No abstract

The present study reports the development of a brief, quantitative, web-based, psychometrically sound measure-the Generalized Acceptance of EvolutioN Evaluation (GAENE, pronounced "gene") in a format that is useful in large and small groups, in research, and in classroom settings. The measure was designed to measure only evolution acceptance-not related knowledge or religious beliefs. Item development was based on extensive student interviews and pretesting followed by multiple rounds of qualitative review and quantitative validity testing based on expert review (Lawshe, 1975) and multiple rounds of item revision, then by reliability testing of over 600 high school (HS) and post-secondary (PS) students (Study 1, GAENE 1.0). Data analysis strongly supported the reliability and validity of GAENE 1.0, principal components analysis supported a two-factor solution. All the negatively worded items (and only those items) loaded on the second factor. Rasch analysis also suggested the need for items that would be endorsed at the lower end of the person-item scale. The negatively worded items were, therefore, reworded as positives, additional items were generated to attract wider endorsement, two additional rounds of qualitative and quantitative expert validation were conducted, and reliability testing was repeated with over 600 HS and PS students (Study 2, GAENE 2.0). Both reliability and validity indices of GAENE 2.0 were strong (Lawshe content validity index ¼ 0.72; Cronbach's alpha HS ¼ 0.940; Cronbach's alpha PS ¼ 0.948; Cronbach's alphacombined ¼ 0.945). Principal components analysis suggests that GAENE 2.0 measures a single factor. Together with Rasch analysis results, these data provide substantial initial evidence to support the validity and psychometric integrity of the GAENE as a measure of the degree to which high school and college students accept the theory of evolution. The rigorous development process can also serve as a model for others interested in measure development. #

BackgroundThe regional distribution of a disease may provide important insights regarding its pathophysiology, risk factors and clinical care. While sepsis is a prominent cause of death in the United States (US), few studies have examined regional variations with this malady. We identified the national variation in sepsis deaths in the US. We conducted a descriptive analysis of 1999-2005 national vital statistics data from the National Center for Health Statistics summarized at the state-level. We defined sepsis deaths as deaths attributed to an infection, classified according to the International Classification of Diseases, Version 10. We calculated national and state age-adjusted sepsis-attributed mortality rates.ResultsNational age-adjusted sepsis mortality was 65.5 per 100,000 persons (95% CI: 65.8 - 66.0). State level sepsis mortality varied more than two-fold (range 41 to 88.6 per 100,000 persons; median 60.8 per 100,000, IQR 53.9-74.4 per 100,000). A cluster extending from the Southeastern to the mid-Atlantic US encompassed states with the highest sepsis mortality.ConclusionsSepsis mortality varies across the US. The states with highest sepsis mortality form a contiguous cluster in the Southeastern and mid-Atlantic US. These observations highlight unanswered questions regarding the characteristics and care of sepsis.