Skip to main content
. 2017 Jun 28;284(1857):20170356. doi: 10.1098/rspb.2017.0356

Table 1.

Descriptive characteristics of the study participants (n = 416). Values are expressed as mean ± s.d., with range in parentheses.

variable men (n = 154) women (n = 262)
right hand
 2nd finger length, mm 73.82 ± 4.19
(61.00 to 87.00)
67.77 ± 4.60
(42.80 to 79.00)
 4th finger length, mm 75.27 ± 4.61
(64.00 to 89.00)
68.31 ± 4.76
(39.80 to 79.90)
 2D : 4D ratio 0.982 ± 0.037 0.993 ± 0.037
 2D : 4D normalized indexa 0.993 ± 0.034 0.986 ± 0.035
 2D : 4D normalized indexb 0.994 ± 0.033 0.986 ± 0.035
 2D : 4D normalized indexc 2.328 ± 0.079 2.310 ± 0.081
 adjusted 2nd finger length, mma 70.37 ± 2.40 69.79 ± 2.46
 adjusted 2nd finger length, mmb 70.34 ± 2.51 69.82 ± 2.38
left hand
 2nd finger length, mm 74.13 ± 4.10
(61.00 to 90.00)
67.46 ± 4.36
(44.20 to 78.00)
 4th finger length, mm 75.42 ± 4.73
(62.90 to 91.00)
68.08 ± 4.79
(38.80 to 80.00)
 2D : 4D ratio 0.984 ± 0.036 0.992 ± 0.037
 2D : 4D normalized indexa 1.000 ± 0.032 0.983 ± 0.033
 2D : 4D normalized indexb 0.996 ± 0.032 0.985 ± 0.034
 2D : 4D normalized indexc 2.775 ± 0.090 2.730 ± 0.091
 adjusted 2nd finger length, mma 70.67 ± 2.28 69.49 ± 2.32
 adjusted 2nd finger length, mmb 70.64 ± 2.36 69.51 ± 2.24

aTwo-parameter power function with normal, homoscedastic error.

bStraight line with intercept and normal, homoscedastic error.

cPower function ratio. The normalized indices (footnotes a and b) were derived directly from the model residuals [12] in raw arithmetic space, with the 2D : 4D ratio or 2D as the dependent variable and 4D and sex as predictors. Each participant's residual was added to the predicted mean ratio for each sex at the mean 4D length in the whole sample, to obtain an adjusted 2D : 4D ‘ratio’ or 2D free from the influence of 4D length. The normalized index (footnote c) was directly derived from the ratio of 2D to 4D raised to the power of 0.80 and 0.76 in the right and left hand, respectively.