Table 3.
Time-lag correlation coefficients between new COVID-19 cases and NAVER relative search volumes related to the coronavirus and coronavirus test in South Korea.
| Day | Coronavirus searches (type of device) | Coronavirus test searches (type of device) | |||||||||||
|
|
Overall | Mobile | Desktop | Overall | Mobile | Desktop | |||||||
| –3 days | |||||||||||||
|
|
r | 0.729 a,b | 0.761 a | 0.546a | 0.770 | 0.756 | 0.677 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| –2 days | |||||||||||||
|
|
r | 0.694 | 0.726 | 0.534 | 0.797 | 0.787 | 0.657 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| –1 day | |||||||||||||
|
|
r | 0.681 | 0.720 | 0.497 | 0.824 | 0.799 | 0.717 a | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| 0 days | |||||||||||||
|
|
r | 0.663 | 0.704 | 0.461 | 0.812 | 0.804 a | 0.638 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| 1 day | |||||||||||||
|
|
r | 0.661 | 0.692 | 0.475 | 0.828 a | 0.804 | 0.705 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| 2 days | |||||||||||||
|
|
r | 0.606 | 0.650 | 0.417 | 0.805 | 0.788 | 0.654 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
| 3 days | |||||||||||||
|
|
r | 0.597 | 0.633 | 0.423 | 0.781 | 0.761 | 0.626 | ||||||
|
|
P value | <.001 | <.001 | <.001 | <.001 | <.001 | <.001 | ||||||
aStrongest correlation for each column.
bItalics represent a strong correlation with r>0.7.