Dataset Information

A practical guide to selecting models for exploration, inference, and prediction in ecology.

ABSTRACT: Selecting among competing statistical models is a core challenge in science. However, the many possible approaches and techniques for model selection, and the conflicting recommendations for their use, can be confusing. We contend that much confusion surrounding statistical model selection results from failing to first clearly specify the purpose of the analysis. We argue that there are three distinct goals for statistical modeling in ecology: data exploration, inference, and prediction. Once the modeling goal is clearly articulated, an appropriate model selection procedure is easier to identify. We review model selection approaches and highlight their strengths and weaknesses relative to each of the three modeling goals. We then present examples of modeling for exploration, inference, and prediction using a time series of butterfly population counts. These show how a model selection approach flows naturally from the modeling goal, leading to different models selected for different purposes, even with exactly the same data set. This review illustrates best practices for ecologists and should serve as a reminder that statistical recipes cannot substitute for critical thinking or for the use of independent data to test hypotheses and validate predictions.

SUBMITTER: Tredennick AT

PROVIDER: S-EPMC8187274 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundActive surveillance (AS) is a management option for men diagnosed with lower risk prostate cancer. There is wide variation in all aspects of AS internationally, from patient selection to investigations and follow-up intervals, and a lack of clear evidence on the optimal approach to AS. This study aimed to provide guidance for clinicians from an international panel of prostate cancer experts.MethodsA modified Delphi approach was undertaken, utilising two rounds of online questionnaires followed by a face-to-face workshop. Participants indicated their level of agreement with statements relating to patient selection for AS via online questionnaires on a 7-point Likert scale. Factors not achieving agreement were iteratively developed between the two rounds of questionnaires. Draft statements were presented at the face-to-face workshop for discussion and consensus building.Results12 prostate cancer experts (9 urologists, 2 academics, 1 radiation oncologist) participated in this study from a range of geographical regions (4 USA, 4 Europe, 4 Australia). Complete agreement on statements presented to the participants was 29.4% after Round One and 69.0% after Round Two. Following robust discussions at the face-to-face workshop, agreement was reached on the remaining statements. PSA, PSA density, Multiparametric MRI, and systematic biopsy (with or without targeted biopsy) were identified as minimum diagnostic tests required upon which to select patients to recommend AS as a treatment option for prostate cancer. Patient factors and clinical parameters that identified patients appropriate to potentially receive AS were agreed. Genetic and genomic testing was not recommended for use in clinical decision-making regarding AS.ConclusionsThe lack of consistency in the practice of AS for men with lower risk prostate cancer between and within countries was reflected in this modified Delphi study. There are, however, areas of common practice and agreement from which clinicians practicing in the current environment can use to inform their clinical practice to achieve the best outcomes for patients.

Project description:BackgroundAs in other fields of medicine, development of new medications for management of neuropathic pain has been difficult since preclinical rodent models do not necessarily translate to the clinics. Aside from ongoing pain with burning or shock-like qualities, neuropathic pain is often characterized by pain hypersensitivity (hyperalgesia and allodynia), most often towards mechanical stimuli, reflecting sensitization of neural transmission.Data treatmentWe therefore performed a systematic literature review (PubMed-Medline, Cochrane, WoS, ClinicalTrials) and semi-quantitative meta-analysis of human pain models that aim to induce central sensitization, and generate hyperalgesia surrounding a real or simulated injury.ResultsFrom an initial set of 1569 reports, we identified and analysed 269 studies using more than a dozen human models of sensitization. Five of these models (intradermal or topical capsaicin, low- or high-frequency electrical stimulation, thermode-induced heat-injury) were found to reliably induce secondary hyperalgesia to pinprick and have been implemented in multiple laboratories. The ability of these models to induce dynamic mechanical allodynia was however substantially lower. The proportion of subjects who developed hypersensitivity was rarely provided, giving rise to significant reporting bias. In four of these models pharmacological profiles allowed to verify similarity to some clinical conditions, and therefore may inform basic research for new drug development.ConclusionsWhile there is no single "optimal" model of central sensitization, the range of validated and easy-to-use procedures in humans should be able to inform preclinical researchers on helpful potential biomarkers, thereby narrowing the translation gap between basic and clinical data.SignificanceBeing able to mimic aspects of pathological pain directly in humans has a huge potential to understand pathophysiology and provide animal research with translatable biomarkers for drug development. One group of human surrogate models has proven to have excellent predictive validity: they respond to clinically active medications and do not respond to clinically inactive medications, including some that worked in animals but failed in the clinics. They should therefore inform basic research for new drug development.

Project description:The National Lung Screening Trial (NLST) demonstrated that low dose computed tomography (LDCT) screening could reduce lung cancer mortality by 20% in high-risk individuals. The United States Preventive Services Task Force (USPSTF) and Centers for Medicare and Medicaid Services (CMS) approved lung cancer screening. The NLST, USPSTF and CMS define high risk as smoking ?30 pack-years, smoking within the past 15 years, and being ages 55-74, 55-80 or 55-77. Retrospective studies demonstrated selection using model-estimated risk is superior to NLST-like criteria: higher sensitivity and positive predictive value (PPV), more deaths averted and higher cost-effectiveness. Projects are underway that may additionally support use of risk to determine eligibility. Firstly, the International Lung Screen Trial (ILST) is prospectively enrolling 4,000 individuals for screening if individuals have PLCOm2012 model risk ?1.5% or are USPSTF+ve. Six-year follow-up will allow comparisons. Interim results support the risk approach. Secondly, Cancer Care Ontario started the Lung Cancer Screening Pilot for People at High Risk in order to find optimal design for province-wide programmatic screening. They are enrolling 3,000 individuals to screening based on PLCOm2012 risk ?2%. Some hesitation to recommend screening selection based on model risk comes from the observation that selected individuals are older, have more comorbidities, are expected to have fewer life years and quality-adjusted life years (QALY) and are more likely to die from competing causes. We show that 25.6% of NLST eligible smokers are at low risk (6-year lung cancer incidence proportion =0.008). This group will not benefit from screening but has lower age, fewer comorbidities and fewer competing causes of death. When they are excluded from the NLST+ve group, age, comorbidity count and competing causes of death are similar to those in the PLCOm2012+ve group. In some jurisdictions, model-based lung cancer screening selection needs to take into consideration the elevated risk in blacks and indigenous peoples.

Dataset Information

A practical guide to selecting models for exploration, inference, and prediction in ecology.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets