RTI Press

We compare traditional survey inference, which is based on probability sample selection and weighting, with a model-based approach based on sampling quotas and model-based weighting. Compared with the traditional approach, the model-based approach more efficiently controls subgroup sample sizes when a large number of rare subgroups are studied. Using data from a national survey of US businesses, we simulated a model-based paradigm and compared estimates with those under the traditional paradigm. In this study, the findings suggest that the model-based approach offers advantages over the traditional sampling approach; however, a hybrid approach capturing the advantages of both paradigms proved best.
Abstract» | External Link» | RTI Press - Methods Report PDF»

In this paper, I use a nationally representative sample of American high school seniors in 1992 to examine change and stability in the employment patterns of youth as they make the transition from high school to college. Students with weak attachments to the labor force in high school tend to remain unemployed during the first year of college. Conversely, students who work in moderation while in high school have the highest odds of enrolling in college and working while doing so. Compared with their nonworking peers, student workers enter college with lower grades and test scores but are equally engaged in school. Socioeconomic factors have little bearing on high school employment, but they are strongly related to postsecondary employment: students who work during the first year of college have fewer socioeconomic resources than nonworking students. The findings highlight the intersection of school and work in young adulthood and its importance when studying the transition from high school to college among contemporary American youth.
Abstract» | External Link» | RTI Press - Research Report PDF»

In the wake of the September 11, 2001, terrorist attacks, Congress provided a fund to help states offset costs for protecting against terrorist attacks and for emergency preparedness. More than one-third of this money is shared equally by all states, with the rest distributed based on the states’ population share, regardless of the potential targets in each state. This paper develops a rational public finance framework for distributing money to states for protecting against terrorist attacks. We propose two allocation criteria: (1) an efficiency criterion that equalizes the marginal expected loss (human and monetary) across all targets and (2) an equity criterion that adjusts payments to states based on their ability to pay for their own protection. These criteria imply a much more concentrated distribution of protection spending in a few highly populated, target-rich states than is now the case. We then explore the additional information required to protect against all types of terrorists. Limiting the set of protected targets to a few that are highly valued by well-funded terrorist groups produces an even more geographically concentrated funding portfolio. Terrorist insurance is preferable for low-likelihood, difficult-to-protect targets, or targets attractive to individual terrorists.
Abstract» | External Link» | RTI Press - Occasional Paper PDF»

Based on the authors' experience, a gap is apparent between technology managers' needs and opportunity assessment resources. Many organizations need a tool for evaluating seed funding or small business investment opportunities, but they lack a means for determining what options are available. This research report describes several opportunity assessment methods and matches the methods to user types in public and private sector organizations in analyzing technical, business, and market information for investment and commercialization decisions. Organizations that may benefit from this review share a common goal of seeking to strengthen the viability of a new technology venture or capitalizing on their investment in new technology. These include venture capitalists who choose investments (in early-stage technology companies) that will return the largest return on investment for their client investors, angel investors placing their own money in pet projects, public or quasi-public new or small business-funding agencies, and academic and government research and development agencies. The authors' focus was to evaluate not only the availability of methods but also the associated analysis framework in order to understand which opportunity assessment methods are optimal in supporting various investment objectives and stakeholders. This report should also provide individuals in the technology transfer and commercialization community with a succinct, useful reference for identifying what methods are available to them and which methods work best in various situations.
Abstract» | External Link» | RTI Press - Methods Report PDF»

Public concern is increasing over the safety of medicines, particularly serious adverse events detected after extensive use of products in the general marketplace. This concern has led to the need for prompt evaluation of safety signals within large populations following drug approval. The most relevant and available data resources primarily include electronic health care claims and electronic medical records and can be used to identify new safety issues and to evaluate known or suspected signals. In this review paper, we (1) summarize the data resources available for detection and evaluation of safety signals and (2) critically describe these resources and methods used in drug safety research. For each type of data resource, we summarize the characteristics and describe the associated applications and appropriate methods. To place each data resource and method in perspective, we provide examples from disease areas with substantial public health impact. We conclude that in certain circumstances these data resources can be valuable for the relatively cost-effective evaluation of serious adverse events in users of specific medications. However, implementation of such research requires a thorough understanding of the strengths and weaknesses of the data sources and the pharmacoepidemiologic methods used for analysis.
Abstract» | External Link» | RTI Press - Method Report PDF» | News Release»

This study estimates the cost burden associated with six major illnesses among Americans age 65 or older: chronic lung disease, ischemic heart disease, stroke, lung cancer, pneumonia, and gastrointestinal illness. These illnesses were selected because of their relatively high impact among older populations and because they include environmental exposures as a significant risk factor. A prevalence-based cost-of-illness approach was applied. Medical costs were estimated from Medicare claims data in 2000. Productivity losses were estimated through regression analyses of National Health Interview Survey data and using age-specific earnings and household production estimates. Estimated costs range from $0.5 billion (gastrointestinal illness) to almost $60 billion (ischemic heart disease). The combined costs of these conditions among the 65-or-older population in 2000 were almost $135 billion. A simple extrapolation of these cost estimates based on population growth and increases in average medical care prices since 2000 implies that the comparable costs in 2007 were more than $196 billion. With the expectation that these costs will continue to increase significantly as the population ages, priority should be given to prevention strategies such as environmental quality improvements.
Abstract» | External Link» | RTI Press - Research Report PDF» | News Release»

Stressful life events have been associated with preterm delivery in some studies but not in others. One cause of this inconsistency may be that different life events have different effects. The author used data collected by the Pregnancy Risk Assessment Monitoring System, a survey of American women with a recent live birth, for 1990–1995 to examine the relationship between individual life events and the risk of preterm delivery overall and by levels of severity. Four events of the 18 examined were associated with an increased risk of at least one category of preterm delivery: being in debt, being injured by a partner, having someone close attempt suicide, and being divorced. Women who reported being in debt had an increased risk of preterm delivery overall and for each level of severity. One event, having a partner who lost his (or her) job, was associated with a decreased risk of preterm delivery. These results provide some support for the theory that increased stress from life events causes preterm delivery. The lack of a pattern by type of stress, expected stressfulness, or severity of prematurity are hard to reconcile with those theories, however.
Abstract» | External Link» | RTI Press - Research Report PDF» | News Release»