Spss 21

The statistical analysis was performed using Statistic Package for Social Science (SPSS) 21.0 software (IBM, Armonk, NY, USA). Quantitative indicators, including the levels of TNF-α, IL-6, and IL-8 in gingival crevicular fluid for both groups, were represented as mean (± standard deviation). Group comparisons for the above-mentioned quantitative indicators were conducted using the t-test. Patient satisfaction survey data and gender distribution fell under the category of categorical variables. Comparative analyses between the two groups for these variables were carried out using the chi-square test or non-parametric tests. A significance level of P<0.05 was considered as statistically significant differences.

All experiments were carried out at least three times with the description of mean ± standard deviation. The significant difference (P < 0.05) was identified with one-way ANOVA and the Duncan test by SPSS software (SPSS 21, IBM, USA).

Statistical analysis used IBM SPSS 21.0.

All statistical data were analyzed in SPSS 21.0 (IBM Corp., Armonk, NY, USA). Scatterplots of each raw neuroanatomical structure volume vs. ICV were created for men and women combined with a superimposed regression line with corresponding confidence intervals. Independent-sample t-tests were run to determine if there were differences in age, ICV, height, and diastolic blood pressure (DBP) between men and women. The participants were stratified in 11 groups according to their ICV, starting from 1000 ml and increasing with 100 ml up to 2100 ml. One-Way ANOVAs were run to determine if the proportion (relative size) of the four different brain tissue types (cortical gray matter, white matter, subcortical gray matter and the ventricles) were different for groups with different ICVs. Two-Way ANOVAs were run to examine the effects of sex and ICV on the relative size of cortical and subcortical gray matter, white matter and the ventricles.
The proportions and residuals methods described above were performed on the ICV-matched subsample and the whole sample. The proportions and residuals methods were also applied to the men- and women-only small vs. large ICV subsamples. To test for significant main effect of sex, as well as sex^*age and sex^*ICV interactions, General linear models (GLMs) were used with age, sex and ICV as between subjects factors and neuroanatomical structure as within-subjects factor. Each neuroanatomical structure was tested separately and the significance threshold was set at p < 0.05, corrected for multiple comparisons using the Bonferroni-Holm method. These analyses were first performed on the neuroanatomical volumes obtained in the ICV-matched subsample, then the raw volumes and the volumes obtained from the different ICV-correction methods in the entire sample, and finally in the men- and women-only small vs. large ICV subsamples. Subsequently, post-hoc t-tests were run to investigate sex differences between the neuroanatomical volumes. The significance threshold was set at p < 0.05, corrected for multiple comparisons using the Bonferroni-Holm method. Boxplots of the standardized residuals for all neuroanatomical structures in the ICV-matched subsample and the different ICV-corrected data were created for men and women separately, and in the men- and women-only small vs. large ICV subsamples boxplots were created for the small and large ICV-groups separately. Lastly, to investigate which ICV-correction method had the best match with the ground truth, i.e., the ICV-matched subsample, a Spearman's rank-order correlation was run between the mean standardized volumes for all 18 neuroanatomical structures obtained with the proportions and residuals method and the ICV-matched subsample.