Figures (1)  Tables (4)
    • Figure 1. 

      Model estimation results.

    • VariablesProsocial behaviorAntisocial behavior
      Self-reportedParental reportedSelf-reportedParental reported
      Book reading0.011***0.006**−0.007***−0.004**
      (0.002)(0.002)(0.002)(0.001)
      Media use variables
      Watching TV−0.008***−0.007**0.005***0.004***
      (0.002)(0.002)(0.001)(0.001)
      Playing games−0.005**−0.004ϯ0.014***0.006***
      (0.002)(0.002)(0.001)(0.001)
      Demographic variablesControlledControlledControlledControlled
      Cognitive abilityControlledControlledControlledControlled
      Family relationship variablesControlledControlledControlledControlled
      Teachers related variablesControlledControlledControlledControlled
      Peers' relationship variablesControlledControlledControlledControlled
      Class fixed effectControlledControlledControlledControlled
      Constant2.668***2.309***1.955***1.726***
      (0.224)(0.236)(0.162)(0.128)
      Observations7,4877,3497,4917,361
      R20.1880.1630.2120.192
      Adj. R20.1600.1340.1850.164
      F20.6117.6125.9621.50
      Standard errors in parentheses. ϯ p < 0.1, ** p < 0.01, *** p < 0.001. For the clarity of the table, the coefficients of each type of control variable were not listed in detail.

      Table 1. 

      Regression results of study 1.

    • MSDPBSNPNRI
      PB3.6470.7851.000
      SN3.6360.8550.5001.000
      PN3.4540.8990.5870.5451.000
      RI3.1150.9200.2600.1670.1841.000
      Note: The p values of all the correlation coefficients < 0.001. PB = prosocial behavior, SN = social norms, PN = personal norms, RI = reading intensity.

      Table 2. 

      The correlation metrics of variables.

    • Dependent variableSocial normsPersonal normsProsocial behavior
      BSE95% CIBSE95% CIBSE95% CI
      Book reading0.100**0.035[0.031, 0.170]0.059ϯ0.031[−0.002, 0.120]0.113***0.023[0.067, 0.159]
      Social norms0.385***0.035[0.316, 0.453]0.178***0.029[0.122, 0.234]
      Personal norms0.207***0.030[0.148, 0.266]
      Female0.150*0.064[0.025, 0.275]−0.0100.056[−0.119, 0.100]−0.0100.042[−0.092, 0.072]
      Age−0.072***0.019[−0.11, −0.034]−0.057**0.017[−0.091, −0.024]−0.0130.013[−0.039, 0.012]
      Parents' education−0.0160.032[−0.078, 0.047]−0.0090.028[−0.064, 0.045]0.0250.021[−0.016, 0.065]
      Income−0.0680.053[−0.173, 0.036]−0.0640.047[−0.155, 0.028]−0.0180.035[−0.086, 0.051]
      Empathy0.451***0.045[0.363, 0.539]0.516***0.042[0.434, 0.599]0.374***0.035[0.305, 0.442]
      constant2.909***0.387[2.149, 3.669]1.037*0.352[0.345, 1.730]0.675*0.266[1.197, 0.152]
      R20.2000.4520.530
      F25.966***73.264***87.834***
      ϯ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.

      Table 3. 

      Regression results of study 2.

    • BSE95% CI
      Total effect0.1510.025[0.101, 0.201]
      Direct effect0.1130.023[0.067, 0.159]
      Indirect effects
      Total indirect effect0.0380.012[0.015, 0.064]
      Book reading → SN → PB0.0180.007[0.005, 0.033]
      Book reading → PN → PB0.0120.007[0.000, 0.027]
      Book reading → SN → PN → PB0.0080.003[0.002, 0.015]
      Note: The standard error (SE) and 95% CI of indirect effects are based on bias-corrected bootstrap samples. SN = social norms, PN = personal norms, PB = prosocial behavior.

      Table 4. 

      Direct and indirect effects.