Table 1.
Granger causality test results for the association between self-reported and Google searching time trends data
| No. of lags used in VAR modelsa | Granger causality test p-values |
||
|---|---|---|---|
| Self-reported data precedes Google searching | Google searching precedes self-reported data | ||
| Mental distress | |||
| Depression | 2 | 0.03 | - |
| Anxiety | 3 | 0.62 | - |
| Suicide ideationb | |||
| Self-harm/suicidal ideation | 3 | 0.74 | - |
| Self harmb | |||
| Self-harm | 2 | 0.28 | 0.43 |
| Suicideb | |||
| Self-harm | 2 | 0.26 | 0.67 |
| Self-harm/suicidal ideation | 4 | 0.04 | - |
| Social distress | |||
| Loneliness | 2 | <0.001 | - |
| Abuse | 3 | 0.11 | - |
a
VAR – Vector autoregressive. The number of lags is the number of days between one time trend (e.g. self-reported) and the other (e.g. Google searches).
b
Google search topics were compared to self-reported self-harm and self-harm/suicidal ideation time trends data