definition: proportionate = equal % change IF we agree on this ... and maybe we don't ... then, since the % change is always UN =, then all changes are DISproportionate
but, given margins of error and the like ... and, just the practical interpretation of the data ... i would say that we could have a pragmatic agreement that if the changes were within P %, then we might "call" the changes = the fact is, if the data are accurate, for both whites and blacks, the after ratings jumped dramatically .. compared to the before ratings ... now we are just quibbling over whether those dramatic jumps should be called = or not thus, the issue in the video and the information that was presented is ... are the changes SO large as to make even tolerant people say that they are different in the case of george w and, the white and black change from pre to post ... i am MORE than willing to concede that they look about the same for the elder bush ... in term of gulf war pre and post ... the changes between approval ratings between whites and blacks i would be less willing to argue that way but, unless we have STANDARD ERRORS OF DIFFERENCES IN PROPORTIONS FOR CORRELATED SAMPLES .. to make the comparisons with, i think it is just an exercise in "whatever you think" about the data i still don't think the person in the video made any egregious misstatements of how the data looked, and in addition ... if you view the data watching the video, which is very clear .... you could make up your own mind anyway perhaps you could elaborate on why you think he should have been saying DISproportionate all the time ... at what threshold "change" value would have to be evidenced in the data for you to think he should have been speaking in opposite terms? At 09:59 PM 1/10/02 +0000, EugeneGall wrote: >His definition of proportionate would mean that if a group's approval of Bush >went from 1% to 31%, that too would be proportionate. The relative odds would >be one way of expressing the changes in proportions, but the absolute >difference (60% to 90% is roughly propotionate to an increase from 33% to 68%) >seems quite wrong. > > >================================================================= >Instructions for joining and leaving this list, remarks about the >problem of INAPPROPRIATE MESSAGES, and archives are available at > http://jse.stat.ncsu.edu/ >================================================================= _________________________________________________________ dennis roberts, educational psychology, penn state university 208 cedar, AC 8148632401, mailto:[EMAIL PROTECTED] http://roberts.ed.psu.edu/users/droberts/drober~1.htm ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =================================================================
