[Air-l] Disadvantage of the Internet (long)

[Denise N. Rall]

To further this discussion, I will bring up something
that I presented 2 years ago at 4S:

I'm much more interested in the notion of cognitive
dissonance - that the way that the internet is
organized does not aid cognitive processes (ways of
thinking) already established in human beings.

Human cognition takes certain very well known pathways
- in particular, in regard to decision-making and also
verification of facts.  Then using the Internet - in
this case - a search engine - which returns millions
of results does not necessarily match generally ways
that people have evolved to recognize truth (or
process data).

However I think truth claims are only one aspect of
this cognitive dissonance. I would like to see more
work or perhaps it is there from the psychologists on
large data-set returns from search engines and how
that helps or hinders people who are learning.  Last
year, Jeremy also brought up extreme multi-tasking
across a variety of media inputs, and how these fit
with the ability that humans have to make sense of
their environment (not trying ot mis-quote him here)
that is another problem.

Bottom line: Do I make sense of a million returns on a
hit? No, I don't.  I scroll through about 5-10 pages
(at most) or I refine my search. Or I ask this list or
etc. I never deal with even a fraction of what is
returned, maybe with a bot I could do better than
that. Usually I seek out hubs or authorities
(Scientific American, 1999) and those answer my
questions rather than millions of hits.

>From this perspective, I then presented a notion of
super-salience of search results (from the
psychological notion of salience) in order to make web
searching more closely match other types of cognitive
processing.

My paper on the aspect that deals with truth claims is
unfortunately still in preparation, 

"Testing hypothesis-testing: Taking Sir Karl Popper
for a spin on the World Wide Web" 

Here's the abstract follows for those who are
interested:

Sir Karl Popper (1902-1994) was the preeminent
philosopher of science in the 20th century, and most
serious scholars in the philosophy of science have
responded in some way to his theories: his attacks on
inductivism, his constructivist stance, and his
integral role in the development of evolutionary
epistemology.  In this century, Popper’s explications
of learning are increasingly claimed by educational
theorists (Bailey, 2000).  Educationalists can learn
from Popper, who insisted that “all life is problem
solving” -and that learners proceed by building
tentative hypotheses to explain their world and use
error elimination to refine those hypotheses (Popper,
1999).

This author takes Popper's words, "[we] understate the
importance of . . . the erroneous trial" (Popper,
1991:101) and applies them to searcher’s results on
the WWW.  The available logic (Boolean) may limit the
number of results but not increase validity (Lawrence,
1999).  Particularly, tradition scientific
null-hypothesis tests do not work well on the Web. 
Either two many results are returned for sensible
testing; contrarily, when no results are returned,
this scarcely proves the condition as true(!).  Null
hypothesis methods are particularly suspect because
cognitive research shows that we do not learn much
from our behaviors when results are nil (Allison &
Messick, 1988).  Therefore, this author turns to human
information processing to locate better techniques for
searchers, especially searching for salience rather
than eliminating error.  Salience will assist WWW
searchers in two cognitive processes: environmental
scanning of cyberspace, and the cognitive structuring
of their impressions based on 'goodness of fit' (Fiske
& Taylor, 1991: 251).  The author concludes that while
Popper correctly recognized error as the crux of
substantiating truth claims, he required more
sophisticated feedback loops to bolster his theories. 
For one example, Norbert Weiner’s (1954) cybernetics
could have provided a better mechanics for the
cognitive processing of errors.  Finally, better
search methods (Clever project, 1999) and especially
the recognition of super-salient ‘hits’ will yield
better results for searchers than error elimination
methods.


Cheers, Denise



=====
Denise N. Rall, PhD candidate, School of Environ. Science,
Southern Cross University, Lismore NSW 2480 - Mobile 0438 233 344
Sustainable Forestry Mentoring Coordinator & Internet Researcher
Presented! 2004 Conf. Association of Internet Researchers: www.aoir.org
http://www.scu.edu.au/schools/rsm/staff/pages/drall/index.html