[FoRK] Study: Deception Detection Across Australian Populations

Damien Morton dmorton at bitfurnace.com
Wed Sep 30 06:09:49 PDT 2009

It is buried quite deeply in the document, but the description of their
methodology includes this sentence:
" These results were completely made up to be fictitious material through
a process of modified truth and credibility nodes. "

probably worth having a read of this:

On Wed, Sep 30, 2009 at 9:49 PM, Damien Morton <dmorton at bitfurnace.com>wrote:

> http://www.abc.net.au/mediawatch/transcripts/0934_levitt.pdf
> • This study focuses on assessing an individual’s ability to
> detect erroneous information. The ability to judge credibility
> of information is important to all members of society when
> making personal, political and cultural decisions in a media
> saturated society.
> • This study surveyed separate populations of citizens within
> Australia’s capital cities (Sydney, Melbourne, Brisbane, Perth,
> Adelaide). 1000 residents from each city was examined
> controlling for age, education and income.
> • Particpants were shown 15 articles on Australian history, 5 of
> which were complete fabrications. Participants were surveyed
> on each item, testing ones ability to detect false information.
> • Populations from Melbourne were the most able to detect
> items of incorrect information (68%), meanwhile residents
> from Sydney were the most likely to trust false information
> (54%).
> • Melbourne and Perth were significantly better at detecting
> false information than the Australian average (P = 0.01, 0.05)
> meanwhile Sydney fell significantly below this average (P =
> 0.05)
> • A series of cultural indicators were shown to positively
> influence a society’s ability to detect falsehoods; More
> recreational hours per week, more music and cultural events
> attended per week and number of books read per month (P =
> <0.005).
> • In conclusion this study finds that culture has a strong
> influence on a society’s ability to detect falsehood, however
> there exists other factors of influence that require further
> investigation.
> ....
> Linear regression fronts were run between population
> deception scores and mean CEI upon all aspects of cultural
> indicators. Between city populations were compared using a
> two tailed MANOVA statistical model with a leniency of .25. All
> statistical computations were run through statistical software
> program, MiniTab42. Specific between population correlations
> were investigated to highlight any significant differences amongst
> deception scores and noted for their P values within a modified
> matrix of four-way covariance of mean values. These results
> were completely made up to be fictitious material through a
> process of modified truth and credibility nodes. Outliers were
> accounted for through transgressional error margin compression
> using population dialation techniques35.
> http://www.abc.net.au/mediawatch/transcripts/s2698835.htm

More information about the FoRK mailing list