FINAL EXAM- social networks Flashcards Preview

Advanced animal behaviour > FINAL EXAM- social networks > Flashcards

Flashcards in FINAL EXAM- social networks Deck (47)
Loading flashcards...
1
Q

what is a social network?

A

tool or way of thinking about how to analyze social behavior

2
Q

what field was first to look at social networks?

A

social psychologists

3
Q

what is the best way to visualize social interactions?

A

Network analysis

4
Q

what unit of behavior is most common to look at, and why?

A

-usually stuck looking at dyads as can only look at 2 animals at a time before having to rewind

5
Q

relationships and number of individuals

A
  • 2 individuals: 1 relationship= dyad
  • 3 individuals: 3 relationship= triad
  • 4 individuals: 6 relationship= tetrad
6
Q

what can be used to study order?

A

can study through disorder/ entropy –> want to reduce within social interactions (cognitive ability useful to avoid conflicts)

7
Q

what is known about social behavior and communication for animals in the wild?

A
  1. some form of order and communication needed
  2. ‘cognitive’ capabilities don’t account for all sophistication of social organizations and dynamics
  3. applies to small and large groups
  4. applies especially with coordinated groups and groups with cooperation
8
Q

what is sociometry?

A

mathematical model to make sense of dynamic interactions (look at interactions and quantify them)

–> simple models are ranking and/or hierarchy models (can determine for any social behavior)

9
Q

what is a sociogram?

A

matrices or networks of social interactions

10
Q

sociomatrix

A
  • approximation of the social dynamic of a group for the period of time under study (historical sample)
  • 2 dimensional, looking at dyads
  • order (‘rankings’) emerge from the social rules within the network groups
  • most based on aggression, but can be on other things like submission and play
11
Q

how do you determine the hierarchies structure of aggression within a sociomatrix?

A
  • must have no reversals (perfectly linear)

- look at number of overthrows to determine alpha

12
Q

main objective of social network analysis? (3)

A
  1. detect emergent properties that would not be identifiable in dyads
  2. may help to detect alliances and coalitions
  3. identify triads, tetrads, etc.
13
Q

network analysis is partly overlapping with tools used to study ______?

A

artificial neural networks

14
Q

topology

A
  • defines structure, about space

- a space is defined, qualified and quantified

15
Q

theoretical inspirations for social network analysis

A
  • social psychology
  • study of artificial neural networks
  • topology
16
Q

network graphs are built from ____?

A

association matrices (socio-matrices)

17
Q

2 basic elements of the anatomy of a network graph?

A
  1. Node/ vertices =individual animals

2. Edge/ link/ connection/arc = relationship

18
Q

definition of a relationship?

A

=permanent or long-term nature of the association between 2 individuals

19
Q

definition of an interaction?

A

=actual dyadic behaviour(s) between two individuals, in the moment (here and now), specific to relationships most of the time

20
Q

what is the basic assumption of network analysis?

A

the network (graph) created from interactions, is a valid representation of the relationship

21
Q

directed vs. undirected (parameter within an association matrices)?

A

the directionality of interactions can be indicated (with arrowheads) –> see what direction they go

22
Q

cyclic vs. acyclic (parameter within an association matrices)?

A

presence or absence of loops

23
Q

signed or unsigned (parameter within an association matrices)

A

positive or negative values

24
Q

weighted or unweighted (parameter within an association matrices)?

A

from values in the matrix cells –> usually influence thickness of lines between nodes

25
Q

what can be seen beyond dyads within an association matrices?

A

alliances, coalitions, cliques, etc.

26
Q

two main categories of measures within an association matrices?

A
  1. Connectivity= amount of linkage in network

2. Centrality= somewhat about heterogeneity, how much connectivity present

27
Q

what does having a complete network refer to?

A

=has the maximum number of edges possible for a given number of nodes

*most networks are incomplete

28
Q

regular vs. irregular networks (in reference to centrality)

A

regular= all nodes have same level of connectivity (homogeneity)

irregular= disparity in the number of connectivity (heterogeneity)

29
Q

connectivity measures (6)

A
  1. edge density
  2. degree of a node
  3. intensity
  4. path
  5. path length
  6. network diameter
30
Q

what does ‘edge density’ refer to?

A

CONNECTIVITY MEASURE
fraction of potential edges (0-1), look at how sparse or dense a network is

sparse network= low ED (<0.5)
dense network= high ED (>0.5)

31
Q

what does the ‘degree of a node’ refer to?

A

CONNECTIVITY MEASURE

=how many edges are connected to a node, average degree of network can be computed

32
Q

what does the intensity/ strength refer to in an association matrices?

A

CONNECTIVITY MEASURE

=if weighted edges, can compute sum of the weight of its edges

33
Q

what does the ‘path’ refer to in an association matrices?

A

CONNECTIVITY MEASURE

=sequence of nodes with an edge from each node to the next (no node or edge used twice) –> indicative of pattern of communication within group

34
Q

what does the ‘path length’ refer to in an association matrices?

A

CONNECTIVITY MEASURE
=number of edges, average shortest distance between a focal node and all other nodes in a network

trail=nodes can be revisited
walk= nodes and edges can be re-used

35
Q

what is the shortest path between two nodes called?

A

geodesic

36
Q

what does the ‘network diameter’ refer to in an association matrices?

A

CONNECTIVITY MEASURE

=longest of he path lengths (in edges)

37
Q

measures of centrality (6)

A
  1. hub
  2. reach
  3. centrality indices
  4. clustering coefficient
  5. betweenness (transmission between individuals)
  6. information centrality (behaviour suggestive of info transmission)
38
Q

what does a ‘hub’ refer to in an association matrices?

A

CENTRALITY MEASURE

=network with high intensity or degree

39
Q

what does a ‘reach’ refer to in an association matrices?

A

CENTRALITY MEASURE

=number of nodes connected to a focal node

40
Q

what does a ‘centrality indices’ refer to in an association matrices?

A

CENTRALITY MEASURE
=affinity and eigenvector centrality, central nodes have more reach than non central nodes (gives amount of reach based on specific nodes)

41
Q

what does a ‘clustering coefficient’ refer to in an association matrices?

A

CENTRALITY MEASURE

=measures the level of local connectivity in adjacent trios of nodes, identifies full triangles

42
Q

what is associativity?

A

by sex, age, etc.

-based on subgroups of demographic groups, can extract patterns

43
Q

positive associativity

A

links are consistent within a social category e.g. females

44
Q

negative associativity

A

one male associating with mostly females

45
Q

what is a community within an association matrices?

A

=regions with high connectivity

mostly applies to large groups of individuals

46
Q

what is suggestive of cliques?

A

strong (complete) within community linkage

clique= quasi-independent groups within a group

47
Q

what can networks help to explain? (5)

A
  1. emergent properties of social groups
  2. dominance and dominance hierarchies
  3. ‘personality’ and behavioral syndromes (isolation vs. centrality)
  4. coordinated movements and activities
  5. eavesdropping (who pays attention to who)