Table 2.
Control variables and independent variables in the survival analysis.
| Variables | Descriptions |
| TotalPosta | The total number of posts a user has published (excluded from the model due to strong correlation with InitPost) |
| InitPosta | The total number of threads a user initiated |
| NumThreada | The number of threads a user contributed to (excluded from the model due to strong correlation with InitPost) |
| PESb | The number of a user’s posts that provided emotional support |
| PISc | The number of a user’s posts that provided informational support |
| SESd | The number of a user’s posts that sought emotional support |
| SISe | The number of a user’s posts that sought informational support |
| COMf | The number of a user’s posts that were related to companionship |
| RISD | Direct informational support received—the number of informational support posts a user received after initiating a support-seeking thread. |
| RESD | Direct emotional support received—the number of emotional support posts a user received after initiating a support-seeking thread. |
| RISIg | Indirect informational support received—the number of informational support posts a user was exposed to in threads that she or he did not initiate but contributed to. |
| RESIg | Indirect emotional support received—the number of emotional support posts a user was exposed to in threads that she or he did not initiate but contributed to. |
| RCOMg | Companionship received—the number of companionship posts a user was exposed to in threads that she or he did not initiate but contributed to. |
adenotes the three control variables.
bPES: providing emotional support.
cPIS: providing informational support.
dSES: seeking emotional support.
eSIS: seeking informational support.
fCOM: companionship.
gFor RISI, RESI, and RCOM, we assumed that a user read others’ replies that were posted within 7 days before the user’s replies in the same thread.