Where's the petri dish thread?

[DCHawk1]

Scientism

[zsn]

Today In: smfh, or, the plural of anecdote (still) isn't data.

an·ec·do·tal

/ˌanəkˈdōdl/
adjective

(of an account) not necessarily true or reliable, because based on personal accounts rather than facts or research.

"while there was much anecdotal evidence there was little hard fact"

Imagine you have the credentials and everything to offer that Dr. Bright does, and a physician/member of congress confronts you with this:

In seriousness, not trying to be a wiseass. What do call "evidence" gathered through observational rather than experimental studies?

"Evidence" gathered through observational rather than experimental studies can still be called "data" as long as you correct for observational bias. One could be a wiseass and say that the difference is the same as between getting your "facts" from Fox vs. PBS or BBC. Kidding aside, if Rep. Burgess had said "I heard from four doctors who analyzed data from 12 clinical trials and they said that hydroxychloroquine is effective against SARS-CoV-2" it would be one thing vs what he actually said.

This very commonly done in clinical research, where there are many who just analyze a bunch of observations, but correct for various factors. I presume this is done in behavioral research, as well.

Don't know if this is what you were asking, but this is my answer to what I think you were asking.

[Deleted User 89]

well said

[DCHawk1]

Today In: smfh, or, the plural of anecdote (still) isn't data.

an·ec·do·tal

/ˌanəkˈdōdl/
adjective

(of an account) not necessarily true or reliable, because based on personal accounts rather than facts or research.

"while there was much anecdotal evidence there was little hard fact"

Imagine you have the credentials and everything to offer that Dr. Bright does, and a physician/member of congress confronts you with this:

In seriousness, not trying to be a wiseass. What do call "evidence" gathered through observational rather than experimental studies?

"Evidence" gathered through observational rather than experimental studies can still be called "data" as long as you correct for observational bias. One could be a wiseass and say that the difference is the same as between getting your "facts" from Fox vs. PBS or BBC. Kidding aside, if Rep. Burgess had said "I heard from four doctors who analyzed data from 12 clinical trials and they said that hydroxychloroquine is effective against SARS-CoV-2" it would be one thing vs what he actually said.

This very commonly done in clinical research, where there are many who just analyze a bunch of observations, but correct for various factors. I presume this is done in behavioral research, as well.

Don't know if this is what you were asking, but this is my answer to what I think you were asking.

So...anecdotes + math.

Got it.

[Shirley]

Today In: smfh, or, the plural of anecdote (still) isn't data.

an·ec·do·tal

/ˌanəkˈdōdl/
adjective

(of an account) not necessarily true or reliable, because based on personal accounts rather than facts or research.

"while there was much anecdotal evidence there was little hard fact"

Imagine you have the credentials and everything to offer that Dr. Bright does, and a physician/member of congress confronts you with this:

In seriousness, not trying to be a wiseass. What do call "evidence" gathered through observational rather than experimental studies?

I'm not sure I understand the question. At first I thought you meant what word would I "call" the observational evidence. If I do understand, to me, it's the difference between association and causation.

Observational studies don’t control any variables, so the results can only be considered associations.

Variables are controlled in a controlled experiment, therefore, you can make conclusions of causation.

Edited to add: I see I'm late to respond. I haven't read the previous answers, cause I'm watching the Dr. Bright hearings on CSpan, and I'm at 12:30 PM. It's actually better than most hearings, if you're into that sort of thing...

[Deleted User 89]

empirical datapoints =\= anecdotes

[Shirley]

In my original post of John Harwood's tweet, I considered adding this to help make my "case", but didn't, because the evidence isn't that strong:

JAMA May 11, 2020 Association of Treatment With Hydroxychloroquine or Azithromycin With In-Hospital Mortality in Patients With COVID-19 in New York State

Question Among patients with coronavirus disease 2019 (COVID-19), is there an association between use of hydroxychloroquine, with or without azithromycin, and in-hospital mortality?

Findings In a retrospective cohort study of 1438 patients hospitalized in metropolitan New York, compared with treatment with neither drug, the adjusted hazard ratio for in-hospital mortality for treatment with hydroxychloroquine alone was 1.08, for azithromycin alone was 0.56, and for combined hydroxychloroquine and azithromycin was 1.35. None of these hazard ratios were statistically significant.

[...]

[zsn]

So...anecdotes + math.

Got it.

Yep....much the same way that "bottle-rocket + math = moon landing"

[Deleted User 89]

So...anecdotes + math.

Got it.

Yep....much the same way that "bottle-rocket + math = moon landing"

lmao

[DCHawk1]

In my original post of John Harwood's tweet, I considered adding this to help make my "case", but didn't, because the evidence isn't that strong:

JAMA May 11, 2020 Association of Treatment With Hydroxychloroquine or Azithromycin With In-Hospital Mortality in Patients With COVID-19 in New York State

Question Among patients with coronavirus disease 2019 (COVID-19), is there an association between use of hydroxychloroquine, with or without azithromycin, and in-hospital mortality?

Findings In a retrospective cohort study of 1438 patients hospitalized in metropolitan New York, compared with treatment with neither drug, the adjusted hazard ratio for in-hospital mortality for treatment with hydroxychloroquine alone was 1.08, for azithromycin alone was 0.56, and for combined hydroxychloroquine and azithromycin was 1.35. None of these hazard ratios were statistically significant.

[...]

I'm not trying to say that this isn't valid (for what it is) or that it isn't done with the best of intentions and the best methodology available.

I'm certain that it is. But the pretense that this is inarguably "science" while other citations of anecdotes are inarguably not is dishonest. And that dishonesty is on Harwood -- not you.

Statistical modeling is math. What this study is -- what any observational study is -- is rigorously applied anecdotal assessment enhanced by statistics.

Again, that doesn't mean it's wrong. I don't doubt that it's right, in fact. The issue is the unironic use of the term "science" as a talisman to ward off the evil spirits of Trumpism.

[DCHawk1]

So...anecdotes + math.

Got it.

Yep....much the same way that "bottle-rocket + math = moon landing"

OK.

Science denier.

[Deleted User 89]

i like the guy with minimal science background trying to tell three people with ample science background how science works and what it is

[seahawk]

Lulz.

[DCHawk1]

Nevar mind.

You right.

[sdoyel]

Lulz

[DCHawk1]

Enabler.

[Deleted User 89]

https://www.npr.org/sections/coronaviru ... virus-test

up to 48% false negatives

[Deleted User 289]

Dumb question/s time. How do they identify a false negative - how do they know it's "false"? They test on a different machine? How do they know the accuracy of the other machine? How do they know what it is that a person has if they do NOT test positive for COVID-19 but are clearly are ill? Is it possible there may be some other "coronavirus" strain?

[Shirley]

Of course!

You have to wonder what took the "Fiscal Responsibility Party" so long, since tax cuts are their go-to answer for...everything.

[zsn]

Dumb question/s time. How do they identify a false negative - how do they know it's "false"? They test on a different machine? How do they know the accuracy of the other machine? How do they know what it is that a person has if they do NOT test positive for COVID-19 but are clearly are ill? Is it possible there may be some other "coronavirus" strain?

Most tests are “validated” using known samples and are continually challenged to demonstrate that they remain in the validated state. The emissions test machine at your local auto shop and various instruments at the clinical lab are examples of these. Typical “system suitability” tests which demonstrate that the system remains in the validated state are performed at set intervals- usually once a day or every time a system is restarted after shutting down. Often the system suitability is run before the first test of the day and after the last test of the day to “bracket” and ensure that the system remained in the validated state throughout the analyses.

For these kinds of +/- tests they run a set of positive and negative controls periodically and assess the rate of false results.

Sorry if it’s too much detail on a quick primer on Good Laboratory Practice (yes, it’s capitalized).