How to Estimate From Random Samples

Obtaining information about a certain population is a resource-demanding task, especially when that population is large. When a researcher wants to know the traits of a population, the only feasible method is often to take a random sample and estimate the traits from the collected data. This task is at the heart of statistics and requires not only basic sampling theory, but also a strong understanding of how the estimates relate to the true parameters of the population of interest.

Instructions

    • 1

      Determine the parameter to be estimated. This parameter is the one you want to know about the population but cannot find because of limited resources. Examples of such parameters are means, ratios, variances or functions of other parameters.

    • 2

      Establish the mathematical formula for the parameter. For example, a mean is given by summing the data points for a group then dividing by the number of data points in that group.

    • 3

      Verify that the mathematical formula for the parameter is unbiased. Take the expectation of the mathematical formula. If the result is equal to the parameter of interest, then the formula you have used is unbiased. Otherwise, it is biased and must be modified.

    • 4

      Apply the data to the unbiased mathematical formula. Plug in the data points from your random sample to yield an estimate for the parameter of interest.

    • 5

      Create a confidence interval for the estimate. Confidence intervals generally are the estimate plus or minus the standard error multiplied by a distribution statistic. Consult a statistics textbook for a list of confidence intervals.

Learnify Hub © www.0685.com All Rights Reserved