Table 3. ANOVA analysis (specialty specific).
Only the percentage of retweets emerges as significant for SJR changes when the independent variable is analyzed while controlling all other variables.
| Df | Sum Sq | Mean Sq | f-value | Pr (>F) | |
| Number of Followers | 1 | 3.905 | 3.905 | 7.882 | 0.009 |
| Number of Tweets | 1 | 0.405 | 0.405 | 0.818 | 0.374 |
| Percentage of Retweets | 1 | 6.839 | 6.839 | 13.805 | 0.001 |
| Twitter Age (Years) | 1 | 1.082 | 1.082 | 2.184 | 0.151 |
| Journal Specialty | 3 | 5.883 | 1.961 | 3.958 | 0.018 |
| Number of Followers: Journal Specialty | 3 | 16.125 | 5.375 | 10.85 | 0.000 |
| Number of Tweets: Journal Specialty | 3 | 1.937 | 0.646 | 1.304 | 0.294 |
| Percentage of Re-tweets: Journal Specialty | 3 | 1.193 | 0.398 | 0.803 | 0.503 |
| Twitter Age (Years): Journal Specialty | 3 | 1.284 | 0.428 | 0.864 | 0.472 |
| Residuals | 27 | 13.376 | 0.495 | NA | NA |