Research talk:Reading time/Work log/2019-01-02
Add topicWednesday, January 2, 2019
[edit]Model selection update.
[edit]I realized that in order to test the hypothesis that readers in the global south are most likely to read longer than readers in the global north in their sessions that we need to include the interaction term for in our model. Correspondingly we need to include in the version of the model with HDI.
| model 2 | model 3 | model 4 | model 5 | ||
|---|---|---|---|---|---|
| Intercept | 8.0375 (0.0083)*** | 8.0276 (0.0083)*** | 8.0188 (0.0083)*** | 8.1658 (0.0084)*** | |
| mobile | 0.1005 (0.0015)*** | 0.1364 (0.0016)*** | 0.1348 (0.0016)*** | -0.1543 (0.0013)*** | |
| mobile : Global South | -0.1396 (0.0024)*** | -0.1490 (0.0024)*** | -0.1957 (0.0024)*** | ||
| mobile : Last in Session | -0.6486 (0.0021)*** | -0.6415 (0.0021)*** | -0.6349 (0.0021)*** | ||
| Global South : Last in Session | -0.0830 (0.0024)*** | -0.1455 (0.0024)*** | |||
| Revision length (bytes) | 0.1762 (0.0004)*** | 0.1758 (0.0004)*** | 0.1758 (0.0004)*** | 0.1762 (0.0004)*** | |
| time to first paint | -0.0171 (0.0006)*** | -0.0172 (0.0006)*** | -0.0171 (0.0006)*** | -0.0165 (0.0006)*** | |
| time to dom interactive | 0.0022 (0.0009)* | 0.0023 (0.0009)** | 0.0024 (0.0009)** | 0.0030 (0.0009)*** | |
| sessionlength | -0.0001 (0.0000)*** | -0.0001 (0.0000)*** | -0.0001 (0.0000)*** | -0.0001 (0.0000)*** | |
| lastinsessionTRUE | 0.9258 (0.0015)*** | 0.9205 (0.0015)*** | 0.9406 (0.0016)*** | 0.6437 (0.0013)*** | |
| nthinsession | 0.0002 (0.0000)*** | 0.0002 (0.0000)*** | 0.0002 (0.0000)*** | 0.0002 (0.0000)*** | |
| dayofweekMon | 0.0925 (0.0020)*** | 0.0925 (0.0020)*** | 0.0926 (0.0020)*** | 0.0991 (0.0020)*** | |
| dayofweekSat | 0.0196 (0.0020)*** | 0.0178 (0.0020)*** | 0.0175 (0.0020)*** | 0.0112 (0.0020)*** | |
| dayofweekSun | 0.0350 (0.0020)*** | 0.0335 (0.0020)*** | 0.0332 (0.0020)*** | 0.0281 (0.0020)*** | |
| dayofweekThu | 0.0552 (0.0019)*** | 0.0550 (0.0019)*** | 0.0548 (0.0019)*** | 0.0530 (0.0020)*** | |
| dayofweekTue | 0.0332 (0.0020)*** | 0.0331 (0.0020)*** | 0.0326 (0.0020)*** | 0.0306 (0.0020)*** | |
| dayofweekWed | 0.0748 (0.0019)*** | 0.0745 (0.0019)*** | 0.0743 (0.0019)*** | 0.0713 (0.0019)*** | |
| usermonth4 | 0.0085 (0.0096) | 0.0084 (0.0096) | 0.0083 (0.0096) | 0.0089 (0.0096) | |
| usermonth5 | 0.0110 (0.0095) | 0.0106 (0.0095) | 0.0104 (0.0095) | 0.0123 (0.0096) | |
| usermonth6 | -0.0100 (0.0097) | -0.0102 (0.0097) | -0.0103 (0.0097) | -0.0082 (0.0098) | |
| usermonth7 | -0.0487 (0.0097)*** | -0.0491 (0.0097)*** | -0.0491 (0.0097)*** | -0.0476 (0.0098)*** | |
| usermonth8 | -0.0115 (0.0097) | -0.0121 (0.0097) | -0.0121 (0.0097) | -0.0098 (0.0097) | |
| usermonth9 | 0.0366 (0.0076)*** | 0.0369 (0.0076)*** | 0.0370 (0.0076)*** | 0.0408 (0.0076)*** | |
| usermonth10 | 0.0029 (0.0075) | 0.0013 (0.0075) | 0.0010 (0.0075) | -0.0049 (0.0075) | |
| GlobalSouthTRUE | 0.1466 (0.0012)*** | 0.2214 (0.0017)*** | 0.2680 (0.0022)*** | 0.3329 (0.0022)*** | |
| R2 | 0.0720 | 0.0724 | 0.0725 | 0.0641 | |
| Adj. R2 | 0.0720 | 0.0724 | 0.0725 | 0.0641 | |
| Num. obs. | 9873641 | 9873641 | 9873641 | 9873641 | |
| RMSE | 14.2331 | 14.2306 | 14.2297 | 14.2940 | |
| ***p < 0.001, **p < 0.01, *p < 0.05 | |||||