Table 2.
Principal Component analysis on the features of behavioural complexity: proportion of variance, cumulative variances, and Eigenvalues of as well as factor loadings on the first two and three principal components for the simulation and the three pilot data-sets, respectively. Numbers in bold indicate loadings for which the absolute value was above a cut-off value of 0.3. We focused on interpreting loadings with values ≥ 0.3, however the remaining loadings were not neglected in the process of component calculation
Data-set Component No. |
Transfor-mation | Simulation | Rats, individuals | Rats, groups | Hens, groups | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
1 A | 2 B | 1 | 2 | 3 | 1 | 2 | 3 | 1 | 2 | 3 | ||
Proportion of variance [%] | 59.64 | 27.68 | 39.74 | 18.92 | 13.39 | 46.86 | 15.13 | 12.78 | 45.54 | 20.54 | 14.60 | |
Cumulative variance [%] | 59.64 | 87.32 | 39.74 | 58.67 | 72.06 | 46.86 | 62.00 | 74.76 | 45.54 | 66.09 | 80.69 | |
Eigenvalues | 7.75 | 3.60 | 4.37 | 2.08 | 1.47 | 4.37 | 1.66 | 1.41 | 5.47 | 2.46 | 1.75 | |
Number of observed behaviours | log | 0.323 | 0.221 | Const. 1 | Const. | Const. | Const. | Const. | Const. | Const. | Const. | Const. |
Number of observed transitions | log | 0.345 | 0.137 | (1) 2 | (1) | (1) | (1) | (1) | (1) | (1) | (1) | (1) |
Proportion of observed behaviours | arcsin–sqrt | Const. | Const. | Const. | Const. | Const. | Const. | Const. | Const. | Const. | Const. | Const. |
Proportion of observed transitions | arcsin–sqrt | 0.020 | –0.417 | 0.416 | 0.160 | 0.177 | 0.407 | 0.044 | 0.092 | 0.327 | 0.131 | 0.155 |
Shannon diversity index (SDI) frequency behaviours | none | 0.351 | 0.072 | 0.356 | 0.305 | 0.221 | 0.409 | 0.006 | 0.097 | 0.374 | –0.035 | 0.271 |
SDI duration behaviours | none | 0.352 | 0.045 | 0.233 | –0.347 | 0.482 | 0.305 | –0.274 | –0.139 | 0.249 | –0.409 | –0.290 |
SDI frequency transitions | none | 0.334 | 0.186 | 0.133 | 0.474 | 0.280 | 0.286 | –0.136 | 0.356 | 0.311 | –0.043 | 0.457 |
SDI duration transitions | none | 0.343 | 0.118 | 0.145 | –0.478 | 0.414 | 0.105 | –0.708 | –0.067 | 0.266 | –0.512 | –0.053 |
Variance within a transition | log | –0.239 | 0.194 | 0.164 | –0.291 | –0.337 | –0.105 | 0.079 | –0.625 | –0.152 | 0.352 | 0.439 |
Variance between transitions | log | –0.242 | 0.319 | –0.156 | 0.437 | 0.141 | –0.220 | 0.307 | 0.516 | –0.114 | –0.268 | 0.376 |
Median number of transitions | none | –0.027 | –0.477 | 0.388 | 0.072 | 0.061 | 0.383 | –0.028 | 0.178 | 0.314 | 0.078 | 0.193 |
IQR 3 of the median number of transitions | none | –0.079 | –0.386 | 0.366 | –0.116 | –0.081 | 0.240 | 0.172 | –0.339 | 0.367 | –0.060 | –0.174 |
Median of the median transition duration | log | –0.293 | 0.250 | –0.387 | –0.112 | 0.332 | –0.294 | –0.464 | 0.160 | –0.247 | –0.448 | 0.146 |
IQR of the median of the median transition | log | –0.252 | 0.352 | –0.350 | 0.001 | 0.428 | –0.362 | –0.237 | 0.057 | –0.205 | –0.369 | 0.424 |
General Complexity
Transition variability
“Const.” indicates features that were (nearly) constant and showed no variability, i.e. all potential behavioural categories or transitions did occur.
(1) The maximum number of possible transitions was constant; therefore, the number of transitions and the proportions of transitions are equivalent
IQR=interquartile range.