To illustrate the interpretation of your term, taking the expit (ex / (1 + ex)) in the edges coefficient from the model containing only the edges term (Table 1, Model E) returns a12Networks composed of valued ties might be binarized for use with ERGM, but this entails 890334415573001 a loss of statistical details. 13A special volume on statnet from the Journal of Statistical Software (Volume 24, 2008) provides a great introduction for the interested user. 14Mutuality, indicating reciprocity in binary directed networks (Wasserman and Faust 1994), really should not be confused with mutualism, a mode of cooperation itself plagued by a confusing proliferation of definitions (Brown 1983, Maynard Smith 1983, Conner 1986, Maynard Smith and Szathmary 1995). The similarity of your two terms within this context is unfortunate.Hum Nat. Author manuscript; offered in PMC 2011 October 1.NolinPageprobability of a tie equal for the density with the network: e-3.440 / (1 + e-3.440) = 0.0311. As in logistic regression, when additional terms are added to the model, the “edges” or PD325901 price intercept term reflects the baseline log-odds of a tie when the values from the other covariates gjhs.v8n9p44 are set to zero. Distance Distance has a considerable impact around the probability of a sharing partnership between two households (Table 1, Model ED). Increasing the distance among two households by 1 km decreases the log-odds of a sharing tie between them by -6.233. More intuitively, odds ratios (OR) could be calculated to evaluate cases. For instance, the odds of a food-giving connection to a household 100m away is twelve instances the odds for a household 500m away (OR = e(-6.233*0.1 ?-6.233*0.5) = 12.1). Kinship A unit enhance in between-household relatedness results inside a 9.612 improve inside the log-odds of a tie (Table 1, Model EK). However, considering that r ordinarily ranges from 0 to 0.5, a “unit increase” in relatedness makes little sense. For comparison, the odds of sharing having a sibling are 37 times the odds of sharing having a initial cousin (OR = e(9.612*0.5 ?9.612*0.125) = 36.eight), and 122 occasions the odds of sharing with an unrelated person (OR = e(9.612*0.5 ?9.612*0) = 122.2)NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author ManuscriptMutuality The mutuality coefficient represents the boost within the log-odds of a sharing tie from household A to household B, provided the presence of a reciprocal tie from B to A. Mutuality (Table 1, Model EM) is really a significant and robust predictor with the log-odds of a sharing relationship among two households. The odds of a tie from A journal.pone.0054688 to B are an impressive 192 instances higher when there is certainly a return sharing partnership from B to A than when a return sharing relationship is absent (OR = e(5.258*1 ?5.258*0) = 192.1). Pairwise Models Models EDK, EDM, EKM in Table 1 present the resulting coefficients from models such as each pair of covariates. When both kinship and distance are entered into the model with each other (Model EKD) there is little modify in the magnitude from the coefficients (kinship: 9.612 vs. 9.604; distance: -6.233 vs. -5.808), suggesting their effects are relatively independent of each and every other.15 Nevertheless, introducing mutuality into a model with either distance or kinship benefits within a modest reduction within the size of the mutuality coefficient (five.258 vs.