Findings

Findings

For the evaluation of this model, CFI, TLI, IFI>0.9，Chi square=147.356, degree of freedom=69, RMSEA<0.5，CHI SQUARE/DF=2.14.

We can see from the structural equation model, the total effect of time spent on SNS on information political participation is 0.074, on civil political participation is 0.06, and radical political participation is 0.02. So we can accept H1: Time spent on SNS is positively linked with political participation. What needs to emphasize here is that there is no direct effect of time spent on SNS on civil political participation and radical political participation, but through information political participation. Also, we can prove H2 by the structural equation model above, there is a positive link between frequency of SNS use and political participation, while time spent on SNS is the mediation.

The total effect of time spent on SNS on information political participation is 0.074, consisting of direct effect 0.13, and the indirect effect -0.057. It indicates that college students spend more time on SNS, they spent more time collecting and sharing public or political information, even though the effect of SNS use is quite limited. Political attitude is important mediation and it is negatively linked between the two variables, time spent on SNS and information political participation. So we accept H3.

We cannot prove H4 in this model, there is no path going through political participation skills. But actually we can see from the model that political participation skills have positive influence on civil political participation and radical political participation.

Table6 Regressions predicting political participation from SNS use variables

Note: * p <0.05， ** p<0.01， *** p < 0.001，**** p < 0.0001

Table7 Regressions predicting political participation from the Internet use variables

Note: * p <0.05， ** p<0.01， *** p < 0.001，**** p < 0.0001

To explore Hypotheses 5a and 5b, the interactions between political participation and SNS and general internet use are examined respectively (Table 6 and 7). Both hypotheses, which predict that the political participation will vary based upon different purposes of use of technology, are supported. Students reporting using the technology for self expression appear to perform more active political participation, while the majority of the users are significantly linked with civil participation and info-participation. It is probably the case that the relatively small number of respondents whose main purpose is playing online games and listening to music makes their answers of political participation vary dramatically, either significantly linked with radical participation or the form of peaceable participation.