Skip to main content
. 2022 Aug 18;1(4):pgac157. doi: 10.1093/pnasnexus/pgac157

Table 2.

Estimated marginal means by gender across language dimensions.

Men Women
Example M SE M SE t df P R2m R2c
Positive emotion terms (%) Brave, success 2.13 0.0327 2.15 0.0327 −4.49 30,637 < 0.001 6.44E-05 0.453
[2.07, 2.20] [2.09, 2.22]
Negative emotion terms (%) Bad, sick 1.39 0.0126 1.41 0.0127 −4.73 37,196 < 0.001 1.65E-04 0.198
[1.36, 1.41] [1.39, 1.44]
Body terms (%) Face, spine 1.90 0.0302 1.86 0.0303 6.28 32,761 < 0.001 3.01E-04 0.400
[1.84, 1.96] [1.80, 1.92]
Impersonal pronouns (%) It, who 0.70 0.0136 0.71 0.0136 −2.35 28,316 .019 2.20E-05 0.281
[0.68, 0.73] [0.68, 0.73]
Analytic thinking 92.88 0.0806 92.80 0.0808 5.94 27,472 < 0.001 9.31E-05 0.305
[92.72, 93.04] [92.64, 92.96]
Cognitive processes (%) Affect, solve 4.63 0.0310 4.67 0.0312 −4.38 30,241 < 0.001 5.64E-04 0.176
[4.57, 4.69] [4.61, 4.73]

Note: these results include various controls reported in the main text and account for ethnicity as a fixed effect in the linear mixed model calculations. Numbers in brackets are 95% CI for the estimated marginal group means and not the difference between the group means, which are used to assess statistical significance. Full model outputs are in the online supplement. R2m refers to the marginal R2, which accounts for variance explained by the fixed effects in linear mixed model calculations (gender and ethnicity). R2c refers to the conditional R2, which accounts for variance explained by the fixed and random effects in linear mixed model calculations.