A regression-based equivalence test for model validation: shifting the burden of proof

Model validation is often realized as a test of how well model predictions match a set of independent observations. One would think that the burden of proof should rest with the model, to force it to show that it can make accurate predictions. Further, one would think that increasing the sample size ought to increase the model's ability to demonstrate its utility. Traditional statistical tools are inappropriate for this because they default to the case that the model and the data are no different, and their ability to detect differences increases with the sample size. These traditional tools are optimized to detect differences, rather than similarities. We present an alternative strategy for model validation that is based on regression and statistical tests of equivalence. Equivalence tests reverse the usual null hypothesis: they posit that the populations being compared are different and use the data to prove otherwise. In this sense, equivalence tests are lumping tests, whereas the traditional statistical tests are splitting tests. To date, model validation with equivalence tests has focused on comparisons of means. Our proposed test checks not only for similarity of means, but also for similarity between individual predictions and observations. The strategy is demonstrated using three case studies that differ in their modeling objectives, and for varied sample sizes. The proposed strategy provides a formal means of model validation that is superior to traditional statistical tests in each case.     

Conservation agriculture practices have changed habitat use by rodent pests: implications for management of feral house mice

Journal of Pest Science - The advent of ‘conservation agriculture’ (CA) farming using zero- or no-tillage practices and an accompanying change in crop rotations in the last...     

Volatile fatty acid profile for grass hay or alfalfa hay fed to alpacas (Vicugna pacos)

The purpose of this study was to determine the diurnal composition and concentration of volatile fatty acids (VFA) and to determine VFA composition and concentration differences between stomach compartment 1 (C1) and caecum of alpacas fed grass and alfalfa hay. The study was divided into two experiments. In Experiment 1 (EXP 1), 10 male alpacas (3+ years old, 65 kg BW) were divided into two groups, housed in drylot pens, provided ad libitum water and fed alfalfa (AH) or grass hay (GH) for 30 days. The alpacas were slaughtered and the digestive tract collected, divided into sub‐tract sections, weighed and digesta sampled for pH, dry matter (DM) and NDF. Volatile fatty acid composition and concentration were determined on C1 and caecal material. Four adult male (3+ years old, 60 kg BW), C1 fistulated alpacas were housed in metabolism crates and divided into two forage groups for Experiment 2 (EXP 2). Alpacas were fed the forages as in EXP 1. Diurnal C1 VFA samples were drawn at 1, 3, 6, 9, 12, 18 and 24 h post‐feeding. There were no differences between forages for tract weight, C1 and caecum digesta DM or NDF. Differences were noted (p < 0.05) for pH between forages and sub‐tract site. Volatile fatty acids concentrations were different (p < 0.05) for forage and site, and total VFA was higher for AH than GH (110.6 and 79.1 mm ) and C1 than caecum (40.7 and 27.6 mm ). Proportion of VFA was significant (p < 0.05) for forage and site, C1 acetate highest for GH (84.8 vs. 74.0 mm ) and caecum acetate 83.7 and 76.2 mm for GH and AH respectively. These data demonstrate the level of VFA produced in C1 and the caecum of alpacas and the diurnal VFA patterns. Composition of VFA is similar to other ruminant species.