Choose a population parameter pertinent to your major area of interest—a population that has an unknown mean, or if the population is binomial, an unknown probability of success. For example, a marketing major may be interested in the proportion of consumers who prefer a certain brand of diet cola. An economics or finance major might want to estimate the mean annual salary of Internet Web site designers. A management major may wish to estimate the proportion of companies that have mandatory sensitivity training for all employees. A premed student might desire an estimate of the average number of patients treated daily in the emergency room. An accounting major may want to find the percentage of audits that result in substantive changes to a company’s accounting system. We could continue with examples, but the point should be clear:

Choose something of interest to you.

Define the parameter you want to estimate and conduct a pilot study to obtain an initial estimate of the parameter of interest and, more importantly, an estimate of the variability associated with the estimator. A pilot study is a small experiment (perhaps 15 to 20 observations) used to gain information about some phenomenon. The purpose of the study is to help plan more elaborate future experiments. Using the results of your pilot study, determine the sample size necessary to estimate the parameter to within a reasonable bound (of your choice) with a 95% confidence interval. Present the results to your class.