Confidence intervals for the activation energy estimated by few experiments

Arrhenius parameters are usually estimated by substituting data of 3–5 experiments into a linearized kinetic equation. The use of both small population and linearization invalidates the normal hypothesis. The study focuses on estimating valid confidence intervals for the activation energy computed by the isoconversional method. For this method, a comparison of confidence intervals estimated by using Studentʹs distribution and a nonparametric (distribution-free) method has shown that Studentʹs estimates tend to be oversized. Realistic 95% confidence intervals can be constructed by using corrected Studentʹs percentiles ϑn−2, 0.975 of 4.0, 2.5, and 2.3 for 3, 4, and 5 heating rates experiments, respectively.