Table 1.
Manual and automated scoring methods were used to analyze twenty-four 15-minute video segments for social behavior, center zone time, and center zone entries. Each scoring method was correlated to each of the others and a Pearson’s R-value was calculated. Bolded R-values met an arbitrarily good correlation cutoff of 0.95. All social contact data were produced by SocialScan 2.0. The values in parentheses denote the “immobility” value used by SocialScan 2.0 to distinguish “immobile social contact” from mere “social contact.” All correlations produced significant linear models (P < 0.001).
| Social time | Pearson’s R | ||
|---|---|---|---|
| Real-time manual huddling | vs | 16X manual huddling | 0.989 |
| Social proximity (EthoVision 3.0) | 0.897 | ||
| Social proximity (SocialScan 2.0) | 0.904 | ||
| Social contact | 0.972 | ||
| Immobile (0.05) social contact | 0.987 | ||
| Immobile (0.04) social contact | 0.992 | ||
| Immobile (0.03) social contact | 0.991 | ||
| Immobile (0.02) social contact | 0.960 | ||
| 16x manual huddling | vs | Social proximity | 0.879 |
| Immobile (0.04) social contact | 0.980 | ||
| Center zone time | Pearson’s R | ||
|
| |||
| Real-time manual | vs | 16x manual | 0.996 |
| EthoVision 3.0 | 0.998 | ||
| SocialScan 2.0 | 0.999 | ||
| 16x manual | vs | EthoVision 3.0 | 0.993 |
| SocialScan 2.0 | 0.980 | ||
| Center zone entries | Pearson’s R | ||
|
| |||
| Real-time manual | vs | 16x manual | 0.925 |
| EthoVision 3.0 | 0.814 | ||
| SocialScan 2.0 | 0.986 | ||
| 16x manual | vs | EthoVision 3.0 | 0.815 |
| SocialScan 2.0 | 0.950 | ||