Table 5 

Egonetwork statistics. 

Size 
Ties 
Pairs 
Density 
2Step Reach 



T1 
Mean 
8.6 
30.7 
51 
0.77 
23.2 
SD 
6.1 
32.4 
63.9 
0.23 
16 

BM^{1} 
7.1 
23.6 
35.4 
0.82 
16.7 

D1 
Mean 
12.6 
55.5 
100.4 
0.7 
41.1 
SD 
7.5 
48.9 
114.7 
0.21 
23.8 

BM^{1} 
10.2 
40.3 
68.3 
0.74 
30.7 

W1 
Mean 
11.4 
45.2 
74.1 
0.71 
36 
SD 
5.5 
34.6 
71.9 
0.2 
19.5 

BM^{1} 
8.7 
29.7 
45.2 
0.76 
23.7 

D2 
Mean 
12.9 
63.7 
106.9 
0.72 
40.6 
SD 
7.8 
66.9 
127.3 
0.19 
24.3 

BM^{1} 
10.4 
46 
73.6 
0.74 
30.2 

W2 
Mean 
7.5 
19.8 
31.9 
0.75 
20.7 
SD 
3.8 
18 
35.3 
0.22 
10.9 

BM^{1} 
6 
14 
20.7 
0.79 
14.1 



^{1}BM is the mean of the bootstrap distribution 'Size' is the number of the subject's (ego's) direct companions, i.e., those individuals seen in a group with her at least once. 'Ties' is the number of ties between a subject's direct companions. 'Pairs' is the number of potential ties in the egonetwork, i.e., the number of ties that would exist if all of subject's direct companions were also each other's direct companions. 'Density' is the ratio of actual to potential connections in the egonetwork (Ties/Pairs), indicating how wellconnected companions are to each other. '2Step Reach' is the number of individuals within two degrees of separation from the subject, i.e., the number of friends and friends of friends. For each season, the first column shows the egonetwork statistic, i.e., the mean of the distribution of egonetwork measures across 88 residents; the second column shows the standard deviation of this distribution, and the third column shows the expected value of the egonetwork statistic derived from 1000 bootstrap samples. Bootstrap distributions were generated as described in Methods. The bootstrap distributions for the Size, Ties, Pairs and 2Step Reach statistics are shifted towards lower values with respect to the actual values in all seasons, and the bootstrap distribution of the Density statistic is shifted towards higher values in all seasons. All bootstrap distributions are significantly different from each other across seasons (Wilcoxon rank sum test, P < 0.001). 

de Silva et al. BMC Ecology 2011 11:17 doi:10.1186/147267851117 