Confidence Intervals For Pairwise Differences Research Articles

Pairwise comparisons for Levene-style variability parameters

The simplest way to test for equality of scale in one-way data is to use the analysis of variance applied to absolute deviations from sample medians in place of the original data. This approach started by Levene (1960, using means instead of medians), appears in most statistical packages and is quite powerful for detecting heterogeneity of scale. However, researchers often want to know where the differences lie and some measure of effect for those differences. Here, the Closed Method of pairwise comparisons combined with F statistics on the absolute deviations is shown to be an excellent method for detecting differences. In addition, the Bonferroni method is combined with pairwise t-statistics to construct confidence intervals for pairwise differences in scale.

Visceral and Subcutaneous Adiposity Measurements in Adults: Influence of Measurement Site

Excess abdominal adiposity is a known risk factor for cardiovascular diseases. Computed tomography can be used to examine the visceral (VAT) and subcutaneous (SAT) components of abdominal adiposity, but it is unresolved whether single-slice or multi-slice protocols are needed. Nine computed tomography scans were obtained in the lumbar spine region of 24 adults. The nine slices were obtained at three intervertebral positions (L2-L3, L3-L4, and L4-L5) and at 7 mm above and below these locations. Intra-site and inter-site differences in SAT, VAT, total adipose tissue, and the VAT/SAT ratio were examined using ANOVA and confidence intervals for pairwise differences between means. Intervertebral SAT values increased from 103.1 +/- 50.9 (standard deviation) cm(2) at L2-L3 to 153.3 +/- 68.8 cm(2) at L4-L5, whereas the corresponding VAT values decreased from 164.3 +/- 125.4 to 126.0 +/- 82.7 cm(2). The VAT/SAT ratio was not constant, decreasing from 1.8 +/- 1.4 to 0.9 +/- 0.7. Repeated-measures ANOVA indicated significant inter- and intra-site differences (p </= 0.02) for SAT, VAT, and the VAT/SAT ratio at L3-L4 and L4-L5 (p < 0.001). These differences show the limitation of using a single-slice assessment of abdominal fat distribution, both for a subject and between subjects. Furthermore, the sizeable differences in the intra-site scans indicate that precise repositioning is needed for longitudinal studies. In summary, our findings suggest that a multi-site imaging protocol may provide a more complete assessment of abdominal fat stores and distribution than use of a single site.

Large-sample pairwise comparisons among multinomial proportions with an application to analysis of mutant spectra

In studies examining patterns of mutational damage, the primary variables of interest are expressed typically as discrete counts within defined categories of damage. The multinomial distribution is a common model for such data settings. Of interest is statistical comparison of the pattern, or spectrum, of mutation among the various mutant categories. A specific question in such instances is the identification of differences among the spectral categories, identified, e.g., via simultaneous confidence intervals for pairwise differences in or ratios of the multinomial response probabilities. Herein, a selection of possible methods for constructing such intervals is described and compared via computer simulations to determine which are most appropriate for practical use in the evaluation of mutant spectra.