“Lies, damn lies, and statistics.” Throw this Disraeli quote in too if you would. “Public thought is moreover public sentiment.” If you ask me any poll conducted by The Washington Post may as well have been conducted by AIPAC...and this one is no exception. But look on the bright side of things. Tom Freidman says he’d rather see a nuclear armed Iran than another mid-east war. How ‘bout the Democrats run him in ‘08? If he ain’t lyin’ I’d vote for him. That, and you guys realize you’ve kicked The Washinton Post in the nuts with this business of “DON’T BELIEVE THE FLAWED NSA SPYING POLL” don’t you? News Papers don’t work anymore. Scheer saw the light.

I don’t believe the poll that most Americans are okay with the NSA collecting millions of phone numbers and seeing who called whom. My fear and it is such a great fear is this. Most of Americans or at least a good number of them are asleep at the wheel and will follow what ever their favorite news media tells them. If they only knew we are about to crash as a nation.

Benjamin Disraeli probably said it best—“There are three kinds of lies: lies, damn lies, and statistics.” I was a psychology major in college—and we were not only required to pass a statistics course in order to graduate but were also given a thorough grounding in the scientific method. Conducting research in the social sciences is particularly tricky because depending on what sort of study is being conducted and how the data is gathered, both the subject’s responses and even the researcher’s reactions to those responses are at least somewhat more subjective and therefore more open to interpretation.  This is only one of the reasons why polls should always be taken with at least one or two grains of salt, or even the whole shaker—and this is only the beginning.

One of the most important tests which can be used to determine whether an experiment (in this case, the poll) is really accurate is that of reliability. This is primarily measured in two ways—the first by running the experimental procedure again and observing whether or not similar results are obtained, the second by examining whether researchers observing the data concur with one another regarding what the data contains and what conclusions can be drawn from it. Failure to meet one or both conditions of reliability casts reasonable doubt on the value of the data since there is a chance that any significant results obtained were due to chance.

However, even a confirmation of reasonable reliability is no guarantee that the study can be trusted. The question of validity—the potential for data to become skewed or biased as a result of unforeseen internal or external variables—must also be considered. One example of a factor which can call validity into question is the lack of a truly random sample. If you want to get a truly accurate measure of how the American people as a whole feel about the NSA database of telephone calls, the subject pool should ideally be composed of roughly equal amounts of Democrats and Republicans (perhaps even Independents as well). Did the pollsters make certain to meet this condition?  We don’t know—they didn’t give us details on the subject pool regarding political affiliation (coincidence?). If there had been a dramatic ratio of Democrats to Republicans—say, five Democrats and thirty Republicans—this would not, strictly speaking, be a random sample and it would be reasonable to expect a data bias in favor of the Republican viewpoint as a result. Demand characteristics are another danger. If there is direct interaction between the subject and the researcher during the experiment (even if only verbally over the telephone), there is always the possibility that the researcher is giving indirect cues to the subject that indicate which response they expect and/or hope to observe—and the researcher may not even be consciously aware that he or she is doing this. In the case of a poll, this doesn’t even have to involve direct interaction—it may be expressed solely in the specific wording of the questions asked. In this case, depending on who specifically conducted this poll and what their own beliefs on this subject may be, it’s not impossible that at least some of this took place here. The right-wing pundits can trumpet all they like that this study PROVES beyond the shadow of a doubt that a majority among the American people approve of the NSA database—but without at least more specific information on the political affiliation of those polled, it actually does nothing of the sort. This poll in fact communicates very little apart from how the specific individuals polled feel about this issue. Additionally, since other polls are coming forward which appear to contest this result, the reliability of this poll is also open to question.

Benjamin Disraeli probably said it best—“There are three kinds of lies: lies, damn lies, and statistics.” I was a psychology major in college—and we were not only required to pass a statistics course in order to graduate but were also given a thorough grounding in the scientific method. Conducting research in the social sciences is particularly tricky because depending on what sort of study is being conducted and how the data is gathered, both the subject’s responses and even the researcher’s reactions to those responses are at least somewhat more subjective and therefore more open to interpretation.  This is only one of the reasons why polls should always be taken with at least one or two grains of salt, or even the whole shaker—and this is only the beginning.

One of the most important tests which can be used to determine whether an experiment (in this case, the poll) is really accurate is that of reliability. This is primarily measured in two ways—the first by running the experimental procedure again and observing whether or not similar results are obtained, the second by examining whether researchers observing the data concur with one another regarding what the data contains and what conclusions can be drawn from it. Failure to meet one or both conditions of reliability casts reasonable doubt on the value of the data since there is a chance that any significant results obtained were due to chance.

However, even a confirmation of reasonable reliability is no guarantee that the study can be trusted. The question of validity—the potential for data to become skewed or biased as a result of unforeseen internal or external variables—must also be considered. One example of a factor which can call validity into question is the lack of a truly random sample. If you want to get a truly accurate measure of how the American people as a whole feel about the NSA database of telephone calls, the subject pool should ideally be composed of roughly equal amounts of Democrats and Republicans (perhaps even Independents as well). Did the pollsters do this?  We don’t know—they didn’t really give us specific details on the subject pool, did they? If there had been a dramatic ratio of Democrats to Republicans—say, five Democrats and thirty Republicans—this would not, strictly speaking, be a random sample and it would be reasonable to expect a data bias in favor of the Republican viewpoint as a result. Demand characteristics are another danger. If there is direct interaction between the subject and the researcher during the experiment (even if only verbally over the telephone), there is always the possibility that the researcher is giving indirect cues to the subject that indicate which response they expect and/or hope to observe—and the researcher may not even be consciously aware that he or she is doing this. In the case of a poll, this doesn’t even have to involve direct interaction—it may be expressed solely in the specific wording of the questions asked. In this case, depending on who specifically conducted this poll and what their own beliefs on this subject may be, it’s not impossible that at least some of this took place here. The right-wing pundits can trumpet all they like that this study PROVES beyond the shadow of a doubt that a majority among the American people approve of the NSA database—but without further information on how the data was gathered and how the poll was conducted, this poll actually communicates very little apart from how the specific individuals polled feel about this issue. Additionally, since other polls are coming forward which appear to contest this result, the reliability of this poll is also open to question.

WHEN DID THEY ‘PARSE’ANYTHING IN THE LAST 6 YEARS?WHEN I ASKED MY (then) STATE SENATOR(EDWARDS OF NC) TO PLEASE SPEAK UP AND OPPOSE THE IRAQ WAR, HE SAID TO ME ,"WE WILL HAVE TO AGREE TO DISAGREE”.

The problem is that when you read the poll closely, they only support it when it might lead to apprehending terrorists. As we know, that is the discredited and highly dubious claim being used by the Busheviks to implement their domestic spying program—and the dismantling of the Constitution in general. So, it is an extremely incompetent, even propagandistic poll.

It is not ‘incompetent’ it is in fact a masterful example of what it is. There is an interesting blog at Thought Theater called Terror Management: A Warning To Democrats that explains the result of this poll in terms of a area of psychological thought called “Terror management theory”.  The Wikipedia article on this subject opens like this

Terror management theory (TMT) is a developing area of study within the academic study of psychology. It looks at what researchers claim to be the implicit emotional reactions of people when confronted with the psychological terror of knowing we will eventually die (it is widely believed that our awareness of mortality is a trait that is unique to humans). The theory was first developed in the late 1980s by Skidmore college psychology professor Sheldon Solomon and others. Solomon was inspired by the theories of Ernest Becker (The Denial of Death,
1973) and Freud, on how potent reminders of one’s own ultimate death often provoke a belief in some form of mystical transcendence (heaven, reincarnation, spiritualism, etc.).
The Theory builds from the assumption that the capability of self-reflection and the consciousness of one’s own mortality, can be regarded as a continuous source for existential anguish. Culture diminishes this psychological terror by providing meaning, organization and continuity to men’s and women’s lives. Compliance with cultural values enhances one’s feeling of security and self-esteem, provided that the individual is capable of living in accordance with whatever particular cultural standards apply to him or her. The belief in the rightness of the cultural values and standards creates the conviction necessary to live a reasonable and meaningful life. . . .
Research has shown that people, when reminded of their own inevitable death, will cling more strongly to their cultural worldviews. The data appears to show that nations or persons who have experienced traumas (e.g. 9/11) are more attracted to strong leaders who express traditional, pro-establishment, authoritarian viewpoints. They will also be hyper-aware of the possibility of external threats, and may be more hostile to those who threaten them.

The terror effect has been shown in experiments demonstrating that making research participants think about death will lead to changes in opinions toward those that protect worldview and self-esteem. 
So just look at the way this poll question was drafted. In post-9/11 America the twice repeated word “terrorism” is a clear way of remaindering us of death.  And we also clearly see in the question a powerful suggestion that the NSA data gathering thing, though unclear as to its exact nature, is some powerful life affirming thing which is being done, and has in the past been done, by our strong leaders to as a successful protection from death. Given this, the somewhat counterintuitive results of this “scientific poll” can be viewed as simply yet another proof of the Terror Effect.
In fact Bush and the Republicans with their “post-9/11” campaign strategy are steeped in terror management. There are some terror management theorists who see the 04 elections as being best explained in those terms.

But exit polls don’t tell the whole story, says Solomon. He and his colleagues believe that they have uncovered a subtle application of a psychological effect—terror management theory—that may have helped tip the election to Bush. According to the theory, Americans traumatized by the 9/11 terrorist attacks turned to Bush in part because, subconsciously, his clear and values-driven message helped assuage their fear of death.

So we should see that poll question as recapitulating the whole of the Rove plan for using 9/11 to leverage the Republicans into a permanent majority by convincing Americans that Republicans have some charm that will ward off death. When Republicans want to turn up the heat of the campaign they will look at the camera and announce darkly that they have saved the lives of many Americans with these things but that they can’t tell us more because if they did our ever-present enemies would hear the secrete of the Republicans’ power and the spell that protects us would be broken.
In a situation like this how could the nation dare bring into power a different political party?  How foolish we would be to let gain power any who do not share the protective magic that is keeping up alive!

So you’re assuming that Democrats can’t or won’t parse the poll language but instead will take it at face value? We’re lost.