计量经济分析 = Econometric analysis

25. Multiple Imputation for Nonresponse in Surveys. By D. B. Rubin. ISBN 0 471 08705 X. Wiley, Chichester, 1987. 258 pp. £30.25.

Multiple Imputation for Nonresponse in Surveys

We discuss a method for improving causal inferences called ‘‘Coarsened Exact Matching’’ (CEM), and the new ‘‘Monotonic Imbalance Bounding’’ (MIB) class of matching methods from which CEM is derived. We summarize what is known about CEM and MIB, derive and illustrate several new desirable statistical properties of CEM, and then propose a variety of useful extensions. We show that CEM possesses a wide range of statistical properties not available in most other matching methods but is at the same time exceptionally easy to comprehend and use. We focus on the connection between theoretical properties and practical applications. We also make available easy-to-use open source software for R, Stata, and SPSS that implement all our suggestions.

/pdf/causal-inference-without-balance-checking-coarsened-exact-1xeumfa548.pdf

Causal Inference without Balance Checking: Coarsened Exact Matching

We explore the effect of dimensionality on the nearest neighbor problem. We show that under a broad set of conditions (much broader than independent and identically distributed dimensions), as dimensionality increases, the distance to the nearest data point approaches the distance to the farthest data point. To provide a practical perspective, we present empirical results on both real and synthetic data sets that demonstrate that this effect can occur for as few as 10-15 dimensions. These results should not be interpreted to mean that high-dimensional indexing is never meaningful; we illustrate this point by identifying some high-dimensional workloads for which this effect does not occur. However, our results do emphasize that the methodology used almost universally in the database literature to evaluate high-dimensional indexing techniques is flawed, and should be modified. In particular, most such techniques proposed in the literature are not evaluated versus simple linear scan, and are evaluated over workloads for which nearest neighbor is not meaningful. Often, even the reported experiments, when analyzed carefully, show that linear scan would outperform the techniques being proposed on the workloads studied in high (10-15) dimensionality!.

When is nearest neighbor meaningful

Contrasting treatment‐specific survival using double‐robust estimators

The objective of this research is to advance the evaluation and monetization of well-being improvement programs, offered by population health management companies, by presenting a novel method that robustly monetizes the entirety of well-being improvement within a population. This was achieved by utilizing two employers’ well-being assessments with medical and pharmacy administrative claims (2010–2011) across a large national employer (n = 50,647) and regional employer (n = 6170) data sets. This retrospective study sought to monetize both direct and indirect value of well-being improvement across a population whose medical costs are covered by an employer, insurer, and/or government entity. Logistic regression models were employed to estimate disease incidence rates and input–output modelling was used to measure indirect effects of well-being improvement. These methodological components removed the burden of specifying an exhaustive number of regression models, which would be difficult in small populations. Members who improved their well-being were less likely to become diseased. This reduction saved, per avoided occurrence, US$3060 of total annual health care costs. Of the members who were diseased, improvement in well-being equated to annual savings of US$62 while non-diseased members saved US$26. The method established here demonstrates the linkage between improved well-being and improved outcomes while maintaining applicability in varying populations.

/pdf/the-well-being-valuation-model-a-method-for-monetizing-the-24apgyc19n.pdf

The well-being valuation model: a method for monetizing the nonmarket good of individual well-being

Objective: The objective of this study is to evaluate effectiveness of a firm’s 5-year strategy toward improving well-being while lowering health care costs amidst adoption of a Consumer-Driven Health Plan. Methods: Repeated measures statistical models were employed to test and quantify association between key demographic factors, employment type, year, individual wellbeing, and outcomes of health care costs, obesity, smoking, absence, and performance. Results: Average individual well-being trended upward by 13.5% over 5 years, monthly allowed amount health care costs declined 5.2% on average per person per year, and obesity and smoking rates declined by 4.8 and 9.7%, respectively, on average each year. The results show that individual well-being was significantly associated with each outcome and in the expected direction. Conclusions: The firm’s strategy was successful in driving statistically significant, longitudinal well-being, biometric and productivity improvements, and health care cost reduction.

http://info.healthways.com/hubfs/Science_and_Research/The_Value_of_a_Well_Being_Improvement_Strategy_.4.pdf?t=1454511884111

The Value of a Well-Being Improvement Strategy: Longitudinal Success across Subjective and Objective Measures Observed in a Firm Adopting a Consumer-Driven Health Plan

Objective:To compare utility of employee well-being to health risk assessment (HRA) as predictors of productivity change.Methods:Panel data from 2189 employees who completed surveys 2 years apart were used in hierarchical models comparing the influence of well-being and health risk on longitudinal c

http://info.healthways.com/hubfs/Comparison_of_the_Utility_of_Two_Assessments_for.12_1.pdf?t=1454510439195

Comparison of the Utility of Two Assessments for Explaining and Predicting Productivity Change: Well-Being Versus an HRA.

The growing prevalence of obesity and associated morbidity indicates the need for effective programs for sustainable lifestyle changes to achieve and maintain meaningful weight loss. The Practice-based Opportunities for Weight Reduction (POWER) trial found that a scalable, remotely-delivered intervention achieved significant and meaningful weight loss over 24 months compared to a usual care control group, with results that were statistically comparable to an in-person-based intervention also tested in the study and designed with the same principles. Given that results from clinical trials in research settings do not necessarily translate to larger scaled delivery of interventions, the objective of this study is to evaluate a pilot-implementation of Innergy, a weight loss program designed to replicate the remote-arm of the POWER trial that provided online support and regular telephonic coaching. An intention-totreat retrospective design evaluates weight loss within 6-months among 265 participants offered the program through their employer. The mean change in weight from baseline for the study group was -6.8 lbs. (P<0.0001) and 23% of participants lost 5.0% or more of their initial starting weight. Greater weight loss was associated with greater levels of participation. Specifically, members who utilized a broader range of program support components, took more coaching calls, regularly used the website, and tracked their weight, food and exercise lost significantly more weight. Similarly, longer duration of active participation was associated with greater weight loss. Participants who interacted with the program into the final month of study lost an average of 11.5 lbs, 5.0% of their starting weight. Study results show that Innergy is an effective weight loss program that can be delivered at scale, and thus can have a broad-spanning impact on obesity.

Initial Evaluation of a Scalable Lifestyle Program for Sustained Weight Loss

Large-scale microarray gene expression studies can provide insight into complex genetic networks and biological pathways. A comprehensive gene expression database was constructed using Affymetrix GeneChip microarrays and RNA isolated from more than 6,400 distinct normal and diseased human tissues. These individual patient samples were grouped into over 700 sample sets based on common tissue and disease morphologies, and each set contained averaged expression data for over 45,000 gene probe sets representing more than 33,000 known human genes. Sample sets were compared to each other in more than 750 normal vs. disease pairwise comparisons. Relative up or down-regulation patterns of genes across these pairwise comparisons provided unique expression fingerprints that could be compared and matched to a gene of interest using the Match/X algorithm. This algorithm uses the kappa statistic to compute correlations between genes and calculate a distance score between a gene of interest and all other genes in the d...

Carter Coberley

Papers

The well-being valuation model: a method for monetizing the nonmarket good of individual well-being

The Value of a Well-Being Improvement Strategy: Longitudinal Success across Subjective and Objective Measures Observed in a Firm Adopting a Consumer-Driven Health Plan

Comparison of the Utility of Two Assessments for Explaining and Predicting Productivity Change: Well-Being Versus an HRA.

Initial Evaluation of a Scalable Lifestyle Program for Sustained Weight Loss

Match/X, A gene expression pattern recognition algorithm used to identify genes which may be related to CDC2 function and cell cycle regulation.