A place to ask questions and add to probative and informative discussions associated with the various aspects of the field of fire investigation. -- FORUM RULES---BE CIVIL AND NO NAME CALLING, NO BELITTLING, NO BERATING, NO DENIGRATING others. Postings in violation of these rules can be removed or editted to remove the offending remarks at the discretion of the moderators and/or site administrator.
Re: Process of Elimination
Posted by:
dcarpenter (IP Logged)
Date: April 22, 2014 09:21AM
Asking more questions to try and understand your methodology or try and answer your questions. See below.
J L Mazerat Wrote:
-------------------------------------------------------
> This is my opinion only for all it is worth, and
> that may not be much. The process of elimination
> is used with the linear step scientific method
> that is shown in NFPA 921. Is it not true that
> once you form you hypothesis you start the testing
> process.
I am not sure what you are saying in the last sentence. Do you start the testing process before you formulate the hypothesis? Can you clarify the meaning of your last sentence?
>If your testing shows the hypothesis is
> false then you go back and develop another
> hypothesis. You then do the testing process again.
> This process continues until you get to the point
> where there is only one hypothesis that cannot be
> proven to be false. Is this not a process of
> elimination?
"Elimination" is the process where you have evidence that disproves your hypothesis. Thus, it is "eliminated." That is one potential outcome of the SM, but it is not the only outcome. What if you do not have evidence that disproves the hypothesis and it remains a valid hypothesis? What if you have a number of hypotheses that cannot be disproved? Thus, while the process may include some aspects that can be described as "elimination," the overall methodology is not characterized by this process.
It is the process of "falsification" that is the cornerstone of the SM, where one tries to disprove a hypothesis instead of "proving" a hypothesis. Yes, "falsification" can result in the disproving of a hypothesis, but it is not the only outcome of the testing of hypotheses.
>
> This process is used in all problem-solving
> methods. One will always have more than one
> possibility as to the cause of a fire. The
> process of elimination is what is used to
> determine what hypothesis is false and which one
> is not.
You are describing the "falsification" process in the testing of hypotheses and not a process of elimination.
>
> The last standing hypothesis is the correct one if
> it cannot be proven to be false. This is exactly
> what 921 calls for when it tell the user to repeat
> the steps if the hypothesis is proven to be
> false.
Read 921. It describes the selection of the final hypothesis as the one being "uniquely consistent" with the facts and evidence in a specific cases. You may not be able to disprove all but one hypothesis, but you are still able to reach a reliable determination with multiple remaining hypotheses.
The use of the word "correct" is not accurate. It is the most reliable determination based on the available evidence in a specific case. It could be correct, but how would you reliably know?
>
> I think one will have multiple hypotheses when
> starting an investigation. I know we start the
> investigation with no predetermined idea as to how
> the fire originated. There is a good chance one
> will not form only one hypothesis after evaluating
> the original data that is collected. There will be
> a good chance of having multiple hypotheses at
> this point. Now through a process of elimination
> you will test each in an attempt to prove it to be
> false. When you are finished there should be only
> one. There is a possibility that all one's
> hypotheses will be proven to be false.
You seem to be formulating hypotheses using the available data, which is good and consistent with the SM.
Your "a process of elimination" is really testing your hypothesis through "falsification."
There is no guarantee that this process will yield only one hypothesis that cannot be disproved. Again, that is one outcome of the testing of hypotheses, but it is not the only outcome.
>
> Yes, the lack of physical evidence can be
> sufficient evidence to eliminate a hypothesis. The
> fact you do not find parts of a airplane in the
> burning building would be sufficient evidence to
> say the hypothesis of a airplane hitting the
> structure and causing the fire has been eliminate.
Now your "disproving" a hypothesis because there is no evidence to support it. That is not in keeping with the SM and 921.
Let's use your example. Say I have a fire occur in a building in the path of a runway at a local airport. With respect to considering alternative hypotheses, one might consider a hypothesis of a plane crashing into the building. Thus, I may want to be looking for data that provides evidence that I could formulate that hypothesis. If I find no data and evidence of such, then I never formulate the hypothesis to begin with.
It makes no sense to formulate the hypothesis based on no evidence and then turn around and disprove the hypothesis because there is no evidence.
In addition, you now have the intractable problem that if one can formulate any hypothesis without evidence, then the number of potential hypotheses become infinite. So are you going to go through this process with this infinite series of hypotheses?
The SM provides a means to avoid testing an infinite number of hypotheses by only formulating those hypotheses with evidence. That is not to say that you did not consider alternative hypotheses based on your knowledge and experience, you just do not formulate them without any evidence. If you formulate a hypothesis based on evidence in the SM (as opposed to consider), it remains valid until and if, it can be disproved with additional evidence.
Douglas J. Carpenter, MScFPE, CFEI, PE, FSFPE
Vice President & Principal Engineer
Combustion Science & Engineering, Inc.
8940 Old Annapolis Road, Suite L
Columbia, MD 21045
(410) 884-3266
(410) 884-3267 (fax)
www.csefire.com