DATA-ANALYSIS IN THE COMPUTER-AGE - A MONTE-CARLO COMPUTER-SIMULATION OF ERROR SENSITIVITY IN THE DETERMINATION OF FORMATION-CONSTANTS

As an illustration of a new generation of computer-based statistics, a Monte Carlo approach is described for performing an analysis of error sensitivity in experimental methods. This approach may be more meaningful to students than traditional statistical methods. Also, it can be applied when traditional techniques become impracticably complicated. An example is given of the assessment of error sensitivity in the determination of formation constants and heats of formation of 1:1 donor-acceptor complexes. Application of these procedures not only yields the uncertainty in the results, but also identifies the most critical experimental parameters, optimum experimental conditions, and a comparison of the reliability of alternate experimental methods.