According to Gallup (1944), through his instrumentality of public opinion polls, fact-finding in the field of public opinion has been met with wide acceptance. Fact-finding methods in the field of public opinion, or for that matter, in any field which deals with human reactions, have not, and never will, achieve the exactness of those employed in the pure sciences. Human beings can not be studied as easily as the elements. But this does not mean that,even as of today, it is impossible to obtain a highly accurate measurement of both quantitative and qualitative public opinion, . Perfection in techniques requires repeated trial and error. Many mistakes will contiue to be made and will contribute toward the refinement and improvement of methods. The young in mind will appraise the work fairly and realistically, and for this reason it is inevitable that facts about public opinion will be of increasing importance to the people and to their government<37>.

Jack Hart (2001) pointed out that even politically sophisticated journalists can sometimes have a difficult time understanding how a couple of thousand individuals can accurately represent the entire U.S. population. Accurate representation is possible, however, if the pollster draws a good sample. The experts have proven their ability to predict voting patterns within a couple of percentage points time and time again.

George Gallup often explained polling with his soup analogy. One spoonful of soup can accurately represent the taste of the whole pot so long as everything is well stirred (Jack Hart, 2001). The stirring is the key. That is what introduces the random element that is so important to scientific sampling. Randomness, in turn, is what brings the laws of probability into play. And the laws of probability are what make accurate polling possible. Everything depends on the "sampling distribution." This idea states that ifis this: If the repeated samples out of the same population is drawn, they distribute themselves in a normal, bell-shaped, curve around the average for all possible samples. And that average will exactly equal the average for the whole population.

Probability theorists also use the Central Limit Theorem. For example, a study might record sampling heights from a sample of 100 men who, on average, are 6 feet tall and take repeated random samples of 10, drawing every one of the hundreds of unique samples possible. More samples will have average heights close to six feet because there are more of those men and they are more likely to be selected. Very few samples will average 5 or 7 feet because it is unlikely to get many extremely short or tall men in any given sample. The total "sampling distribution" will form a normal curve that falls away on both sides of the point representing the average of all samples, which is exactly 6 feet:<37>:

The magic of this is that the average for the sampling distribution, 6 feet, is exactly the same as the average for the whole population. Of course, a study never actually measures the entire sampling distribution because every sampling distribution forms a normal curve that can calculates the "standard deviation," This calculation describes the shape of the normal curve and is based on the total amount that each individual differs from the average for the whole group. In ther words, it can measure how flat or steep any given normal curve happens to be<37>.

The critical fact of nature is that, in every normal curve, about two-thirds of the things being measured will be within one standard deviation of the average. About 95 percent will be within two standard deviations. So every time random sample has a 95 percent chance of falling within two standard deviations of the average for all samples, which is the same as the average for the whole population. This information reflects how much faith should be put into the sample<37>.

First of all, this is a 95 percent chance that the results are within plus or minus two standard deviations of the true figures for the sampled population. That 95 percent figure is the "confidence level." Second, the value of one standard deviation for this sampling distribution can be calculated to further reveal the number of percentage points that two standard deviations represent in either direction. This is the margin of error. The two figures, the confidence level, and the margin of error are all critical to assessing the value of any poll<37>.

If the average man in one random sample of 10 is 5-foot-10, the standard deviation in heights for a sampling distribution of 10-man samples is drawn from a 100-man population, and if it happens to be 2 inches, that means that 95 percent of the possible samples will average between 5 feet 8 inches tall and 6 feet 4 inches tall. The chances are 95 out of a hundred that the average height in our sample, 5-foot-10, is within plus or minus 4 inches of the true average height for the whole population. An occasional wild sample will still be way off the mark. This occurs 5 percent of the time. And it does not mean that other sources of error will not bias the results. A flawed measurement technique can skew the results, for example, the confidence level and the margin of error are absolutely certain<37>.

Those result from the Central Limit Theorem. And they are just as reliable as the Theory of Relativity, the formula for pior any other law of the universe. And that's why George Gallup has been able to call every presidential election within 3 percentage points since 1952<37>.