[Top] [All Lists]

Re: [ontolog-forum] Polysemy and Subjectivity in Ontolgies - theHDBIexam

To: <doug@xxxxxxxxxx>, "'[ontolog-forum] '" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "Rich Cooper" <rich@xxxxxxxxxxxxxxxxxxxxxx>
Date: Wed, 17 Nov 2010 10:43:23 -0800
Message-id: <20101117184338.283D4138D03@xxxxxxxxxxxxxxxxx>

Hi Doug,


Sorry for the delay in responding - I finally have time to do so.  Please see below,





Rich Cooper


Rich AT EnglishLogicKernel DOT com

9 4 9 \ 5 2 5 - 5 7 1 2


-----Original Message-----
From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of doug foxvog
Sent: Saturday, November 06, 2010 9:08 PM
To: [ontolog-forum]
Subject: Re: [ontolog-forum] Polysemy and Subjectivity in Ontolgies - theHDBIexample


Doug Foxvog wrote:


On Sat, November 6, 2010 16:22, Rich Cooper said:

> ...

> I'm focusing on the METHOD of starting with a set of data points based on

> a theory (-good, bad, indifferent theory ok-),


Note that the data points are based on subjective classification of each of the answers for each of the questions.  This data set is not based on one dimension for each question, but on these subjective classifications, which may have been refined during analysis of the results.


True.  But that is also true of most human linguistics interactions.  All human perception is subjective, and all human action is subjective.  Certainly language is subjective, and yet we are able to communicate conceptually complex thoughts with all that ambiguity going both ways in the conversations.  


Medicine is almost completely based on subjective interpretations of physical conditions, which can be mapped by means of blood tests, etc, into syndromes that have similar issues.  So the multidimensional decision making process is performed by doctors every day.  It is only objective when certain measurements are taken that all doctors agree on - white blood cell count, various protein titrations, other numerical measurements that can conceivably be called objective (but only if you have the training of an MD in dealing with these subjective classifications).  


The test designers defined what they deemed to be opposite attributes,

such that if one was scored high for one attribute, they scored low for an opposite attribute and vice versa.  This made it hard for test takers to score high on two opposite attributes.


Yes, that is an aspect of the four-pole theories that the Myers-Briggs team did after studying Jung's EXTREMELY subjective musings on personal drives.  These are the E-I, S-N, T-F, J-P poles of the measurement questionnaires they used.  


The same thing is true of electricity (positive and negative poles), gravity (helium goes up, rocks go down, in the absence of other forces to stop them from doing what comes naturally), and many other things that are talked about in NLP utterances.  I don't see why you are objecting to my assertion of objective measurements by observers for phenomenon demonstrating subjective perceptions by experimental subjects.  


There is no first-principles explanation for the clusters - they have to be subjectively explained after being objectively classified by the observers.  


> and CLASSIFYing based on how

> the set of points clusters in the multidimensional volume of said data > points.  Each said cluster I can interpret as a CLASS in the metalanguage > of sets.


Yes, they are classes based upon the classes defined by the test makers.


Yes, and presumably the test makers wanted the test takers to regurgitate approximately the same classes.  How did they know that the subjects would demonstrate clustering patterns?  That theory, though vague and subjective, still has very valuable implications.  Most business processes are just as subjective, or even more so, and just as in need of organized analysis by the business principals.  


> So regardless of whether the theory fits other tests, choosing clusters of > points in a multi-D volume is equivalent to CLASSIFYing subclasses of the > set of all points - the Universe.


It seems that the clusters were selected in 2D spaces. 


For that paper, yes, 2D plots were published, but typically, pattern recognition methods and algorithms are based on any number N of dimensions, with Multi-D projections into 2D very common in the discipline.  


Yes, defining clusters either according to some rule, or totally arbitrarily results in classifying the elements of the cluster as members of that cluster.  If a rule is defined specifying that any test that meets the rule is a member of the class, then it does define a subclass of the set of tests, which itself is a subclass of the universal set.


> Do we still agree, or do you think otherwise?


I initially thought this discussion was about the automatic

of new classes.  Now it is being discussed as an example of polysemy

of adjectives used by the sample subjects. 


It is about both polysemy of language used in the test, and about the placement of decision surfaces in the multi-D space which the observers can agree on as meaningful, and that is the point of this exercise.  Most communication, even when followed by precise measurement of supposedly meaningful concepts, is about a variety of those meaningful concepts AS PERCEIVED IN THE OBSERVERS' MINDS.  


Not only does this seem very different, but polysemy of word usage does not require polysemy of ontological terms.


In fact, if a cluster is defined by a set of rules applied to test

result, then the cluster meanings are monosemous:  those tests which

match the rule set are in the cluster, those which do not match are not.


That describes the decision surfaces used to wall off clusters from one another; such a surface distinguishes those in the cluster from those outside the cluster, but has no more meaning than that simple decision.  The observers are the ones who, in their own subjective ways, came to the conclusion that the clusters are meaningful in their conversations.  


There can certainly be different theories as to the meaning of the

clusters or what someone's test matching one of the clusters implies

about that person, but that is quite a different matter than the

monosemy of the defined clusters -- or of words used to name or

describe the clusters.


True, but not relevant to my own point of showing how personal and subjective the meanings are to each observer.  




-- doug foxvog


> -Rich




> Sincerely,


> Rich Cooper


> EnglishLogicKernel.com


> Rich AT EnglishLogicKernel DOT com


> 9 4 9 \ 5 2 5 - 5 7 1 2


>   _____


> From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx

> [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Simon Spero

> Sent: Saturday, November 06, 2010 1:02 PM

> To: [ontolog-forum]

> Subject: Re: [ontolog-forum] Polysemy and Subjectivity in Ontolgies -

> theHDBIexample




> I would caution against placing too much reliance on studies based on this

> instrument.




> See e.g.   "Review of the Herrmann Brain Dominance Instrument [Revised]"

> by

> GABRIELE van LINGEN, In  Plake, B. S., & Impara, J. C. (Eds.). (2001).

> Fourteenth Mental measurements yearbook. Lincoln, NE: Buros Institute of

> Mental Measurements.

> <http://buros.unl.edu/buros/jsp/reviews.jsp?item=07001170>

> http://buros.unl.edu/buros/jsp/reviews.jsp?item=07001170




> The HBDI, although reportedly receiving a uniformly positive response from

> participants of Herrmann's management training workshops, cannot make

> claims

> for meeting test standards that would recommend it to the public domain.

> Appropriate reliability and validity studies are not available. The

> instrument's format itself would appear to present problems for

> establishing

> the former, whereas the variety of claimed purposes and constructs would

> present difficulties for demonstrating the latter. Despite decades of

> research, there is only minimal credible evidence that the HBDI results in

> scores that are temporally stable and that the scores relate to meaningful

> nontest behavior. Otherinstruments, with established psychometric

> properties, are better suited for the individual applications that the


> claims. For instance, the Myers-Briggs Type indicator is a better

> instrument

> for assessing more general personality type or style. For career or

> occupational decision-making, the Hogan Personality Inventory (Hogan &

> Hogan, 1995) or the System of Interactive Guidance Information (SIGI-Plus;

> Katz, 1993), updated for adults, are better suited. Cognitive style is

> perhaps best assessed by the more established perceptual tasks such as the

> Embedded Figures Tests (Witkin, Oltman, Raskin & Karp, 1971) or by

> exploring

> more complex processes, such as Sternberg's thinking styles (1994).


> [...]


> References:


> Hogan, R., Hogan, J., & Roberts, B. W. (1996). Personality measurement and

> employment decisions. American Psychologist, 51(5), 469-477.


> Katz, M. R. (1993). Computer-assisted career decision making: The guide in

> the machine. Hillsdale, NJ: Erlbaum.


> Sternberg, R. J. (1994). Thinking styles: Theory and assessment at the

> interface between personality and intelligence. In R. J. Sternberg & P.

> Ruzgis (Eds.), Intelligence and Personality. New York: Cambridge

> University

> Press.


> Witkin, H. A., Oltman, P. K., Raskin, E., & Karp, S. A. (1971). A manual

> for

> the Embedded Figures Tests. Palo Alto, CA: Consulting Psychologists Press.



> Simon



doug foxvog    doug@xxxxxxxxxx   http://ProgressiveAustin.org


"I speak as an American to the leaders of my own nation. The great

initiative in this war is ours. The initiative to stop it must be ours."

    - Dr. Martin Luther King Jr.





Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/ 

Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/ 

Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx

Shared Files: http://ontolog.cim3.net/file/

Community Wiki: http://ontolog.cim3.net/wiki/

To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx


Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (01)

<Prev in Thread] Current Thread [Next in Thread>