Table 6.
Usage variables across Birch engagement clusters for the COVID cohort.
| Variable | C0 | C1 | C2 | C3 |
|---|---|---|---|---|
| COVID Birch clusters | ||||
| Signups | 7,099 | 3,847 | 2,597 | 298 |
| Dependent variables used for clustering | ||||
| Journal entry | 6,147 (86.59% [85.78, 87.36]) | 3,141 (81.65% [80.39, 82.84]) | 2,237 (86.14% [84.76, 87.41]) | 150 (50.34% [44.69, 55.97]) |
| Personal goal created | 1,170 (16.48% [15.64, 17.36]) | 642 (16.69% [15.54, 17.9]) | 585 (22.53% [20.96, 24.17]) | 21 (7.05% [4.66, 10.53]) |
| Article created | 184 (2.59% [2.25, 2.99]) | 177 (4.6% [3.98, 5.31]) | 100 (3.85% [3.18, 4.66]) | 5 (1.68% [0.72, 3.87]) |
| Discussion created | 1,186 (16.71% [15.86, 17.59]) | 916 (23.81% [22.49, 25.18]) | 623 (23.99% [22.39, 25.67]) | 16 (5.37% [3.33, 8.54]) |
| Comment created | 2091 (29.45% [28.41, 30.53]) | 1,626 (42.27% [40.71, 43.83]) | 991 (38.16% [36.31, 40.04]) | 64 (21.48% [17.19, 26.49]) |
| Message sent | 831 (11.71% [10.98, 12.47]) | 692 (17.99% [16.81, 19.23]) | 647 (24.91% [23.29, 26.61]) | 67 (22.48% [18.11, 27.56]) |
| Ad-hoc chat | 3,529 (49.71% [48.55, 50.87]) | 2,331 (60.59% [59.04, 62.13]) | 1,697 (65.34% [63.49, 67.15]) | 259 (86.91% [82.61, 90.28]) |
| Booked chat (no value of p) | 1 (0.01% [0.0, 0.08]) | 3 (0.08% [0.03, 0.23]) | 101 (3.89% [3.21, 4.7]) | 298 (100.0% [98.73, 100.0]) |
| Observational variables not used for clustering | ||||
| Successful chat | 794 (11.18% [10.47, 11.94]) | 831 (21.6% [20.33, 22.93]) | 717 (27.61% [25.92, 29.36]) | 269 (90.27% [86.37, 93.14]) |
| Failed chat | 1,066 (15.02% [14.2, 15.87]) | 769 (19.99% [18.76, 21.28]) | 610 (23.49% [21.9, 25.16]) | 230 (77.18% [72.09, 81.58]) |
Reported values are absolute counts and percentages with 95% confidence intervals. Logistic regression models adjusted for demographic variables returned p < 0.001 for all variables unless otherwise specified. Cases where ‘no value of p’ is stated indicate where the algorithm did not converge.