Data contributing to double crossovers were re-examined and, if suspect, regenerated for second analysis.
The overall error rates for LDA and KNN are 0.17 and 0.1, resp., for the first analysis and 0.23 and 0.21, respectively, for the second analysis.
For this first analysis the decay parameter for each gene was set to k i =1.
For the first analysis, ANOVA-1, intensity levels for each QMT were standardized to Z-scores across all samples.
Data were censored if no BE occurred before the patient started ERT for the first analysis and until the closing date for the second analysis.
For the first analysis, we calculated a composite FA-index for each individual based on all traits.
The analyses were performed using a Poisson regression analysis, adjusted for over-dispersion, with treatment as factor included for the first analysis.
For the first analysis, pairs of strains were selected to be representative for their species.
For the first analysis, we computed the odds ratios with 95% confidence intervals for low SES as a function of Type D personality, using logistic regression analysis.
For the first analysis, the measured maximum tumour diameters were compared with the maximum diameter obtained from pathology.
For the first analysis, there were in total three syllables that had multiple tokens, jī, ji̍p, and jû, and could thus be examined for intra-speaker consistency.
