Data analysis with ratings
- From: Marco <greeniguana@xxxxxxxxx>
- Date: Wed, 27 Apr 2005 22:53:10 +0200
I compare information quality of two different sources, i.e. printed and electronic content. To this end, information to nine questions was gathered from each source. This was done for five countries (the domain chosen is tourism). The information was then presented to independent raters, such that each of the 90 information units was rated three times (it was not always the same three raters judging an information unit).
A rating (dependent variable) consists of both a mark on a one to six scale and a binary judgement (+/-). Independent variables are source, country and question.
Now several questions arise:
1)Do I need to aggregate the three ratings per question before comparing means between countries and media? What would be an appropriate aggregation for the binary (nominal) values?
2)I want to reject the null-hypothesis that the ratings are random values. What would be appropriate methods to do this?
I really appreciate your help, I have been stuck with this for quite some time now.
Thank you, marco .
- Prev by Date: Re: MatLab randn and Simulation Step Numbers
- Next by Date: Useful statistics groups
- Previous by thread: logistic regression and model misspecification
- Next by thread: Useful statistics groups
- Index(es):
Relevant Pages
|
|
