Factors affecting Adolescents' behavior and attitudes toward destructive rock lyrics

In a study of rock music preferences, listening/watching behavior, and views on destructive rock lyrics, 894 adolescents in grades 9 through 12 in rural, urban, suburban public, and metropolitan parochial schools were administered a questionnaire. Demographic information on parents was also collected. It was found that 17.5% of the students were fans of rock music with lyrics that promote homicide, suicide, or satanic practices (HSSR). Loglinear and multiple regression analyses were used to analyze the data. Parents' marital status, student sex, race, and school environment (urban, rural, suburban, etc.) were found to be significant predictors of HSSR status. As compared with non-HSSR fans, the HSSR fans were more likely to have parents who were “Never married” or “Remarried” and less likely to have parents in the “married” category. The HSSR fans were more likely to be male, white, and enrolled in urban schools but not parochial schools than expected; HSSR status also significantly predicted other music-related attitudes and behaviors. The HSSR fans reported liking both the sound and the lyrics of rock music more often than did non-HSSR fans, as did females and urban students. The HSSR fans more often felt that children under 10 years of age should be allowed to listen to HSSR lyrics, and they more frequently expressed the conviction that HSSR music does not affect adolescents' homicidal or self-destructive behavior. The HSSR fans and females reported more nearly complete knowledge of lyrics than did non-HSSR fans and males. The HSSR fans reported that they watch more MTV than did any other group except rural students.     

Policy Perspective on Housing and HIV/AIDS

For persons battling HIV/AIDS a stable place to live may decide the length and quality of life itself. It is nearly impossible for a person on the streets to engage in a needed continuous AIDS treatment regimen when the very basic question of where that person will rest his or her head when darkness comes in just a few hours is unresolved. When danger lurks on the streets, when cold numbs the limbs, when tiredness overwhelms the mind, when fear breaks the spirit, a place to call home would make all the difference.     

Effect of short-term erythropoietin therapy in anemic premature infants.

OBJECTIVE: To determine the effectiveness of a 10-day subcutaneous erythropoietin (rHuEpo) course of 300 units per kg per dose plus oral iron compared to oral iron alone in anemic infants during their convalescent phase of illness. STUDY DESIGN: Prospective, randomized trial performed at a 40-bed, teaching, referral, level III, neonatal intensive care unit. Infants with a gestational age at birth of less than 32 weeks, hematocrit of less than or equal to 28% with a corrected reticulocyte count of less than or equal to 5%, postconceptual age of less than 48 weeks or 5 months chronological age, and a diagnosis of anemia of prematurity were considered for inclusion. Major outcome parameters included hematocrit, corrected reticulocyte count and red cell transfusion requirements. RESULTS: A total of 60 infants were enrolled (n=30 per group). Infants randomized to rHuEpo had a significantly higher post-treatment hematocrit and corrected reticulocyte count than infants in the iron only group (p<0.001). There was a trend towards fewer red cell requirements in the rHuEpo group. CONCLUSIONS: The rHuEpo regimen studied here was associated with an acute improvement in hematocrit and corrected reticulocyte counts. This study did not demonstrate a statistically significant decrease in transfusion therapy, in part related to increased subsequent use of rHuEpo in the control group. Taken together, these data demonstrate that this regimen can effectively treat anemia in convalescent premature infants.     

Bayesian estimation of a random effects heteroscedastic probit model

Summary Bayesian analysis is given of a random effects binary probit model that allows for heteroscedasticity. Real and simulated examples illustrate the approach and show that ignoring heteroscedasticity when it exists may lead to biased estimates and poor prediction. The computation is carried out by an efficient Markov chain Monte Carlo sampling scheme that generates the parameters in blocks. We use the Bayes factor, cross‐validation of the predictive density, the deviance information criterion and Receiver Operating Characteristic (ROC) curves for model comparison.