Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: For years social scientists in general and IR researchers
in particular who utilized time-series cross-sectional (TSCS) data
produced overly optimistic findings because the methods they used
failed to account for the manner in which their observations were
interrelated. Observations in TSCS datasets are generally correlated
over time and space. Failure to take these correlations into account
leads to inconsistent standard errors and hypothesis tests. Beck and
Katz 1995 (BK) offer a corrective that has been widely used but has two
limiting features. The BK correction requires the analyst to specify a
precise time-series structure, and it requires that the cross-sectional
correlation structure not change over time. Both of these features
imply substantive assumptions that are undesirable in IR research. For
example, while research on institutions suggests that correlations
among policy outputs across states both exist and may change over time,
the BK solution does not permit the cross-sectional correlation
structure to change over time. Similarly, while the precise form of
serial correlation is typically difficult to establish and may vary
across countries, the BK solution requires an explicit specification of
a common AR structure. We offer a method of calculating consistent TSCS
standard errors that is robust to arbitrary cross-sectional correlation
that changes over time and to arbitrary serial correlation that changes
across units. We offer Monte Carlo simulations and reanalyses of
published work in IR to demonstrate its utility.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: In Arizona v Fulminante (1991), the Supreme Court gave appellate courts the power to subject confession evidence to harmless error analysis. In the first test of actual judges’ abilities to do so, the present study found that judges are capable of accurate harmless error analysis. Participants’ assessment of a confession’s coerciveness was not unduly influenced by other evidence, and they provided guilt judgments identical to control conditions when asked to imagine that the confession were never introduced into evidence.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This article examines published corrections in the New York Times and Washington Post in 2007 and 1997. The Times corrected more than twice as many errors in 2007 than in 1997. Wrong descriptions and spelling/typographical errors constituted the bulk of these additional errors. The Post corrected slightly more errors in 2007 than in 1997. Objective errors, mostly notably wrong descriptions, wrong names, wrong numbers and wrong explanations, were the most corrected errors in both years.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: What happens when there’s an imperfect fit between everyday terms and the special definitions for them respondents are supposed to apply in a survey? Everyday concepts are often quite complicated and may allow for multiple senses, but surveys may impose definitions that depart from these everyday meanings. We examined two concepts — residence and disability — that appear in many surveys. Our first study used several methods to explore the everyday senses of these concepts. In our second experiment, respondents got definitions for residence and disability and classified vignettes describing concrete instances. For each concept, one definition reflected the everyday sense of the concept and the second departed from it. The vignettes varied in how well the instance matched each definition’s requirements. Participants who got the technical definition for residence applied it with reasonable accuracy, but those who got the technical definition for disability fell back on their everyday sense of the term. In addition, participants were better at classifying vignettes that closely matched the definition (central instances) than ones that matched it only partly (peripheral instances). Our third experiment encouraged participants to pay more attention to the definitions by giving the concepts unfamiliar labels (e.g., calling a residence an enumeration unit). Although the use of unfamiliar terminology increased the proportion of respondents who consulted the definitions, respondents still relied on their everyday sense of the concepts, especially for disability, and they were better at classifying central than peripheral instances. Respondents have difficulty using concepts in unfamiliar ways and, when they have to, they often make errors. Special definitions for everyday concepts are common in surveys; our results suggest that they can produce reporting errors because respondents don’t realize that the technical sense of a concept differs from the everyday sense and because some situations don’t map neatly onto either sense.

Publication Type: Abstract
Abstract: Objectives: Approximately 1.5 million preventable adverse drug events (ADEs) occur each year in the United States (IOM, 2006). For each preventable ADE in a hospital setting, an estimated $8,750 is added to the cost of a hospital stay (IOM, 2006). Medication Errors is a newly developed course at the University of Wyoming School of Pharmacy designed to educate multidisciplinary students as to why the error(s) occurred, and instruct them in prevention techniques. In addition, the students become familiar with available medication error reporting systems. Methods: The course is a one credit hour elective that includes a combination of lecture and student discussion. The student discussion revolves around pertinent medication error cases relating to the weekly lecture topic. Following small group discussions, all case-specific results and recommendations are shared. Students complete small-group presentations discussing a medication error encountered at one of the student’s work sites. These presentations include how the medication error occurred, and the means taken to prevent the medication error from reoccurring. Results: There are nine pharmacy students (four second-year students, five third-year students) and one pre-nursing student enrolled. This interprofessional enrollment has facilitated education of fellow healthcare providers’ roles in patient care. Further results will be available upon completion of the course in May 2007. Implications: Establishing an interprofessional medication errors elective promotes an understanding that various healthcare providers working collaboratively will enhance patient care and medication safety, and potentially reduce extraneous costs associated with preventable medication errors.