|
While certainly containing accurate descriptions of least-squares analysis, I think that de Levie's letter misses the point of the visualization, which is to see how minimizing the sum of the squares of the deviations leads to the best least-squares line. I leave it to the authors of the application (1) to discuss the issue of the inclusion or exclusion of the 0 value of the origin point: it would be easy to rewrite the spreadsheet to include this if the authors elect to do so.
In response to de Levie's concerns about my use of the word "trendline" (2), I would like to point out that this was a deliberate choice on my part. The majority of introductory chemistry students first encounter regression analysis in this context; using the JCE Webware application (1) demystifies the algorithm behind the Trendline routine in Excel (3). Students should definitely be exposed to the details of the statistics of regression analysis, and de Levie is correct that the function LinEst and the macro Regression in Excel's software application provide more data (uncertainty estimates for the coefficients). However, these least-squares routines do not provide the visual insight into the fundamental principle of regression analysis—the minimization of the sum of the squares of the deviation—that the JCE Webware application in question does (1). Literature Cited- Kim, M. S.; Burkart, M.; Kim, M.-H. J. Chem. Educ. 2006, 83, 1884.
- Coleman, W. F.; Fedosky, E. W. J. Chem. Educ. 2006, 83, 1884.
- Microsoft Office Excel Home Page (accessed Dec 2007).
|
