Appendix 2: Tables 14 and 15 reports decomposition results on contributions to the male-female test score gap at mean and selected percentiles by grade for Reading (Numeracy).
The main parameters used to describe mean location and spread of the data were the geometric mean of the log-normal distribution (identical to its median), which is less susceptible to extreme values than the arithmetic mean, and selected quantiles.
Tables 3 and 4 report the estimated total male-female test score gap, together with its contributing factors at the mean and selected percentiles, separated by grades for reading and numeracy, respectively.
The τ is determined by the Beta-binomial method with defined marginal mean and selected intraclass correlation.
Six other analytes were detected at lower frequencies: PFDeA (31.3%), Me-PFOSA-AcOH (27.5%), PFOSA (22.2%), PFUA (9.7%), PFHpA (6.2%), and Et-PFOSA-AcOH, (3.4%); their geometric mean and selected percentile concentrations are given as Supplemental Material in Tables S1 S6 (online at http://www.ehponline.org/docs/2007/10598/suppl.pdf).org/docs/2007/10598/suppl.pdf
From the available pixels at the same coordinate, we retained those within 70% of the mean brightness, and selected the pixel exhibiting the median Normalized Difference Vegetation Index (NDVI).
Genes were grouped based on their expression vectors using the k-means algorithm, and selected groups of interest were again organized for similar expression patterns using Euclidean distance and hierarchical clustering with the average linkage clustering method JCVI MeV (<http://www.pfgrc.tigr.org/tools.shtml>, JCVI, Rockville, MD).
The mean energy and selected nutrient intake of lactating women and the control group of women is shown in Table 4.
Summary statistics, including the mean, SD, and selected quantiles (q01, q05, q95, and q99), were calculated for log EC10) (see Supplemental Material, Table S1).
The analytic methods consisted of univariate statistics, including means, standard deviations, and selected percentiles presented as tables of descriptive statistics.
