The physiology of ticks supports a diverse community of non-pathogenic and pathogenic organisms. This study aims to initially characterize the microbial community present within colony-reared Amblyomma americanum using PCR of the variable region 5 of the 16S rRNA gene followed by semiconductor sequencing and classification of sequence data using the Ribosomal Database Project and MG-RAST analysis tools. Comparison of amplicon library datasets revealed changes in the microbiomes in newly engorged nymphs, newly-molted adults, and aged adults, as well as ticks exposed to different environmental conditions. These preliminary data support the concept that microbe survivorship and diversity are partially dependent upon environmental variables and the sequence of blood feeding, molting, and aging. The maintenance and/or emergence of pathogens in ticks may be dependent in part on temporal changes in the microbial community of the tick microbiome.

No abstract

The COMAT subject examination scores were moderately intercorrelated, as hypothesized, with higher correlations between COMAT and COMLEX-USA Level 2-CE scores. The COMAT performance was predictive of COMLEX-USA Level 2-CE performance.

It is not uncommon to use unidimensional item response theory models to estimate ability in multidimensional data with computerized adaptive testing (CAT). The current Monte Carlo study investigated the penalty of this model misspecification in CAT implementations using different item selection methods and exposure control strategies. Three item selection methods—maximum information (MAXI), a-stratification (STRA), and a-stratification with b-blocking (STRB) with and without Sympson–Hetter (SH) exposure control strategy—were investigated. Calibrating multidimensional items as unidimensional items resulted in inaccurate item parameter estimates. Therefore, MAXI performed better than STRA and STRB in estimating the ability parameters. However, all three methods had relatively large standard errors. SH exposure control had no impact on the number of overexposed items. Existing unidimensional CAT implementations might consider using MAXI only if recalibration as multidimensional model is too expensive. Otherwise, building a CAT pool by calibrating multidimensional data as unidimensional is not recommended.

No abstract