Table 2 Model results.
From: Sociodemographic characteristics of missing data in digital phenotyping
Fixed Effects | Accelerometer | GPS | ||
|---|---|---|---|---|
\({\mathrm{e}}^{\upbeta }\) (95% CI) | SD | \({\mathrm{e}}^{\upbeta }\) (95% CI) | SD | |
Intercept | 0.191 (0.089, 0.485) | 0.435 | 0.269 (0.168, 0.459) | 0.255 |
Time (weeks) | 1.005 (1.004, 1.007) | 0.001 | 1.009 (1.007, 1.010) | 0.001 |
iOS user | 1.301 (0.803, 2.114) | 0.246 | 0.660 (0.453, 0.948) | 0.188 |
Male | 0.821 (0.576, 1.171) | 0.179 | 0.822 (0.607, 1.106) | 0.152 |
4-year degree or higher | 0.786 (0.332, 1.839) | 0.437 | 0.688 (0.339, 1.416) | 0.364 |
Non-Hispanic Black | 1.638 (1.059, 2.517) | 0.223 | 1.329 (0.907, 1.953) | 0.196 |
Asian | 0.724 (0.486, 1.100) | 0.205 | 0.898 (0.630, 1.295) | 0.183 |
American Indian | 1.137 (0.637, 2.047) | 0.298 | 1.241 (0.758, 2.044) | 0.255 |
Other/Multiple | 0.978 (0.232, 4.074) | 0.727 | 0.926 (0.257, 3.240) | 0.649 |
Age (10 years) | 1.010 (0.973, 1.048) | 0.019 | 1.011 (0.982, 1.042) | 0.015 |
Random Effects | SD (95% CI) | Groups | SD (95% CI) | Groups |
|---|---|---|---|---|
Level-1: Participant (\({\sigma }_{\gamma }\)) | 1.012 (0.911, 1.127) | 197 | 0.888 (0.799, 0.983) | 197 |
Level-2: Study (\({\sigma }_{\delta }\)) | 0.721 (0.222, 1.729) | 6 | 0.295 (0.014, 0.906) | 6 |
Model | ||||
|---|---|---|---|---|
Observations (N) | 28,218 | 28,053 | ||
Shape \(\omega\) (95% CI) | 0.53 (0.52, 0.54) | 0.64 (0.63, 0.65) | ||
Bayes R2 (95% CI) | 0.384 (0.365, 0.403) | 0.415 (0.391, 0.439) | ||
WAIC (SE) | 430,853.5 (710.8) | 216,522.5 (549.5) | ||
LOO (SE) | 430,856.1 (710.9) | 216,524.8 (549.6) | ||
