Misinterpreting Data

Every day, we witness numerous submissions of data, statistics, etc., that some organization or other, interprets in some way. We often see different polling organizations doing polls that they claim to be statistically valid that are widely different. One of the glaring examples of that is in political polling, where one polling organization sometimes distributes polling numbers that widely differ from another, while both polls claim to be accurate within a statistical variation, how could they both possibly be when they show far different results.

We often witness surveys that indicate certain types of results. However, these results are often skewed because of the wording, or timing of the specific questions. How many times have you, or anyone you know, been polled, so one might ask, who exactly are these surveys polling?

Financial and stock market analysts often present numbers in a rather confusing manner. Even if a particular company met or exceeded the companies stated objectives, we often hear analysts state that the numbers were below or worse than many analysts expected. Wouldn't that make one suspect that perhaps many of these analysts do not fully understand the situation?

When there are fewer individuals applying for first time unemployment benefits, is that a good or a bad think from an economic and financial standpoint. Aren't there a number of factor that might be involved? How does seasonal employment impact these numbers? For example, when many Americans were hired for temporary Census employment, the employment rate improved. When those jobs ended, therefore, wouldn't and shouldn't it have been suspected that the unemployment rate would go up again?

If a company had a bad previous financial period, and the numbers are not as bad in the following period, it can be reported that a company was still having difficulties, or it can be "spun" to indicate that there was improvement.

When the Obama administration states that a certain number of jobs were created because of his economic stimulus programs, how does anyone know how many would have been created anyway? If the economy is indeed cyclical, which historically it has always been, won't there always be modifications and improvements?

Many organizations use surveys to find out feedback from its members, or attendees at one of its programs or events. However, aren't the people filling out the surveys probably a somewhat skewed audience, firstly because they were one of the attendees, and secondly because they agreed to fill out the survey. Of course, there is also the difficulty in knowing how honest and accurate the information was.

Many seminars ask attendees to evaluate the program. While the intent of this might be fine, isn't it a fact that the reason most people attend a seminar is because they do not have a thorough understanding of the topic? If that is the case, how would that attendee know if what he learned was important and accurate, or trivial and inaccurate? It has been indicated that many people complete surveys based on either emotion, entertainment or the need to please?

Several years ago, Michael J. Fox starred on a television program, "Spin City," which was about how politicians "spin" the facts for maximum impact. While this is true in politics, it is also true in almost every other type of data interpretation.

I always recommend that someone study the raw data carefully, understanding how it was gathered, rather than relying upon a so-called expert's analysis. Data can be valuable or worthless, or somewhere in-between, dependent upon how and by whom it was gathered, how it's used, and how it is interpreted.