Table 2. Multivariate regression on citation count for 85 publications.
| Percent increase in citation count (95% confidence interval) | p-value | |
| Publish in a journal with twice the impact factor | 84% (59 to 109%) | <0.001 |
| Increase the publication date by a month | −3% (−5 to −2%) | <0.001 |
| Include a US author | 38% (1 to 89%) | 0.049 |
| Make data publicly available | 69% (18 to 143%) | 0.006 |
We calculated a multivariate linear regression over the citation counts, including covariates for journal impact factor, date of publication, US authorship, and data availability. The coefficients and p-values for each of the covariates are shown here, representing the contribution of each covariate to the citation count, independent of other covariates.