< Return to Video

34C3 - Science is broken

  • 0:00 - 0:15
    34c3 intro
  • 0:15 - 0:19
    Hanno Böck: Yeah, so many of you probably
    know me from doing things around IT
  • 0:19 - 0:25
    security, but I'm gonna surprise you to
    almost not talk about IT security today.
  • 0:25 - 0:32
    But I'm gonna ask the question "Can we
    trust the scientific method?". I want to
  • 0:32 - 0:39
    start this by giving you which is quite a
    simple example. So if we do science like
  • 0:39 - 0:45
    we start with the theory and then we are
    trying to test if it's true, right? So I
  • 0:45 - 0:50
    mean I said I'm not going to talk about IT
    security but I chose an example from IT
  • 0:50 - 0:57
    security or kind of from IT security. So
    there was a post on Reddit a while ago,
  • 0:57 - 1:01
    a picture from some book which claimed that
    if you use a Malachite crystal that can
  • 1:01 - 1:06
    protect you from computer viruses.
    Which... to me doesn't sound very
  • 1:06 - 1:11
    plausible, right? Like, these are crystals and
    if you put them on your computer, this book
  • 1:11 - 1:19
    claims this protects you from malware. But
    of course if we really want to know, we
  • 1:19 - 1:24
    could do a study on this. And if you say
    people don't do Studies on crazy things:
  • 1:24 - 1:29
    that's wrong. I mean people do studies on
    homeopathy or all kinds of crazy things
  • 1:29 - 1:35
    that are completely implausible. So we can
    do a study on this and what we will do is
  • 1:35 - 1:40
    we will do a randomized control trial,
    which is kind of the gold standard of
  • 1:40 - 1:46
    doing a test on these kinds of things. So
    this is our question: "Do Malachite
  • 1:46 - 1:52
    crystals prevent malware infections?" and
    how we would test that, our study design
  • 1:52 - 1:58
    is: ok, we take a group of maybe 20
    computer users. And then we split them
  • 1:58 - 2:06
    randomly to two groups, and then one group
    we'll give one of these crystals and tell
  • 2:06 - 2:11
    them: "Put them on your desk or on your
    computer.". Then we need, the other group
  • 2:11 - 2:16
    is our control group. That's very
    important because if we want to know if
  • 2:16 - 2:21
    they help we need another group to compare
    it to. And to rule out that there are any
  • 2:21 - 2:27
    kinds of placebo effects, we give these
    control groups a fake Malachite crystal so
  • 2:27 - 2:32
    we can compare them against each other.
    And then we wait for maybe six months and
  • 2:32 - 2:39
    then we check how many malware infections
    they had. Now, I didn't do that study, but
  • 2:39 - 2:45
    I simulated it with a Python script and
    given that I don't believe that this
  • 2:45 - 2:50
    theory is true I just simulated this as
    random data. So I'm not going to go
  • 2:50 - 2:55
    through the whole script but I'm just like
    generating, I'm assuming there can be
  • 2:55 - 3:00
    between 0 and 3 malware infections and
    it's totally random and then I compare the
  • 3:00 - 3:05
    two groups. And then I calculate something
    which is called a p-value which is a very
  • 3:05 - 3:11
    common thing in science whenever you do
    statistics. A p-value is, it's a bit
  • 3:11 - 3:17
    technical, but it's the probability that
    if you have no effect that you would get
  • 3:17 - 3:24
    this result. Which kind of in another way
    means, if you have 20 results in an
  • 3:24 - 3:29
    idealized world then one of them is a
    false positive which means one of them
  • 3:29 - 3:35
    says something happens although it
    doesn't. And in many fields of science
  • 3:35 - 3:41
    this p-value of 0.05 is considered that
    significant which is like these twenty
  • 3:41 - 3:49
    studies. So one error in twenty studies
    but as I said under idealized conditions.
  • 3:49 - 3:53
    So and as it's the script and I can run it
    in less than a second I just did it twenty
  • 3:53 - 4:00
    times instead of once. So here are my 20
    simulated studies and most of them look
  • 4:00 - 4:06
    not very interesting so of course we have
    a few random variations but nothing very
  • 4:06 - 4:12
    significant. Except if you look at this
    one study, it says the people with the
  • 4:12 - 4:17
    Malachite crystal had on average 1.8
    malware infections and the people with the
  • 4:17 - 4:25
    fake crystal had 0.8. So it means actually
    the crystal made it worse. But also this
  • 4:25 - 4:32
    result is significant because it has a
    p-value of 0.03. So of course we can
  • 4:32 - 4:36
    publish that, assuming I really did these
    studies.
  • 4:36 - 4:41
    applause
    B.: And the other studies we just forget
  • 4:41 - 4:46
    about. I mean they were not interesting
    right and who cares? Non significant
  • 4:46 - 4:53
    results... Okay so you have just seen that
    I created a significant result out of
  • 4:53 - 5:01
    random data. And that's concerning because
    people in science - I mean you can really do
  • 5:01 - 5:08
    that. And this phenomena is called
    publication bias. So what's happening here
  • 5:08 - 5:13
    is that, you're doing studies and if they
    get a positive result - meaning you're
  • 5:13 - 5:19
    seeing an effect, then you publish them
    and if there's no effect you just forget
  • 5:19 - 5:27
    about them. We learned earlier that with
    this p-value of 0.05 means 1 in 20 studies
  • 5:27 - 5:33
    is a false positive, but you usually don't
    see the studies that are not significant,
  • 5:33 - 5:39
    because they don't get published. And you
    may wonder: "Ok, what's stopping a
  • 5:39 - 5:44
    scientist from doing exactly this? What's
    stopping a scientist from just doing so
  • 5:44 - 5:48
    many experiments till one of them looks
    like it's a real result although it's just
  • 5:48 - 5:55
    a random fluke?". And the disconcerning
    answer to that is, it's usually nothing.
  • 5:57 - 6:04
    And this is not just a theoretical
    example. I want to give you an example,
  • 6:04 - 6:09
    that has quite some impact and that was
    researched very well, and that is a
  • 6:09 - 6:18
    research on antidepressants so called
    SSRIs. And in 2008 there was a study, the
  • 6:18 - 6:23
    interesting situation here was, that the
    US Food and Drug Administration, which is
  • 6:23 - 6:29
    the authority that decides whether a
    medical drug can be put on the market,
  • 6:29 - 6:35
    they had knowledge about all the studies
    that had been done to register this
  • 6:35 - 6:40
    medication. And then some researchers
    looked at that and compared it with what
  • 6:40 - 6:46
    has been published. And they figured out
    there were 38 studies that saw that these
  • 6:46 - 6:51
    medications had a real effect, had real
    improvements for patients. And from those
  • 6:51 - 6:57
    38 studies 37 got published. But then
    there were 36 studies that said: "These
  • 6:57 - 7:00
    medications don't really have any
    effect.", "They are not really better than
  • 7:00 - 7:07
    a placebo effect" and out of those only 14
    got published. And even from those 14
  • 7:07 - 7:11
    there were 11, where the researcher said,
    okay they have spent the result in a way
  • 7:11 - 7:18
    that it sounds like these medications do
    something. But they were also a bunch of
  • 7:18 - 7:22
    studies that were just not published
    because they had a negative result. And
  • 7:22 - 7:26
    it's clear that if you look at the
    published studies only and you ignore the
  • 7:26 - 7:29
    studies with a negative result that
    haven't been published, then these
  • 7:29 - 7:34
    medications look much better than they
    really are. And it's not like the earlier
  • 7:34 - 7:38
    example there is a real effect from
    antidepressants, but they are not as good
  • 7:38 - 7:40
    as people have believed in the past.
  • 7:43 - 7:46
    So we've learnt in theory with publication bias
  • 7:46 - 7:51
    you can create result out of nothing.
    But if you're a researcher and you have a
  • 7:51 - 7:55
    theory that's not true but you really want
    to publish something about it, that's not
  • 7:55 - 8:00
    really efficient, because you have to do
    20 studies on average to get one of these
  • 8:00 - 8:06
    random results that look like real
    results. So there are more efficient ways
  • 8:06 - 8:13
    to get to a result from nothing. If you're
    doing a study then there are a lot of
  • 8:13 - 8:17
    micro decisions you have to make, for
    example you may have dropouts from your
  • 8:17 - 8:22
    study where people, I don't know they move
    to another place or they - you now longer
  • 8:22 - 8:26
    reach them, so they are no longer part of
    your study. And there are different things
  • 8:26 - 8:30
    how you can handle that. Then you may have
    cornercase results, where you're not
  • 8:30 - 8:35
    entirely sure: "Is this an effect or not
    and how do you decide?", "How do you
  • 8:35 - 8:40
    exactly measure?". And then also you may
    be looking for different things, maybe
  • 8:40 - 8:47
    there are different tests you can do on
    people, and you may control for certain
  • 8:47 - 8:52
    variables like "Do you split men and women
    into separate?", "Do you see them
  • 8:52 - 8:56
    separately?" or "Do you separate them by
    age?". So there are many decisions you can
  • 8:56 - 9:02
    make while doing a study. And of course
    each of these decisions has a small effect
  • 9:02 - 9:10
    on the result. And it may very often be,
    that just by trying all the combinations
  • 9:10 - 9:15
    you will get a p-value that looks like
    it's statistically significant, although
  • 9:15 - 9:21
    there's no real effect. So and there's
    this term called p-Hacking which means
  • 9:21 - 9:26
    you're just adjusting your methods long
    enough, that you get a significant result.
  • 9:27 - 9:33
    And I'd like to point out here, that this
    is usually not that a scientist says: "Ok,
  • 9:33 - 9:36
    today I'm going to p-hack my result,
    because I know my theory is wrong but I
  • 9:36 - 9:42
    want to show it's true.". But it's a
    subconscious process, because usually the
  • 9:42 - 9:47
    scientists believe in their theories.
    Honestly. They honestly think that their
  • 9:47 - 9:52
    theory is true and that their research
    will show that. So they may subconsciously
  • 9:52 - 9:58
    say: "Ok, if I analyze my data like this
    it looks a bit better so I will do this.".
  • 9:58 - 10:05
    So subconsciously, they may p-hack
    themselves into getting a result that's
  • 10:05 - 10:11
    not really there. And again we can ask:
    "What is stopping scientists from
  • 10:11 - 10:22
    p-hacking?". And the concerning answer is
    the same: usually nothing. And I came to
  • 10:22 - 10:26
    this conclusion that I say: "Ok, the
    scientific method it's a way to create
  • 10:26 - 10:32
    evidence for whatever theory you like. No
    matter if it's true or not.". And you may
  • 10:32 - 10:36
    say: "That's a pretty bold thing to say.".
    and I'm saying this even though I'm not
  • 10:36 - 10:42
    even a scientist. I'm just like some
    hacker who, whatever... But I'm not alone
  • 10:42 - 10:48
    in this, like there's a paper from a
    famous researcher John Ioannidis, who
  • 10:48 - 10:52
    said: "Why most published research
    findings are false.". He published this in
  • 10:52 - 10:57
    2005 and if you look at the title, he
    doesn't really question that most research
  • 10:57 - 11:03
    findings are false. He only wants to give
    reasons why this is the case. And he makes
  • 11:03 - 11:08
    some very possible assumptions if you look
    at that many negative results don't get
  • 11:08 - 11:12
    published, and that you will have some
    bias. And it comes to a very plausible
  • 11:12 - 11:17
    conclusion, that this is the case and this
    is not even very controversial. If you ask
  • 11:17 - 11:23
    people who are doing what you can call
    science on science or meta science, who
  • 11:23 - 11:28
    look at scientific methodology, they will
    tell you: "Yeah, of course that's the
  • 11:28 - 11:32
    case.". Some will even say: "Yeah, that's
    how science works, that's what we
  • 11:32 - 11:38
    expect.". But I find it concerning. And if
    you take this seriously, it means: if you
  • 11:38 - 11:43
    read about a study, like in a newspaper,
    the default assumption should be 'that's
  • 11:43 - 11:51
    not true' - while we might usually think
    the opposite. And if science is a method
  • 11:51 - 11:56
    to create evidence for whatever you like,
    you can think about something really
  • 11:56 - 12:01
    crazy, like "Can people see into the future?",
    "Does our mind have
  • 12:01 - 12:10
    some extra perception where we can
    sense things that happen in an hour?". And
  • 12:10 - 12:16
    there was a psychologist called Daryl Bem
    and he thought that this is the case and
  • 12:16 - 12:20
    he published a study on it. It was titled
    "feeling the future". He did a lot of
  • 12:20 - 12:25
    experiments where he did something, and
    then something later happened, and he
  • 12:25 - 12:30
    thought he had statistical evidence that
    what happened later influenced what
  • 12:30 - 12:35
    happened earlier. So, I don't think that's
    very plausible - based on what we know
  • 12:35 - 12:42
    about the universe, but yeah... and it was
    published in a real psychology journal.
  • 12:42 - 12:47
    And a lot of things were wrong with this
    study. Basically, it's a very nice example
  • 12:47 - 12:51
    for p-hacking and just even a book by
    Daryl Bem, where he describes something
  • 12:51 - 12:55
    which basically looks like p-hacking,
    where he says that's how you do
  • 12:55 - 13:04
    psychology. But the study was absolutely
    in line with the existing standards in
  • 13:04 - 13:09
    Experimental Psychology. And that a lot of
    people found concerning. So, if you can
  • 13:09 - 13:14
    show that precognition is real, that you
    can see into the future, then what else
  • 13:14 - 13:19
    can you show and how can we trust our
    results? And psychology has debated this a
  • 13:19 - 13:22
    lot in the past couple of years. So
    there's a lot of talk about the
  • 13:22 - 13:30
    replication crisis in psychology. And many
    effects that psychology just thought were
  • 13:30 - 13:35
    true, they figured out, okay, if they try
    to repeat these experiments, they couldn't
  • 13:35 - 13:41
    get these results even though entire
    subfields were built on these results.
  • 13:44 - 13:48
    And I want to show you an example, which
    is one of the ones that is not discussed so
  • 13:48 - 13:56
    much. So there's a theory which is called
    moral licensing. And the idea is that if
  • 13:56 - 14:01
    you do something good, or something you
    think is good, then later basically you
  • 14:01 - 14:05
    behave like an asshole. Because you think
    I already did something good now, I don't
  • 14:05 - 14:11
    have to be so nice anymore. And there were
    some famous studies that had the theory,
  • 14:11 - 14:18
    that people consume organic food, that
    later they become more judgmental, or less
  • 14:18 - 14:28
    social, less nice to their peers. But just
    last week someone tried to replicate this
  • 14:28 - 14:33
    original experiments. And they tried it
    three times with more subjects and better
  • 14:33 - 14:39
    research methodology and they totally
    couldn't find that effect. But like what
  • 14:39 - 14:44
    you've seen here is lots of media
    articles. I have not found a single
  • 14:44 - 14:51
    article reporting that this could not be
    replicated. Maybe they will come but yeah
  • 14:51 - 14:57
    there's just a very recent example. But
    now I want to have a small warning for you
  • 14:57 - 15:01
    because you may think now "yeah these
    psychologists, that all sounds very
  • 15:01 - 15:05
    fishy and they even believe in
    precognition and whatever", but maybe your
  • 15:05 - 15:10
    field is not much better maybe you just
    don't know about it yet because nobody
  • 15:10 - 15:16
    else has started replicating studies in
    your field. And there are other fields
  • 15:16 - 15:22
    that have replication problems and some
    much worse for example the pharma company
  • 15:22 - 15:27
    Amgen in 2012 they published something
    where they said "We have tried to
  • 15:27 - 15:33
    replicate cancer research and preclinical
    research" that is stuff in a petri dish or
  • 15:33 - 15:39
    animal experiments so not drugs on humans
    but what happens before you develop a drug
  • 15:39 - 15:45
    and they were only able to replicate 47
    out of 53 studies. And these were they
  • 15:45 - 15:50
    said landmark studies, so studies that
    have been published in the best journals.
  • 15:50 - 15:54
    Now there are a few problems with this
    publication because they have not
  • 15:54 - 15:59
    published their applications they have not
    told us which studies these were that they
  • 15:59 - 16:03
    could not replicate. In the meantime I
    think they have published three of these
  • 16:03 - 16:07
    replications but most of it is a bit in
    the dark which points to another problem
  • 16:07 - 16:11
    because they say they did this because
    they collaborated with the original
  • 16:11 - 16:16
    researchers and they only did this by
    agreeing that they would not publish the
  • 16:16 - 16:22
    results. But it still sounds very
    concerning so but some fields don't have a
  • 16:22 - 16:27
    replication problem because just nobody is
    trying to replicate previous results I
  • 16:27 - 16:34
    mean then you will never know if your
    results hold up. So what can be done about
  • 16:34 - 16:43
    all this and fundamentally I think the
    core issue here is that the scientific
  • 16:43 - 16:50
    process is tied together with results, so
    we do a study and only after that we
  • 16:50 - 16:55
    decide whether it's going to be published.
    Or we do a study and only after we have
  • 16:55 - 17:01
    the data we're trying to analyze it. So
    essentially we need to decouple the
  • 17:01 - 17:10
    scientific process from its results and
    one way of doing that is pre-registration
  • 17:10 - 17:14
    so what you're doing there is that before
    you start doing a study you will register
  • 17:14 - 17:20
    it in a public register and say "I'm gonna
    do a study like on this medication or
  • 17:20 - 17:26
    whatever on this psychological effect" and
    that's how I'm gonna do it and then later
  • 17:26 - 17:34
    on people can check if you really did
    that. And yeah that's what I said. And this
  • 17:34 - 17:41
    is more or less standard practice in
    medical drug trials the summary about it
  • 17:41 - 17:47
    is it does not work very well but it's
    better than nothing. So, and the problem
  • 17:47 - 17:52
    is mostly enforcement so people register
    study and then don't publish it and
  • 17:52 - 17:57
    nothing happens to them even though they
    are legally required to publish it. And
  • 17:57 - 18:02
    there are two campaigns I'd like to point
    out, there's the all trials campaign which
  • 18:02 - 18:08
    has been started by Ben Goldacre he's a
    doctor from the UK and they like demand
  • 18:08 - 18:13
    that like every trial it's done on
    medication should be published. And
  • 18:13 - 18:19
    there's also a project by the same guy the
    compare project and they are trying to see
  • 18:19 - 18:25
    if a medical trial has been registered and
    later published did they do the same or
  • 18:25 - 18:29
    did they change something in their
    protocol and was there a reason for it or
  • 18:29 - 18:37
    did they just change it to get a result,
    which they otherwise wouldn't get.But then
  • 18:37 - 18:41
    again like these issues in medicine they
    offer get a lot of attention and for good
  • 18:41 - 18:47
    reasons because if we have bad science in
    medicine then people die, that's pretty
  • 18:47 - 18:53
    immediate and pretty massive. But if you
    read about this you always have to think
  • 18:53 - 18:59
    that these issues in drug trials at least
    they have pre-registration, most
  • 18:59 - 19:04
    scientific fields don't bother doing
    anything like that. So whenever you hear
  • 19:04 - 19:08
    something about maybe about publication
    bias in medicine you should always think
  • 19:08 - 19:13
    the same thing happens in many fields of
    science and usually nobody is doing
  • 19:13 - 19:19
    anything about it. And particularly to
    this audience I'd like to say there's
  • 19:19 - 19:24
    currently a big trend that people from
    computer science want to revolutionize
  • 19:24 - 19:30
    medicine: big data and machine learning,
    these things, which in principle is ok but
  • 19:30 - 19:35
    I know a lot of people in medicine are
    very worried about this and the reason is,
  • 19:35 - 19:39
    that these computer science people don't
    have the same scientific standards as
  • 19:39 - 19:44
    people in medicine expect them and might
    say "Yeah we don't need really need to do
  • 19:44 - 19:50
    a study on this it's obvious that this
    helps" and that is worrying and I come
  • 19:50 - 19:54
    from computer science and I very well
    understand that people from medicine are
  • 19:54 - 20:01
    worried about this. So there's an idea
    that goes even further as pre-registration
  • 20:01 - 20:05
    and it's called registered reports. There
    is a couple of years ago some scientists
  • 20:05 - 20:11
    wrote an open letter to the Guardian where
    they.. that was published there and the idea
  • 20:11 - 20:16
    there is that you turn the scientific
    publication process upside down, so if you
  • 20:16 - 20:21
    want to do a study the first thing you
    would do with the register report is, you
  • 20:21 - 20:27
    submit your design your study design
    protocol to the journal and then the
  • 20:27 - 20:33
    journal decides whether they will publish
    that before they see any result, because
  • 20:33 - 20:37
    then you can prevent publication bias and
    then you prevent the journals only publish
  • 20:37 - 20:43
    the nice findings and ignore the negative
    findings. And then you do the study and
  • 20:43 - 20:46
    then it gets published but it gets
    published independent of what the result
  • 20:46 - 20:54
    was. And there of course other things you
    can do to improve science, there's a lot
  • 20:54 - 20:59
    of talk about sharing data, sharing code,
    sharing methods because if you want to
  • 20:59 - 21:04
    replicate a study it's of course easier if
    you have access to all the details how the
  • 21:04 - 21:11
    original study was done. Then you could
    say "Okay we could do large
  • 21:11 - 21:15
    collaborations" because many studies are
    just too small if you have a study with
  • 21:15 - 21:20
    twenty people you just don't get a very
    reliable outcome. So maybe in many
  • 21:20 - 21:26
    situations it would be better get together
    10 teams of scientists and let them all do
  • 21:26 - 21:32
    a big study together and then you can
    reliably answer a question. And also some
  • 21:32 - 21:36
    people propose just to get higher
    statistical thresholds that p-value of
  • 21:36 - 21:42
    0.05 means practically nothing. There was
    recently a paper that just argued which
  • 21:42 - 21:48
    would just like put the dot one more to
    the left and have 0.005 and that would
  • 21:48 - 21:55
    already solve a lot of problems. And for
    example in physics they have they have
  • 21:55 - 22:01
    something called Sigma 5 which is I think
    zero point and then 5 zeroes and 3 or
  • 22:01 - 22:08
    something like that so in physics they
    have much higher statistical thresholds.
  • 22:08 - 22:13
    Now whatever if you're working in any
    scientific field you might ask yourself
  • 22:13 - 22:20
    like "If we have statistic results are
    they pre registered in any way and do we
  • 22:20 - 22:26
    publish negative results?" like we tested
    an effect and we got nothing and are there
  • 22:26 - 22:32
    replications of all relevant results and I
    would say if you answer all these
  • 22:32 - 22:36
    questions with "no" which I think many
    people will do, then you're not really
  • 22:36 - 22:42
    doing science what you're doing is the
    alchemy of our time.
  • 22:42 - 22:50
    Applause
    Thanks.
  • 22:50 - 22:54
    Herald: Thank you very much..
    Hanno: No I have more, sorry, I have
  • 22:54 - 23:03
    three more slides, that was not the
    finishing line. Big issue is also that
  • 23:03 - 23:10
    there are bad incentives in science, so a
    very standard thing to evaluate the impact
  • 23:10 - 23:16
    of science is citation counts for you say
    "if your scientific study is cited a lot
  • 23:16 - 23:19
    then this is a good thing and if your
    journal is cited a lot this is a good
  • 23:19 - 23:22
    thing" and this for example the impact
    factor but there are also other
  • 23:22 - 23:27
    measurements. And also universities like
    publicity so if your study gets a lot of
  • 23:27 - 23:33
    media reports then your press department
    likes you. And these incentives tend to
  • 23:33 - 23:40
    favor interesting results but they don't
    favor correct results and this is bad
  • 23:40 - 23:45
    because if we are realistic most results
    are not that interesting, most results
  • 23:45 - 23:50
    will be "Yeah we have this interesting and
    counterintuitive theory and it's totally
  • 23:50 - 24:00
    wrong" and then there's this idea that
    science is self-correcting. So if you
  • 24:00 - 24:05
    confront scientists with these issues with
    publication bias and peer hacking surely
  • 24:05 - 24:12
    they will immediately change that's what
    scientists do right? And I want to cite
  • 24:12 - 24:16
    something here with this sorry it's a bit
    long but "There are some evidence that
  • 24:16 - 24:21
    inferior statistical tests are commonly
    used research which yields non significant
  • 24:21 - 24:29
    results is not published." That sounds
    like publication bias and then it also
  • 24:29 - 24:32
    says: "Significant results published in
    these fields are seldom verified by
  • 24:32 - 24:38
    independent replication" so it seems
    there's a replication problem. These wise
  • 24:38 - 24:47
    words were set in 1959, so by a
    statistician called Theodore Sterling and
  • 24:47 - 24:52
    because science is so self-correcting in
    1995 he complained that this article
  • 24:52 - 24:56
    presents evidence that published result of
    scientific investigations are not a
  • 24:56 - 25:01
    representative sample of all scientific
    studies. "These results also indicate that
  • 25:01 - 25:07
    practice leading to publication bias has
    not changed over a period of 30 years" and
  • 25:07 - 25:13
    here we are in 2018 and publication bias
    is still a problem. So if science is self-
  • 25:13 - 25:21
    correcting then it's pretty damn slow in
    correcting itself, right? And finally I
  • 25:21 - 25:27
    would like to ask you, if you're prepared
    for boring science, because ultimately, I
  • 25:27 - 25:32
    think, we have a choice between what I
    would like to call TEDTalk science and
  • 25:32 - 25:41
    boring science..
    Applause
  • 25:41 - 25:47
    .. so with tedtalk science we get mostly
    positive and surprising results and
  • 25:47 - 25:53
    interesting results we have large defects
    many citations lots of media attention and
  • 25:53 - 26:00
    you may have a TED talk about it.
    Unfortunately usually it's not true and I
  • 26:00 - 26:04
    would like to propose boring science as
    the alternative which is mostly negative
  • 26:04 - 26:12
    results, pretty boring, small effects but
    it may be closer to the truth. And I would
  • 26:12 - 26:18
    like to have boring science but I know
    it's a pretty tough sell. Sorry I didn't
  • 26:18 - 26:35
    hear that. Yeah, thanks for listening.
    Applause
  • 26:35 - 26:38
    Herald: Thank you.
    Hanno: Two questions, or?
  • 26:38 - 26:41
    Herald: We don't have that much time for
    questions, three minutes, three minutes
  • 26:41 - 26:45
    guys. Question one - shoot.
    Mic: This isn't a question but I just
  • 26:45 - 26:49
    wanted to comment Hanno you missed out a
    very critical topic here, which is the use
  • 26:49 - 26:53
    of Bayesian probability. So you did
    conflate p-values with the scientific
  • 26:53 - 26:57
    method which isn't.. which gave the rest
    of you talk. I felt a slightly unnecessary
  • 26:57 - 27:02
    anti science slant. On p, p-values isn't
    the be-all and end-all of the scientific
  • 27:02 - 27:07
    method so p-values is sort of calculating
    the probability that your data will happen
  • 27:07 - 27:11
    given that no hypothesis is true whereas
    Bayesian probability would be calculating
  • 27:11 - 27:16
    the probability that your hypothesis is
    true given the data and more and more
  • 27:16 - 27:20
    scientists are slowly starting to realize
    that this sort of method is probably a
  • 27:20 - 27:26
    better way of doing science than p-values.
    So this is probably a a third alternative
  • 27:26 - 27:30
    to your sort of proposal boring science is
    doing the other side's Bayesian
  • 27:30 - 27:34
    probability.
    Hanno: Sorry yeah, I agree with you I
  • 27:34 - 27:38
    unfortunately I only had
    half an hour here.
  • 27:38 - 27:41
    Herald: Where are you going after this
    like where are we going after this lecture
  • 27:41 - 27:46
    can they find you somewhere in the bar?
    Hanno: I know him..
  • 27:46 - 27:51
    Herald: You know science is broken but
    then scientists it's a little bit like the
  • 27:51 - 27:55
    next lecture actually that's waiting there
    it's like: "you scratch my back and I
  • 27:55 - 27:59
    scratch yours for publication". Hanno:
    Maybe two more minutes?
  • 27:59 - 28:05
    Herald: One minute.
    Please go ahead.
  • 28:05 - 28:12
    Mic: Yeah hi, thank you for your talk. I'm
    curious so you've raised, you know, ways
  • 28:12 - 28:16
    we can address this assuming good actors,
    assuming people who want to do better
  • 28:16 - 28:21
    science that this happens out of ignorance
    or willful ignorance. What do we do about
  • 28:21 - 28:26
    bad actors. So for example the medical
    community drug companies, maybe they
  • 28:26 - 28:30
    really like the idea of being profitably
    incentivized by these random control
  • 28:30 - 28:35
    trials, to make out essentially a placebo
    do something. How do we begin to address
  • 28:35 - 28:41
    them current trying to maliciously p-hack
    or maliciously abuse the pre-reg system or
  • 28:41 - 28:44
    something like that?
    Hanno: I mean it's a big question, right?
  • 28:44 - 28:51
    But I think if the standards are kind of
    confining you so much that there's not
  • 28:51 - 28:56
    much room to cheat that's way out right
    and a basis and also I don't think
  • 28:56 - 29:00
    deliberate cheating is that much of a
    problem, I actually really think the
  • 29:00 - 29:07
    bigger problem is people honestly
    believe what they do is true.
  • 29:07 - 29:16
    Herald: Okay one last, you sir, please?
    Mic: So the value in science is often an
  • 29:16 - 29:21
    account of publications right? Account of
    citations so and so on, so is it true that
  • 29:21 - 29:25
    to improve this situation you've
    described, journals of whose publications
  • 29:25 - 29:31
    are available, who are like prospective,
    should impose more higher standards so the
  • 29:31 - 29:37
    journals are those who must like raise the
    bar, they should enforce publication of
  • 29:37 - 29:43
    protocols before like accepting and etc
    etc. So is it journals who should, like,
  • 29:43 - 29:49
    do work on that or can we regular
    scientists do something also? I mean you
  • 29:49 - 29:53
    can publish in the journals that have
    better standards, right? There are
  • 29:53 - 29:59
    journals that have these registered
    reports, but of course I mean as a single
  • 29:59 - 30:03
    scientist is always difficult because
    you're playing in a system that has all
  • 30:03 - 30:07
    these wrong incentives.
    Herald: Okay guys that's it, we have to
  • 30:07 - 30:13
    shut down. Please. There is a reference
    better science dot-org, go there, and one
  • 30:13 - 30:16
    last request give really warm applause!
  • 30:16 - 30:24
    Applause
  • 30:24 - 30:29
    34c3 outro
  • 30:29 - 30:46
    subtitles created by c3subtitles.de
    in the year 2018. Join, and help us!
Title:
34C3 - Science is broken
Description:

more » « less
Video Language:
English
Duration:
30:46

English subtitles

Revisions