Model validation: an annotated bibliography

This paper contains an annotated bibliography of 316 papers and books relevant to model validation, with emphasis on articles of potential interest to statisticians. The bibliography is accompanied by a brief introduction, a list of some important areas for statistical research, a short glossary, and a key word index.     

Cocoa Pod Borer (Conopomorpha cramerella Snellen) in Papua New Guinea: Biosecurity Models for New Ireland and the Autonomous Region of Bougainville

Cocoa Pod Borer (Conopomorpha cramerella Snellen) (CPB) is an important pest of cocoa. Following its emergence as a pest in East New Britain, Papua New Guinea, in 2006, it was considered relevant to assess its potential spread to other cocoa growing regions. Its likelihood of introduction to the islands of Bougainville and New Ireland from East New Britain Province, Papua New Guinea, was modeled using Monte Carlo simulation. This dispersal model was based around different scenarios, identifying trends rather than explicitly attempting to encapsulate true values. The model suggested that CPB is far more likely to establish on New Ireland than on Bougainville. More important, incertitude resulting from incomplete knowledge of the amount and frequency of cocoa transported between islands had a significant effect on model outputs. Quarantine and agriculture officials will be able to refine these parameter values, and then use the relevant scenarios from those presented here as a guide to develop quarantine procedures. In addition, a contingency model was employed to estimate the optimal sampling effort to use following an incursion of CPB into Bougainville or New Ireland and the seemingly successful implementation of an initial eradication program. The model suggests that at a 1% infestation level, sampling should continue for 2.5–2.7 years (90% CI) after claiming eradication, and this estimate changed little for higher infestation levels. Through modeling variations in sampling intensity, the model also suggested that determining the full spread of CPB is more important than increased sampling within one region.     

DDP—Managerial opportunities and consequences

Applications of DDP are examined via two case studies to determine how the technology was used by management to aid decision-making; the level of management involvement in the MIS (Management Information Systems) design process and the nature of the relationship between managers and data processing departments were also considered. The study showed that because the technology was largely under end user control, better tailored information systems resulted and system output was more directly used by middle and top managers. This only occurred, however, if certain moderators, such as computing and modelling skills, were present in functional departments. DDP also offers opportunities for better relationships with data processing departments but the consequences include the need for better management training and functional support groups for systems implementation. A conceptual model is presented as a theoretical framework to guide organisations when using DDP.     

Configurations of informal social support among older lesbians

Limited research has been conducted on the needs and concerns of older lesbians in the American population. The purpose of this paper is to present findings from a qualitative study of older (aged 55 and older) lesbians' experiences of and perspectives on aging, with a focus on their discussions of social support. Through an in-depth qualitative study with 25 older lesbians, we found that aging lesbians may be less likely to access formal social services due to perceptions of bias within the services and a potential lack of connection with those who do use services. However, participants were creative at formulating informal networks of support. Four heuristic categories (empowered, secure, independent, and precarious) of informal social support were developed to characterize the various ways in which participants configured their support systems. Social support within each category was conceptualized both in terms of intentionality and degree of connectedness. Being partnered did not necessarily equate with a strong sense of social support. Further research needs to be conducted on older lesbians in order to better characterize and respond to their needs as they age.     

Statistical methods for handling missing data to align with treatment policy strategy

The International Council for Harmonization (ICH) E9(R1) addendum recommends choosing an appropriate estimand based on the study objectives in advance of trial design. One defining attribute of an estimand is the intercurrent event, specifically what is considered an intercurrent event and how it should be handled. The primary objective of a clinical study is usually to assess a product's effectiveness and safety based on the planned treatment regimen instead of the actual treatment received. The estimand using the treatment policy strategy, which collects and analyzes data regardless of the occurrence of intercurrent events, is usually utilized. In this article, we explain how missing data can be handled using the treatment policy strategy from the authors' viewpoint in connection with antihyperglycemic product development programs. The article discusses five statistical methods to impute missing data occurring after intercurrent events. All five methods are applied within the framework of the treatment policy strategy. The article compares the five methods via Markov Chain Monte Carlo simulations and showcases how three of these five methods have been applied to estimate the treatment effects published in the labels for three antihyperglycemic agents currently on the market.     

Growing up Without Status: The Integration of Children in Mixed‐Status Families

Over the past three decades, a central new challenge confronting millions of children of immigrants has emerged: growing up in a mixed‐status family in which at least one member lacks legal authorization to live and work in the United States. A body of recent research argues that unauthorized immigrant status is the fundamental determinant of integration for unauthorized immigrants, with intergenerational consequences for their U.S.‐born children. We discuss the immigration and other policies that create the particular social context within which unauthorized immigration status becomes so detrimental for integration. Specifically, we focus on federal and state policies that undermine the very factors thought to protect children and support the integration of new generations of Americans: families and social networks, economic resources and opportunities, and health. We conclude with recommendations for future research.     

The feasibility of goal attainment scaling to measure case resolution in elder abuse and neglect adult protective services intervention

Objectives: This pilot study describes implementation procedures of goal attainment scaling (GAS) and examines the feasibility of using GAS to measure the multifarious intervention outcome of case resolution in elder mistreatment (EM) adult protective services (APS).

Methods: Substantiated EM victims (n = 27) were recruited prospectively from the State of Maine APS. An adapted GAS approach was implemented involving development of a pre-populated goal scale menu and web-based GAS application.

Results: The GAS menu comprised 18 goals and corresponding scales spanning several domains of case resolution: social support, service access, health/functioning, enhancing independence, and protective measures. The overall GAS process had mean length 33.8 min per case. The mean GAS summary t-score (54.3) aligned with theoretical expectations.

Discussion: Without a measure of case resolution, research cannot compare the effectiveness of different EM intervention models. Findings suggest that GAS is a feasible, client-centered strategy to measure the multifarious EM intervention case resolution outcome.     

Where are They from and How Far Must They Go? Examining Location and Travel Distance in U.S. Abortion Fund Patients

The current study examines a secondary data analysis of 3,452 administrative cases from a national abortion fund representing patients who received pledges for financial assistance to pay for an abortion from 2010 to 2015 in the United States, where abortion costs are not covered under federal public health programs. Case data were analyzed to assess patients' geographic origin, and whether or not cases were more likely to originate from states with Medicaid and private insurance restrictions. The anticipated travel distance to the provider and whether or not travel distances had been increasing over time were also examined. Results indicate that the majority of pledges are made to residents of the South, Midwest, states without expanded Medicaid access to abortion, and states that have private insurance restrictions on abortion coverage. Results further indicate that those who receive funding pledges anticipated traveling approximately 225 kilometers (140 miles) on average to access the abortion. This distance increased from 2010 to 2015, with patients seeking second trimester procedures expecting to travel nearly 3 times farther than patients in their first term. Abortion fund patients travel great distances to access an abortion and those distances are increasing over time. It is recommended that policy restricting public funding of abortion be repealed to improve access to abortion not only in the United States, but also in other countries where policy restrictions may impede access to abortion, even when it is legal.     

The effects of food and density of the movement ofArion ater andAriolimax columbianus (Pulmonata: Stylommatophora) between habitats

Summary Experiments were conducted to determine if slug density, or food quantity and quality could cause dispersal by slugs.Arion ater dispersed quickly from overcrowded areas, where high mortality was incurred by this species. Poor food conditions, caused more dispersal than good food conditions, but food had no effect on the mortality ofA. ater. Neither food nor density factors had significant effects on the dispersal ofAriolimax columbianus. However, moreA. columbianus died in the high- than low-density treatments.A. columbianus mortality was unaffected by the food factor. DispersingArion were heavier than non-dispersers, but there was no difference between the weights of dispersing and non-dispersingAriolimax. There were significant seasonal changes in the dispersal rate, mortality and weight of dispersingArion, whereas only the mortality ofAriolimax changed significantly during the experiments. SinceArion ater did not markedly reduce its activity during unsuitable weather it was able to leave areas of poor food and high slug density fairly rapidly.Ariolimax columbianus' large size allowed it to remain inactive during unfavourable weather and permitted this species to be relatively unresponsive to the experimental treatments.