Collection of Data — Economics STD 11 Commerce — Question

Statistical errors can be of two types: Sampling and non sampling. Sampling Errors: When sample method is used, we cannot expect to get the same result as we would have got if we would have used census method. There is always a difference in conclusions drawn from sample and conclusions of the universe. This is called sampling error. These errors are minimum when random sampling techniques are used and size of sample is fairly large. As we increase size of sample, sampling error keeps falling and become zero at 100% i.e. census method. Therefore, these errors exist only in sample method.Sampling errors may be biased or unbiased:

1. Biased Errors: Biased errors are those errors which arise because conclusions have been drawn from a biased sample. These errors can be eliminated by choosing unbiased sample.
2. Unbiased errors: These errors occur due to usage of sample method. If on the basis of height of five students in a class of 15 we are calculating average height, obviously results will vary from actual. These errors cannot be eliminated.

Non Sampling Errors: These errors are committed due to human mistake or clerical reasons. These are also called measurement errors. These errors are found in sampling as well as census method rather their occurrence is more expected