COVID. Trials found both vaccines to be 100% effective at preventing severe……...


    Thanks for the detailed explanation.


    Unfortunately those that could most learn, won’t because that kind of math is all Greek to them, so they will simply continue to believe whatever ignorant bullshit their preselected media and political sources tell them.




    Sent from my iPhone using TGR Forums
    Quote Originally Posted by Benny Profane View Post
    Keystone is fucking lame. But, deadly.

  18. #6918
    Join Date
    Nov 2005
    Posts
    6,122
    Quote Originally Posted by CS2-6 View Post
    Of the 18,198 people who were vaccinated, 8 developed symptoms; so we can expect ~1 out of every 2,000 vaccinated people to have breakthrough disease (with Wuhan Classic).
    Good stuff. This bit here requires the caveat that we would expect 1/2000 over the same time period and given the same level of exposure. So how many will get a breakthrough case in their lifetime or even a year is higher and depends on prevalence.

  19. #6919
    Join Date
    Nov 2003
    Location
    none
    Posts
    7,385
    I got the J&J early March, because it was the first available.
    Thinking of possibly getting a single dose of Phizer or Moderna.

  20. #6920
    Join Date
    Jan 2008
    Location
    truckee
    Posts
    18,413
    Quote Originally Posted by CS2-6 View Post
    I know it's not as fun as arguing about freedom and the tyranny of nose+mouth coverings; but again,


    is not what vaccine trial efficacy means. It means how much the prevalence of the disease was reduced in the vaccinated group versus the placebo group in a clinical trial.
    As long as we're getting technical--your second link says the efficacy is how much the vaccine reduces INCIDENCE (new cases), not PREVALENCE (total active cases). The two words aren't interchangeable.

  21. #6921
    Join Date
    May 2014
    Location
    Santiago Chile
    Posts
    1,207

  22. #6922
    Join Date
    Aug 2019
    Posts
    1,083
    Quote Originally Posted by jono View Post
    Thanks for that example. It's a little redundant in that it's mainly a failure to translate between math and prose again, but it's also an example of misapplying data on one thing (influenza) to something different (beta coronavirus), which is kind of a higher level mistake, but you can commit that one, too, it definitely adds to the richness of your tapestry.
    Both respiratory viruses, both of similar size.

    I found this in the actual text of the study you referred to:

    "The effect of hand hygiene combined with face masks on laboratory-confirmed influenza was not statistically significant (RR 0.91, 95% CI 0.73–1.13; I2 = 35%, p = 0.39). Some studies reported being underpowered because of limited sample size, and low adherence to hand hygiene interventions was observed in some studies."

    What those numbers mean is that masks plus hand hygiene showed evidence of helping, but in the studies they picked they didn't have enough confidence to call it a statistically significant finding. Also, they admit that the studies they picked were both flawed and underpowered (that's the word we use to describe studying 4 people for 2 weeks to check for new skin cancer, too) but they left it to the reader to understand that being underpowered is a really very super effective way to guarantee no statistical significance can be found. Because significance requires large enough samples to show up, particularly for relatively rare events.

    After that they cited a couple studies that disagreed with the above statement, but they neglected those. Cherry picking is a problem, too, like when you find a study that says what you want and claim it's the best without reading past the abstract. So that's pretty good--it really looks like you come by that example honestly, too, or at least that's what I choose to believe for as long as you can keep the veil up. Carry on soldier.
    You pulled that from the section on hand washing. Key word: combined.

    "The effect of hand hygiene combined with face masks on laboratory-confirmed influenza was not statistically significant..."

  23. #6923
    Join Date
    Aug 2019
    Posts
    1,083
    It's just amazing to watch you guys try to dismiss me as some deplorable troll when you can't find anything I've posted that wasn't valid. Some interesting psychology on display.

  24. #6924
    Join Date
    Aug 2019
    Posts
    1,083
    Click image for larger version. 

Name:	zuby.jpeg 
Views:	80 
Size:	249.2 KB 
ID:	380354

    Points #1,4,8,11,15,20 are especially relevant to people in this thread.

  25. #6925
    Join Date
    Oct 2003
    Location
    closer
    Posts
    3,835
    Thus thread delivers.
    Name:  Screenshot_20210724-192351.jpeg
Views: 233
Size:  28.7 KB
    It's a war of the mind and we're armed to the teeth.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •