ethno_bayes_consensus: Gives a measure of the confidence we can have in the answers...

In ethnobotanyR: Calculate Quantitative Ethnobotany Indices

Gives a measure of the confidence we can have in the answers in the ethnobotany data.

Description

Determine the probability that informant citations for a given use are 'correct' given informant responses to the use category for each plant, an estimate of each person's prior_for_answers with this plant and use, and the number of possible answers about this plant use.

Usage

ethno_bayes_consensus(data, answers = 2, prior_for_answers, prior = -1)

Arguments

data	is an ethnobotany data set with column 1 'informant' and 2 'sp_name' as row identifiers of informants and of species names respectively. The rest of the columns are the identified ethnobotany use categories. The data should be populated with counts of uses per person (should be 0 or 1 values).
answers	The number of answers available for each question. These are set to '2' because the way the package works at the moment users can have either a '1' or a '0' in the table. It is however, possible to have more.
prior_for_answers	A matrix representing the probability of a given use being more likely. If not provided the function assumes a uniform distribution across uses.
prior	a prior distribution of probabilities over all answers as a matrix. If this is not provided the function assumes a uniform distribution (prior = -1).

Value

A matrix, where columns represent plant use categories and rows represent responses per person and plant (matching the data). Each value represents the bayes_consensus that an answer was 'correct' for a particular use, within the cultural consensus framework.

Warning

Identification for informants and species must be listed by the names 'informant' and 'sp_name' respectively in the data set. The rest of the columns should all represent separate identified ethnobotany use categories. These data should be populated with counts of uses per informant (should be 0 or 1 values).

Application

ethnobotanyR users often have a large number of counts in cells of the data set after categorization (i.e one user cites ten different ‘food’ uses but this is just one category). Most quantitative ethnobotany tools are not equipped for cases where the theoretical maximum number of use reports in one category, for one species by one informant is >1. This function and the bayes_boot function may be useful to work with these richer datasets for the Bayes consensus analysis.

References

Oravecz, Z., Vandekerckhove, J., & Batchelder, W. H. (2014). Bayesian Cultural Consensus Theory. Field Methods, 1525822X13520280. doi: 10.1177/1525822X13520280

Romney, A. K., Weller, S. C., & Batchelder, W. H. (1986). Culture as Consensus: A Theory of Culture and Informant Accuracy. American Anthropologist, 88(2), 313-338.

Alastair Jamieson Lane and Benjamin Grant Purzycki (2016), AnthroTools: Some custom tools for anthropology.

Examples

#Use built-in ethnobotany data example
#assign a non-informative prior to prior_for_answers with 'prior_for_answers=0.5'
ethno_bayes_consensus(ethnobotanydata, answers = 2, prior_for_answers = 0.5, prior = -1)

#Generate random dataset of three informants uses for four species

eb_data <- data.frame(replicate(10,sample(0:1,20,rep=TRUE)))
names(eb_data) <- gsub(x = names(eb_data), pattern = "X", replacement = "Use_")  
eb_data$informant <- sample(c('User_1', 'User_2', 'User_3'), 20, replace=TRUE)
eb_data$sp_name <- sample(c('sp_1', 'sp_2', 'sp_3', 'sp_4'), 20, replace=TRUE)

#assign a non-informative prior to prior_for_answers
eb_prior_for_answers <- rep(0.5, len = nrow(eb_data))

ethno_bayes_consensus(eb_data, answers = 5, prior_for_answers = eb_prior_for_answers)

ethnobotanyR documentation built on Dec. 28, 2022, 2:15 a.m.