Table 1.
Spearman Correlation Coefficients Relating Tweets and Google Searches With COVID-19 Incidence Between January 1 and April 8, 2020, in the United States.a
| Parameter | Entire Study Duration | P Value | Excluding March 22-24 | P Value | Difference | P Value |
|---|---|---|---|---|---|---|
| Tweets about anosmia | 0.539 | <.001 | 0.483 | <.001 | −0.056 | .299 |
| Tweets about anosmia with URLs, retweets, and replies included | 0.240 | .016 | 0.553 | <.001 | 0.313 | .004 |
| Tweets about nonsmell symptoms | 0.761 | <.001 | 0.756 | <.001 | −0.005 | .467 |
| Tweets about all symptoms (anosmia and nonsmell symptoms) | 0.765 | <.001 | 0.760 | <.001 | −0.005 | .467 |
| Tweets about COVID-19 | 0.848 | <.001 | 0.851 | <.001 | 0.003 | .470 |
| Google searches about anosmia | 0.564 | <.001 | 0.524 | <.001 | −0.040 | .346 |
| Google searches about cough | 0.629 | <.001 | 0.612 | <.001 | −0.017 | .424 |
| Google searches about fatigue | 0.052 | .611 | 0.065 | .531 | 0.013 | .464 |
| Google searches about shortness of breath | 0.732 | <.001 | 0.716 | <.001 | −0.016 | .407 |
| Google searches about fever | 0.749 | <.001 | 0.739 | <.001 | −0.010 | .438 |
| Google searches about all nonsmell symptoms | 0.744 | <.001 | 0.733 | <.001 | −0.011 | .433 |
| Google searches about nasal rinses or sinus irrigations | 0.307 | .002 | 0.272 | .007 | −0.035 | .395 |
| Google searches about COVID-19 | 0.899 | <.001 | 0.893 | <.001 | −0.006 | .416 |
| Google searches about dysgeusia | 0.512 | <.001 | 0.467 | <.001 | −0.045 | .341 |
| Google searches about anosmia from 2019 | −0.223 | .027 |
Abbreviation: COVID-19, coronavirus disease 2019.
a
Significance of differences between Spearman correlations were assessed after performing Fisher r-to-z transformation.