Table 2: The results of the regression analysis Equation (1) of each communication system

From: Limitations of time resources in human relationships determine social structures

Communication System Coefficient Estimate Standard Error t-value P-value
Twitter a 1.189567 0.023256 8.15 4.4×10−16
  b 1.309346 0.006815 192.12 Less than 2.0×10−16
755 Group chat a 1.214229 0.004640 46.17 Less than 2.0×10−16
  b 1.269766 0.002294 553.5 Less than 2.0×10−16
755 Wall communication a 1.562142 0.006250 89.94 Less than 2.0×10−16
  b 1.476393 0.002769 533.2 Less than 2.0×10−16
Ameba Pigg a 1.0954104 0.0007440 128.24 Less than 2.0×10−16
  b 1.0939529 0.0003137 3487 Less than 2.0×10−16
Mobile phone a 1.07332 0.15756 0.47 3.2×10−1
  b 1.25628 0.04689 26.795 Less than 2.0×10−16
SMS a 1.24089 0.07815 3.08 3.5×10−3
  b 1.21949 0.02995 40.72 Less than 2.0×10−16
  1. The t-values and the p-values of a measuring the statistical uncertainty in coefficient a are larger than 1. The t-values and the p-values of b measuring the statistical uncertainty in coefficient b are not equal to 0. The coefficient a was larger than 1, that is, the user behaviour data did not obey the null model (C=Nm; a=1). Their adjusted R2 values were 0.990 (Twitter), 0.974 (755 group chat), 0.959 (755 wall communication), 0.997 (Ameba Pigg), 0.994 (mobile phone) and 0.990 (SMS)