|
Most current analytical textbooks (1) describe the statistical concepts and formulas on which data processing in analytical chemistry is based. Classical univariate statistics provides well-established equations (2) that thoroughly deal with these issues. Most science students are familiar with Excel spreadsheets. However, in the author’s experience, translating statistical equations into Excel formulas is not straightforward for students. In particular, assessing which of Excel’s statistical functions perform the calculation that corresponds to the classical equations and determining how to calculate errors from Excel statistical function outputs are not trivial exercises. To help students make the jump from theory to practical data processing, I have used this teaching approach. First step, list all the concepts and formulas involved and prepare a lesson in which all these concepts are explained as though they were completely new. The explanation should be simplified and yet still rigorous. Second step, write a data processing worksheet and show the students how the equations explained in the lesson correspond to the worksheet’s input and output cells. As a practical example, I have created an Excel spreadsheet for applying the calibration straight-line method and the standard-addition straight-line method. Generic data are used and attention is focused on data processing. A lesson giving the complete explanation of the two worksheets included in the spreadsheet is also provided. The spreadsheet and the relevant lesson are available in the online material. These teaching tools are currently used in the first-semester analytical chemistry course, in which students learn the basics on univariate statistics and how to apply them to quantitative analytical chemistry. It is my personal experience that the teaching approach I propose helps students make the transition from theory to application. At the end of the course, at least 80% of the students are able to correctly use two basic statistical Excel functions: LINEST for calculating regression lines and TINV for calculating Student’s t values. LINEST outputs a 3 × 2 matrix: finding the meaning of the elements of the matrix using the Excel help function is not straightforward. TINV is easy and powerful because is avoids wasting time searching through statistical handbooks. Literature Cited- Skoog, D. A.; West, D. M.; Holler, F. J.; Crouch, S. R. Fundamentals of Analytical Chemistry, 8th ed.; Brooks/Cole: Belmont, CA, 2004.
- Miller, J. C.; Miller, J. N. Statistics for Analytical Chemistry, 1st ed.; John Wiley and Sons: New York, 1984.
|
