Jump to content


Photo

Is the Scientific Method dead?

  • Please log in to reply
22 replies to this topic

#1 Charlie B

Charlie B

    Viking 1

  • -----
  • Posts: 970
  • Joined: 22 Mar 2008
  • Loc: Sterling, Virginia

Posted 06 August 2013 - 05:40 PM

“The End of Theory: The Data Deluge Makes The Scientific Method Obsolete,” “correlation supercedes causation,”– Chris Anderson, editor-in-chief of Wired.

“With four parameters I can fit an elephant, and with five I can make him wiggle his trunk” – John von Neumann.

“Cancer causes smoking” – R. A. Fischer of Fischer’s linear discriminant (poss. apocryphal).

Here is Chris Anderson's article suggesting the scientific method is dead. I don't agree, but that may be because I'm a reactionary old man too well trained in the scientific method.

Regards,

Charlie B

#2 Jay_Bird

Jay_Bird

    Surveyor 1

  • *****
  • Posts: 1701
  • Joined: 04 Jan 2006
  • Loc: USA

Posted 06 August 2013 - 05:55 PM

Correlation still needs some framework, preferably "multiple working hypotheses"

I don't see it as the end of scientific method but instead a boost for empiricism; however some theory still is needed for understanding, what my favorite prof called 'intuitive conceptual grasp'.

#3 EJN

EJN

    Vanguard

  • *****
  • Posts: 2266
  • Joined: 01 Nov 2005
  • Loc: 53 miles west of Venus

Posted 06 August 2013 - 06:09 PM

Working in IT, I keep thinking of the phrase GIGO (garbage in garbage out).
Sure, the ability exists now to store and manipulate enormous amounts
of data. But is it meaningful and accurate data?

#4 llanitedave

llanitedave

    Humble Megalomaniac

  • *****
  • Posts: 22667
  • Joined: 25 Sep 2005
  • Loc: Amargosa Valley, NV, USA

Posted 06 August 2013 - 06:12 PM

It's still conjecture and refutation.

#5 Charlie B

Charlie B

    Viking 1

  • -----
  • Posts: 970
  • Joined: 22 Mar 2008
  • Loc: Sterling, Virginia

Posted 06 August 2013 - 06:32 PM

I see that I'm not the only reactionary old man around.

Charlie B :jump:

#6 UND_astrophysics

UND_astrophysics

    Ranger 4

  • -----
  • Posts: 323
  • Joined: 19 Jan 2013

Posted 06 August 2013 - 07:20 PM


Just glanced at the article... obviously written by someone who lacks a firm understanding of empiricism judging by the examples and comparisons he tried to use.

#7 derangedhermit

derangedhermit

    Apollo

  • -----
  • Posts: 1158
  • Joined: 07 Oct 2009
  • Loc: USA

Posted 06 August 2013 - 07:50 PM

Wired, although aging, is the epitomy of the tech hype machine as money maker. This article seems to have succeeded to the extent it got a post here, but failed in my case when I visited their site due to my ad and tracker strippers. The page barely runs when those are blocked, by they way.

It would make as much sense to spend time considering this from Wired as if it were from the Watchtower or DC Comics.

#8 UND_astrophysics

UND_astrophysics

    Ranger 4

  • -----
  • Posts: 323
  • Joined: 19 Jan 2013

Posted 06 August 2013 - 07:55 PM

It would make as much sense to spend time considering this from Wired as if it were from the Watchtower or DC Comics.


:roflmao: :roflmao: :roflmao:
.
.
.


#9 Otto Piechowski

Otto Piechowski

    Apollo

  • -----
  • Posts: 1467
  • Joined: 20 Sep 2005
  • Loc: Lexington, KY

Posted 06 August 2013 - 08:23 PM

What is the connection between vast amounts of data and the scientific method? Wouldn't the method be the same if only a few data points were considered or many more data points?

Otto

#10 Rick Woods

Rick Woods

    Voyager 1

  • *****
  • Posts: 14938
  • Joined: 27 Jan 2005
  • Loc: Inner Solar System

Posted 06 August 2013 - 09:56 PM

Yeah, but it's more, innit?

#11 llanitedave

llanitedave

    Humble Megalomaniac

  • *****
  • Posts: 22667
  • Joined: 25 Sep 2005
  • Loc: Amargosa Valley, NV, USA

Posted 07 August 2013 - 12:01 AM

What is the connection between vast amounts of data and the scientific method? Wouldn't the method be the same if only a few data points were considered or many more data points?

Otto


Yep.


BTW, I've pretty much renounced "The Scientific Method" as being a formal stepwise heuristic. Popper did that much for me. I've taken the title of one of his compilations, "Conjectures and Refutations" as being perfectly adequate to describe the process of scientific exploration. I like that you can also apply it pretty much universally, to any attempt at gaining knowledge on any level. (I don't necessarily agree with everything he has to say about what is important in scientific discovery, but I do like his summation of it.)

#12 Charlie B

Charlie B

    Viking 1

  • -----
  • Posts: 970
  • Joined: 22 Mar 2008
  • Loc: Sterling, Virginia

Posted 07 August 2013 - 06:41 AM

It would make as much sense to spend time considering this from Wired as if it were from the Watchtower or DC Comics.



Careful! I grew up on DC Comics.

Charlie B ;)

#13 Otto Piechowski

Otto Piechowski

    Apollo

  • -----
  • Posts: 1467
  • Joined: 20 Sep 2005
  • Loc: Lexington, KY

Posted 07 August 2013 - 08:00 AM

Dave,

What does "formal stepwise heuristic" mean?

Otto

#14 derangedhermit

derangedhermit

    Apollo

  • -----
  • Posts: 1158
  • Joined: 07 Oct 2009
  • Loc: USA

Posted 07 August 2013 - 03:19 PM

Careful! I grew up on DC Comics.

Charlie B ;)


If it had come from them back then, there would have been good illustrations. Honest entertainment, not disguising itself.

#15 llanitedave

llanitedave

    Humble Megalomaniac

  • *****
  • Posts: 22667
  • Joined: 25 Sep 2005
  • Loc: Amargosa Valley, NV, USA

Posted 07 August 2013 - 08:22 PM

Dave,

What does "formal stepwise heuristic" mean?

Otto


"Formal" may have been too strong a word, since there is some variation in how the steps are defined. The basic sequence that I was taught in elementary school is roughly this:

1. Make an observation about something that happens.
2. Question how it happened
3. Create a hypothesis to explain what is going on as an answer to the question.
4. Conduct an experiment to see if your hypothesis was right.
5. Make a conclusion from the results of the experiment
6. Profit!!!

More generally, something along those lines still goes on, but the process is actually pretty fluid. The Conjectures and Refutations approach is more like:

1. Consider some system that affects the world.
2. Propose a conjectural hypothesis attempting to explain that system or some part of it.
3. Attempt to refute that hypothesis, through logic, experiment, historical records, or any other means that might seem appropriate. Attack it as deeply and forcefully as possible.
4. If the conjecture survives all attempts at refutation, it may be provisionally accepted. Otherwise, it must be either rejected or modified.

Step 3 is not really a step per se, since here is no time limit on refutation. An idea may survive every prior attempt at a refutation, but when new tools and concepts appear, those can be applied to attack the idea from a different angle. Thus, there is no real "conclusion" in any sense of closure -- it's always subject to future attempts at falsification. Unlike the steps we learned in school, this one doesn't give you neatly tied up ends to your theories. It's always a "what have you done for me lately" situation.

This approach works not only for science, but just about every tangible problem category I can think of.

#16 derangedhermit

derangedhermit

    Apollo

  • -----
  • Posts: 1158
  • Joined: 07 Oct 2009
  • Loc: USA

Posted 07 August 2013 - 09:51 PM

The notion that the process is linear or synchronous is obviously an oversimplification. People don't even make dinner in such a regimented fashion.

(Unordered list, can occur simultaneously or in any order.)
- Collect data about some part of reality. The data collection must be repeatable by other parties.
- Create or modify a model that describes some part of reality. The model must contain statements that can be compared to collectible data.
- Compare an existing model to existing data; if it is the model, or one of the models, that best matches the data, then note it as such, until it isn't.

Sometimes the data collectors get ahead of the theorists.
Sometimes the theorists get ahead of the data collectors.

Logic isn't enough.

#17 Otto Piechowski

Otto Piechowski

    Apollo

  • -----
  • Posts: 1467
  • Joined: 20 Sep 2005
  • Loc: Lexington, KY

Posted 08 August 2013 - 01:26 AM

Thank you Dave. That was clear and helpful.

Otto

#18 GregLee1

GregLee1

    Viking 1

  • -----
  • Posts: 669
  • Joined: 21 Jul 2013
  • Loc: Waimanalo, HI

Posted 10 August 2013 - 09:58 PM

The author would have profited from reading Braithwaite's Scientific Explanation, which distinguishes between model, essentially a data summary, and theory, which explains something to us. Science is not about summarizing facts, but about understanding them.

#19 Rick Woods

Rick Woods

    Voyager 1

  • *****
  • Posts: 14938
  • Joined: 27 Jan 2005
  • Loc: Inner Solar System

Posted 11 August 2013 - 02:13 PM

6. Profit!!!


You're watching too much South Park, Dave.

#20 llanitedave

llanitedave

    Humble Megalomaniac

  • *****
  • Posts: 22667
  • Joined: 25 Sep 2005
  • Loc: Amargosa Valley, NV, USA

Posted 11 August 2013 - 08:12 PM

I've never, ever seen South Park, not even once.

#21 Rick Woods

Rick Woods

    Voyager 1

  • *****
  • Posts: 14938
  • Joined: 27 Jan 2005
  • Loc: Inner Solar System

Posted 12 August 2013 - 02:40 AM

I've never, ever seen South Park, not even once.


Well, the Underwear Gnomes (who came in the night and stole your underwear) had a business plan:

Phase 1: Collect underwear
Phase 2: ?
Phase 3: Profit!

Nobody actualy knew what phase 2 was. (They probably hired a consulting firm to create the plan.)

#22 russell23

russell23

    Aurora

  • *****
  • Posts: 4506
  • Joined: 31 May 2009
  • Loc: Upstate NY

Posted 12 August 2013 - 07:47 AM

Dave,

What does "formal stepwise heuristic" mean?

Otto


"Formal" may have been too strong a word, since there is some variation in how the steps are defined. The basic sequence that I was taught in elementary school is roughly this:

1. Make an observation about something that happens.
2. Question how it happened
3. Create a hypothesis to explain what is going on as an answer to the question.
4. Conduct an experiment to see if your hypothesis was right.
5. Make a conclusion from the results of the experiment
6. Profit!!!

More generally, something along those lines still goes on, but the process is actually pretty fluid. The Conjectures and Refutations approach is more like:

1. Consider some system that affects the world.
2. Propose a conjectural hypothesis attempting to explain that system or some part of it.
3. Attempt to refute that hypothesis, through logic, experiment, historical records, or any other means that might seem appropriate. Attack it as deeply and forcefully as possible.
4. If the conjecture survives all attempts at refutation, it may be provisionally accepted. Otherwise, it must be either rejected or modified.

Step 3 is not really a step per se, since here is no time limit on refutation. An idea may survive every prior attempt at a refutation, but when new tools and concepts appear, those can be applied to attack the idea from a different angle. Thus, there is no real "conclusion" in any sense of closure -- it's always subject to future attempts at falsification. Unlike the steps we learned in school, this one doesn't give you neatly tied up ends to your theories. It's always a "what have you done for me lately" situation.

This approach works not only for science, but just about every tangible problem category I can think of.


Excellent post Dave!

As a high school science teacher I have never liked or taught the "cookbook" scientific method. Instead I provide my students with a list of science vocabulary terms they should have learned in elementary and middle school (observation, inference, measurement, hypothesis ... etc). I also provide them with a sheet titled "Ten Important Science Attitudes" which is based off an article I read somewhere titled "Twenty Important Science Attitudes". I condensed the ideas in the original article for my students. It includes things like - understanding that measurements have uncertainty, scientists look for natural explanations, ... and so on.

I have the students contrast some of the vocabulary words that students often get confused. Throughout the year I refer to the ten attitudes sheet.

There are a few standard ways the "scientific method" is taught at the younger grade levels that have to be undone in high school. The first is the "cookbook" approach to how science works. Students get this drilled into them in elementary and middle school.

The second is the definition of "hypothesis". All they learn in the younger grades is that a hypothesis is an "educated guess" - which is a definition that reduces the meaning of hypothesis to something totally useless. I prefer to define a hypothesis as "a proposed explanation for an observation or observed phenomenon that may be tested with experimentation". Not perfect, but better than "educated guess". The problem is that students don't know what "educated" or "guess" mean in the context of science. Educated means that the researcher is basing the hypothesis upon what science has already established as scientific fact. The "guess" is really the testable proposed explanation.

If you ask students what a hypothesis is most will parrot "educated guess". If you ask them "What does that mean?" you will get blank stares because they don't know.

Another point I emphasize with my students is the difference between a Natural Law and a Theory. Students often think Natural Laws become Theories with enough evidence - which is incorrect.

Dave

#23 llanitedave

llanitedave

    Humble Megalomaniac

  • *****
  • Posts: 22667
  • Joined: 25 Sep 2005
  • Loc: Amargosa Valley, NV, USA

Posted 12 August 2013 - 10:02 PM

Good point about the "educated guess". Everyone knows what a guess is, but nobody seems really sure how you go about educating one.

The hardest concept to learn in science is the requirement to construct all guesses, conjectures, hypotheses, claims, etc. in testable form. And even if you know the concept, it's not always easy to put into practice, either.

This is another area where the scientific approach can be said to apply to everything. If a claim is not testable, at least in some fundamental sense, it's meaningless. I don't know of any field where this fails to apply.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics