Wednesday, June 9, 2010

12like, visualDNA and a HOAX

12like claims to be scientifically proven, but it is a HOAX

"The bases is the scientifically proven and psychologically sound 12like Image Personality Test. In this fun, accurate, and easy test, questions are answered exclusively with images. Test results are then summarized in a personality type and high quality individual Matches are proposed."


"Basis ist der wissenschaftlich und psychologisch fundierte 12like-Bilder-Persönlichkeitstest. Dabei werden Fragen ausschließlich mit Bildern beantwortet. Der Test macht Spaß, ist besonders einfach in der Durchführung und bringt treffsichere Ergebnisse. Das Testergebnis wird in einem Persönlichkeitstyp zusammengefasst und entsprechend passende hochwertige, individuelle Partnervorschläge unterbreitet."

"scientifically proven" should mean the online dating site offering a compatibility matching method can prove its matching algorithm can match prospective partners who will have more stable and satisfying relationships than couples matched by chance, astrological destiny, personal preferences, searching on one's own, or other technique as the control group in a peer_reviewed Scientific Paper.

The test used at 12like is powered by visualDNA

VisualDNA: At the overviews of how VisualDNA integrates with Portals and Ad Networks,
page 6 says "Personality Type: Attitudes, Big 5 Categorisation (Extraversion, Agreeableness, Conscientiousness, Neuroticism, Openness)"

Who independently validated and independently certified how that profiling technology correlates with the Big5? I want to see a paper showing VisualDNA correlates with the Big5 (like the one for the Dewey Color System showing its correlation with the 16PF and the distortion it adds.)

How does VisualDNA compare one pattern with others?
Using solely a linear or logistic multivariate regression equation?

Profiling by music preferences to assess personality
video preferences to assess personality
color preferences to assess personality
bookmarks preferences to assess personality
handwriting analysis to assess personality
purchases and buying trends from credit cards to assess personality
and other methods like them
add distortion to the measurement.

I had invented 2 fake (dummy) profiles to test 12like.

I had accidentally discovered that something is wrong with 12like's compatibility matching algorithm, there is a glitch when it first shows your compatible matches.

See for example, the list of compatible men for the "German" female profile (Bibi) sorted by "Match %"

The screen shows CPUX4FUM is 100% compatible (Er, 36 Jahre, 175 cm, IT professional, Potsdam, Deutschland).

I opened that profile and says
CPUX4FUM is 87% compatible (Er, 36 Jahre, 175 cm, IT professional, Potsdam, Deutschland)

then I had pressed the back button and appeared other compatible persons with high percentage than 87%.

There are 10 personality types and 12like matchs you using some kind of "personality" "similarity".
Some screenshots

No comments:

Post a Comment