Statistical data are obtained either by measurement or by observation. Hence to think of perfect accuracy is only a delusion or a myth, It is not possible to ensure perfect accuracy. This is one of the vital limitation inherent in statistical studies. Dr. A .L. Bowley has rightly said, this is not inexistence a perfectly accurate measurement physical or economical just as there is no perfectly straight line or a perfect fluid.
In the science of statistics, the difference between the observed value and the true value is called error for example. In a population census there is no guarantee that not a single person is left out. While measuring the weight of the students, no one can guarantee that there will be perfect accuracy in the measurement. These inaccuracies are called error in statistics. It does not mean mistake.
If there are 62 students in a class room, and we count themes 60 it is a mistake. But on some basis if approximate them to be60 it is called an error in statistics, According to Prof. L. R. Connor an error in the statistical sense is merely a different between an estimate and the true value or idea value which it may be impossible to determine exactly. According to prof. Boddington. The statistical error is the difference between the true figure and the approximation Prof, Boddington has clarified that the error does not impel an arithmetic miscalculation, but arises simply because in collecting data a perfect coverage in impossible as with sampling.