roachboy

problems noted above aside, let's think about these questions and numbers..

in the excerpted results, the first question is the key one because it establishes the logic from which all the other responses follow. linking the iraq adventure to the "war on terror" defines the stakes. the strange thing about it is that it does not appear in this position in the poll results themselves--if you look at the .pdf and check out question 4, when folk are asked directly about the "meaning" of the iraq aventure, 32% responded that the americans should set a strict timetable for withdrawal no matter the situation in iraq, and only 27% agreed with a longer version of what is excerpted as the major premise of the list of results.

in the .pdf, it seems that the order and wording of questions 6-15 is much more determinant of the outcomes---i think the poll measures the effects of its own design as much or more as anything else-----this because the results for question 10 (which is the question bit for press release purposes and positioned first in the list of results) are so different from those to question 4.
the press release figure also aggregates "strongly agree" and "somewhat agree" responses to a question that is not designed to allow for questions about the question itself (number 4 is more open in this respect)...so even on its own terms, the way the poll results are presented is odd, and the sequence misleading (to my mind anyway) so the meaning of these results seems ambiguous at best.

the logic of questions 5 through 15 is built entirely around administration claims concerning iraq and are geared to getting respondents to derive political consequences from administration premises. but again, the questions are designed as they are--which is simply a choice for poll design, neither entirely right or wrong in itself--but it certainly has effects---this section of the poll seems to me particularly badly designed, and the results are a function of that bad design.

which surprised me because i started looking at the poll results after i started thinking about what is bit for the op--i wasn't surprised by these results in particular because i took the first question as preceding the others in the poll and thought that you could work out from that why the results were as they were--as reflecting anxiety about the consequences of making (still more) bad choices on/around/about iraq--and i assumed that at least some anxiety about what is presently going on would be abroad in the land and that a poll could catch fluctuations in relation to various questions that reflected that anxiety if it was designed well. so i would have found nothing surprising in the results as they appear in the press release, but was curious...

but that's not how the poll worked, and it is not what the results say.
so the presentation of the data is problematic.

another strange bit of infotainment comes in the demographic section of the poll results: the self-characterization question---look for yourself--but i find it strange the way in which the pollsters chose to group the 39% of respondents who described themselves as "moderates" as separate from the 39% conservative, and with the 21% of respondents who characterized themselves as "liberal" or "very liberal"---it seems to me that the profile of respondents should look differently.
say split the moderate number on half, attribute half to the conservatives half to the liberals.
if you group conservatives and moderates by self-definition, you get 78% conservative/moderate and 21% liberal.
each choice makes the results look differently, doesn't it?

a very odd poll, then.

__________________
a gramophone its corrugated trumpet silver handle
spinning dog. such faithfulness it hear

it make you sick.

-kamau brathwaite