Need some help with "Need for cognition" and my diploma thesis

[NillePille]

Hi, I spent some time now looking for the sample statistics of the Need For Cognition Test from Cacioppo. Im not sure how this is called in english (maybe that is why Im not finding them..) but I think you`ll understand. Ive used the scale in my diploma thesis and now I want to compare my data with a more representative population then the psychology students and academics in my sample:-) The problem is that I´ve no idea where to cut my group to get the lowest and highest "needer of cognition" to make an comparison of the extreme groups.

I`m looking forward to your answers.
Greetings from Germany.
Niels A.

Pd.: If Cacioppo didn`t publish the needed data maybe you know a big study (with a more representative sample than mine) that used the NFC-Test. Maybe I can find some ideas there which score "cuts" the lowest and the highest 25%.

[NillePille]

In another forum I got this short answer, which I`m going to post here to get your opinions on that issue. I will also post my answer to that post.

Answer:
You'll get terrible comments from reviewers if you use some arbitrary method of defining the extremes.
Use a median split or a regression.

My answer to that post:
First, thanks for your answer! Could you explain why you think that way? Ill try to explain to you what I´m trying to do with another example of my study. I´ve got a sample of 300 people, mainly students. So the scores I got for deductive reasoning are significantly higher than in the normal population. If Ill do a median split, a regression or quartiles to define the extremes, the "lowest" group contains a lot of people (and scores) which you would normally consider to be average deductive reasoners. So I looked up the adequat (in my eyes not arbitrary) sample statistics (by age) of the intelligence test to get the "cut-points". This way i can compare two groups which are actually low/ high in deductive reasoning. The same thing I want to do with Need For cognition. It would be just an aditional analysis, but I can`t really see why this would be a questionable methodology. If you test someones skills with a test it seems to be a quite normal procedure to compare his/her results with a representative population.
I`m looking forward to your explanation.



I`m really getting confused... is it really that wrong what I`m trying to do?

Thanks in anticipation for your help.