A comparison of model choice strategies for logistic regression

Purpose:The purpose of this study is to develop and compare model choice strategies in context of logistic regression.Model choice means the choice of the covariates to be included in the model.Design/methodology/approach:The study is based on Monte Carlo simulations.The methods are compared in terms of three measures of accuracy:specificity and two kinds of sensitivity.A loss function combining sensitivity and specificity is introduced and used for a final comparison.Findings:The choice of method depends on how much the users emphasize sensitivity against specificity.It also depends on the sample size.For a typical logistic regression setting with a moderate sample size and a small to moderate effect size,either BIC,BICc or Lasso seems to be optimal.Research limitations:Numerical simulations cannot cover the whole range of data-generating processes occurring with real-world data.Thus,more simulations are needed.Practical implications:Researchers can refer to these results if they believe that their data-generating process is somewhat similar to some of the scenarios presented in this paper.Alternatively,they could run their own simulations and calculate the loss function.Originality/value:This is a systematic comparison of model choice algorithms and heuristics in context of logistic regression.The distinction between two types of sensitivity and a comparison based on a loss function are methodological novelties.

Optimisation of chemical purification conditions for direct application of solid metal salt coagulants:Treatment of peatland-derived diffuse runoff

The drainage of peatland areas for peat extraction, agriculture or bioenergy requires affordable, simple and reliable treatment methods that can purify waters rich in particulates and dissolved organic carbon. This work focused on the optimisation of chemical purification process for the direct dosage of solid metal salt coagulants. It investigated process requirements of solid coagulants and the influence of water quality, temperature and process parameters on their performance. This is the first attempt to provide information on specific process requirements of solid coagulants. Three solid inorganic coagulants were evaluated： aluminium sulphate, ferric sulphate and ferric aluminium sulphate. Pre-dissolved aiuminium and ferric sulphate were also tested with the objective of identifying the effects of in-line coagulant dissolution on purification performance. It was determined that the pre-dissolution of the coagulants had a significant effect on coagulant performance and process requirements. Highest purification levels achieved by solid coagulants, even at 30% higher dosages, were generally lower （5%-30%） than those achieved by pre-dissolved coagulants. Furthermore, the mixing requirements of coagulants pre-dissolved prior to addition differed substantially from those of solid coagulants. The pH of the water samples being purified had a major influence on coagulant dosage and purification efficiency. Ferric sulphate （70 mg/L） was found to be the best performing solid coagulant achieving the following load removals： suspended solids （59%-88%）, total organic carbon （56%-62%）, total phosphorus （87%-90%）, phosphate phosphorus （85%-92%） and total nitrogen （33%-44%）. The results show that the use of solid coagulants is a viable option for the treatment of peatland-derived runoff water if solid coagulant-specific process requirements, such as mixing and settling time, are considered.