Multilevel modeling analysis of dyadic network data with an application to Ye'kwana food sharing

Behavioral ecologists have recently begun using multilevel modeling for the analysis of social behavior. We present a multilevel modeling formulation of the Social Relations Model that is well suited for the analysis of dyadic network data. This model, which we adapt for count data and small datasets, can be fitted using standard multilevel modeling software packages. We illustrate this model with an analysis of meal sharing among Ye'kwana horticulturalists in Venezuela. In this setting, meal sharing among households is predicted by an association index, which reflects the amount of time that members of the households are interacting. This result replicates recent findings that interhousehold food sharing is especially prevalent among households that interact and cooperate in multiple ways. We discuss opportunities for human behavioral ecologists to expand their focus to the multiple currencies and cooperative behaviors that characterize interpersonal relationships in preindustrial societies. We discuss possible extensions to this statistical modeling approach and applications to research by human behavioral ecologists and primatologists. Am J Phys Anthropol 157:507–512, 2015. © 2015 Wiley Periodicals, Inc.     

Alternative statistical approaches to the use of data as evidence for hypotheses in human behavioral ecology

In their ambitious Evolutionary Anthropology paper, Winterhalder and Smith 1 review the history, theory, and methods of human behavioral ecology (HBE). In establishing how HBE differs from traditional approaches within sociocultural anthropology, they and others laud its hypothetical‐deductive research method. 1 - 3 Our aim is to critically examine how human behavioral ecologists conduct their research, specifically how they analyze and interpret data as evidence for scientific hypotheses. Through computer simulations and a review of empirical studies of human sex ratios, we consider some limitations of the status quo and present alternatives that could strengthen the field. In particular, we suggest that because human behavioral ecologists often consider multiple hypotheses, they should use statistical approaches that can quantify the evidence in empirical data for competing hypotheses. Although we focus on HBE, the principles of this paper apply broadly within biological anthropology.     

Network measures for dyadic interactions: stability and reliability

Social network analysis (SNA) is a general heading for a collection of statistical tools that aim to describe social interactions and social structure by representing individuals and their interactions as graph objects. It was originally developed for the social sciences, but more recently it was also adopted by behavioral ecologists. However, although SNA offers a full range of exciting possibilities for the study of animal societies, some authors have raised concerns about the correct application and interpretation of network measures. In this article, we investigate how reliable and how stable network measures are (i.e. how much variation they show under re-sampling and how much they are influenced by erroneous observations). For this purpose, we took a data set of 44 nonhuman primate grooming networks and studied the effects of re-sampling at lower re-sampling rates than the originally observed ones and the inclusion of two types of errors, "mis-identification" and "mis-classification," on six different network metrics, i.e. density, degree variance, vertex strength variance, edge weight disparity, clustering coefficient, and closeness centrality. Although some measures were tolerant toward reduced sample sizes, others were sensitive and even slightly reduced samples could yield drastically different results. How strongly a metric is affected seems to depend on both the sample size and the structure of the specific network. The same general effects were found for the inclusion of sampling errors. We, therefore, emphasize the importance of calculating valid confidence intervals for network measures and, finally, we suggest a rough research plan for network studies.     

Data-Mining Discovery of Pattern and Process in Ecological Systems

ABSTRACT Most ecologists use statistical methods as their main analytical tools when analyzing data to identify relationships between a response and a set of predictors; thus, they treat all analyses as hypothesis tests or exercises in parameter estimation. However, little or no prior knowledge about a system can lead to creation of a statistical model or models that do not accurately describe major sources of variation in the response variable. We suggest that under such circumstances data mining is more appropriate for analysis. In this paper we 1) present the distinctions between data-mining (usually exploratory) analyses and parametric statistical (confirmatory) analyses, 2) illustrate 3 strengths of data-mining tools for generating hypotheses from data, and 3) suggest useful ways in which data mining and statistical analyses can be integrated into a thorough analysis of data to facilitate rapid creation of accurate models and to guide further research.     

Information processing in social insect networks

Investigating local-scale interactions within a network makes it possible to test hypotheses about the mechanisms of global network connectivity and to ask whether there are general rules underlying network function across systems. Here we use motif analysis to determine whether the interactions within social insect colonies resemble the patterns exhibited by other animal associations or if they exhibit characteristics of biological regulatory systems. Colonies exhibit a predominance of feed-forward interaction motifs, in contrast to the densely interconnected clique patterns that characterize human interaction and animal social networks. The regulatory motif signature supports the hypothesis that social insect colonies are shaped by selection for network patterns that integrate colony functionality at the group rather than individual level, and demonstrates the utility of this approach for analysis of selection effects on complex systems across biological levels of organization.     

Inference in ecology and evolution

Most ecologists and evolutionary biologists continue to rely heavily on null hypothesis significance testing, rather than on recently advocated alternatives, for inference. Here, we briefly review null hypothesis significance testing and its major alternatives. We identify major objectives of statistical analysis and suggest which analytical approaches are appropriate for each. Any well designed study can improve our understanding of biological systems, regardless of the inferential approach used. Nevertheless, an awareness of available techniques and their pitfalls could guide better approaches to data collection and broaden the range of questions that can be addressed. Although we should reduce our reliance on significance testing, it retains an important role in statistical education and is likely to remain fundamental to the falsification of scientific hypotheses.     

Distinguishing technology from biology: a critical review of the use of GPS telemetry data in ecology

In the past decade, ecologists have witnessed vast improvements in our ability to collect animal movement data through animal-borne technology, such as through GPS or ARGOS systems. However, more data does not necessarily yield greater knowledge in understanding animal ecology and conservation. In this paper, we provide a review of the major benefits, problems and potential misuses of GPS/Argos technology to animal ecology and conservation. Benefits are obvious, and include the ability to collect fine-scale spatio-temporal location data on many previously impossible to study animals, such as ocean-going fish, migratory songbirds and long-distance migratory mammals. These benefits come with significant problems, however, imposed by frequent collar failures and high cost, which often results in weaker study design, reduced sample sizes and poorer statistical inference. In addition, we see the divorcing of biologists from a field-based understanding of animal ecology to be a growing problem. Despite these difficulties, GPS devices have provided significant benefits, particularly in the conservation and ecology of wide-ranging species. We conclude by offering suggestions for ecologists on which kinds of ecological questions would currently benefit the most from GPS/Argos technology, and where the technology has been potentially misused. Significant conceptual challenges remain, however, including the links between movement and behaviour, and movement and population dynamics.     

Network-based diffusion analysis: a new method for detecting social learning

Social learning has been documented in a wide diversity of animals. In free-living animals, however, it has been difficult to discern whether animals learn socially by observing other group members or asocially by acquiring a new behaviour independently. We addressed this challenge by developing network-based diffusion analysis (NBDA), which analyses the spread of traits through animal groups and takes into account that social network structure directs social learning opportunities. NBDA fits agent-based models of social and asocial learning to the observed data using maximum-likelihood estimation. The underlying learning mechanism can then be identified using model selection based on the Akaike information criterion. We tested our method with artificially created learning data that are based on a real-world co-feeding network of macaques. NBDA is better able to discriminate between social and asocial learning in comparison with diffusion curve analysis, the main method that was previously applied in this context. NBDA thus offers a new, more reliable statistical test of learning mechanisms. In addition, it can be used to address a wide range of questions related to social learning, such as identifying behavioural strategies used by animals when deciding whom to copy.     

Bayesian inference in ecology

Bayesian inference is an important statistical tool that is increasingly being used by ecologists. In a Bayesian analysis, information available before a study is conducted is summarized in a quantitative model or hypothesis: the prior probability distribution. Bayes’ Theorem uses the prior probability distribution and the likelihood of the data to generate a posterior probability distribution. Posterior probability distributions are an epistemological alternative to P‐values and provide a direct measure of the degree of belief that can be placed on models, hypotheses, or parameter estimates. Moreover, Bayesian information‐theoretic methods provide robust measures of the probability of alternative models, and multiple models can be averaged into a single model that reflects uncertainty in model construction and selection. These methods are demonstrated through a simple worked example. Ecologists are using Bayesian inference in studies that range from predicting single‐species population dynamics to understanding ecosystem processes. Not all ecologists, however, appreciate the philosophical underpinnings of Bayesian inference. In particular, Bayesians and frequentists differ in their definition of probability and in their treatment of model parameters as random variables or estimates of true values. These assumptions must be addressed explicitly before deciding whether or not to use Bayesian methods to analyse ecological data.     

社会网络分析法及其在动物行为学研究中的应用

近年来,社会网络分析法被广泛用于动物行为学研究。它通过量化动物社会关系中的特定属性 (如中心度和中心势),可辨识动物群体中的关键个体及其在群体中的作用,揭示动物社会交往的形成机制,深化人们对动物社会性起源、个体行为与种群动态格局间关系的理解。本文简要介绍了社会网络分析法的发展史和基本概念,然后阐述了如何建立网络及选择常用指标,强调了创建原模型的必要性及途径。最后着重评述了社会网络分析法在动物行为学研究中的应用现状,并提出未来应当关注直接身体接触 (如打斗和理毛) 和非肢体接触 (如声音通讯) 等不同交往形式的变动过程,以此构建社会交往的动态网络。同时也需要加强种间社交网络的研究,以便增进人们对社会行为的生态功能以及合作理论等问题的理解。     

Infectious disease transmission and contact networks in wildlife and livestock

The use of social and contact networks to answer basic and applied questions about infectious disease transmission in wildlife and livestock is receiving increased attention. Through social network analysis, we understand that wild animal and livestock populations, including farmed fish and poultry, often have a heterogeneous contact structure owing to social structure or trade networks. Network modelling is a flexible tool used to capture the heterogeneous contacts of a population in order to test hypotheses about the mechanisms of disease transmission, simulate and predict disease spread, and test disease control strategies. This review highlights how to use animal contact data, including social networks, for network modelling, and emphasizes that researchers should have a pathogen of interest in mind before collecting or using contact data. This paper describes the rising popularity of network approaches for understanding transmission dynamics in wild animal and livestock populations; discusses the common mismatch between contact networks as measured in animal behaviour and relevant parasites to match those networks; and highlights knowledge gaps in how to collect and analyse contact data. Opportunities for the future include increased attention to experiments, pathogen genetic markers and novel computational tools.     

Music notation: a new method for visualizing social interaction in animals and humans

Background

Researchers have developed a variety of techniques for the visual presentation of quantitative data. These techniques can help to reveal trends and regularities that would be difficult to see if the data were left in raw form. Such techniques can be of great help in exploratory data analysis, making apparent the organization of data sets, developing new hypotheses, and in selecting effects to be tested by statistical analysis. Researchers studying social interaction in groups of animals and humans, however, have few tools to present their raw data visually, and it can be especially difficult to perceive patterns in these data. In this paper I introduce a new graphical method for the visual display of interaction records in human and animal groups, and I illustrate this method using data taken on chickens forming dominance hierarchies.

Results

This new method presents data in a way that can help researchers immediately to see patterns and connections in long, detailed records of interaction. I show a variety of ways in which this new technique can be used: (1) to explore trends in the formation of both group social structures and individual relationships; (2) to compare interaction records across groups of real animals and between real animals and computer-simulated animal interactions; (3) to search for and discover new types of small-scale interaction sequences; and (4) to examine how interaction patterns in larger groups might emerge from those in component subgroups. In addition, I discuss how this method can be modified and extended for visualizing a variety of different kinds of social interaction in both humans and animals.

Conclusion

This method can help researchers develop new insights into the structure and organization of social interaction. Such insights can make it easier for researchers to explain behavioural processes, to select aspects of data for statistical analysis, to design further studies, and to formulate appropriate mathematical models and computer simulations.     

The effects of data collection and observation methods on uncertainty of social networks in wild primates

In social species, network centralities of group members shape social transmission and other social phenomena. Different factors have been found to influence the measurement of social networks, such as data collection and observation methods. In this study, we collected data on adults and juveniles and examined the effect of data collection method (ad libitum sampling vs. focal animal sampling) and observation method (interaction—grooming; play—vs. association—arm-length; 2 m; 5 m proximities—) on social networks in wild vervet monkeys. First, we showed using a bootstrapping method, that uncertainty of ad libitum grooming and play matrices were lesser than uncertainty of focal matrices. Nevertheless, grooming and play networks constructed from ad libitum and focal animal sampling were very similar and highly correlated. We improved the certainty of both grooming and play networks by pooling focal and ad libitum matrices. Second, we reported a high correlation between the proximity arm-length network and the focal grooming one making an arm-length proximity network a reasonable proxy for a grooming one in vervet monkeys. However, we did not find such a correlation between proximity networks and the play one. Studying the effects of methodological issues as data collection and observation methods can help improve understanding of what shapes social networks and which data collection method to choose to study sociality.     

Consider a Spherical Lizard: Animals, Models, and Approximations

Ecologists and physiologists have used biophysical models toanswer questions and investigate hypotheses about animal biologyfor over 20 years, but many investigators do not use such techniquesbecause such modelling is perceived as an arcane art. Indeed,there is no magic strategy to allow all ecologists to modelany biophysical problem accurately by means of simple recipes.In practice, biophysical ecology depends heavily on mathematicaland engineering principles. But, it need not be impenetrable.Here we discuss relatively simple models that can be incorporatedinto many ecological studies. We also discuss some of the importantapproximations and assumptions inherent in our treatments ofradiative, convective, evaporative, and conductive heat transfer.In so doing, we hope to encourage the use of such models, andto engender an appreciation of when and under what conditionspredictions from such models are most likely to be misleading.Thus, we hope to help ecologists to get into and, hopefully,out of trouble in biophysical ecology.     

The irreducible uncertainty of the demography-environment interaction in ecology

The interpretation of ecological data has been greatly improved by bridging the gap between ecological and statistical models. The major challenge is to separate competing hypotheses concerning demography, or other ecological relationships, and environmental variability (noise). In this paper we demonstrate that this may be an arduous, if not impossible, task. It is the lack of adequate ecological theory, rather than statistical sophistication, which leads to this problem. A reconstruction of underlying ecological processes can only be done if we are certain of either the demographic or the noise model, which is something that can only be achieved by an improved theory of stochastic ecological processes. Ignoring the fact that this is a real problem may mislead ecologists and result in erroneous conclusions about the relative importance of endogenous and exogenous factors in natural ecosystems. The lack of correct model identification may also have far-reaching consequences for population management and conservation.     

How to make methodological decisions when inferring social networks

Social network analyses allow studying the processes underlying the associations between individuals and the consequences of those associations. Constructing and analyzing social networks can be challenging, especially when designing new studies as researchers are confronted with decisions about how to collect data and construct networks, and the answers are not always straightforward. The current lack of guidance on building a social network for a new study system might lead researchers to try several different methods and risk generating false results arising from multiple hypotheses testing. Here, we suggest an approach for making decisions when starting social network research in a new study system that avoids the pitfall of multiple hypotheses testing. We argue that best edge definition for a network is a decision that can be made using a priori knowledge about the species and that is independent from the hypotheses that the network will ultimately be used to evaluate. We illustrate this approach with a study conducted on a colonial cooperatively breeding bird, the sociable weaver. We first identified two ways of collecting data using different numbers of feeders and three ways to define associations among birds. We then evaluated which combination of data collection and association definition maximized (a) the assortment of individuals into previously known “breeding groups” (birds that contribute toward the same nest and maintain cohesion when foraging) and (b) socially differentiated relationships (more strong and weak relationships than expected by chance). This evaluation of different methods based on a priori knowledge of the study species can be implemented in a diverse array of study systems and makes the case for using existing, biologically meaningful knowledge about a system to help navigate the myriad of methodological decisions about data collection and network inference.     

Masting,mixtures and modes: are two models better than one?

A key hypothesis in population ecology is that synchronous and intermittent seed production, known as mast seeding, is driven by the alternating allocation of carbohydrates and mineral nutrients between growth and reproduction in different years, i.e. ‘resource switching’. Such behaviour may ultimately generate bimodal distributions of long‐term flower and seed production, and evidence of these patterns has been taken to support the resource switching hypothesis. Here, we show how a widely‐used statistical test of bimodality applied by many studies in different ecological contexts may fail to reject the null hypothesis that focal probability distributions are unimodal. Using data from five tussock grass species in South Island, New Zealand, we find clear evidence of bimodality only when flowering patterns are analyzed with probabilistic mixture models. Mixture models provide a theory oriented framework for testing hypotheses of mast seeding patterns, enabling the different responses underlying medium‐ and high‐ versus non‐ and low‐flowering years to be modelled more realistically by associating these with distinct probability distributions. Coupling theoretical expectations with more rigorous statistical approaches will empower ecologists to reject null hypotheses more often.