By Wib Leonard (Illinois State University)
Information
In an introductory statistics course of typically 40 sociology undergraduate students large university, students engage in data collection, data entry and data manipulation and the conceptual definitions of modeling/law of large numbers/central limit theorem are revealed. We will demonstrate an activity in which empirical data on police salaries in a medium-size midwestern city were collected by students to see how well they modeled the normal curve and empirical rule and/or Chebyshev's theorem. This was done for the population (N=65) and sampling fractions of .10, .25, .50, .75, and .90. All data were entered into an SPSS file. Means and standard deviations were calculated and visually presented in a series of histograms. The normal curve was superimposed on the histograms. The results indicated that the larger the n (sampling fraction) the closer the data modeled the normal curve (law of large numbers/central limit theorem), demonstrating the practical connection between data collection and theory.