Table 2.
Results of the hierarchical clustering analysis performed on the principal components extracted from a global principal component analysis based on social behaviors at 21 and 42 days of age.
Principal component and variables | Coordinatea | Cluster 1 ‘Low-solicited inactive animals’ |
Cluster 2 ‘Highly-solicited avoiders’ |
Cluster 3 ‘Socially active animals’ |
---|---|---|---|---|
n = 58 (d21: 29; d42: 29) |
n = 24 (d21: 14; d42: 10) |
n = 38 (d21: 17; d42: 21) |
||
v.testb | v.testb | v.testb | ||
Principal component 1 | – | − 8.45*** | 3.45*** | 6.11*** |
Agonistic given | 0.39 | – | – | – |
Social nosing given | 0.70 | − 5.26*** | – | 4.96*** |
Other social given | 0.75 | − 5.62*** | – | 5.14*** |
Agonistic received | 0.43 | − 3.72*** | – | 2.35* |
Social nosing received | 0.21 | − | – | – |
Other social received | 0.29 | − | – | – |
Social play | 0.66 | − 5.70*** | – | 5.16*** |
Avoidance | 0.29 | − | – | − |
Total (supplementary quantitative) | 0.92 | − 7.52*** | – | 6.69*** |
Principal component 2 | – | – | 6.76*** | − 5.31*** |
Agonistic given | − 0.54 | − 4.01*** | – | 5.16*** |
Social nosing given | − 0.10 | – | – | – |
Other social given | − 0.04 | – | – | – |
Agonistic received | − 0.03 | – | – | |
Social nosing received | 0.31 | – | – | – |
Other social received | − 0.66 | − 3.80*** | 6.94*** | – |
Social play | − 0.23 | – | – | – |
Avoidance | 0.77 | − 2.17* | 4.87*** | – |
Total (supplementary quantitative) | − 0.17 | – | – | – |
Principal component 3 | – | – | 2.70** | |
Agonistic given | 0.48 | – | – | |
Social nosing given | − 0.32 | – | – | – |
Other social given | 0.04 | – | – | – |
Agonistic received | − 0.38 | – | – | – |
Social nosing received | 0.77 | − 2.85** | 5.11*** | – |
Other social received | 0.26 | – | – | – |
Social play | 0.09 | – | – | – |
Avoidance | − 0.23 | – | – | – |
Total (supplementary quantitative) | − 0.08 | – | – |
The table presents coordinates and characterizes clusters with active and supplementary variables.
aCoordinates (loadings) of the variable on the principal component. Variables with high coordinates on the principal component (> 0.50 or < − 0.50, in bold) are positively and negatively correlated to the principal component, respectively;
bOnly significant v.test values (v.test > 1.96 or v.test < − 1.96; * p value < 0.05, ** p value < 0.01, *** p value < 0.001) are shown. Test values for variables are shown within the principal component section for which the variable had the highest coordinate. Principal components and variables with significant v.test values described the clusters obtained from the hierarchical clustering analysis. The sign of the v.test value indicates if the mean of the cluster is significantly lower or greater than the overall mean of all clusters.