netfacs.reciprocity: Calculate reciprocity of probabilities that two elements...

In NetFACS: Network Applications to Facial Communication Data

Calculate reciprocity of probabilities that two elements appear together

Description

For all dyadic combinations that ever appear, this function calculates how reciprocal the conditional probabilities (i.e. probability of A given B, and B given A) of the two elements are. Combinations that are highly reciprocal indicate that the two elements always occur together and might represent a fixed combination, while low reciprocity might indicate that one element is an extension of the other. Values approaching -1 indicate that one element is strongly dependent on the other, but this is not reciprocated; values around 0 indicate that neither is conditional on the other; and values approaching 1 indicate that both values are conditional on each other. If P[A|B] is the larger conditional probability, the reciprocity is calculated as reciprocity = ((P[B|A]/P[A|B]) - (P[A|B] - P[B|A])) * P[A|B].

Usage

netfacs.reciprocity(netfacs.data)

Arguments

netfacs.data

object resulting from netfacs() function

Value

Function returns a data frame with each combination, the reciprocity of conditional occurrence from -1 (one element entirely depends on the other, but not vice versa) to 1 (both elements always occur together)

The directions and conditional probabilities of both elements are also returned

Examples

### how do angry facial expressions differ from non-angry ones?
data(emotions_set)
angry.face <- netfacs(
  data = emotions_set[[1]],
  condition = emotions_set[[2]]$emotion,
  test.condition = "anger",
  ran.trials = 100,
  combination.size = 2
)

netfacs.reciprocity(angry.face)

NetFACS documentation built on Dec. 7, 2022, 1:12 a.m.