Teri Peterson

Model selection for ecologists: the worldviews of AIC and BIC

Ecologists frequently ask questions that are best addressed with a model comparison approach. Under this system, the merit of several models is considered without necessarily requiring that (1) models are nested, (2) one of the models is true, and (3) only current data be used. This is in marked contrast to the pragmatic blend of Neyman-Pearson and Fisherian significance testing conventionally emphasized in biometric texts (Christensen 2005), in which (1) just two hypotheses are under consideration, representing a pairwise comparison of models, (2) one of the models, H0, is assumed to be true, and (3) a single data set is used to quantify evidence concerning H0. As Murtaugh (2014) noted, null hypothesis testing can be extended to certain highly structured multi-model situations (nested with a clear sequence of tests), such as extra sums of squares approaches in general linear models, and drop in deviance tests in generalized linear models. This is especially true when there is the expectation that higher order interactions are not significant or nonexistent, and the testing of main effects does not depend on the order of the tests (as with completely balanced designs). There are, however, three scientific frameworks that are poorly handled by traditional hypothesis testing. First, in questions requiring model comparison and selection, the null hypothesis testing paradigm becomes strained. Candidate models may be non-nested, a wide number of plausible models may exist, and all of the models may be approximations to reality. In this context, we are not assessing which model is correct (since none are correct), but which model has the best predictive accuracy, in particular, which model is expected to fit future observations well. Extensive ecological examples can be found in Johnson and Omland (2004), Burnham and Anderson (2002), and Anderson (2008). Second, the null hypothesis testing paradigm is often inadequate for making inferences concerning the falsification or confirmation of scientific claims because it does not explicitly consider prior information. Scientists often do not consider a single data set to be adequate for research hypothesis rejection (Quinn and Keough 2002:35), particularly for complex hypotheses with a low degree of falsifiability (i.e., Popper 1959:266). Similarly, the support of hypotheses in the generation of scientific theories requires repeated corroboration (Ayala et al. 2008). Third, ecologists and other scientists are frequently concerned with the plausibility of existing or default models, what statistician would consider null hypotheses (e.g., the ideal free distribution, classic insular biogeography, mathematic models for species interactions, archetypes for community succession and assembly, etc.). However, null hypothesis testing is structured in such a way that the null hypothesis cannot be directly supported by evidence. Introductory statistical and biometric textbooks go to great lengths to make this conceptual point (e.g., DeVeaux et al. 2013:511, 618, Moore 2010:376, Devore and Peck 1997:300–303).

Toxicity of the Piscicide Rotenone to Columbia Spotted Frog and Boreal Toad Tadpoles

Abstract The piscicide rotenone is commonly used to remove nonnative fishes from natural aquatic systems. While the effects of rotenone on fish are well documented, the effects of this chemical on amphibians are less well known. We determined the toxicity of the rotenone formulation CFT Legumine (5% rotenone) to three ages—Gosner age ranges 21–25, 30–35, and 40–45—of tadpoles of the Columbia spotted frog Rana luteiventris and the boreal toad Anaxyrus boreas under laboratory conditions. Tadpoles of both species were exposed to 0.1, 0.5, and 1.0 mg/L CFT Legumine (0.005, 0.025, and 0.050 mg/L rotenone, respectively) in static, 96-h exposure trials; surviving individuals were placed in rotenone-free water and raised until metamorphosis. In an additional experiment, Columbia spotted frog tadpoles were exposed to 1.0 mg/L CFT Legumine for 1, 2, 3, or 4 h before being placed in rotenone-free water for the duration of a 96-h exposure period. Tadpole mortality increased with increases in CFT Legumine concentratio...

Blood pressure control and factors predicting control in a treatment-compliant male veteran population.

Study Objective. To estimate blood pressure control and identify treatment variables predicting control in treatment‐compliant, hypertensive, male veterans.

Oral health champions in long-term care facilities-a pilot study.

This preintervention/postintervention pilot study examined impact of onsite support by a dental hygiene champion (DHC) on oral health and quality of life (QOL) of elderly residents in three long-term care facilities (LTCFs) in Arkansas. Oral health and oral health-related QOL were operationalized using the Oral Health Assessment Tool (OHAT) and Geriatric Oral Health Assessment Index (GOHAI), respectively. CNAs in Facility A received standardized oral health education/materials with onsite DHC support. Facility B received education/materials only. Facility C served as control. Data analyses included Wilcoxon-signed rank tests (OHAT) and repeated measures ANOVA (GOHAI) (p ≤ .05). OHAT postintervention data in Facility A showed significant improvements in three measured areas (tongue health, denture status, and oral cleanliness); in Facility B, one area (tongue health); and none in Facility C. No significant differences were found in GOHAI scores across facilities. Findings suggest that the presence of DHCs in LTCFs may positively impact the oral health of CNA-assisted residents.This preintervention/postintervention pilot study examined impact of onsite support by a dental hygiene champion (DHC) on oral health and quality of life (QOL) of elderly residents in three long-term care facilities (LTCFs) in Arkansas. Oral health and oral health-related QOL were operationalized using the Oral Health Assessment Tool (OHAT) and Geriatric Oral Health Assessment Index (GOHAI), respectively. CNAs in Facility A received standardized oral health education/materials with onsite DHC support. Facility B received education/materials only. Facility C served as control. Data analyses included Wilcoxon-signed rank tests (OHAT) and repeated measures ANOVA (GOHAI) (p ≤ .05). OHAT postintervention data in Facility A showed significant improvements in three measured areas (tongue health, denture status, and oral cleanliness); in Facility B, one area (tongue health); and none in Facility C. No significant differences were found in GOHAI scores across facilities. Findings suggest that the presence of DHCs in LTCFs may positively impact the oral health of CNA-assisted residents.

Prior Experience and Physicians Intentions to Adopt EMR

Electronic Medical Records (EMR) are expected to increase efficiency, quality of care, and patient safety, yet most studies find that fewer than 25% of hospitals and physicians have adopted them. Theories posit that prior experience with EMR in medical school and residency programs will lead to greater acceptance of EMR among physicians. This work explores the potential relationship between prior experience and adoption of EMR technology among physicians.

Gender interactions and success

Does gender by itself, or does genders interaction with career variables, better explain the difference between women and mens careers in healthcare management? US healthcare managers were surveyed regarding career and personal experiences. Gender was statistically interacted with explanatory variables. Multiple regression with backwards selection systematically removed non-significant variables. All gender interaction variables were non-significant. Much of the literature proposes that work and career factors impact working women differently than working men. We find that while gender alone is a significant predictor of income, it does not significantly interact with other career variables.

A graphical framework for model selection criteria and significance tests: refutation, confirmation and ecology

Summary 1.In this paper we use a novel graphical heuristic to compare the way four methods: significance testing, two popular information-theoretic approaches (AIC and BIC), and Goods Bayes/Non-Bayes compromise (an underutilized hypothesis testing approach whose demarcation criterion adjusts for n) evaluate the merit of competing hypotheses, e.g., H0 and HA. 2.A primary goal of our work is to clarify the concept of strong consistency in model selection. Explicit considerations of this principle (including the strong consistency of BIC) are currently limited to technical derivations, inaccessible to most ecologists. We use our graphical framework to demonstrate, in simple terms, the strong consistency of both BIC and Goods compromise. 3.Our framework also locates the evaluated metrics (and ICs in general) along a conceptual continuum of hypothesis refutation/confirmation that considers n, parameter number, and effect size. Along this continuum, significance testing, and particularly AIC are refutative for H0, whereas Goods compromise, and particularly BIC are confirmatory for the true hypothesis. 4.Our work graphically demonstrates the well-known asymptotic bias of significance tests for HA, and the incorrectness of using statistically non-consistent methods for point hypothesis testing. To address these issues we recommend: 1) dedicated confirmatory methods with strong consistency like BIC for use in point hypothesis testing and confirmatory model selection; 2) significance tests for use in exploratory/refutative hypothesis testing, particularly when conjoined with rational approaches (e.g., Goods compromise, power analyses) to account for the effect of n on P-values; and 3) asymptotically efficient methods like AIC for exploratory model selection. This article is protected by copyright. All rights reserved.

Model Selection and Regression t-Statistics

ABSTRACT It is shown that dropping quantitative variables from a linear regression, based on t-statistics, is mathematically equivalent to dropping variables based on commonly used information criteria.

The Use of Health Information Technology in Ambulatory Surgery Centers

Many of the more than 1.2 billion ambulatory care visits in the United States in 2011 resulted in a patient being handed-off from one outpatient provider to another. As patients transition from one outpatient provider to another, information gaps present a challenge to safe, continuous, and efficient patient care [1]. Among the most common outpatient transitions are those between referring physicians and ambulatory surgery centers (ASC). The use of Health Information Technology (HIT) is imperative to timely and excellent patient care. Despite the projected benefits of HIT and the frequency of physician hand-offs to ASCs, it has not been at all clear how, or even if, HIT is used internally by ASCs, or if it is used during patient transitions to and from ASCs. We conclude that ASCs in this study have low usage of HIT, both within their Centers and between outpatient settings.

Perceived risk, severity of abuse, expectations, and needs of women experiencing intimate partner violence.

ABSTRACT This prospective, descriptive, correlational study examined perceived risk, severity of abuse, expectations, and needs of women experiencing intimate partner violence (IPV) with arrest of the offender occurring at the time of incident. This study builds on previous research completed on fear and expectations of female victims/survivors of IPV that come to the attention of police, to expand knowledge of women’s experiences once they enter the criminal justice system and to create a comprehensive response to this recognized public health problem (Apsler, Cummins, & Carl, 2002). Forty-three women were interviewed regarding the incident, relationship, and experience. Most of the women in this study reported experiencing mild violence and varied forms of threats. There was a significant relationship between the experience of mild violence, serious violence, sexual violence, threats to victims, threats to objects, and others and nonverbal threats with fear of the offender. However, there were no significant correlations between levels of violence or threats with perceived risk of future physical abuse. As the criminal justice response to this crime has changed with the development of legislation and laws aimed at keeping women safe and holding offenders accountable, further research is needed to understand the experience of IPV victims and support an informed response. Forensic nurses are critical interdisciplinary team members in these efforts and play a significant role in providing expertise, sharing of knowledge, and application of evidence fostering victim-centered approaches to addressing IPV.