Visual assessment of camouflaged targets with different background similarities

The study investigated the effectiveness of different camouflage designs using a computational image quality index. Camouflaged human targets were presented on a natural landscape and the targets were designed to be similar to the landscape background with different levels of background similarity as estimated by the image index. The targets were presented in front of the observer (central 0 degrees) or at different angles in the left (-7 degrees, -14 degrees, -21 degrees) or right (+7 degrees, +14 degrees, +21 degrees) visual fields. The observer had to detect the target using peripheral vision if the target appeared in the left and right visual fields. The camouflage effectiveness was assessed by detection hit rates, detection times, and subjective ratings on detection confidence and task difficulty. The study showed that the psychophysical measures correlated well with the image similarity index, suggesting a potentially more efficient camouflage effectiveness assessment tool if the relationship between the psychophysical results and the index can be quantified in the future.     

Reactions to a survey among those who were and were not sexually assaulted while serving in the military

Surveys are among the most common methods for evaluating military sexual assault experiences among members of the U.S. military; however, little research has examined how receiving surveys about such sexual assaults might affect recipients. In the present sample of 530 active duty and veteran military personnel, just 10% reported unexpected upset, 11% reported regretting participation in the survey, and 22% reported benefitting from that participation overall. A minority of respondents with a history of sexual assault while in the military reported unexpected upset, although the prevalence was three times higher than that of participants without such history (24% vs 8%). There were no statistically significant differences in perceived regret and benefit of participation in the survey between those with and without a history of sexual assault while in the military. Although limited in number, male military sexual assault survivors (n = 8) were significantly more likely than female survivors to report being more upset by the survey than they had anticipated. Implications for future research are discussed.     

Keith Lehrer on Compatibilism

Keith Lehrer has been publishing on free will and compatiblism since 1960. Our concern here is to present an account of the development on his work on the subject.     

An analysis of U.S. parachuting fatalities: 2000-2004

Between 2000 and 2004, 125 recreational parachuting fatalities were reported in the USA. Each was categorized using a previously reported taxonomy. While patterns of skydiving fatality were largely unchanged since the last such analysis of fatalities from 1993-1999, some variations were noted. As in previous analyses, most of the recent skydiving fatalities were the direct result of human error. Some examples of commonplace human error fatalities were turning low to the ground resulting in a hard landing, accidentally colliding with other skydivers resulting in entangled parachutes, and failure to deploy the reserve parachute correctly in the event of malfunctioning main parachute.     

A General Family of Limited Information Goodness-of-Fit Statistics for Multinomial Data

Maydeu-Olivares and Joe (J. Am. Stat. Assoc. 100:1009–1020, 2005; Psychometrika 71:713–732, 2006) introduced classes of chi-square tests for (sparse) multidimensional multinomial data based on low-order marginal proportions. Our extension provides general conditions under which quadratic forms in linear functions of cell residuals are asymptotically chi-square. The new statistics need not be based on margins, and can be used for one-dimensional multinomials. We also provide theory that explains why limited information statistics have good power, regardless of sparseness. We show how quadratic-form statistics can be constructed that are more powerful than X ² and yet, have approximate chi-square null distribution in finite samples with large models. Examples with models for truncated count data and binary item response data are used to illustrate the theory.     

Religious Involvement among Caribbean Blacks in the United States

This study examined demographic and denominational differences in religious involvement (i.e., organizational, non-organizational, subjective) among Caribbean Blacks (Black Caribbeans) residing in the U.S. using data from the National Survey of American Life. Caribbean Blacks who were born in the U.S. had lower levels of religious involvement than those who immigrated and respondents originating from Haiti (as compared to Jamaica) had higher levels of religious involvement, while persons from Trinidad-Tobago reported lower service attendance than did Jamaicans. Older persons, women and married persons generally demonstrated greater religious involvement than their counterparts, while highly educated respondents expressed lower levels of self-rated religiosity. Denominational differences indicated that Baptists reported high levels of religious involvement; however, in several cases, Pentecostals and Seventh Day Adventists reported greater involvement.     

Assessing Approximate Fit in Categorical Data Analysis

A family of Root Mean Square Error of Approximation (RMSEA) statistics is proposed for assessing the goodness of approximation in discrete multivariate analysis with applications to item response theory (IRT) models. The family includes RMSEAs to assess the approximation up to any level of association of the discrete variables. Two members of this family are RMSEA₂, which uses up to bivariate moments, and the full information RMSEA_n. The RMSEA₂ is estimated using the M₂ statistic of Maydeu-Olivares and Joe (2005, 2006), whereas for maximum likelihood estimation, RMSEA_n is estimated using Pearson's X² statistic. Using IRT models, we provide cutoff criteria of adequate, good, and excellent fit using the RMSEA₂. When the data are ordinal, we find a strong linear relationship between the RMSEA₂ and the Standardized Root Mean Squared Residual goodness-of-fit index. We are unable to offer cutoff criteria for the RMSEA_n as its population values decrease as the number of variables and categories increase.